Understanding animal welfare: the science in its cultural context

Understanding Animal Welfare Dfraser_C000.indd i 6/28/2008 9:33:22 AM The Universities Federation for Animal Welfare...

Author: David Fraser

93 downloads 3364 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Understanding Animal Welfare

Dfraser_C000.indd i

6/28/2008 9:33:22 AM

The Universities Federation for Animal Welfare

UFAW, founded 1926, is an internationally recognised, independent, scientific and educational animal welfare charity concerned with promoting high standards of welfare for farm, companion, laboratory and captive wild animals, and for those animals with which we interact in the wild. It works to improve animals’ lives by: •

Promoting and supporting developments in the science and technology that underpin advances in animal welfare;

•

Promoting education in animal care and welfare;

•

Providing information, organising meetings, and publishing books, videos, articles, technical reports and the journal Animal Welfare;

•

Providing expert advice to government departments and other bodies and helping to draft and amend laws and guidelines;

•

Enlisting the energies of animal keepers, scientists, veterinarians, lawyers and others who care about animals.

“Improvements in the care of animals are not now likely to come of their own accord, merely by wishing them: there must be research … and it is in sponsoring research of this kind, and making its results widely known, that UFAW performs one of its most valuable services.” Sir Peter Medawar CBE FRS, 8th May 1957 Nobel Laureate (1960), Chairman of the UFAW Scientific Advisory Committee (1951–1962) UFAW relies on the generosity of the public through legacies and donations to carry out its work improving the welfare of animal now and in the future. For further information about UFAW and how you can help promote and support its work, please contact us at the address below. Universities Federation for Animal Welfare The Old School, Brewhouse Hill, Wheathampstead, Herts AL4 8AN, UK Tel: 01582 831818 Fax: 01582 831414 Website: www.ufaw.org.uk Email: [email protected]

Dfraser_C000.indd ii

6/28/2008 9:33:22 AM

Understanding Animal Welfare The Science in its Cultural Context David Fraser Animal Welfare Program University of British Columbia Vancouver Canada

Dfraser_C000.indd iii

6/28/2008 9:33:22 AM

This edition ﬁrst published 2008 © 2008 by UFAW Series editor James K. Kirkwood and Robert C. Hubrecht Blackwell Publishing was acquired by John Wiley & Sons in February 2007. Blackwell’s publishing programme has been merged with Wiley’s global Scientiﬁc, Technical, and Medical business to form Wiley-Blackwell. Registered ofﬁce John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom Editorial ofﬁce 9600 Garsington Road, Oxford, OX4 2DQ, United Kingdom 2121 State Avenue, Ames, Iowa 50014-8300, USA For details of our global editorial ofﬁces, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell. The right of the author to be identiﬁed as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Library of Congress Cataloging-in-Publication Data Fraser, David (David F.) Understanding animal welfare : the science in its cultural context / David Fraser. p. cm. Includes bibliographical references and index. ISBN-13: 978-1-4051-3695-2 (pbk. : alk. paper) ISBN-10: 1-4051-3695-2 (pbk. : alk. paper) 1. Animal welfare. 2. Animal experimentation. I. Title. HV4708.F745 2008 179′.3--dc22 2008006202 A catalogue record for this book is available from the British Library. Set in 10/12.5 pt Sabon by Newgen Imaging Systems Pvt. Ltd, Chennai, India Printed in Singapore by Markono Print Media Pte Ltd 1

Dfraser_C000.indd iv

2008

6/28/2008 9:33:23 AM

Contents

Foreword Preface Dedication

vi vii xi

Part I Animal Welfare in Context Introduction 1 Animals and Moral Concern 2 Animals in the Human Mind 3 A Good Life for Animals 4 A Science of Animal Welfare?

1 2 9 24 41 61

Part II Studying Animal Welfare Introduction 5 Disease, Injury and Production 6 ‘Stress’ 7 Abnormal Behaviour 8 Affective States 9 ‘Natural’ Living 10 Preferences and Motivation

79 80 84 104 125 146 169 191

Part III Drawing Conclusions about Animal Welfare Introduction 11 How Do the Different Measures Relate to Each Other? 12 Selecting and Combining Criteria of Animal Welfare 13 Animal Welfare, Values and Mandated Science

217 218 222 241 260

Coda References Index

275 287 309

v

Dfraser_C000.indd v

6/27/2008 5:35:56 PM

Foreword

The welfare of animals has, in recent years, become a matter of widespread and prominent concern around the world. Although through history in many societies there have been traditions for respect of and kindness to animals, what is new is that, to a remarkable extent, these matters have come to be on the centre stage. So now – as it has become very clear that, in our growing billions, we directly or indirectly inﬂuence the quality of the lives of very many other animals – the natures of all these interactions are, one-by-one and from North to South and East to West, being sifted through, re-examined and reconsidered. What are their impacts from the animals’ points of view? Can they be justiﬁed? How can adverse effects on welfare be prevented or ameliorated? As this process of radical review progresses, so animal welfare considerations are increasingly informing the ways we should conduct all our dealings with other animals and being formally factored-in to animal management systems. Although it has gathered momentum only recently, animal welfare science – that directed at determining animals’ needs and how these can bet met – has already proved to be powerful in changing attitudes and practices and seems likely to become increasingly inﬂuential. At this stage it is helpful and constructive – towards charting the best way forward – to reﬂect upon how and why the current interest in animal welfare has come about, on how welfare science can contribute to tackling problems (which often have major cultural or non-technical aspects), and also on its limitations. Undertaking broad syntheses is difﬁcult and in his preface to this book, David Fraser mentions some early misgivings in embarking on an introduction and overview of this broad multidisciplinary topic. However, where the various threads of complex subjects can be drawn together to provide a thorough but accessible perspective (that is, where there is someone with the rare combination of knowledge, skills and determination to do it) such synthesis is extremely worthwhile and valuable as this stylish and excellent book demonstrates. We are most grateful to David Fraser for this book and proud to include it in the UFAW/Blackwell series. James K. Kirkwood April 2008 vi

Dfraser_C000.indd vi

6/27/2008 5:35:56 PM

Preface

In a ﬁeld of science that draws on a number of different disciplines, it may seem unwise for any one person to attempt an introduction and overview of the entire topic. Surely the discussion of stress physiology should be written by a stress physiologist, the health-related parts by a veterinary scientist and so on. I feel, however, that there is also a need for an integrative work that explores the connections between the different types of knowledge we use when trying to understand animal welfare. Especially in a ﬁeld where different types of scientiﬁc information are sometimes used to draw different conclusions, we need to see an overview of the forest even at the expense of expert examination of certain important trees. The book is divided into three parts. Part I (Chapters 1–4) is about the cultural context in which the ﬁeld of animal welfare science arose. It tries to show how the ideas in the ﬁeld were inﬂuenced by different modes of thought and by certain historical events at the time when the science began. These chapters also go much farther back in history to argue that animal welfare science should be viewed as one (distinctly contemporary) attempt to solve the ancient moral dilemma of how we ought to treat animals. A sub-theme in these chapters is the mutual inﬂuence that occurs between science and other elements of culture. Part II (Chapters 5–10) is about the methods of animal welfare science. Each chapter deals with a different set of methods: studies of basic health, studies of physiological ‘stress’ responses, studies of abnormal behaviour and so on. These are more conventional review essays, although here too I have tried to bring out some of the context, history and development of the scientiﬁc approaches to studying animal welfare, rather than focusing on the most recent or technically advanced examples. I hope that these chapters will serve as accessible introductions to contemporary methods and debates in the ﬁeld. With the various methods and their limitations discussed in Part II, Part III (Chapters 11–13) explores the logic involved when we try to draw conclusions about animals welfare, often in complex situations where different types of evidence may point in different directions. These chapters examine some current debates vii

Dfraser_C000.indd vii

6/27/2008 5:35:56 PM

viii

Preface

and applications of the science to policy issues, and use these as talking points to explore some of the ways that ‘facts’ and ‘values’ interact in the conduct and interpretation of animal welfare science and of science generally. I hope that the book will be of value to several groups of readers. The principal audience consists of those who want an introduction to animal welfare science. I have in mind students, veterinarians, scientists, animal producers, and others in the animal care professions, together with corporate and government workers who are involved in animal welfare and its application. To keep the book accessible to these readers, I have tried to outline and illustrate the key methods and debates of the ﬁeld without a welter of technical detail. A second audience consists of scientists and graduate students already working in the ﬁeld. For them, I hope that Parts I and III will set the ﬁeld in a cultural and historical context that they will ﬁnd thought-provoking, explore the origins of some beliefs and assumptions that have become embedded in the ﬁeld, and examine how debates and disagreements among scientists sometimes boil down to different value-based beliefs and assumptions rather than disagreements on technical matters. Part II will be of less interest to these readers; undoubtedly they themselves could have written more thorough reviews of their speciﬁc areas of research. However, I hope these chapters may provide some historical context and integration of ideas that may be of interest even to specialists. A third audience might be captured under the term ‘science studies’. Animal welfare science is a small, emerging and multi-disciplinary ﬁeld. It is also an example of ‘mandated science’ – science that has been brought into existence to guide action and policy. I believe that a study of animal welfare science makes points about the place of science in society, the inﬂuence of culture and language on science, the interplay of ‘facts’ and ‘values’, and the complexity of interpretation in multi-disciplinary ﬁelds. Thus a study of animal welfare science may function as a case study of science and society, in much the same way that examining a small and complex star cluster can serve as an introduction to astronomy. I HAVE MANY PEOPLE to thank for their support and assistance in writing this book. First and foremost I am grateful to my wife Nancy who made this project possible through her truly extraordinary support, not only during the two years when the writing was a daily preoccupation, but also during the 37 years when her ability to create a happy home environment, even amid mosquito-infested moose swamps, gave me the freedom to pursue the scientiﬁc interests that ultimately resulted in this book. It is also a pleasure to thank my colleagues in the University of British Columbia Animal Welfare Program, especially my exceptionally supportive co-workers Dan Weary and Marina von Keyserlingk, plus many other valued colleagues in the Faculty of Land and Food Systems, the W. Maurice Young Centre for Applied Ethics, and further aﬁeld, most notably historian Rod Preece, and many colleagues who have served with me on animal welfare policy, advisory and funding bodies

Dfraser_C000.indd viii

6/27/2008 5:35:56 PM

Preface

ix

including the Animal Welfare Working Group of the World Organization for Animal Health, the Food and Agriculture Organization of the United Nations, the Burger King Corporation, the National Council of Chain Restaurants, the Food Marketing Institute and the Animal Welfare Foundation of Canada. In an important sense, the book had a gestation period of some 35 years, and many friends and co-workers along the way have made important contributions to my understanding of the ﬁeld. I would mention in particular Peter Phillips, Brian Thompson, Ed Pajor, Jeff Rushen and others in the former Centre for Food and Animal Research, Ottawa; Harry Lumsden, Ed Addison, Charles MacInnes and Hank Hristienko in the Wildlife Research Section of the Ontario Ministry of Natural Resources; Colin Whittemore, Andrew Fraser, Ian Duncan, Barry Hughes, Mike Gentle, John Savoury and the late David Wood-Gush and Frank Elsley during my years in Edinburgh; S.A. Barnett and Michael Hansell during my years at the University of Glasgow; and Jerry Hogan, Nicholas Mrosovsky, Sarah Shettleworth, and the late I.M. Spigel during my years at the University of Toronto. I also want to acknowledge the many people who have contributed to the ﬁeld of animal welfare science whose work I have not cited but who have nonetheless made important contributions to our understanding of the subject. Because my aim was to illustrate and discuss key concepts, rather than provide an exhaustive review, I have selected certain examples to make the points, and am painfully aware of the large amount of good work I have had to pass over. I also want to acknowledge some colleagues whose friendship and hospitality I have made a life of working in this ﬁeld particularly enjoyable, especially Bo Algers, Mike Appleby, Don Broom, Marian Dawkins, Ian Duncan, Sandra Edwards, Andrew Fei, Andrew Fraser, Bob Friendship, Harold Gonyou, Temple Grandin, Paul Hemsworth, Per Jensen, Jin Suk Kim, Jan Ladewig, Andrew Luescher, Vonne Lund, Guy-Pierre Martineau, Joy Mench, John Patience, Janice Swanson, Joe Regenstein, Bernard Rollin, Paul Thompson and John Webster. The book proﬁted greatly from the suggestions of several friends and colleagues. Drs. Ed Pajor, Evan Fraser and Dan Weary kindly read the manuscript and made many valuable comments. Many individuals provided helpful suggestions and comments on passages or chapters. These include John Barnett, Marc Bracke, Ron Broglio, Robert Dantzer, Marian Dawkins, Ian Duncan, Ingvar Ekesbo, Alan Hein, Paul Hemsworth, Georgia Mason, Jill Mellen, David Mellor, Dana Miles, Elisabeth Ormandy, Viktor Reinhardt, Janeen Salak-Johnson, Ernest Sanford, Chris Sherwin, Ragnar Tauson, Tina Widowski, Nadja Wielebnowski and Yasushi Kiyokawa. I am grateful to all these colleagues for their kindness and attention even when I have not taken their advice. Several people have helped me ﬁnd illustrations. These include Ingvar Ekesbo, the children of the late Ruth Harrison, Marlene Halvorsen, Carol Knicely, Hal Markowitz, Nadja Wielebnowski, Yasushi Kiyokawa, Chris Sherwin, Ian Dohoo, Hank Hristienko, Robert Zingg of the Zurich Zoo, and Jim Schulz of the Brookﬁeld Zoo who took the stunning photograph that appears on the cover.

Dfraser_C000.indd ix

6/27/2008 5:35:56 PM

x

Preface

I have been very fortunate to have the research assistance of Jane Orihel, Nicole Fenwick and Anna Drake, all of whom patiently found sources, drew ﬁgures, checked details, and graciously tolerated the wild-goose chases I set them on when trying to locate dimly recalled publications. Finally, in a service far beyond any call of duty, my cousin Susan Simons read the entire manuscript and gave me the beneﬁt of her expertise in written English. In various places I have used or reworked material from my own earlier essays. It would be too tedious for the reader if I attempted to put quotation marks around all the phrases or sentences taken from these sources, so I have chosen instead merely to indicate in the notes where I have drawn on previously published material. In some cases these were from jointly authored essays, and I am grateful to Rod Preece, Dan Weary, Ed Pajor, Barry Milligan, Joy Mench, Suzanne Millman, Ian Duncan and Lindsay Matthews for kindly allowing me to pilfer bits from our joint publications. Finally I need to express my gratitude to many animals, especially of two species – pigs and moose – whose animal welfare challenges created the scientiﬁc questions that have kept me engaged during much of my research career, and have taught me much of what I know about animal welfare. I refer to them repeatedly throughout the book, much as others might cite inﬂuential human mentors. If the book seems to rely too much on these species, I can only say that were it otherwise, it would not be my book.

Dfraser_C000.indd x

6/27/2008 5:35:56 PM

For Nancy

xi

Dfraser_C000.indd xi

6/27/2008 5:35:57 PM

Dfraser_C000.indd xii

6/27/2008 5:35:57 PM

Part I Animal Welfare in Context

Dfraser_C001.indd 1

6/27/2008 4:56:20 PM

Introduction

In 1964, the Vancouver Aquarium commissioned Mr. Sam Burich, a local sculptor who also had experience as a commercial ﬁsherman, to kill an orca. Orcas, or ‘killer whales’, are impressive predators that can reach nine metres in length and weigh over eight tonnes. Burich’s task was to use the carcass of an orca to make a lifelike replica which would hang from the ceiling in the new foyer of the Aquarium as an impressive display to greet visitors.1 In May of that year Burich, with ample assistance from scientists, photographers and Aquarium staff who were keenly interested in the project, set up a harpoon gun on a coastal island near a stretch of water where orcas were known to pass. They waited for many weeks, but few orcas came into view and the team had no success in harpooning those that did. Gradually, the scientists and other personnel returned to their normal duties leaving Burich and one assistant to keep up the watch. Finally, on 16 July the Aquarium received an urgent message. Burich had sunk a harpoon into the body of an orca which was now struggling vigorously on the line but showed no sign of expiring. A hasty decision was made to tow the orca some 60 kilometers to a makeshift enclosure in the port of Vancouver. There the orca, named Moby Doll by its captors who mistook it for a female, quickly became a celebrity. An estimated 20 000 people ﬂocked to see it on the ﬁrst day when public viewing was allowed. Stories about it appeared in Time, Newsweek, Life, The New York Times and a host of other publications. A ﬁlm about the orca and its capture was shown in 43 countries. The orca died only 75 days after it had been captured, but the experience was enough to demonstrate the huge public interest in a live orca and the unexpected docility of a species that had previously been considered too dangerous to keep in captivity. On this basis, the Aquarium decided to construct a tank large enough to accommodate a live orca display, and for the next three decades a series of wildcaught orcas became the star attractions of the Aquarium. 1This story is related in Newman, M.A. 1993. The History of the Vancouver Aquarium. Vancouver Public Aquarium Association, Vancouver. I am grateful to Peter Hamilton for bringing this book to my attention.

2

Dfraser_C001.indd 2

6/27/2008 4:56:20 PM

Introduction

3

By the 1990s, however, the Aquarium found itself facing increasing pressure from critics, the media, and some of their own paying customers who questioned the ethics of keeping captive orcas. Surely, the critics argued, such an intelligent and social animal must live a miserable and unhealthy life swimming in a concrete tank and putting on daily shows for the amusement of spectators. After a lengthy debate, the Aquarium decided that it would no longer keep wild-caught orcas in their facility. Releasing a long-captive orca to the ocean was out of the question because it would likely die of starvation. Instead, in 2001 the Aquarium transferred its last orca to a facility in San Diego where she would at least have more space and the company of other orcas. The Vancouver orcas provided one small example of a profound change in human attitudes toward animals that occurred in the second half of the twentieth century, especially in the European and English-speaking countries. The change was paralleled in virtually every aspect of the human use of animals. A few examples follow. In the 1950s, many jurisdictions in North America paid out public funds as ‘bounties’ to encourage citizens to kill wolves as a public service, either to protect livestock or to increase populations of deer and other wild ruminants that formed the basis of recreational hunting. However, research in the relatively new scientiﬁc ﬁelds of ecology and animal behaviour had already begun portraying wolves as intelligent animals that live in tight-knit families and serve the vital ecological function of keeping natural prey populations healthy.2 Wildlife biologist Aldo Leopold even used a gruesome encounter with a family of wolves to communicate his developing respect for wild nature: We saw what we thought was a doe fording the torrent, her breast awash in white water. When she climbed the bank toward us and shook out her tail, we realized our error: it was a wolf. A half-dozen others, evidently grown pups, sprang from the willows and all joined in a welcoming melee of wagging tails and playful maulings … . In those days we had never heard of passing up a chance to kill a wolf. In a second we were pumping lead into the pack, but with more excitement than accuracy; how to aim a steep downhill shot is always confusing. When our riﬂes were empty, the old wolf was down, and a pup was dragging a leg into impassable side-rocks … . We reached the old wolf in time to watch a ﬁerce green ﬁre dying in her eyes. I realized then, and have known ever since, that there was something new to me in those eyes – something known only to her and to the mountain.3

Faced with such depictions of wolves by scientists, public perception of wolves underwent a remarkable change, to the point that bounties were eliminated in 2Dunlap,

T.R. 1988. Saving America’s Wildlife: Ecology and the American Mind, 1850–1990. Princeton University Press, Princeton. 3Leopold, A. 1948. A Sand County Almanac. Republished 1987, Oxford University Press, New York. The quotation is from pages 129–132.

Dfraser_C001.indd 3

6/27/2008 4:56:21 PM

4

Animal Welfare in Context

most areas, and public funds were used instead to protect wolves, and even to re-establish them in areas where they had been exterminated as a matter of public policy barely a generation before. In Britain during roughly the same period, hunting itself was at the centre of a raging controversy. The practice of hunting deer and foxes by running them with dogs and mounted hunters had been a target for animal protectionists for many decades, but there was little government support for any action against this favoured sport of the wealthy land-owning class. In 1951 a committee appointed by the British government considered the issue, but it recommended that hunting should be permitted to continue as long as the amount of suffering inﬂicted on the animals is not greater than what would be caused by other means of controlling animal numbers.4 In the 1990s, a study was ﬁnally commissioned whose goal was, in essence, to address the issue raised by this committee. The study looked for evidence of suffering caused to deer by hunting with dogs and horses, and compared that to evidence of suffering caused by other means of killing the animals such as stalking them and shooting them from a distance. The report concluded that, ‘All available evidence strongly suggests that hunting with hounds poses a greater welfare problem for individual deer than stalking’.5 The day after the results of the study were announced, the hunting of deer with dogs was banned on the lands of Britain’s National Trust (the organization that had commissioned the study), and within a few years, the British House of Commons passed legislation to ban all such hunting throughout the country. Changing attitudes toward animals inﬂuenced biomedical research as well. In the 1950s, scientists used hunters in Africa to shoot female chimpanzees so as to capture their infants who were then raised in steel cages and used as subjects – even as living test-crash dummies – in various sorts of research. The most famous of these was nick-named ‘Ham’, short for the Holloman Aerospace Medical Center, who was used to test the safety of a spacecraft before it carried a human into space. Then in the late 1960s, Jane Goodall began publishing her stunning ﬁeld research on chimpanzees, with widespread public exposure through television and magazines together with her eminently successful book In the Shadow of Man.6 Through Goodall, people were exposed to the real-life story of ‘MacGregor’, a chimpanzee who was stricken with polio in adulthood and tried pathetically to re-establish friendly relations with his old group members despite being partly paralysed; and ‘Mike’ an undistinguished member of his troop who learned to intimidate the older males by charging into their midst while banging parafﬁn cans together and thus catapulted himself to the top of the dominance hierarchy. 4Turner,

E.S. 1964. All Heaven in a Rage. Michael Joseph, London. E.L. and Bateson, P. 2000. Welfare implications of culling Red Deer (Cervus elaphus). Animal Welfare 9: 3–24. The quotation is on page 21. 6Goodall, J. 1971. In the Shadow of Man. William Collins, London. 5Bradshaw,

Dfraser_C001.indd 4

6/27/2008 4:56:21 PM

Introduction

5

The result of the observations by Goodall and others was that the chimpanzee was transformed from a mere curiosity – a smudgy and disposable carbon copy of a human being – to a precious cousin of humanity, similar yet different in interesting ways, whose quality of life and very survival were being tragically threatened by human actions. In 1985 the United States amended its Animal Welfare Act to require, among other stipulations, that steps be taken to provide for the ‘psychological well-being’ of captive primates.7 By the 1990s, harmful research on chimpanzees and other great apes had become highly controversial,8 to the point that two countries (New Zealand and the Netherlands) passed legislation to prohibit it. It was in agriculture, however, that the changes were most remarkable, if only because of the vast numbers of animals involved. In the decades after the Second World War, farm animal production underwent a major revolution in the industrialized countries. Previously, most production had used fairly traditional methods that relied on human labour for routine tasks such as collecting eggs and removing manure. Then animal agriculture in the industrialized countries began a massive move toward more automated production methods, generally involving ‘conﬁnement’ housing systems. These included tiers of cages for laying hens, narrow ‘gestation stalls’ where sows were conﬁned during most of pregnancy, and individual crates for calves raised for veal. Almost as soon as they were invented, however, the industrialized methods were attacked by critics who alleged that animals could not possibly live happy, healthy lives under such unnatural conditions. Near the end of the century, backed up by a growing body of scientiﬁc research, the European Union passed three agreements on farm animal welfare. These required their member countries to phase out the use of crates for veal calves, to enlarge and improve cages for laying hens, and to severely limit the use of stalls for pregnant sows,9 thereby over-hauling some of the predominant technology of the vast animal-production industry throughout most of Europe. In all the above examples, we see that a half-century that began in one cultural climate – a climate where it seemed modern and progressive to take orcas from the wild for display, to exterminate wolves, to capture infant chimpanzees for research, and to keep laying hens in tiers of cages – ended in a very different climate where such actions were increasingly a subject for debate, disagreement and sometimes reform. It was during this period of changing attitudes that the scientiﬁc study of animal welfare began, initially as a response to public concern about the welfare of animals, and then increasingly as a force that guided and sometimes motivated reforms.

7Rowan,

A.N. and Rosen, B. 2005. Progress in animal legislation: Measurement and assessment. Pages 79–94 in State of the Animals III (D.J. Salem and A.N. Rowan, editors). Humane Society of the United States, Washington. 8Cavalieri, P. and Singer, P. (editors). 1993. The Great Ape Project: Equality beyond Humanity. St. Martin’s Press, New York. 9Stevenson, P. 2004. European Union Law on the Welfare of Farm Animals. Compassion in World Farming Trust, Petersﬁeld, UK.

Dfraser_C001.indd 5

6/27/2008 4:56:21 PM

6

Animal Welfare in Context

Animal welfare was, however, a most unusual subject for scientiﬁc research. A good deal of science has traditionally been motivated by a simple desire to understand the world around us. For Galileo (1564–1642), a desire to understand the movements of the heavenly bodies led to the research that formed the basis of astronomy. For the engineer-geologist William Smith (1769–1839), a fascination with fossilized marine organisms led to the research that became the ﬁeld of stratigraphy. Other ﬁelds of science grew from a more practical motivation. The studies of Louis Pasteur (1822–1895), which contributed so much to the germ theory of disease, began with his attempt to ﬁnd ways to prevent wine from spoiling. In contrast to these various cases, animal welfare science did not occur because people suddenly became curious about the well-being of animals, or because they were pursuing a practical goal such as creating ‘cruelty-free’ cosmetics, but rather as a response to ethical concerns about the treatment of animals and debate about the kind of life they should be allowed to live. If this made animal welfare science an oddity, to some people it was also an impossibility. For one thing ‘welfare’, and roughly similar terms such as ‘wellbeing’ and ‘quality of life’, are rather nebulous concepts whose meaning will vary from person to person and from culture to culture. It is hard enough to agree on how to deﬁne quality of life for human beings, let alone for laboratory mice. Moreover, animal welfare is at least partly a ‘mentalistic’ concept – a concept that includes mental states such as pain, distress and comfort; yet many scientists in the twentieth century held that the mental states of animals are not open to scientiﬁc enquiry. Worst of all, animal welfare is a morally charged concept, intimately linked to debates about how we ought to treat animals; yet Western thought has long favoured the view that we cannot give empirical, scientiﬁc answers to ethical questions. How, then, could there possibly be a ‘science’ of animal welfare? TO UNDERSTAND THIS PARADOX, in addition to looking at the technical aspects of animal welfare research, we also need to reﬂect on the nature of science and its role in culture. People seem to recognize that the arts are cultural activities that draw on (or react against) certain cultural traditions, certain shared understanding, and certain values and ideas that are characteristic of the time and place in which the art is created. In the case of science, however, opinions differ. Some scientists, like the great biologist J.B.S. Haldane (1892–1964), see science in a similar light – as a historical activity that occurs in a particular time and place, and that needs to be understood within that context.10 Others, however, see science as a purely ‘objective’ pursuit, uninﬂuenced by the cultural viewpoint and values of those who create it. In describing this view of science, philosopher Hugh Lacey speaks of the belief that there is

10For example, Haldane, J.B.S. 1923. Daedalus, or Science and the Future. Republished 1930, Kegan Paul, Trench, Trubner & Co, London.

Dfraser_C001.indd 6

6/27/2008 4:56:21 PM

Introduction

7

an ‘underlying order of the world’ which is ‘simply there to be discovered – the world of pure “fact” stripped of any link with value’. The aim of science (according to this view) is ‘to represent this world of pure “fact”, … independently of any relationship it might bear contingently to human practices and experiences’.11 A similar debate arises over the relation of science and ethics. One view, expressed in 1975 by a committee of scientists in the United States, is that the activities of scientists ‘are conditioned and directed at every turn by considerations of human values’.12 In the nineteenth century, however, when there was still active debate about the boundary between science and ﬁelds like theology and philosophy, a number of inﬂuential scientists proposed a clear separation of science from ethics and other areas that involve values. Sociologist Max Weber (1864–1920), whose scientiﬁc studies were fundamental to social policy, held (in the words of sociologist Ralf Dahrendorf) that ‘statements of fact are one thing, statements of value another, and any confusing of the two is impermissible’.13 The French physicist and mathematician Henri Poincaré (1854–1912) proposed: Ethics and science have their own domains, which touch but do not interpenetrate. The one shows us to what goal we should aspire, the other, given the goal, teaches us how to attain it. So they never conﬂict since they never meet.14

Thus, in Poincaré’s terms, we might say that an archeologist’s decision to excavate an ancient grave site is ‘touched’ by ethical issues related to the importance of preserving historic artefacts and showing respect for human remains; however, the ethical issues do not ‘penetrate’ into the actual scientiﬁc issues being investigated such as the time when the burial occurred and the signiﬁcance of the artefacts that were buried with the dead. Poincaré’s view may make a plausible ﬁt with much so-called ‘curiosity-driven’ science that is done primarily to understand the world around us, but today a good deal of science comes from a different mould. The term ‘mandated science’ refers to science that is done for a particular social purpose, for example to guide action, policy or legislation. Scientiﬁc studies on topics such as food safety, occupational health, biological diversity and agricultural sustainability are not done primarily out of curiosity but to answer questions of importance to society, often because people are concerned that the right course of action is not being followed, or that 11Lacey,

H. 1999. Is Science Value Free? Values and Scientiﬁc Understanding. Routledge, London. The quotation is on page 3. 12Edsall, J.T. 1975. Scientiﬁc Freedom and Responsibility: A Report of the AAAS Committee on Scientiﬁc Freedom and Responsibility. American Association for the Advancement of Science, Washington. The quotation is on page 6. 13Dahrendorf, R. 1987. Max Weber and modern social science. Pages 574–580 in Max Weber and His Contemporaries (W.J. Mommsen and J. Osterhammel, editors). Allen & Unwin, London. The quotation is on page 577. 14Lacey, 1999, page 1.

Dfraser_C001.indd 7

6/27/2008 4:56:21 PM

8

Animal Welfare in Context

important safeguards are not in place. People are concerned, for example, about the safety of the food they eat, about health problems that may result from their working environments, about the loss of species from the planet, and about the ability of agricultural systems to continue producing adequate food in the future. These concerns, and the debates that arise, lead society to ‘mandate’ research to explore the concerns and often to make recommendations about whether and how certain changes should be made. In such ﬁelds, public concerns and associated debates play a key role in causing the research to be undertaken, and they also help to shape exactly what science is done and how it is applied. Animal welfare science is a ‘mandated’ ﬁeld. It began because ethical concerns arose in society about the welfare of animals, and these led to debates and disagreements that involved researchable issues such as: are hens frustrated when conﬁned in cages? how can we maintain the psychological well-being of primates in laboratories? what are the long-term consequences for orcas of living in aquaria? do deer suffer more when chased by dogs than when killed in other ways? and are farm animals healthier and more content if they are kept outdoors? When these questions and many others arose, the ﬁeld of animal welfare science emerged as a means of providing answers. This book is intended as an overview of this emerging ﬁeld – its methods, insights, contributions, and limitations. I hope it will also serve as a case study in mandated science by bringing out how the ﬁeld developed in response to social concerns, how it was shaped by the cultural context in which it emerged, and how it is applied to issues of practice and policy in the everyday world. In this ﬁrst part of the book, I focus on the cultural context. Chapter 1 uses two case studies to describe some elements of the historic debate about the proper treatment of animals; I see these as setting the stage because animal welfare science represents one modern attempt, which to a degree competes with earlier attempts, to grapple with the age-old problem of animal ethics. Chapter 2 looks at how our understanding of animals has changed over the centuries and how these changes have been accompanied by evolving views of what constitutes proper treatment of animals. In this chapter I argue that there has been a complex conversation between scientiﬁc knowledge and popular understanding of animals, and that each has inﬂuenced the other. Chapter 3 discusses four world-views in Western thought which I believe inﬂuence contemporary ideas about what constitutes a good life for animals. Here we encounter some of the tension in Western culture between, for example, values based on rationality and control of nature on the one hand, versus emotion and respect for nature on the other. Chapter 4 describes the concerns about the treatment of animals that emerged in the twentieth century and how they gave rise to, and helped to shape, scientiﬁc research on animal welfare.

Dfraser_C001.indd 8

6/27/2008 4:56:21 PM

Animals and Moral Concern

1

When debate emerged about the proper treatment of animals in the late twentieth century, it was not the ﬁrst time that the issue had arisen, but it was the ﬁrst time that science had been called upon to clarify the issues and guide the resulting reforms. To understand how science contributed to the process, it helps to see the twentieth century debate in light of some earlier instances when issues of animal ethics were approached in quite different ways. PERHAPS THE FIRST WELL RECORDED debate about the proper treatment of animals occurred in Greece, beginning in the sixth century BC. In its use of animals, ancient Greece had much in common with modern Europe roughly a century ago. Horses were used for transportation, racing and warfare, and oxen for tilling the land. Sheep were raised for wool and also, like cattle and goats, for milk and cheese. Meat was eaten from these species as well as pigs. A good deal of the slaughtering of animals occurred in the course of religious sacriﬁce followed by feasting on the carcass. Dogs, as well as being used for guarding and warfare, were kept as companions by people of all social classes, and some of these animals received funerals and tombstones carved with touching epitaphs that spoke about the mutual affection between the dog and its owner.1 Animals were also the subject of scientiﬁc research. Aristotle (384–322 BC), as the foremost natural historian of ancient Greece, maintained a collection of wild animals as part of the reference material of his school in Athens, and was supported

1I

am drawing this information mainly from Works and Days by Hesiod, published as pages 9–30 (D. Grene, translator) in: Nelson, S.A. 1998. God and the Land: The Metaphysics of Farming in Hesiod and Vergil. Oxford University Press, Oxford; and from Georgics by the Latin poet Virgil, published as: Lembke, J. (translator). 2005. Virgil’s Georgics. Yale University Press, New Haven. Dog tombstones are described by: Bodson, L. 2000. Motivations for pet-keeping in ancient Greece and Rome: a preliminary survey. Pages 27–41 in Companion Animals and Us (A.L. Podberscek, E.S. Paul and J.A. Serpell, editors). Cambridge University Press, Cambridge.

9

Dfraser_C001.indd 9

6/27/2008 4:56:21 PM

10

Animal Welfare in Context

in this by his most famous pupil, Alexander the Great, who brought back exotic animals from his military campaigns. Aristotle’s History of the Animals records a remarkable amount of information (and some mis-information) about both zoology and animal husbandry. On the subject of pigs, for example, he made a variety of observations that modern science has borne out: that the number of piglets born in a litter will be reduced if the boar is required to mate too often; that allowing the animals to stay cool in hot weather is important to maintain their appetite; and that over-feeding a sow during pregnancy can result in poor milk production during lactation.2 The ancient Greeks also engaged in vigorous debate about how animals ought to be used and treated. The earliest radical voice was that of Pythagoras (born about 580 BC), who is remembered today mainly for his contributions to mathematics, including the geometric theorem that bears his name. In his own day, however, Pythagoras was a diverse and highly inﬂuential thinker who held strong views on the ethical treatment of animals. None of his own writing survives, but several centuries after his death the Roman writer Ovid (43 BC–AD 17) wrote a long poem on ‘The Teachings of Pythagoras’. In the poem Pythagoras proposes close connections between humans and animals, and he declares that it is ‘wicked as human bloodshed to draw the knife across the throat of the calf’ (Box 1.1).3 A more modern translation might read ‘Meat is murder’ – a slogan that is sometimes scrawled across the walls of butcher shops by vegetarian protesters armed with aerosol paint cans. The medium has changed, but the message hardly at all. The debate that pitted the followers of Pythagoras against certain opposing views has been carefully documented by classicist Richard Sorabji in his book, Animal Minds and Human Morals: The Origins of the Western Debate.4 As Sorabji notes, the followers of Pythagoras saw ‘kinship’ as the key to determining the proper objects of moral concern, and they advanced strong arguments for perceiving kinship between humans and other species: we are made from the same elements, we are permeated by the same breath, and animals and humans alike are animated by the same reincarnated souls. On this basis the Pythagoreans rejected the killing of animals for food or religious sacriﬁce, and Pythagoras (according to legend) once stopped a man from beating a dog on the grounds that he could recognize the voice of a dead friend in the dog’s cries.5

2Fraser,

D., Friendship, R.M. and Martineau, G.-P. 1994. Aristotle on pigs: husbandry, health and natural history of pigs in ancient Greece. Pig News and Information 15: 77N–80N. 3Ovid. The Teachings of Pythagoras. Republished 1955 as pages 367–379 in Ovid’s Metamorphoses (R. Humphries, translator). Indiana University Press, Bloomington. The quotation is on lines 465–466. 4Sorabji, R. 1993. Animal Minds and Human Morals: The Origins of the Western Debate. Cornell University Press, Ithaca, USA. 5Burnet, J. 1930. Early Greek Philosophy, 4th edition. Adam and Charles Black, London. The anecdote is related on page 84.

Dfraser_C001.indd 10

6/27/2008 4:56:22 PM

Animals and Moral Concern

11

Box 1.1 A passage from ‘The Teachings of Pythagoras’ by the Latin poet Ovid (43 BC–AD 17) expressing the view that the bodies of both humans and animals are the dwelling places for reincarnated souls, and urging people to avoid killing or consuming animals. The heavens and all below them, earth and her creatures, All change, and we, part of creation, also Must suffer change. We are not bodies only, But wingèd spirits, with the power to enter Animal forms, house in the bodies of cattle. Therefore, we should respect those dwelling-places Which may have given shelter to the spirit Of fathers, brothers, cousins, human beings At least, and we should never do them damage, Not stuff ourselves like the cannibal Thyestes. An evil habit, impious preparation, Wicked as human bloodshed, to draw the knife Across the throat of the calf, and hear its anguish Cry to deaf ears! And who could slay The little goat whose cry is like a baby’s, Or eat a bird he has himself just fed? One might as well do murder; he is only The shortest step away. Let the bull plow And let him owe his death to length of days; Let the sheep give you armor for rough weather, The she-goats bring full udders to the milking. Have done with nets and traps and snares and springes, Bird-lime and forest-beaters, lines and fish-hooks. Kill, if you must, the beasts that do you harm, But, even so, let killing be enough; Let appetite refrain from flesh, take only A gentler nourishment. From ‘The Teachings of Pythagoras‘ by Ovid. Republished 1955 as pages 367–379 in Metamorphoses (R. Humphries, translator). Indiana University Press, Bloomington.

The view that souls migrate between human and animal bodies continued in Hindu and other Eastern thought, but the idea was soon dropped in the West. By about 300 BC, however, the Greek Theophrastus proposed a concept of kinship between species that has a much more modern, Western ring. He noted that people who are born from the same ancestors are naturally kin, but so too, he claimed,

Dfraser_C001.indd 11

6/27/2008 4:56:22 PM

12

Animal Welfare in Context

are fellow citizens sharing the same land, fellow Greeks sharing the same nation, and fellow humans sharing the same nature. By viewing kinship as expanding in these widening circles – a metaphor commonly used by animal ethics philosophers today – Theophrastus argued that humans are also kin to animals because they have the same bodily organs, the same tissues and ﬂuids, and the same appetites, emotions, perceptions and reason.6 The debate, however, was far from one-sided. Aristotle, through his philosophy and natural history, had concluded that although humans and animals share many characteristics such as perception and emotion, humans alone have the capacity for logos or reason. For Aristotle (Sorabji noted) this was simply a factual conclusion about the mental capabilities of animals. However, thinkers of the Stoic school – a rival to the Pythagoreans – made it the basis for their ethical position on animals. The Stoics saw justice as rooted in the concept of mutual ‘belonging’.7 Some Stoics applied the notion of belonging narrowly, to ourselves and our offspring; others applied it more widely to all virtuous people, or even to all fellow humans. But the Stoics considered that no such community of belonging can exist between rational and non-rational beings. Hence, what had been for Aristotle a purely factual conclusion about the mental powers of animals was used by the Stoics as the basis for the ethical conclusion that animals fall outside the sphere of human justice and moral concern. Another rival theory was that of the Epicureans. Epicurus (c. 341–271 BC) maintained that a good life is a happy, hedonically pleasant life to be achieved not through the pursuit of transitory pleasures, but by avoiding pain and suffering and by fulﬁlling natural and wholesome desires. The Epicureans viewed justice as a contract or agreement between different people to avoid causing harm to each other. Justice, because it requires a measure of agreement about what constitutes acceptable behaviour, could not be applied to animals because animals lack the powers of reason needed to enter into such a contract. Thus, Epicurean theory, like Stoic theory, denied that the principles of justice apply to animals on the grounds that animals are irrational.8 These arguments put the onus on those who sought to protect animals to show that Aristotle’s original conclusion was incorrect, and Plutarch (AD 46–119), a prominent Latin essayist and biographer, took up this cause with gusto. In an essay on ‘the cleverness of animals’, he produced many anecdotes to argue that animals use reason. He noted, for example, that in Thrace, people use a fox to test whether it is safe to venture onto ice. The fox walks warily on the ice and listens carefully. If it hears running water, it deduces that the ice is not thick and returns to shore, but if there is no sound, then it proceeds ahead. Plutarch also told the story of a mule that was employed to carry bags of salt. Upon falling down while fording a river, the mule discovered that if the bags became soaked, the load would become lighter because some of the salt would dissolve away. The mule then began sinking down deliberately in any water that it crossed. This bad habit was ﬁnally cured when the owner of the mule secretly 6Sorabji,

1993, pages 177–178. 1993, page 184. 8Sorabji, 1993, page 124. 7Sorabji,

Dfraser_C001.indd 12

6/27/2008 4:56:22 PM

Animals and Moral Concern

13

ﬁlled the bags with sponges rather than salt, and the animal thereafter took great care not to wet its load. Plutarch also recounted a famous case of a dolphin that befriended a boy, and would let the boy ride on its back to the delight of the local people. One day, however, in stormy weather, the boy fell off and drowned. The dolphin recovered the boy’s body and brought it to shore, and then lay beside the boy, refusing to leave until it too had died. Plutarch’s explanation was that the dolphin saw itself as partly to blame for the boy’s death and thought it right to share his fate.9 In another essay, Plutarch invented a satirical dialogue between Ulysses, the Greek hero of the Iliad, and Gryllus, one of Ulysses’ sailors whom the enchantress Circe had turned into a pig. Ulysses was determined to release his men from Circe’s evil spell, but Gryllus, having experienced life as both a human and an animal, was not at all sure that he wanted to be human again. The reason: animals are more rational. When animals meet, he noted, they are not unduly impressed by another’s ﬁne clothing; they follow natural and necessary desires, uncontaminated by a lust for wealth; they mate only in the proper season and in a natural manner; their sexual appetites are awakened by the natural odours of the body, not by artiﬁcial ointments and perfumes; they eat simple food that is easily obtained, thus avoiding the indigestion that befalls people from an excessive quantity and variety of foods; and animals teach their children useful skills while avoiding the human penchant for knowledge ‘that has no point or purpose’.10 Apart from the issue of rationality, debate also arose over whether it is natural for humans to eat meat. In the fourth century BC, the non-vegetarian Heraclides had concluded that meat-eating must be natural for humans because the practice has been universal since the invention of ﬁre. Nor can meat be bad for us, he claimed, judging from the prowess of such strictly carnivorous animals as wolves and lions.11 Untrue, claimed Plutarch in ‘On the Eating of Meat’. Nature has obviously not equipped us to eat meat because we ﬁnd meat disagreeable unless we transform it by cooking and by adding spices.12 As for humans being natural carnivores, how many human meat-eaters could catch animals with their teeth and eat them alive? Theophrastus contributed to the debate by refuting the anti-vegetarian argument (heard then as now) that if we avoid harming animals, then logically we should also avoid harming plants. This is not so, Theophrastus argued, because we can more justly claim ownership of plants given the labour that we put into cultivating them, because plants are not unwilling to give up their fruit, and because we are so much more similar to animals than to plants.13 9Plutarch. The cleverness of animals, both of the sea and of the land. Republished 1971 as pages 97–158 in Plutarch Moral Essays (R. Warner, translator). Penguin Books, Harmondsworth, UK. The fox story is on page 121, the mule on page 126, and the dolphin on page 156. 10Plutarch. On the use of reason by ‘irrational’ animals. Republished 1992 as pages 383–399 in Plutarch Essays (R. Waterﬁeld, translator). Penguin Books, London. The quotation is from page 397. 11Sorabji, 1993, page 178. 12Sorabji, 1993, page 178. 13Sorabji, 1993, page 176.

Dfraser_C001.indd 13

6/27/2008 4:56:22 PM

14

Animal Welfare in Context

Plutarch’s cause was later taken up by the philosopher Porphyry (AD 232–309) in a book-length treatise, On Abstinence from Killing Animals.14 The book took the form of a letter to a fellow philosopher in which Porphyry criticized his colleague for abandoning vegetarianism. Porphyry argued that the purity and self-discipline of a vegetarian diet is important for those who devote themselves to an intellectual life, but he also assembled many arguments, some repeated from Plutarch and other sources, to show that animals are rational beings and that killing them for the pleasure of the palate in unjust. Like Plutarch he noted that animals live ordered, rational lives; for example, they mate to produce offspring, and cease mating when the female is pregnant, unlike humans who are driven by mere lust. He noted the complexity of their communication: Animals are heard to speak differently when they are afraid, when they are calling, when they are asking to be fed, when they are friendly and when they are challenging to a ﬁght. The diversity is so great that even those who have given their life to observing animals ﬁnd it very difﬁcult to distinguish the variations, because there are so many.

He noted that rabid dogs are observed to become mad when they contract rabies, but how could an animal be considered mad unless their normal lives are governed by reason and intelligence? Most signiﬁcantly, Porphyry added an argument that was to become pre-eminently important in modern times: that animals deserve moral consideration because they, like us, have the capacity ‘to feel distress, to be afraid, to be hurt, and therefore to be injured’. But should we treat animals well simply out of benevolence, or does justice demand it? Do animals, in fact, have rights? Plutarch argued that even if we refuse to apply the principles of justice to animals, at least we should be benevolent toward them.15 However, Porphyry based his call for vegetarianism not on human kindness but on the properties of animals themselves – speciﬁcally their many similarities to humans. As Sorabji noted, this emphasis ‘makes his call for justice look more like an assertion of their rights’.16 This classical debate was so comprehensive that I ﬁnd it difﬁcult to identify arguments advanced today that were not touched on in ancient Greece. Hermarchus (third century BC) anticipated the ecological arguments of modern hunters: that we must kill animals or they would become too numerous and bring destruction on themselves and the environment. Porphyry, foreshadowing the anti-hunt lobby, refuted this claim on the grounds that nature is self-regulating and that other species would restore a natural balance if only humans would withdraw.17 14Porphyry. On Abstinence from Killing Animals. Republished 2000 (G. Clark, translator). Duckworth, London. The quotations are from pages 82 and 91. 15Sorabji, 1993, pages 118 and 125. 16Sorabji, 1993, page 156. 17Sorabji, 1993, page 184.

Dfraser_C001.indd 14

6/27/2008 4:56:22 PM

Animals and Moral Concern

15

Plotinus (AD 205–270) anticipated modern veganism by refusing medicines made with animal ingredients.18 And if we can perceive a rough functional analogy between modern biomedical testing and the classical use of animals in sacriﬁce and divination – activities that were also viewed as ways to obtain useful knowledge and prevent future harm to people – then even modern opposition to the use of animals in science had a parallel in the ancient world. IN THE RATIONAL WORLD of ancient Greece, theories of justice and principles of ethics were major elements in the debate over the treatment of animals. However when the debate was replayed in England during the eighteenth and nineteenth centuries, the proper treatment of animals was decided less by logical debate (although that was certainly present) and more by extending to animals an ethic of the heart coupled with a zeal for social reform. In a book with the homey title, Love for Animals and How it Developed in Great Britain, philosopher Dix Harwood described how concern for animals (for which Britain is famous today) developed from the most unpromising beginnings.19 As late as the 1600s, Harwood noted, brutality to both humans and animals was sufﬁciently commonplace in Britain to provoke surprised comment by visitors from continental Europe. The practice of boiling criminals in oil was used only brieﬂy in the 1500s, but hot-iron branding of criminals and vagrants, amputating hands for petty theft, and severing ears for failure to attend church, persisted long after. Capital punishment by hanging or dismemberment was so common that it formed a regular public spectacle. The 38-year reign of Henry VIII, from 1509 to 1547, saw 72 000 hangings, generally involving a slow death by strangulation unless, as a German visitor to England noted, friends had been engaged to pull at the legs of the dangling victim in order to speed the process.20 Against such a background, cruelty to animals was simply an aspect of daily life. In the mid-1700s, the British artist William Hogarth (1697–1764) produced a series of four engravings that illustrated the uses and abuses of animals that ﬂourished at the time. In one of the pictures, called the ‘Second Stage of Cruelty’, a callous coach driver is beating a delicate horse that has stumbled when trying to pull a coach over-ﬁlled with corpulent passengers (Figure 1.1). Nearby, a drover is driving sheep through the street to a slaughterhouse, and he clubs one of the animals to death for failing to stay bunched with the ﬂock. Further back two men are goading an over-loaded donkey. All of these were no doubt common occurrences in the streets of London at the time.

18Sorabji,

1993, page 172. D. 1928. Love for Animals and How it Developed in Great Britain. Republished 2002 as Dix Harwood’s Love for Animals and How it Developed in Great Britain (1928) (R. Preece and D. Fraser, editors). Edwin Mellen Press, Lewiston, USA. Page numbers cited are based on this edition. 20Harwood, 1928, page 50. 19Harwood,

Dfraser_C001.indd 15

6/27/2008 4:56:22 PM

Figure 1.1 ‘Second Stage of Cruelty’: one of William Hogarth‘s depictions of cruelty to animals in the streets of London. Bull-baiting, with a dog being thrown into the air, is shown in the upper right-hand portion of the picture. Elsewhere a coach-driver is beating his horse who has stumbled while trying to pull a coach over-filled with corpulent lawyers, a drover is clubbing a sheep that did not remain bunched with the others on its way to slaughter, men are goading an over-loaded donkey, and, to introduce the point that callous treatment of animals goes hand-in-hand with harm to people, a careless cart-driver is about to run over a child. Reproduced with permission, © the Trustees of The British Museum.

Dfraser_C001.indd 16

6/27/2008 4:56:22 PM

Animals and Moral Concern

17

But Hogarth’s purpose in creating these grim scenes was not merely descriptive. He deliberately made the engravings in a coarse-grained style so that many copies could be produced cheaply for wide distribution, because the pictures were intended as a kind of visual sermon to the masses. Hogarth was a believer in moral progress, and he considered that stamping out cruelty to animals was important for improving the moral tone of society. One of his arguments was that callousness toward animals creates a spirit of callousness toward people. To make this point, Hogarth depicted the cruel coach driver of the Second Stage as a hideous murderer in the third engraving of the series. But to foreshadow the idea that cruelty to animals goes hand in hand with harm to people, Hogarth included a careless carter in the Second Stage, who is about to run down a child that has fallen in the street. At the very back of the Second Stage of Cruelty, Hogarth also included a form of animal abuse that was to become one of the ﬁrst targets of humane reform. This was the ancient sport of bull-baiting which reached its peak of popularity around 1600 and was not ﬁnally abolished by law until 1835. Harwood described the procedure: A bull selected and trained at great expense was ﬁrst tethered in the baiting ring – the village green in the provinces or privately owned gardens in London. Sometimes his own horns were cut off and the great horns of an ox were fastened to his head, though tipped with leather to save the dogs a goring. The bull was usually given rope enough to turn with ease and watch the stealthy approach of his opponents. The object of the game was for the dog to catch the bull by the nose and if possible make him roar. The most exciting moment in the baiting usually came when a dog got a ﬁrm hold on the bull and refused to let go till his teeth were knocked out or until his master pried him loose with a crowbar.21

Over the seven hundred years that it ﬂourished in England, animal baiting and ﬁghting underwent many reﬁnements. Bear-baiting made for a diverting change – but a costly one owing to the scarcity of bears and the greater likelihood that the dogs would be killed. Other variations involved lions, monkeys, and horses. Some sports allowed the human audience to take part in tormenting the animals. In one variation a blinded bear was secured by a chain and whipped by a circle of ﬁve or six men. In another, a chicken was buried in the ground with only the head protruding, and human contestants attempted to knock the bird’s head off with a well-aimed blow from a stick. As a French visitor commented in the late 1600s, ‘Our neighbours the English like blood in their games’.22 Despite this chilling history, English attitudes toward animals underwent a gradual shift during the 1700s as part of a general awakening of feelings of pity, kindness and moral sense – an attitude which came to be known as ‘sensibility’. 21Harwood, 22Harwood,

Dfraser_C001.indd 17

1928, pages 45–46. 1928, page 50.

6/27/2008 4:56:27 PM

18

Animal Welfare in Context

An early harbinger of the new attitude was a book entitled Characteristics of Men, Manners, Opinions, Times, published in 1711 by Anthony Ashley Cooper (1671–1713), the third Earl of Shaftesbury. To Shaftesbury, people have an inherent moral sense by which they intuitively recognize justice, generosity and sympathy as good; and the essence of moral behaviour is to follow the dictates of these tender feelings. ‘To love and to be kind’, wrote Shaftesbury, ‘… is to feel immediate satisfaction and genuine content’.23 But if people are so naturally inclined to kindness and sympathy, why was the world such a brutal place? Despite their natural affection for good, thought Shaftesbury, humans sometimes succumb to ‘unnatural passions’ for revenge, luxury and delight at inﬂicting suffering. Thus, he claimed, the cruelty that he saw in everyday life was a perversion of human nature that needed to be stamped out: To delight in the Torture and Pain of other Creatures indifferently, Natives or Foreigners, of our own or of another Species, Kindred or no Kindred, known or unknown; to feed, as it were, on Death, and be entertain’d with dying Agonys; this … is wholly and absolutely unnatural, as it is horrid and miserable.24

And in this quotation, even as he stated his view that cruelty is a perversion, Shaftesbury made it clear that he viewed cruelty to humans and cruelty to those ‘of another Species’ as products of the same defect of moral character. However, Shaftesbury’s sensibility was not the only ethical innovation of the century. As the ‘English Enlightenment’ unfolded during the 1700s, philosophers rejected traditional morals that were rooted in the authority of the church and the law, and looked instead for a rational basis for ethical behaviour. Jeremy Bentham (1748–1832) was one of the early champions of the view that we should judge the rightness or wrongness of an action, not by the virtuous intentions from which it springs, or by whether it conforms to established rules, but according to the consequences that ﬂow from it.25 Good acts, Bentham maintained, are those that promote the greatest amount of good (and conversely prevent the greatest amount of evil) for the greatest number of those concerned. In other words, we should judge the rightness or wrongness of an action by its ‘utility’ in causing good outcomes, and the theory came to be called ‘Utilitarianism’. Moreover, Bentham had very speciﬁc deﬁnitions of good and evil. For Bentham (echoing the ideas of the Greek Epicurus) good meant happiness, and evil meant pain and suffering.

23Harwood,

1928, page 146. Third Earl of (Anthony Ashley Cooper). 1711. Characteristics of Men, Manners, Opinions, Times. Republished 1964 (J.M. Robertson, editor). Bobbs-Merrill, Indianapolis. The quotation appears in Treatise IV, Book II, Part II, Section III, on page 331 of the Robertson edition. 25Bentham, J. 1789. Introduction to the Principles of Morals and Legislation. Republished 1961 as pages 5–398 in The Utilitarians. Dolphin Books, Garden City, USA. 24Shaftesbury,

Dfraser_C001.indd 18

6/27/2008 4:56:28 PM

Animals and Moral Concern

19

Hence, in the phrase later coined by John Stuart Mill (1806–1873), a good action is one that causes ‘the greatest happiness of all those whose interest is in question’.26 Like Shaftesbury, even as he expounded his ideas, Bentham realized that his criterion for good outcomes could be applied not only to humans but to all animals that can experience happiness and suffering. As he put it, the question we should ask, when deciding whether to include other beings within the scope of moral concern, is not ‘Can they reason?’ (the criterion used by the Stoics), ‘nor, Can they talk? but, Can they suffer?’27 The approaches to ethics proposed by Shaftesbury and Bentham had profound implications for the proper treatment of animals, and these became a common theme in works of moral philosophy during the 1700s. William Wollaston’s Religion of Nature Delineated, published in seven editions from 1722 to 1750, proposed that animals are less sensitive than human beings because they, living only in the present, lack the reﬂection on the past and future that plays so great a role in the subjective lives of people; yet where physical pain is involved, we ought to take the greatest care not to cause needless anguish to animals.28 Henry St. John Bolingbroke (1678–1751) went much further in proposing similarities between humans and other species. An ‘absurd and impertinent vanity’ he called the human tendency to dismiss animals as mere automatons or to claim that their behaviour is governed only by instinct when their intelligence and ours obviously share important elements.29 David Hartley, in Observations on Man (1749) claimed that animals are like humans in ‘the Formation of their Intellects, Memories and Passions, and in Signs of Distress, Fear, Pain, and Death’, and that we owe greater consideration to the pain and pleasure that animals experience.30 By the end of the 1700s, whole books were appearing on animal ethics, some with a decidedly radical tone.31 George Nicholson’s On the Conduct of Man to Inferior Animals (1797) and Joseph Ritson’s An Essay on Abstinence from Animal Food as a Moral Duty (1802) urged a major change in our dealings with animals including the complete abandonment of meat-eating. John Lawrence, in his Philosophical and Practical Treatise on Horses (1791) even called for legal recognition of animal rights: No human government, I believe, has ever recognized the jus animalium which surely ought to form a part of the jurisprudence of every system founded on the principles 26Mill, J.S. 1863. Utilitarianism. Republished 1961 as pages 399–472 in The Utilitarians. Dolphin Books, Garden City, USA. The quotation is on page 291. 27Bentham, 1789, page 381. 28Harwood, 1928, page 158. 29Harwood, 1928, page 158. 30Harwood, 1928, page 159. 31Preece, R. 2001. Introduction (pages 1–37) in An Essay on Humanity to Animals (1798) by Thomas Young (R. Preece, editor). Edwin Mellen Press, Lewiston, USA. The Nicholson book was republished in 1999 as George Nicholson’s On the Primeval Diet of Man (1801): Vegetarianism and Human Conduct Toward Animals (R. Preece, editor). Edwin Mellen Press, Lampeter, UK.

Dfraser_C001.indd 19

6/27/2008 4:56:28 PM

20

Animal Welfare in Context

of justice and humanity … I therefore propose that the Rights of Beasts be formally acknowledged by the State and that a law be framed upon that principle to guard and protect them from acts of ﬂagrant and wanton cruelty, whether committed by their owners or others.32

Some of the most inﬂuential writing, however, came from reformers who maintained a more conservative stance and sought to change the treatment of animals in ways that were more in line with the established norms of the day. Among the reformers were several English priests who, in the mid-1700s, wrote sermons and essays with titles like Free Thoughts upon the Brute Creation, An Essay on the Future Life of Brutes, An Apology for the Brute Creation and The Duty of Mercy and the Sin of Cruelty to Brute Animals.33 One of the most inﬂuential of these, in the view of historian Rod Preece, was the Reverend Thomas Young’s An Essay on Humanity to Animals, initially published in 1798 and reprinted in abridged form in 1804, 1809 and 1822 to support attempts to pass animal protection legislation in those years.34 Young was neither a vegetarian nor a promoter of animal rights but rather a sober clergyman who based his arguments on the authority of Christian scripture and by calling on his fellow citizens to exercise conventional Christian virtues which were too often forgotten in humankind’s dealings with other species. Many of the points Young stressed remain key elements of animal welfare reforms today. He pointed out the importance of understanding sentience and sensitivity to pain in animals, and of accommodating the needs of animals in practical ways. He commented on the friendship that arises between people and animals, and the duties implied by that relationship. He noted that those who abuse animals often go on to commit violence toward humans. Even his approach to animal experimentation – calling on scientists not to abandon all use of animals in research but to minimize animal suffering, avoid duplication of experiments, and to use animals only in pursuit of ‘some great and public good’ – is very much in line with current thinking about the use of animals in science. The reasonableness of Young’s views, aided no doubt by his ‘eminent respectability’ as an Anglican clergyman and Fellow of Trinity College, Cambridge,35 made him a particularly effective voice for change. With such respected ﬁgures arguing the cause of animals, it became feasible to attempt legislative reform. The ﬁrst attempt was a bill to ban bull-baiting, proposed in 1800. However, opponents defeated the bill by arguing that bull-baiting should remain because it was traditional, it helped to build character, it provided 32Turner,

1964, page 74. essays, listed by Preece 2001, were John Hildrop, Free Thoughts upon the Brute Creation, 1742; Richard Dean, An Essay on the Future Life of Brutes, 1767; James Granger, An Apology for the Brute Creation, 1772; and Humphry Primatt, The Duty of Mercy and the Sin of Cruelty to Brute Animals, 1776. 34Preece, 2001. 35Preece, 2001, page 8. 33The

Dfraser_C001.indd 20

6/27/2008 4:56:28 PM

Animals and Moral Concern

21

amusement for the poor, and because laws should not meddle with the lives of people unless other people were harmed.36 A second attempt was defeated in 1802, and another in 1809 even though the proponent of the bill claimed that he had received ‘three trunk-loads’ of letters of support from the public. In 1821 Richard Martin, a wealthy land-owner from rural Ireland, introduced his Ill-Treatment of Horses bill. The attitude of some members of Parliament is captured in E.S. Turner’s description of the proceedings. When an alderman suggested, that protection should be given to asses, there were such howls of laughter that The Times reporter could hear little of what was said. When the chairman repeated this proposal, the laughter intensiﬁed. Another Member said Martin would be legislating for dogs next, which caused a further roar of mirth, and a cry ‘And cats!’ sent the House into convulsions.37

The next year, however, Martin tried again with an expanded bill which included cattle as well as horses, and this time it was passed. But when prosecutors attempted to use the new law to bring bull-baiting to an end, they found that the courts did not consider bulls to be ‘cattle’. Various other attempts at legislation occurred in the intervening years until baiting was ﬁnally made illegal in 1835.38 The British debates of the eighteenth and nineteenth centuries took place in a highly stratiﬁed society where traditional authority and the class system were important for the success of social reforms. The involvement of aristocrats and land-owners such as Martin, combined with the support of many clergy, played a key role in the movement for reform of animal treatment. And in 1840, when the 21-year-old Queen Victoria allowed the ﬂedgling Society for the Prevention of Cruelty to Animals to add the preﬁx ‘Royal’ to its name, the cause of humane treatment of animals had itself come of age as an established element of British society. IN THE TWO EXAMPLES we have followed – Greece in classical times and England during the eighteenth and nineteenth centuries – we see that the proper treatment of animals is an ancient ethical dilemma that has resurfaced in different cultures and different times, and on each occasion people have approached it in a manner distinctive of their society. When the issue arose in classical times, the Greeks treated it with their characteristic mixture of logic and philosophical theory. When it arose in England during the eighteenth and nineteenth centuries, it was approached with that culture’s characteristic mix of moralizing, sermon-writing and legislative reform, aided by the authority of the church and the class system.

36Turner,

1964, pages 110–114. 1964, page 127. 38Turner, 1964, page 137. 37Turner,

Dfraser_C001.indd 21

6/27/2008 4:56:28 PM

22

Animal Welfare in Context

Today, the proper treatment of animals is once again a topic of passionate debate, and many of the themes that occurred in classical Greece and Enlightenment England have returned: speculation about the mental powers of animals, philosophical debate about whether the principles of justice should apply to other species, the necessity of hunting, the beneﬁts of a vegetarian diet and so on. Today, however, the cultural context includes an additional element that had not been nearly so prominent before. This is a respect for the authority of science and a belief that scientiﬁc research can help resolve the difﬁcult issues of the day. Thus, confronted with an ancient ethical quandary, people in the West during the late twentieth century did something that had not happened in earlier iterations of the debate: they turned (in part) to science for guidance. But not only was the approach different; the central issue of the debate had also shifted. In Greece, if we could identify a single, central issue of the debate it would be justice: whether the principles of justice apply to animals, and if so, how justice demands that we should treat them. In the English Enlightenment, the central concern had to do more with human virtue and moral progress. The reformers wanted to improve the moral tone of society, and they saw eliminating cruelty to animals as one avenue for achieving this goal. In the debate today the key themes of the earlier iterations – justice, cruelty – are still very much present. However, the central issue in the modern debate has shifted toward emphasizing the quality of life of the animals themselves: that animals should be happy and healthy, that they should not suffer unnecessary pain, that they should not be put in situations that are so unnatural as to prevent them from having a good life. These concerns are not principally about justice or human virtue but about the quality of life of animals, and they came to be captured under the term ‘animal welfare’. Thus, when science was invoked in the animal ethics debates of the late twentieth century, its commission was speciﬁcally to investigate the ‘welfare’ or ‘well-being’ or ‘quality of life’ of animals as a fundamental component of society’s attempts to resolve questions about how animals ought to be treated. What lessons does this historical perspective hold for the scientiﬁc study of animal welfare? Let me brieﬂy introduce three points that will be developed throughout this book. First, we see that the scientiﬁc approach of investigating the welfare of animals is one approach, but not the only approach, that people have used when engaging in social debate about the proper treatment of animals. Science plays a distinct role, but some people continue to conceptualize the ethical problem in quite different ways. For those who see the central issue as one of rights, or justice, or religious tradition, or of fostering a caring attitude toward animals, the scientiﬁc study of animal welfare may seem irrelevant or even a harmful distraction. Here science ﬁnds itself competing for authority with other, and often older, modes of thought. Rather than assuming that the science simply trumps or replaces these other modes of thought, scientists need to understand, and be able to articulate, how the science ﬁts into the social debate.

Dfraser_C001.indd 22

6/27/2008 4:56:28 PM

Animals and Moral Concern

23

Second, we see that the study of animal welfare developed in response to certain kinds of problems, speciﬁcally ethical problems. Hence, when scientists study animal welfare, they need to keep the ethical application of their work visible, at least out of the corner of their eye. If they lose sight of the ethical context, their work may drift into irrelevance. They may, for example, try to redeﬁne animal welfare in a way that makes it amenable to scientiﬁc study but that does not correspond to the meaning (or meanings) that the term has in the social debate. In such a case, the scientists’ conclusions about what they call ‘animal welfare’ may simply confuse rather than clarify the issues that caused the research to be done in the ﬁrst place. For the science to be relevant to the social debate, scientists must be in touch with that debate, at least to the point of understanding the nature of the concerns that gave rise to the research, and the meaning (or meanings) that animal welfare holds for the rest of society. And third, we see that the scientiﬁc study of animal welfare developed in a particular historical and cultural context characterized by certain historically and culturally conditioned beliefs and values. When people discuss what constitutes a good life for animals, their ideas will be inﬂuenced by their understanding of ‘a good life’ and ‘animals’ – concepts which (as we will discuss in the next two chapters) have meant quite different things to different people. And when scientists enter their laboratories to study animal welfare, we need to understand how these same beliefs and values are likely to inﬂuence the questions they ask, the methods they use and the conclusions they draw.

Dfraser_C001.indd 23

6/27/2008 4:56:28 PM

Animals in the Human Mind

2

How should people treat animals? The answer will depend in part on who or what we perceive animals to be. Are they fundamentally different from people, or fundamentally similar? Are they so inferior to humans that we should not waste our compassion on them, or are they our equals, or even superior beings whom we should approach with respect and humility? In this chapter we will look brieﬂy at some different traditional beliefs about the nature of animals, and how these ‘factual’ beliefs (about what animals are like) have inﬂuenced ethical beliefs (about how animals should be treated). We will also consider how our beliefs about animals have been reshaped by modern science, and the ethical implications of this evolving scientiﬁc understanding. Thus the chapter also introduces a discussion, to be pursued later, about the interplay between science and ethics. BUT SINCE THAT SOUNDS like a rather academic programme, let’s start with a story: Now, it happened that the daughter of a great chief fell from heaven, down into this world which at that time was completely covered in water. Some waterfowl saw the young woman falling and they joined their bodies together and carried her until they became tired. Then the great Turtle carried her, and when he became tired, the animals asked themselves what they should do to provide her with a permanent dwelling place. Finally it was decided to prepare an island where she could live. The Turtle agreed to let its carapace serve as the base, and the toad was persuaded to dive to the bottom of the water and bring up soil. When the soil was spread over the carapace of the Turtle, it began to grow in size and depth and it formed the dry land on which people live today.1

1This is one of many versions of the story. I have adapted this text from: http://groups.msn.com/ TalesFromtheSmokehouse/vthewomanwhofellfromthesky.msnw, accessed January 2007. For another version see Johnston, B. 1976. Ojibway Heritage. University of Nebraska Press, Lincoln.

24

Dfraser_C002.indd 24

6/27/2008 12:41:40 PM

Animals in the Human Mind

25

This story, together with its many variants, describes the start of human life on earth according to the Ojibway, the Seneca, and many of the related cultures of the woodlands of central North America. These were people who survived harsh northern winters by killing and eating animals, and the story was one of many that reminded them how profoundly their existence depended on the assistance – even the pity – of other creatures.2 The creation story of the Bible arose in a culture not of hunters but of pastoralist herders for whom the tending of animals was a major cultural and economic activity. According to the opening chapter of the Bible, life on earth was made possible not by animals but by God. God began by separating the water from the sky, and then the dry land from the sea. Next God created the plants to cover the land, and the sun, moon and stars for light. Having thus prepared the world, God then created animals to inhabit it, and as the ﬁnal act of creation God formed the ﬁrst humans – both male and female – and commissioned them to rule over the other species in God’s place. Much later in the narrative, when God caused an enormous ﬂood to purge the world of the evil that had developed, the humans were responsible for saving the animals from drowning. This was roughly the opposite of how the Ojibway saw the relationship. There are, however, two creation stories in the Bible, probably written by different authors perhaps at quite different times.3 In the ﬁrst, just described, the humans were created after the animals and were to serve as God’s lieutenants by ruling over the other species. The second story, in Genesis 2, has a different ﬂavour. According to this version, God began by creating the ﬁrst man and fashioning a garden where the man could live. Then, wanting the man to have companionship, God created the various animals and brought them to him so that the man could give them names. When none of the animals proved to be a fully satisfactory companion, God made the ﬁrst woman, not de novo as he had done for the other species, but from a part of the man’s own body. Here the relationship between humans and animals was much less hierarchical, based more on companionship than authority. Thus, even in the opening passages of the Bible we see the beginning of a tension that would later become a dominant theme in the debate about our relationship to other species: are we to view animals as our subjects, to be ruled by humans as a king might rule over a nation of people, or are animals (as Robert Burns would put it much later) our earth-born companions and fellow mortals?4 2Johnston,

1976. biblical scholars believe that the opening chapters of the Bible include works by four different writers or groups of writers living at different times. Genesis 1 is thought to have come from the Priestly source, abbreviated ‘P’, and Genesis 2 from the Yahwist source, abbreviated ‘J’. See Hiebert, T. 1957. The Yahwist’s Landscape: Nature and Religion in Early Israel. Oxford University Press, New York. 4Paraphrasing Burns, R. 1786. To a mouse. Originally published in Poems, Chieﬂy in the Scottish Dialect. Republished 1955 as pages 111–112 in Poems and Songs of Robert Burns (J. Barke, editor). Collins, London. 3Many

Dfraser_C002.indd 25

6/27/2008 12:41:40 PM

26

Animal Welfare in Context

The Jain religion arose in a culture where sedentary agriculture was already established, and where there was no practical necessity either to raise animals as the Biblical herders had done, nor to hunt them for food. The Jain view of animals is captured not in a creation narrative but in a complex cosmology that sees humans, animals and the natural world as parts of a continuous and interwoven process of life, death and rebirth. As Jain scholar Christopher Chapple has explained, Jain taxonomy sees all life forms as falling into ﬁve levels which are distinguished by their sensory capacities.5 The lowest, which includes earth, air, microorganisms and plants, possess only the sense of touch. The next, which includes worms, leeches, oysters and snails, also have the sense of taste. The third, including most insects and spiders, add the sense of smell. The fourth, including butterﬂies, ﬂies and bees, also possess the ability to see. And the highest, which includes reptiles, birds, mammals and humans, have all four of these senses plus the ability to hear. According to this view of the world, all life forms have value, but the more complex beings have more value than the simplest. Humans remain the highest of all: a virtuous animal may be rewarded by being reborn in human form, and only from human form can one achieve the ultimate state of spiritual liberation. Nonetheless, as Chapple notes, the Jain faith ‘seeks to uphold and respect animals as being fundamentally in reality not different from ourselves’. Associated with each of these views about the nature of animals, and about the relationship between humans and animals, were certain correlated ethical beliefs about how animals should be treated. Among the North American hunting people, the Innu of northern Quebec and Labrador provide one of the most authentic examples because their culture survived intact long after the more southern cultures had been inﬂuenced by European contact. As recently as the 1930s, an anthropologist recorded and photographed the elaborate lengths to which the Innu went in order not to offend the animals they had killed. The most noble of their prey, and the one requiring the most elaborate demonstration of respect, was the bear. Out of courtesy, people referred to a slain bear not by the common name for bear, but by polite euphemisms such as the ‘Great Food’. Unmarried women were not allowed to look upon the bear for fear of insulting him. For the same reason, the tail must not be cut off, the right arm must not be severed from the paw, the meat must not be eaten out of doors, and only the oldest man in the community could eat the head and the right arm. The bear’s skull had to be treated with particular reverence: it was carefully cleaned, provided with beads and other gifts, and erected on a pole so that the bear could continue to see.6 5Chapple, C.K. 2006. Inherent value without nostalgia: Animals and the Jaina tradition. Pages 241–249 in A Communion of Subjects: Animals in Religion, Science and Ethics (P. Waldau and K. Patton, editors). Columbia University Press, New York. The quotation is on page 248. 6Speck, F.G. 1935. Naskapi, the Savage Hunters of the Labrador Peninsula. University of Oklahoma Press, Norman. The culture that Speck called Naskapi/Montaignais is now commonly called Innu.

Dfraser_C002.indd 26

6/27/2008 12:41:40 PM

Animals in the Human Mind

27

These attitudes and practices of the Innu are sometimes said to reﬂect ‘reverence’ for animals, although the use of the term is controversial given that much of the veneration occurred after deliberate killing. Nonetheless, the customs of the Innu showed a clear recognition that animals are ethically signiﬁcant beings and that harms done to them are important enough to require appropriate expressions of gratitude and respect. This deferential treatment of animals was also seen as serving a practical function. The Innu people depended utterly on animals for their own survival, and in their belief system successful hunting required the cooperation of an animal who would allow itself to be killed. If animals took offence at humans for treating their members badly, then hunting people could starve. Worse yet, animals could be dangerous; alone in the wilderness, hoping to subdue a bear with simple technology, an Innu hunter could not assume the power imbalance that modern hunters take for granted. Thus, showing reverence for animals was a manner of making peace with the animal world, of reconciling other species to the human need to exploit them, and thus securing the animals’ continued cooperation in the hunt. For the Jains, with their view of the continuity of life forms, the guiding ethical principle was ahimsa, a term roughly equivalent to non-violence or not causing harm. To the Jains, taking care not to harm other beings is a key virtue. Devout Jains will sweep a path in front of themselves to avoid stepping on ants and other small creatures; they will not eat outdoors so as not to consume ﬂying insects by accident; they may even breathe through a mask and ﬁlter their water before they drink it. The principle applies to a reduced degree to plants; thus, it is better to eat cherries, which can be harvested without harming the tree, than carrots whose consumption causes the plant to die.7 When walking outdoors monks and nuns will even avoid raising their arms so as not to frighten or disturb animals.8 Naturally their world-view made it impossible for devout Jains to be farmers who, in tilling the soil, would harm countless creatures ranging from burrowing rodents and ground-nesting birds to the tiny soil organisms that play a lesser but still signiﬁcant role in the Jain world. The pastoralist culture of the Bible, to which we will return in the next chapter, had a very different view of the moral status of animals. In a herding culture, it was essential that animals could be owned, traded and killed for human purposes. At the same time, human prosperity depended on providing domestic animals with appropriate care. In Biblical terms animals had to be rested in green pastures and led beside still waters; they also had to be defended when in danger and nursed back to health when injured. These practical demands of pastoralist life were reinforced by a culture that attached great value to the diligent care of animals. The biblical character Noah, who was singled out by God as the one virtuous man of 7Shah, H. 2005. Jainism. Available at: http://www.jcnc.org/jainism/reference_detail.asp? categoryNon%2Dviolence%2DQuestions%26Answers, accessed July, 2006. 8Chapple, 2006.

Dfraser_C002.indd 27

6/27/2008 12:41:40 PM

28

Animal Welfare in Context

his time, was given the task of preserving all the types of animals during the great ﬂood.9 Rebecca, who was chosen by God to be the wife of Isaac and the mother of her nation, was ﬁrst identiﬁed by her offer to provide water for the camels of a stranger ‘until they have done drinking’ – a not insubstantial task given the prodigious thirst for which camels are famous.10 David’s suitability to become king was ﬁrst indicated by the care and courage he showed in protecting his father’s sheep.11 Indeed, so strong was the positive value attached to diligent care of animals that a conscientious shepherd protecting a ﬂock of sheep was used as a metaphor for divine goodness.12 The message from these various examples – the Innu, the Jains and the Biblical herders – is that human cultures have what we might call an ‘animal mythology’. By mythology I do not mean incorrect or outmoded beliefs, but a set of fundamental assumptions and values that we see in the enduring stories, art and ideas of a culture, and which serve (in the words of historian Ronald Wright) as ‘the maps by which cultures navigate through time’.13 This animal mythology involves two elements. One consists of ‘factual’ beliefs about the nature of animals and their historical relationship to people. The other is a correlated set of evaluative and ethical beliefs about the importance of animals and appropriate conduct toward them. As we will see, if the factual beliefs evolve and change, the evaluative and ethical beliefs are likely to change as well. IT IS TEMPTING TO imagine that our modern Western culture, with its emphasis on science and rational thought, has abandoned any form of animal mythology and adopted a purely scientiﬁc understanding of animals and purely rational conclusions about how animals should be treated. To disabuse ourselves of this notion, let us brieﬂy consider how differently we portray and treat three types of animals. In modern Western culture, the most revered animal is, of course, the domestic dog. Dogs appear in countless stories and works of art, both traditional and contemporary, as the chief animal companion and ‘best friend’ of humans. Greyfriar’s Bobby, a Skye Terrier owned by an ageing shepherd, became one of Edinburgh’s most celebrated citizens by spending the rest of his life guarding the gravesite of his deceased owner.14 Lassie, the faithful collie sold by an impoverished family to a wealthy land-owner, demonstrated both her loyalty and her intelligence by making

9The

Bible, Genesis 6. Bible, Genesis 24:19. The quotation follows the King James Version, Oxford University Press, London. 11The Bible, 1 Samuel 17:35. 12The Bible, Psalms 23:1–4. 13Wright, R. 1992. Stolen Continents: Conquest and Resistance in the Americas. Houghton Mifﬂin, Boston. The quotation is on page 5. 14Atkinson, E. 1912. Greyfriar’s Bobby. Harper and Brothers, New York. 10The

Dfraser_C002.indd 28

6/27/2008 12:41:40 PM

Animals in the Human Mind

29

the seemingly impossible journey back to her original home.15 And in real life, Western practices are largely consistent with this uniquely positive portrayal: dogs are treated as members of human families, given distinctive names, rescued from abuse by public institutions, and totally exempted from slaughter for human food. Ironically, the animal at the lowest end of the scale of value has traditionally been the dog’s close relative the wolf. For much of Western history, the wolf has been cast as the arch-enemy of humans. In traditional folk-tales wolves connive to eat children, the elderly and the domestic animals that are cared for by humans, and the death of a wolf, no matter how gruesome, is invariably a source of satisfaction. When the Three Little Pigs killed the wolf by boiling him alive, this was merely a satisfactory ending to the tale; had the reverse occurred, the story would have been horriﬁc, even by the grim standard of traditional children’s literature. And again, real-life treatment of wolves in the West has ﬁt this negative image: for centuries we have hunted, trapped and poisoned wolves with few scruples. The animals kept for food production in the West fall between these two extremes. As thousands of children’s stories attest, the farm animals are a source of great interest and sometimes sympathy, pride and friendship, and they are essential elements of the rural landscape. They are certainly seen as worthy of care, and bear none of the negative loading associated with wolves. Nonetheless, they are valued more for their usefulness than for their loyalty, intelligence or individuality. Hence, Jack (who climbed the beanstalk) remained a sympathetic if gullible ﬁgure after selling the family cow for a few beans; had he sold the family dog, we would regard him as heartless. Thus, on a farm in Canada for example, a person who is deemed to be perfectly rational might begin the day by taking a sick and ageing dog to a veterinarian to prolong its life; then come home and ship a group of six-month-old pigs for slaughter, taking care not to cause them unnecessary distress; and then set out a leg-hold trap to do away with some annoying coyotes. Objectively, those three animal species are roughly similar in their level of mental functioning, their capacity for suffering, and probably most other attributes which (rationally) might make animals worthy of moral concern. The fact that we treat them so differently shows that Western culture does have an animal mythology which is captured in its stories and art, and which has a pervasive inﬂuence on human behaviour. IF SCIENCE HAS NOT simply replaced our traditional animal mythology in the manner of fact replacing ﬁction, it has certainly not been without inﬂuence either. Indeed science, as a valued component of Western culture, has led to important changes in our factual beliefs about animals, and in so doing has (I believe) had an important inﬂuence on our ethical beliefs as well.

15Knight,

Dfraser_C002.indd 29

E.M. 1940. Lassie Come-Home. Holt, Rinehart and Winston, New York.

6/27/2008 12:41:41 PM

30

Animal Welfare in Context

One of the longest-running debates in Western thought centres on the nature of animals. Dix Harwood nicely summarized the two poles. One body of thought, he said, holds that humans and animals ‘are very much alike, with the same emotions and similar mental processes’. The opposing view considers ‘that an unbridgeable chasm yawns between the human race and the other species’.16 The debate has been made more persistent and more passionate because of its implications for human self-deﬁnition: are we little more than beasts, or little less than gods? We have already seen two iterations of the tension: in ancient Greece between the Pythagoreans versus the Stoics and Epicureans, and in the two creation stories of the Bible. Among medieval Christian thinkers we see a similar polarization. St. Francis of Assisi (1181–1226), who saw a cosmic unity joining humankind with all of nature, addressed the birds and animals as his brothers and sisters. St. Thomas Aquinas (1225–1274), although encouraging kindly treatment of other creatures, saw them as fundamentally different because humans, unlike animals, have immortal souls.17 The same debate has extended into more recent times. In France, René Descartes (1596–1650) famously claimed that there is a fundamental difference between humans and animals in that humans have a unique capacity for rational thought which, in Descartes’ view, is the essence of human life. On this basis, Descartes (or at least his followers) saw animals as machine-like entities acting without thought or feeling.18 In contrast, Descartes’ countryman Voltaire (1694–1778) vehemently denounced this view. In his Philosophical Dictionary he opened the entry on animals with the words: ‘What a pitiful, what a sorry thing to have said that animals are machines bereft of understanding and feeling’, and he went on to ask, ‘ … has nature arranged all the means of feeling in this animal, so that it may not feel?’19 Again, in the German Enlightenment Immanuel Kant (1724–1804) emphasized the difference between humans and animals. His approach to ethics was based on the claim that we should treat our fellow humans not as means to our ends but as ends in themselves, whereas animals (he claimed) ‘are not self-conscious and are

16Harwood,

1928, page 6. R. 1999. Animals and Nature: Cultural Myths, Cultural Realities. University of British Columbia Press, Vancouver. But note Preece’s comment that ‘soul’ had a somewhat different meaning in St. Thomas’ time than it does today. 18Descartes, R. 1637. Discourse on the method for properly conducting reason and searching for truth in the sciences. Republished 1984–1991 in The Philosophical Writings of Descartes, Volume 1 (J. Cottingham, R. Stoothoff, D. Murdoch and A. Kenny, editors and translators). Cambridge University Press, Cambridge. There has been controversy over whether Descartes, in denying thought to animals, also denied feeling. For different views see Cottingham, J. 1993. A Descartes Dictionary. Blackwell, Oxford; and Steiner, G. 1998. Descartes on the moral status of animals. Archiv für Geschichte der Philosophie 80: 268–291. 19Voltaire, F.-M.A. 1764. A Philosophical Dictionary. Republished 1924 (H.I. Woolf, translator). Knopf, New York. The quotation is from the entry entitled ‘Animals’ on page 18. 17Preece,

Dfraser_C002.indd 30

6/27/2008 12:41:41 PM

Animals in the Human Mind

31

there merely as a means to an end. The end is man’.20 In contrast, the poet Johann Wolfgang von Goethe (1749–1832), believing strongly in a continuity linking humans and the other species, proposed that ‘Each animal is an end in itself ’.21 And a remarkably similar tension persists in philosophical debate today, with some contemporary philosophers emphasizing our similarity to animals while others emphasize our differences.22 But against this background of on-going disagreement, I believe we can also discern a general shift in how animals are perceived in Western culture. If we look at Medieval depictions of the creation story (Figure 2.1), I believe we can imagine the vast difference between humans and animals in the minds of the artists who created those images. First was a difference in appearance: the animals had four legs and fur, or wings and feathers, or ﬁns and scales. They looked nothing like the smooth-skinned biped that had been fashioned to resemble God. Second, humans had a different origin. People had not been created simply as one species among many, but in a separate act of creation that set them apart from the natural world. And third, to the extent that people of the time accepted the views of the ofﬁcial Christian church which sided with St. Thomas Aquinas on this issue, humans possessed a unique spiritual and intellectual nature that animals did not share. It would be wrong to stereotype the medieval view of animals. Presumably then as now, the ordinary women and men who lived with animals and looked after them on a daily basis recognized animals as beings with individuality and emotion. Yet inasmuch as the writers and artists of the period can be trusted to reﬂect the mythology of the time, the Medieval period seems to have been one when the pendulum had swung very far in the direction of emphasizing the differences between people and animals. Over the centuries, however, the beliefs and perceptions that supported the view of humans as unique have gradually been chipped away, and scientiﬁc research has played a large part in wielding the hammer. The ﬁrst perception to fall was that humans and animals differ fundamentally in appearance. The modern study of anatomy in Europe dates back at least to the fourteenth century,23 and by the 1500s cadavers were being researched with an intensity that other natural scientists devoted to cataloguing the heavens. Anatomists, like geographical explorers, staked their claims by giving their names to new-found anatomical entities: the tubes of Fallopius, the canals of Eustachius, the ﬁssure of Sylvius.

20Preece,

1999. The quotation is on page 123. R. 2003. Darwinism, Christianity, and the great vivisection debate. Journal of the History of Ideas 64: 399–420. The quotation is on page 409. 22The contrast and some examples are discussed by Fraser, D. and Preece, R. 2004. Animal ethics and the scientiﬁc study of animals: bridging the ‘is’ and the ‘ought’. Essays in Philosophy 5, available at: http://sorrel.humboldt.edu/essays/fraser.html. 23Singer, C. 1957. A Short History of Anatomy from the Greeks to Harvey. Dover Publications, New York. 21Preece,

Dfraser_C002.indd 31

6/27/2008 12:41:41 PM

32

Animal Welfare in Context

Figure 2.1 God bringing forth Adam after having created the natural world. From the ‘Morgan Crusader Bible’, believed to have been created around 1250, likely in Paris. I am indebted to Professor Carol Knicely, Department of Art History, University of British Columbia, for kindly identifying the source of the picture. Reproduced with permission, The Morgan Library and Museum.

And this scientiﬁc investigation was no ivory tower affair. ‘Dissecting theatres’ (Figure 2.2) sprouted up across Europe in the major centres of learning, and they allowed the public to witness the dissection of an animal or, better yet, of a human criminal cut down fresh from the gallows. In fact, some of the very criminals whose execution provided street-corner entertainment for the masses likely served double duty by providing the material for anatomical demonstrations. Through such research, and this remarkably direct form of public education, it came to be

Dfraser_C002.indd 32

6/27/2008 12:41:41 PM

Animals in the Human Mind

33

Figure 2.2 View of the Leiden anatomy theatre, depicted by Bartholomeus Dolendo c. 1609. Reproduced from Figure 6 in Sawday, J. 1995. The Body Emblazoned: Dissection and the Human Body in Renaissance Culture. Routledge, London. Reproduced by the kind permission of the Department of Special Collections, Leiden University Library.

recognized that humans are built on the same anatomical template as the other vertebrate animals. According to Harwood, the anatomical resemblance between humans and other species ‘was obvious to comparatively few in 1600; in 1700 nearly everybody recognized it’.24 With the anatomical similarity thus established, the 1700s and 1800s saw scientists and writers alike struggling to identify the deeper implications of this new knowledge. As early as 1734 the English poet Alexander Pope (1688–1744) used the common metaphor of the ‘Great Chain of Being’ to propose that all life is interconnected, from God and ‘natures ethereal’ down to those tiny creatures that ‘no eye can see’: Vast chain of being, which from God began, Natures ethereal, human, angel, man 24Harwood,

Dfraser_C002.indd 33

1928, page 156.

6/27/2008 12:41:42 PM

34

Animal Welfare in Context

Beast, bird, ﬁsh, insect! what no eye can see, No glass can reach! from Inﬁnite to thee … Nothing is foreign; parts relate to whole; One all-extending, all-preserving, soul Connects each being, greatest with the least; Made beast in aid of man, and man of beast; All serv’d all serving: nothing stands alone; The chain holds on, and where it ends unknown.25

If for Pope the similarity of humans and other species was a cause for such cosmic speculation, for Carl Linnaeus, the Swedish scientist who created the binomial system of classifying plants and animals according to genus and species, it created a scientiﬁc conundrum. In 1747 he wrote to a fellow naturalist that he could not ﬁnd any ‘generic character … by which to distinguish between Man and Ape’. But realizing the delicate nature of the issue, Linnaeus decided to place humans into the separate genus ‘Homo’ noting ‘if I had called man an ape, or vice-versa, I would have fallen under the ban of the ecclesiastics’.26 By the end of the 1700s, the discussion had shifted from metaphysical claims about the interconnectedness of life, to the more concrete idea that the anatomical similarities might be due to a common evolutionary ancestry. In his work Zoonomia, the brilliant and diverse English intellectual Erasmus Darwin (1731–1802) questioned whether species are immutable entities. Might not one species be somehow transformed into another? In fact, he wrote (some years before the birth of his famous grandson Charles), ‘ … would it be too bold to imagine, that all warmblooded animals have arisen from one living ﬁlament … ?’27 But by what means might species have evolved? Early evolutionists imagined a process of gradual transformation whereby characteristics acquired by one animal during its life are somehow passed on to its offspring.28 Then, in the 1850s Charles Darwin and Alfred Russel Wallace proposed the mechanism that is now generally accepted to explain the transformation of one species to another. Nature, they suggested, produces an over-abundance of animals varying in certain ways; some variants are better adapted to survive and reproduce; and the better adapted leave more descendants than the others. Through this process of ‘natural selection’, 25Pope, A. 1734. An Essay on Man. Republished 1948 as pages 97–137 in Alexander Pope: Selected Works (L. Kronenberger, editor). The Modern Library, New York. The quotation (following Preece, 1999, page 121) combines Epistle 1, Section 8, lines 5–8, and Epistle 3, Section 1, lines 21–26. 26Sagan, C. and Druyen, A. 1992. Shadows of Forgotten Ancestors: A Search for Who We Are. Random House, New York. The quotation is on page 274. 27Darwin, E. 1796. Zoonomia; Or, the Laws of Organic Life, Volume 1. Republished 1801, J. Johnson, London. The quotation is in ‘Generation’, section 39. 28Preece, R. 1999, especially pages 157–159; Lamarck, J.B. 1809. Philosophie Zoologique. Republished 1984 as Zoological Philosophy: An Exposition with Regard to the Natural History of Animals (H. Elliot, translator). University of Chicago Press, Chicago.

Dfraser_C002.indd 34

6/27/2008 12:41:47 PM

Animals in the Human Mind

35

they claimed, new species gradually emerge. And the massive amount of evidence collected by Darwin gave the idea of evolution a plausibility that previous evolutionary theories had lacked. Thus, by the late 1800s the perception that humans possess a unique physical form had long been abandoned, and the seeds had been sown for a belief in a common origin. Where medieval thinkers had seen a ‘Great Chain of Being’ with God and other heavenly beings at the top, animals (from higher to lower) beneath, and humans as a special kind of being that bridges the natural and supernatural realms, evolutionists saw instead a kind of family tree wherein Homo sapiens counted as one species among countless others, most closely related to a minor group of primates with opposing thumbs and large brains, and sharing its more distant ancestry with all the animal kingdom. But even if humans shared their femurs and their distant origins with other species, surely they were still unique in being rational, emotional and spiritual beings as evidenced especially by their use of language. Yet even this claim came to be challenged. One challenge came from Charles Darwin himself. In 1872, the same year that he published the ﬁnal edition of the Origin of the Species, Darwin produced a much less famous work entitled The Expression of the Emotions in Animals and Man, in which he proposed that many species share similar emotional experiences – fear, pain, pleasure, affection, anger – and often express them in similar ways.29 Darwin’s contemporary George Romanes took up this theme with a book entitled Animal Intelligence. Noting that ‘within the last twenty years the facts of animal intelligence have suddenly acquired a new and profound importance, from the proved probability of their genetic continuity with those of human intelligence’,30 Romanes set out to classify the mental powers of the different animal species, much as a comparative anatomist might classify variations in anatomical traits. His method was to collect narrative accounts illustrating the mental abilities of animals, but limiting himself to accounts recorded by observers whom he considered reliable or to observations that he considered to have been ‘corroborated by similar or analogous observations made by other and independent observers’.31 Romanes amassed a large body of evidence showing the various intellectual powers of animals arranged from mollusks, insects and other invertebrates, through reptiles, ﬁsh, birds, mammals, and, as the ﬁnal chapter, monkeys, apes and baboons. As one example, Romanes described the behaviour of an elephant that had developed a disease of the eyes and had been blind for several days. A local

29Darwin, C. 1872. The Expression of the Emotions in Man and Animals. Republished 1965, University of Chicago Press, Chicago. 30Romanes, G.J. 1891. Animal Intelligence. D. Appleton and Company, New York. The quotation is on page vi. 31Romanes, 1891, page ix.

Dfraser_C002.indd 35

6/27/2008 12:41:47 PM

36

Animal Welfare in Context

doctor agreed to try treating one of the eyes with silver nitrate, a remedy commonly used for similar eye ailments in humans. The animal was accordingly made to lie down, and when the nitrate of silver was applied, uttered a terriﬁc roar at the acute pain which it occasioned. But the effect of the application was wonderful, for the eye was in a great degree restored, and the elephant could partially see. The doctor was in consequence ready to operate similarly on the other eye on the following day; and the animal, when he was brought out and heard the doctor’s voice, lay down of himself, placed his head quietly on one side, curled up his trunk, [and] drew in his breath like a human being about to endure a painful operation.32

More experimental approaches to understanding animal intelligence were also undertaken. In 1927 primatologist Robert M. Yerkes published his work The Mind of a Gorilla based on his experiences in studying ‘Congo’, a captive gorilla aged about ﬁve years. Yerkes devised an extensive series of experiments requiring the gorilla to use a stick or to stack one wooden box on top of another in order to reach food. He noted in particular that Congo’s behaviour often failed to follow the pattern one would expect if she learned simply by trial and error. Instead of attempting a large number of different responses until she hit upon a solution, Congo often seemed to solve a problem more by observation, reﬂection and insight. But Yerkes also expressed doubt over the ability of such experiments to reveal the true mental capacity of other species. ‘The experiment may be a human masterpiece’, he wrote, ‘but as likely as not it is so contrived as to give the animal meager opportunity to utilize its peculiar adaptive or expressive capacities’.33 Despite these intriguing beginnings, during the middle decades of the twentieth century the scientiﬁc study of animal behaviour moved in different directions.34 Under the inﬂuence of a school of thought known as Positivism, whose effects we will discuss in a later chapter, the early attempts to understand the thoughts and emotions of animals were largely abandoned, and people looked instead for more mechanistic explanations – brain mechanisms, hormonal changes, ecological triggers – that could account for behaviour without any need to invoke the animals’ thoughts, emotions and other experiential states. Although most scientists who studied animal behaviour did not speciﬁcally deny that animals have mental lives, such states were largely by-passed in the search for other types of explanations. By about 1970, however, a different generation of scientists was beginning to restore the cognitive and emotional processes of animals as a subject for scientiﬁc study. One of the pioneers (as we noted earlier) was Jane Goodall, who studied chimpanzees not simply to collect numerical data so as to calculate norms and averages for the species, but using methods closer to cultural anthropology, studying 32Romanes,

1891, page 399. R.M. 1927. The Mind of a Gorilla. Genetic Psychology Monographs 2: 1–193. The quotation is on page 137. 34Rollin, B.E. 1990. The Unheeded Cry. Oxford University Press, Oxford. 33Yerkes,

Dfraser_C002.indd 36

6/27/2008 12:41:47 PM

Animals in the Human Mind

37

animals more as persons with individuality, unique life histories, and complex social and mental lives. One of Goodall’s most touching narratives concerned the chimpanzee that she called ‘Flint’, a young male who, even at the mature age of eight years, was still strongly attached to his ageing mother, Flo. Here, as an example of her work, is Goodall’s description of Flint’s reaction when his mother died: So far as we know, Flint was the only one with Flo when she died; he was in a tree overhead when we found her. After a while he approached the body, bending right down to stare into her dead eyes. He reached to touch her, brieﬂy groomed her arm, and then moved away … After Flo’s death [Flint] became increasingly depressed and lethargic … . By the second week Flint was spending most of his time lying on the ground, often under thick clumps of vegetation, always close to where he had last seen Flo. His eyes, which sank ever deeper in his head, acquired a glossy lustre and sometimes he stared unwinkingly ahead with a gaze that gave the impression of insanity. He ate seldom, and by the end of the third week had lost more than a third of his weight. Within a few days he was dead. Flint’s death was a tragedy in every way; at the same time it is an amazing testimony to the depth and signiﬁcance of the affectionate bond which can unite a chimpanzee child to his mother.35

Accompanying these natural history observations were remarkable developments in other aspects of the study of animal behaviour. In a series of three books beginning with The Question of Animal Awareness, behavioural scientist Donald Grifﬁn helped to restimulate scientiﬁc interest in the mental lives of animals.36 Other scientists took up the challenge by producing studies of the cognitive powers and stages of mental development of animals. Before the end of the century scientiﬁc books began to appear with such frankly mentalistic titles as How Monkeys See the World: Inside the Mind of Another Species and Reaching into Thought: The Minds of the Great Apes – titles that harked back to Yerkes but that would have been unthinkable for scientiﬁc books during the intervening decades.37 The result of all this scientiﬁc activity has been a shift in human perception of animals to the point that some species, at least, are now widely seen as experiencing rich mental and emotional lives. 35Goodall, 1971. The quotation is from ‘Postscript 1972’ on pages 261–262 of the edition published by Fontana Books, London, 1973. 36Grifﬁn, D.R. 1976. The Question of Animal Awareness. Rockefeller University Press, New York; Grifﬁn, D.R. 1984. Animal Thinking. Harvard University Press, Cambridge; Grifﬁn, D.R. 1992. Animal Minds. University of Chicago Press, Chicago. 37Cheney, D.L. and Seyfarth, R.M. 1990. How Monkeys See the World: Inside the Mind of Another Species. University of Chicago Press, Chicago; Russon, A.E., Parker, S.T. and Bard, K.A. (editors). 1996. Reaching into Thought: The Minds of the Great Apes. Cambridge University Press, New York.

Dfraser_C002.indd 37

6/27/2008 12:41:47 PM

38

Animal Welfare in Context

This further shrinking of the human–animal divide in the late twentieth century was accompanied by a growing concern for animals that rivalled, and then arguably surpassed, what had been seen in some European countries a century earlier, during the ﬁrst great wave of animal protectionism. Clearly, science was not the only factor that brought about the growing concern for animals. The change from horses to automobiles, and the demographic shift from rural to urban living, meant that most people experienced animals as companions and family members rather than in utilitarian roles as providers of transportation or food. The end of two World Wars and the Great Depression freed people in the industrialized countries to be concerned about matters other than personal security and the necessities of life. And the media, from nature ﬁlms to animated cartoons, depicted animals sometimes as fascinating natural beings and sometimes as sympathetic humanized ones. But whatever the role of these other factors, the altered scientiﬁc understanding of animals reinforced and gave weight to the altered popular understanding. To summarize, Western perceptions of animals, which have always been pulled between the two poles of emphasizing our similarities with other species and emphasizing our differences, have been evolving slowly over several centuries and perhaps more rapidly over the past 50 years. By the later part of the twentieth century, the gap that people in the West perceived between humans and other species had narrowed substantially. In particular, humans and other species were seen as sharing a common anatomical form, a common evolutionary ancestry and, in the case of some species, a complex mental and emotional life. As the perceived gap separating humans and animals became progressively smaller, people directed more attention and sympathy toward animals, and this set the stage for a remarkable reshaping of ethical beliefs about how animals should be treated. LET US PAUSE AT this point and reﬂect on the role played by science in the new perception of animals and their ethical signiﬁcance, and (more broadly) on the relationship between science and culture. Science is sometimes depicted as an alternative – even a kind of opposite – to the sort of popular beliefs that I have been calling mythology. There are, of course, important characteristics of scientiﬁc understanding that set it apart from some other types of beliefs. For example, the beliefs that constitute science are subject to a particular kind of critical scrutiny, based on certain types of evidence. Karl Popper termed the process ‘falsiﬁcation’: if the Earth is ﬂat, we would predict that we could not travel always in one direction and return to our starting point; the fact that we can do so falsiﬁes the hypothesis that the Earth is ﬂat.38 The criterion of evidence-based falsiﬁcation creates an important difference between scientiﬁc beliefs and those that are accepted on the basis of tradition or authority. When Galileo was accused of heresy by a church court for stating that the Earth orbits the sun, he saw the issue as a scientiﬁc one – as a matter to be decided on the basis of 38Popper,

Dfraser_C002.indd 38

K.R. 1959. The Logic of Scientiﬁc Discovery. Basic Books, New York.

6/27/2008 12:41:47 PM

Animals in the Human Mind

39

evidence – whereas the church ofﬁcials assumed that the matter could be settled by appealing to the authority of the church. By the twentieth century, such issues were clearly recognized as the province of science. Thus when Linus Pauling proposed a three-strand model for the structure of DNA, even the immense authority of a scientist with two Nobel Prizes could not hide the fact that the evidence actually supported a double-helix model proposed soon after by two young upstarts.39 Nonetheless, it would be a mistake to think that our understanding of animals and nature has simply passed from a time of traditional beliefs to a time of scientiﬁc beliefs. Science is an important source of beliefs within our culture, but it is only one source. It is also a relatively recent source which so far has dealt with only some of the questions that concern us. Science has greatly reshaped our beliefs about how to prevent infectious diseases but has told us remarkably little about how to prevent feelings of guilt; science has reshaped our thinking about the movement of stars but has left our understanding of love and ambition largely to the realm of literature. And in some cases, the insights that we gain from science are slow to enter popular understanding perhaps because they seem less immediate than ideas that can be passed on in art, literature and the media. Thus, for example, people may avoid camping in bear habitat because of a traditional fear of wild animals, but continue in activities such as over-eating which (science indicates) involve much greater risks. The West’s mythology – including its cultural beliefs about animals – has not disappeared in the face of scientiﬁc understanding; rather, like a landscape that has undergone volcanic activity, it has been variously augmented, modiﬁed, overlain or, in some cases, left untouched by the beliefs we obtain from science. It would also be a mistake to see science as an isolated element of modern culture, sealed off from values and ethical beliefs. Instead, by reshaping factual beliefs, science also reshapes what people see as interesting, important, and hence worthy of attention and respectful treatment. Science does not answer ethical questions, but it inﬂuences the kind of ethical questions we ask and the kind of answers we ﬁnd satisfying. This inﬂuence is to some extent cross-cultural. Modern science is an off-shoot of Western thought, but its inﬂuence has become increasingly global. The view of animals arising from science will almost certainly not be limited to the industrialized countries of the West. Rather, inasmuch as science is becoming a global activity, its view of animals may well erode some of the cultural differences in attitudes toward animals that have existed in the past. Finally, we need to ask whether the interplay of science and culture is simply a one-way street whereby science inﬂuences cultural beliefs and values, or whether intellectual trafﬁc also moves in the opposite direction – whether changes in cultural beliefs and values also help to shape science. Did scientists in the late twentieth century just happen to study the communication and social behaviour 39Watson, J.D. 1968. The Double Helix: A Personal Account of the Discovery of the Structure of DNA. Atheneum, New York. The passage about Pauling’s model is on pages 92–95.

Dfraser_C002.indd 39

6/27/2008 12:41:48 PM

40

Animal Welfare in Context

of chimpanzees and gorillas, and then, with the data duly collected and published, people’s attitudes toward these animals changed? Or was there rather a cultural change – perhaps a growing tendency to see such animals as interesting beings whose communication and social lives are worthy of study – which led to, and helped to shape, the scientiﬁc study of these species? Did Darwin simply arrive at the theory of evolution because of the rich biological information he had collected, or was his thinking inﬂuenced by a culture – indeed, a family – which had already embraced the concept of a fundamental unity between humans and other species? My tentative conclusion is that the relationship between science and popular culture is a rich and complex one: that science inﬂuences popular beliefs and values, and that these in turn inﬂuence science. Later we will explore this interplay, including how cultural values and beliefs inﬂuenced animal welfare science and vice versa. But ﬁrst let us go back to the issue of animal ethics and complete our scan of the various modes of thought which have guided Western beliefs about the proper treatment of animals, and which (I will argue) inﬂuenced the kind of science that people created when trying to understand animal welfare.

Dfraser_C002.indd 40

6/27/2008 12:41:48 PM

A Good Life for Animals

3

What constitutes a good life for animals? I think we can discern at least four major modes of thought in Western culture that inﬂuence how people answer this question today. Some of these modes of thought are directly concerned with animals and our relationship to them; others are concerned with such basic themes as ‘nature’ and ‘progress’. For convenience I will label these modes of thought Pastoralism, Agrarianism, Romanticism and Industrialism, but recognizing that the labels are imperfect. All four, I will argue, have helped to shape current thinking about what kind of life animals ought to be allowed to live.1 OF THE VARIOUS VIEWS of animals that continue to inﬂuence Western people today, perhaps the most traditional is the pastoralist ethic of care derived from the Bible and introduced brieﬂy in Chapter 2. Given its importance, and the controversies that have raged over it, let us think carefully about the elements of this pastoralist world-view and its implications for the modern debate about animal welfare. The Bible has often been portrayed as providing a kind of licence for unrestrained exploitation of animals and nature. In one famous expression of the argument, the American medievalist Lynn White claimed that the historical roots of today’s environmental crisis lie in the Jewish and Christian scriptures. Our misuse of the environment, according to White, stems from ‘the Christian dogma of man’s 1In

this chapter as in the previous ones, I am wearing the hat of a historian of ideas. It is not a hat that I wear comfortably. However, despite the excellent work of such historians as Rod Preece, John Passmore, K.V. Thomas and Harriet Ritvo (listed in the bibliography) who have dealt ably with certain periods and themes, I do not believe that historians have yet provided a comprehensive view of the diversity of our ideas about animals. I offer this chapter as a beginning, and hope that it will soon be superseded by a more thorough and knowledgeable analysis. For the purpose of this chapter I have omitted the hunter–gatherer view because I do not see it exerting a major and continuing inﬂuence in Western thought today. It is a pleasure to acknowledge the debt I owe to Rod Preece in the development of these ideas.

41

Dfraser_C003.indd 41

6/27/2008 12:42:01 PM

42

Animal Welfare in Context

transcendence of and rightful mastery over nature’. According to this view, animals, as part of nature, had been created ‘explicitly for man’s beneﬁt and rule’ and were available for humans to dominate and use as they saw ﬁt.2 Such interpretations of Biblical thought have been criticized on many grounds,3 but for my purposes, their key failing is that they do not acknowledge the pastoralist ethic which formed an important element of the Biblical world-view. The pastoralist economy of the Hebrew people required that domestic animals be owned, traded and used for human purposes, but at the same time they had to be given appropriate care.4 With this mixture of demands, the Bible placed animals in a special moral category: they were not seen as equal to humans, but neither were they mere objects. Instead, they were viewed as beings that had been created by God and assigned to people in a relationship called rada or ‘dominion’, a term that was variously used to describe the relationship of God to the world and of conquering people to conquered people.5 It was a relationship of unequal power, but it did not imply the lesser party to be of no moral signiﬁcance. The pastoralist ethic captured in the Bible included at least two elements that guided how people should deal with animals.6 First, it allowed animals to be used for certain purposes as long as appropriate conventions were observed. Domestic animals could be eaten, but they were to be slaughtered and prepared in a ritually correct manner. Animals could be used for labour, but they, like human servants, were to be given the customary day of rest, and certain inappropriate muzzling and harnessing practices were expressly forbidden. Second, as we have seen earlier, the pastoralist ethic attached great value to the diligent care of animals. Rescuing animals, like healing sick humans, was one of the few tasks permitted on the Sabbath when other work was not allowed. The occupation of shepherd was one of the few (along with king, teacher and judge) that could be used when people tried to describe the attributes of God. In fact, as we see in Box 3.1, some descriptions of divine love sounded remarkably like lessons in animal husbandry. The Bible did not, of course, present a consistent view of how animals should be treated, just as it was not monolithic on many themes. Occasional passages provided different views of appropriate conduct toward animals by condemning 2White,

L. 1967. The historical roots of our ecological crisis. Science 155: 1203–1207. The quotations are on pages 1206 and 1205 respectively. 3Some of these are summarized in Preece, R. and Fraser, D. 2000. The status of animals in Biblical and Christian thought: A study in colliding values. Society and Animals 8: 245–263. Examples, listed in the bibliography, include Moncrief, 1970; Dobel, 1977; Bratton, 1984; and Linzey, 1991. 4Schochet, E.J. 1984. Animal Life in Jewish Tradition: Attitudes and Relationships. KTAV Publishing House, New York. 5Schochet, 1984; Preece and Fraser, 2000. Examples of the diverse use of ‘dominion’ include Psalm 72:8 and Judges 14:4 from the King James Version of the Bible. 6The biblical sources cited in this paragraph are (respectively) Deuteronomy chapters 12–14; Exodus 20:10; Deuteronomy 25:4; Luke 13:15; Psalm 23.

Dfraser_C003.indd 42

6/27/2008 12:42:01 PM

A Good Life for Animals

43

Box 3.1 A Biblical passage from the prophet Ezekiel (likely composed around 590 BC) using the behaviour of a diligent shepherd to describe God‘s love for the people of Israel. For these are the words of the Lord God: Now I myself will ask after my sheep and go in search of them. As a shepherd goes in search of his sheep when his flock is dispersed all around him, so will I go in search of my sheep and rescue them no matter where they were scattered in dark and cloudy days … . I will graze them on the mountains of Israel, by her streams and in all her green fields. I will feed them on good grazing-ground, and their pasture shall be the high mountains of Israel … . I myself will tend my flock, I myself pen them in their fold, says the Lord God. I will search for the lost, recover the straggler, bandage the hurt, strengthen the sick, leave the healthy and strong to play, and give them their proper food. The Bible, Ezekiel 34: 11–16, The New English Bible. Oxford University Press and Cambridge University Press, 1970.

cruelty and by claiming a fundamental kinship between humans and other species.7 Nor were all the stories about animals positive. For example, the Greek philosopher Porphyry noted with disapproval that when Christ cast a devil out of a possessed man, he transferred the devil into a herd of swine who then hurled themselves off a cliff. This story may have had a certain cosmic resonance to the Hebrew mind – unclean spirits transferred to unclean animals who then met an unclean death – but this made no impression on Porphyry who took the story literally as a sign of Christian disregard for pigs.8 The Bible’s pastoralist ethic of legitimate use combined with diligent care has remained an inﬂuential moral idea long after pastoralist herding has ceased being the dominant form of animal production in the West. In fact, when we listen to conscientious animal producers today, we often hear them expressing much the same combination of use and care that we observe in Biblical texts. As one Canadian dairy farmer put it: Our life’s work is to make the life of our cows the best possible … . Our animals are never hungry, thirsty, homeless and never really at a loss for company … I feel that they live a full, productive and useful life. Ultimately I have to feel responsible for them for I was responsible for their arrival. And when my old cows complete their life here

7The

Bible: Proverbs 12:10 and Ecclesiastes 3:18–20. 1993, page 181.

8Sorabji,

Dfraser_C003.indd 43

6/27/2008 12:42:01 PM

44

Animal Welfare in Context

I have to think that it was better for them to have lived the life I helped provide for them, rather than not to have lived at all.9

What, then, constitutes a good life for animals? In the world-view of the Biblical pastoralists, the answer must lie in diligent and skilful care of animals by their human keepers. Within this world-view the key issue is not the ‘welfare’ of animals as such. Indeed, some modern animal producers see the very term ‘welfare’ as being the language of their critics and foreign to their own mode of thought. To them the key issue is whether animals are receiving appropriate care. MUCH AS TRADITIONAL PASTORALIST herding became linked to an ethic of animal care, settled ‘agrarian’ agriculture became linked to the view that living close to the land brings out the best in humankind. In Agrarianism in American Literature, literary scholar Thomas Inge shows how persistent this idea has been in Western thought.10 As early as the fourth century BC, Aristotle had proposed that: The best common people are the agricultural population, so that it is possible to introduce democracy as well as other forms of constitution where the multitude lives by agriculture or by pasturing cattle.11

In Rome, writers such as Cicero (106–43 BC), Cato the Elder (234–149 BC), Horace (65–8 BC) and Virgil (70–19 BC) extolled agriculture as the noblest of occupations and the most likely to foster virtuous conduct. Within English literature, Inge argues, ‘The contrast between the corrupt, disorderly life of the city and the virtuous, simple life in the country became… one of the most common thematic motifs’.12 The idea was taken up by Geoffrey Chaucer (1343–1400), who contrasted the virtuous simplicity of rustic living versus the decadence brought on by commerce. In the eighteenth century the theme found natural soul-mates among Romantic writers and artists who looked back to a Golden Age of simple, rural living and contrasted it with the shallow, artiﬁcial

9Davidson, A.B. 1995. It isn’t all black and white. Pages 25–30 in Decision Making and Agriculture: The Role of Ethics (K.B. Beesley, S. Burns, M. Campbell and P. Sanger, editors). Rural Research Centre, Truro, Canada. The quotation is on page 28. Elsewhere I have discussed some of the practical challenges facing farmers who subscribe to this pastoralist ethic; see Fraser, D. 2006a. Caring for farm animals: Pastoralist ideals in an industrialized world. Pages 547–555 in A Communion of Subjects: Animals in Religion, Science and Ethics (P. Waldau and K. Patton, editors). Columbia University Press, New York. 10Inge, T.M. (editor). 1969. Agrarianism in American Literature. The Odyssey Press, New York. 11Inge, 1969. The quotation is on page xv. 12Inge, 1969. The quotation is on pages xvi–xvii.

Dfraser_C003.indd 44

6/27/2008 12:42:02 PM

A Good Life for Animals

45

and unhealthy life being created by the Industrial Revolution. The poet Robert Burns (1759–1796) ended an admiring poem about a Scottish peasant family, with the prayer: And, O, may Heaven their simple lives prevent From luxury’s contagion, weak and vile! Then, howe’er crowns and coronets be rent, A virtuous populace may rise the while …13

And Oliver Goldsmith (1728–1774), in lamenting what he saw as the decline brought about by the industrialization of England, imagined an earlier and better time of agrarian living: A time there was, ere England’s griefs began, When every rood of ground maintained its man; For him light Labor spread her wholesome store, Just gave what life required, but gave no more: His best companions, innocence and health; And his best riches, ignorance of wealth.14

But if industrialization and urbanization were destroying health and virtue in the Old World, the New World still offered hope. With its vast spaces and potential for agriculture, the New World seemed the ideal place to recreate virtuous agrarian society. Thomas Jefferson (1743–1826) took up the theme in his Notes on the State of Virginia (Box 3.2) with the famous pronouncement that, ‘Those who labour in the earth are the chosen people of God, if ever he had a chosen people …’. And Jefferson proposed that America, with its abundance of land, should remain an agricultural nation, sending raw products to Europe for further manufacturing rather than risk infecting the new republic with the degenerate morals that accompany city life and an industrial economy. In fact, as philosopher Paul Thompson notes, under the inﬂuence of Jefferson and others, agrarian living came to be seen as fundamental to the success of democracy.15 With the French Revolution in Europe and the American Revolution in the New World, both continents were searching for new forms of government, and democracy was an obvious candidate. At the time, however, democracy was a largely unknown option and many found it thoroughly disturbing. The very word 13Burns, 1786. The Cotter’s Saturday Night. Republished as pages 105–110 in the Barke edition. The quotation is from stanza 20. 14Goldsmith, O. 1770. The Deserted Village, A Poem. N. Douglas, London. The quotation is from lines 57–62. 15Thompson, P.B. 1998. Agricultural Ethics: Research, Teaching, and Public Policy. Iowa State University Press, Ames.

Dfraser_C003.indd 45

6/27/2008 12:42:02 PM

46

Animal Welfare in Context

Box 3.2 Thomas Jefferson‘s famous passage from Notes on the State of Virginia, claiming that agrarian living (or ‘husbandry’) promotes virtue and good citizenship, and that America should focus on producing agricultural products to be sent to Europe for further manufacturing so that the degenerate morals of industrial life would not undermine virtue and good government in the new republic. Those who labour in the earth are the chosen people of God, if ever he had a chosen people, whose breasts he has made his peculiar deposit for substantial and genuine virtue. It is the focus in which he keeps alive that sacred fire, which otherwise might escape from the face of the earth. Corruption of morals in the mass of cultivators is a phaenomenon of which no age nor nation has furnished an example … . While we have land to labour then, let us never wish to see our citizens occupied at a workbench, or twirling a distaff. Carpenters, masons, smiths, are wanting in husbandry [that is, are needed in agriculture]: but, for the general operations of manufacture, let our workshops remain in Europe. It is better to carry provisions and materials to workmen there, than bring them to the provisions and materials, and with them their manners and principles. The loss by the transportation of commodities across the Atlantic will be made up in happiness and permanence of government. The mobs of great cities add just so much to the support of pure government, as sores do to the strength of the human body. It is the manners, and spirit of a people which preserve a republic in vigour. A degeneracy in these is a canker which soon eats to the heart of its laws and constitution. Thomas Jefferson, 1787. Notes on the State of Virginia, Second American Edition, Query XIX: The present state of manufactures, commerce, interior and exterior trade? Republished 1969 as pages 9–10 in Agrarianism in American Literature (T. Inge, editor). The Odessey Press, New York.

tended to be pejorative – something closer to mob-rule than to government by the people.16 Surely, critics argued, ordinary citizens could not be trusted to govern a nation wisely because they would take too selﬁsh and short-term a view. What would stop them from voting themselves beneﬁts that the nation could not sustain? In the United Kingdom a second tier of government, occupied by wealthy landowning families with a long history of involvement with the land and affairs of state, was seen as counter-balancing the short-sightedness that might arise among elected

16Williams, R. 1958. Culture and Society 1780–1950. Harper Torchbook edition (published 1966), Harper & Row, New York. The idea appears on page xii.

Dfraser_C003.indd 46

6/27/2008 12:42:02 PM

A Good Life for Animals

47

representatives. But in the New World (so the advocates of democracy claimed) ordinary citizens would exercise their democratic powers prudently because they would own and work the land. Thus, as Thompson concludes, in America smallscale agrarian agriculture came to be seen not merely as the well-spring of virtue but as essential to the success of democratic government.17 Unlike pastoralism, agrarianism was not deﬁned principally in terms of human care of animals, yet animals were still an integral part of both the perception and reality of agrarian life. In the rural landscapes of Thomas Gainsborough (1727–1788), John Constable (1776–1837), and many other painters, farm animals drinking from a stream or resting in ﬁelds or woodlands, often accompanied by their human attendants, formed one of the most persistent and positive themes in the visual arts. For painters such as George Morland (1763–1804), Edwin Henry Landseer (1802–1873) and others, animals were an essential element of the domestic life of simple country people. Other painters, such as Jean-François Millet (1814–1875), used the care of animals – people leading cattle to drink or carrying a new-born calf in from the ﬁelds – as a way of showing the gentleness and goodness of country folk. In a practical sense too, as Paul Thompson notes, farm animals were an integral part of the ecology and economy of the farm, with the different species serving important and complementary roles.18 No small farm would be complete without a horse to provide traction, a cow for milk, and a few chickens and pigs to turn surplus grain and food waste into eggs and bacon. Like pastoralism, agrarianism has remained a powerful moral idea long after it ceased to be the dominant model for agricultural production, and it has inﬂuenced people’s views on animals in at least two ways. First, much as the pastoralist world-view put the diligent animal keeper on a moral pedestal, the agrarian world-view did the same for the rural family living and producing food from their own land. Hence, the raising and killing of animals has continued to be seen as an acceptable activity, even a virtuous activity, as long as it is carried out within the context of family farming. This perception has had an important inﬂuence on the debate about modern animal production. Critics of animal production realize that they must counteract the positive image of the family farm if they are to be effective with the public, and they go to lengths to portray modern animal production as ‘factory farming’ as distinct from family farming, and to claim that corporations have now replaced families in the control of animal production. In response, defenders of animal production do the opposite by emphasizing the level of family ownership that remains in many aspects of animal production.19 17Thompson,

1998, pages 161–165. 1998, pages 165–169. 19I have summarized some of these contrasting depictions of animal production in Fraser, D. 2001. The ‘new perception’ of animal agriculture: legless cows, featherless chickens, and a need for genuine analysis. Journal of Animal Science 79: 634–641. 18Thompson,

Dfraser_C003.indd 47

6/27/2008 12:42:02 PM

48

Animal Welfare in Context

But secondly, and more importantly for purposes of understanding animal welfare, the agrarian world-view sees animals on the traditional farm (like the family members themselves) as living a good life. It is not a life of ease or pleasure, or a life devoid of pain and struggle, but a life that is wholesome because it is lived in harmony with the cycles of nature and rural life.20 ROUGHLY THREE HUNDRED YEARS ago, Western culture saw the dramatic growth of another world-view which provided an alternative interpretation of how animals ought to be viewed and treated. The new ideas were seen most clearly in the visual arts. In European painting before about 1600, animals sometimes appeared in portraits and religious scenes, but they served mainly as decorations, symbols, and allegorical elements to enhance a visual effect or reinforce the human or religious subject-matter of the work. If an elegant lady was painted with one hand stroking a lap dog, the dog was intended as a symbol of her ﬁdelity, not of her tenderness toward animals; indeed, the other hand might well be toying with an ermine stole which the artist included as a symbol of purity. The portrait of a duke might have a hunting dog looking up at him with an expression of fawning subservience, but the dog was principally an embellishment whose obsequious attention suggested the kind of tactful approach that human suitors might be advised to show to so eminent a person. In the 1600s, however, animals themselves began to emerge as the subjectmatter of art. Some of the earliest examples occurred in paintings by Dutch, Flemish and German artists. Paulus Potter (1625–1654) painted scenes that remain famous today, showing clusters of cattle, perhaps with a human attendant as a minor element to one side of the picture. Aelbert Cuyp (1620–1691) produced a vast output of countryside scenes with horses, cattle, sheep and dogs painted in minute and appreciative detail. Frans Snyders (1579–1657) became famous for rather whimsical canvasses ﬁlled with many species of birds perched in a tree like an avian chorus, sometimes even with an owl conducting the music (Figure 3.1). In British painting, the change occurred gradually and almost a century later. In the 1700s, the traditional hack-work paintings of horses, commissioned by the wealthy to commemorate their most successful racing animals, was subtly transformed by George Stubbs (1724–1806) into a sensitive celebration of the animals themselves, independent of any illustrious owner and often with little or no background to detract from the attributes – indeed the personality – of the horses. Animals included in portraits of human subjects were no longer mere symbols or embellishments but companions and friends. By the 1800s, artistic attention to animals had become so strong that (to take just one example) the illustrious Edwin Henry Landseer became the most famous British painter of his generation largely on the basis of his depictions of animals: noble stags living wild in 20Thompson,

Dfraser_C003.indd 48

1998, pages 167–169.

6/27/2008 12:42:02 PM

A Good Life for Animals

49

Figure 3.1 Bird’s Concert, by Frans Snyders, c. 1630–1640. Reproduced with permission, Museo Nacional del Prado, Madrid.

the highlands of Scotland, a loyal dog grieving by the cofﬁn of a dead shepherd, and Landseer’s immensely popular paintings of animals with subtly humanized facial expressions. This change in the treatment of animals in art was an early harbinger of the Romantic movement, and it heralded a transformation in how artists saw and depicted the world. The Romantics valued ordinary life: to them, peasants living in a humble cottage were as worthy a subject-matter as wealthy patrons dressed in their ﬁnery, and a country girl leading a cow to the riverside was as deserving of the artist’s attention as St. Mary visited by an angel. With this new mind-set, animals – as a quintessential example of life uncorrupted by the conventions of society – were a tremendously popular theme. The Romantic painters also attached great value to ‘nature’. Country roads and hillsides were as interesting to the Romantic artists as Mount Olympus and the heavens had been to their predecessors. Ruins, where the forces of nature could be seen reclaiming former centres of civilization, were especially favoured. And throughout the wild or rural settings depicted by Romantic artists, animals were often an important focus of attention as the natural inhabitants of the countryside. The Romantic movement also involved a focus on the emotions. George Stubbs painted canvas after canvas of encounters between horses and lions – an improbable scene but one that allowed Stubbs to depict his beloved horses showing emotions

Dfraser_C003.indd 49

6/27/2008 12:42:02 PM

50

Animal Welfare in Context

Figure 3.2 Miss Jane Bowles, by Sir Joshua Reynolds, c. 1775. Reproduced with kind permission of the Trustees of The Wallace Collection, London.

of surprise, terror and agony. Likewise, Théodore Géricault (1791–1824) ensured that his viewers recognized the emotions of animals; he depicted a horse frightened by lightening, others railing at being captured by people, others sharing the emotions of their human companions. With this emphasis on feeling, human empathy for animals and friendship with animals were portrayed as positive and enriching. In a portrait of a small girl by Joshua Reynolds (1723–1792), the charm of the painting comes from the girl’s obvious affection for her dog (Figure 3.2). The artist Hogarth even gave his dog pride of place in his own self-portrait. A similar change occurred in literature. As Dix Harwood points out, apart from Shakespeare and other occasional exceptions, English literature contained few expressions of concern over animals before about 1700. During the next century, however, animals increased in prominence in English literature much as they did in the visual arts, and the key elements of the Romantic movement were present there as well. Nature became a common theme. The changing seasons, a quiet churchyard, the colour of the dawn sky became the focus of literary attention. Daffodils, daisies and other simple ﬂowers were eulogized as having the power to restore happiness to a troubled mind. And with these, the animate inhabitants of

Dfraser_C003.indd 50

6/27/2008 12:42:03 PM

A Good Life for Animals

51

nature – nightingales, skylarks, stags, even crickets – became sources of solace and objects of admiration and empathy. Like painters, writers also portrayed animals as possessing human-like emotions. In his celebration of nature The Seasons poet James Thomson (1700–1748) empathized with hares forced by winter to overcome their fear and seek food in human gardens, and he pitied the sheep who ‘with looks of dumb despair’ resorted to digging for withered forage through freshly fallen snow.21 Perhaps the ultimate depiction of the subjective lives of animals came from literature written from the animal’s own viewpoint or even in the animal’s own words. This genre became immensely popular in the late 1800s with Black Beauty, Beautiful Joe and many other animal narratives.22 However, Robert Burns used this convention a century earlier by composing a conversation between two dogs and the dying words of a sheep to its owner.23 With this valuing of animals and nature, some writers even proclaimed a fundamental unity between humans and other species. When he saw a mouse ﬂeeing after his plow had destroyed its nest, Burns lamented that ‘nature’s social union’, which should bind human and beast together, had been broken by the ‘dominion’ of humankind.24 And William Blake (1757–1827), on accidentally harming a ﬂy, proposed an equality between people and other creatures that exceeds even modern-day animal rights philosophy, asking: Am not I A ﬂy like thee? Or art not thou A man like me?25

The close and sympathetic attention to animals seen in the Romantic movement no doubt had the effect of making people see animals as more worthy of moral 21Thomson,

J. 1728. Winter, in The Seasons: A Poem. Republished 1951 as pages 184–244 in The Complete Poetical Works of James Thomson (J.L. Robertson, editor). Oxford University Press, London. The passage is lines 257–265. 22Sewell, A. 1877. Black Beauty: The Autobiography of a Horse. Republished 1945, Grosset & Dunlap, New York; Saunders, M. 1893. Beautiful Joe: An Autobiography. C.H. Banes, Philadelphia. Black Beauty is arguably the most popular animal story of all time. Beautiful Joe, which sold seven million copies, was written to do for dogs what Black Beauty had done for horses. The author, Margaret Marshall Saunders (writing initially as Marshall Saunders to conceal her female identity) went on to write many novels about animals. Beautiful Joe (the dog) is still commemorated in his home town of Meaford, Canada, by the Beautiful Joe Heritage Society at www.beautifuljoe.org. 23Burns, R. 1786. The Twa Dogs, and The Death and Dying Words of Poor Mailie. Republished as pages 44–50 and 72–74 in the Barke edition. 24Burns, R. 1786. To a Mouse. Republished as pages 111–112 in the Barke edition. 25Blake, W. 1794. The Fly. In Songs of Innocence and of Experience. Republished 1989 as pages 220–221 in Blake: The Complete Poems, Second edition (W.H. Stevenson, editor). Longman, London. The quotation is from lines 5–8.

Dfraser_C003.indd 51

6/27/2008 12:42:07 PM

52

Animal Welfare in Context

concern, but sometimes the moral messages were explicit in the art itself. William Cowper (1731–1800) gave a clear warning to his acquaintances: I would not enter on my list of friends (Though graced with polished manners and ﬁne sense, Yet wanting sensibility) the man Who needlessly sets foot upon a worm.26

William Blake raged against the mistreatment of animals: A robin redbreast in a cage Puts all Heaven in a rage. A dove house ﬁll’d with doves and pigeons Shudders Hell thro’ all its regions. A dog starv’d at his master’s gate Predicts the ruin of the state. A horse misus’d upon the road Calls to Heaven for human blood.27

James Thomson even urged people to observe a vegetarian diet. Wolves and lions, he argued, can be forgiven for eating ﬂesh because they act out of necessity and do not feel pity. ‘But man, whom Nature formed of milder clay/With every kind emotion in his heart’ should not stoop to the level of a prowling predator ‘and dip his tongue in gore’.28 The Romantic movement heralded a view of animals quite different from what had gone before. Seen through the lens of the Romantic world-view, animals are not living possessions to which we owe proper pastoralist care, nor integral elements of a wholesome agrarian life. Rather, animals are fellow beings possessing individuality, emotional and experiential lives, the capacity for friendship with each other and with humans, and at the same time possessing a distinct non-human nature that causes us to respect and admire them as the different kinds of beings that they are. Before shifting our focus, let us try to be clear about the troublesome term ‘Romantic’. ‘Romantic’ is sometimes used in a pejorative sense, to dismiss ideas as unduly imaginative and unconnected to the real world. When I write of a 26Cowper, W. 1785. Winter Walk at Noon. In The Task, Book VI. Republished 1994 as The Task, and Selected Other Poems (J. Sambrook, editor). Longman, New York. The quotation is on lines 560–564. 27Blake, W. about 1805. Auguries of Innocence. Republished 1989 as pages 589–592 in Blake: The Complete Poems, Second edition (W.H. Stevenson, editor). Longman, London. The quotation is from lines 5–12. 28Thomson, J. 1728. Spring, in The Seasons: A Poem. Republished 1951 as pages 3–46 in the Robertson edition. The quotation is on lines 349–357.

Dfraser_C003.indd 52

6/27/2008 12:42:07 PM

A Good Life for Animals

53

Romantic world-view (with a capital R) I intend a more technical and certainly not pejorative meaning which identiﬁes the set of values and attitudes discussed above. However, ‘Romantic’ is also used by scholars to refer to a particular period or even particular artists. In that sense, the term is time-bounded and limited to the past. If instead we think of the Romantic world-view as a mode of thought which values the natural and simple, which focuses on experiential and emotional life, and which sees animals and nature as enriching and inherently valuable, then I believe we can see a Romantic element present in ﬁgures as separate in time as St. Francis of Assisi and Albert Schweitzer.29 In this sense, a Romantic world-view is not limited to the works or times of certain writers and artists. Rather, it represents a major and continuing mode of human thought and (I will argue) it plays a key role in shaping attitudes to animals today. WHEN THE POET WILLIAM WORDSWORTH (1770–1850), from his cottage in the English Lake District, complained: Getting and spending, we lay waste our powers. Little we see in nature that is ours30

he was objecting to a world that had become too commercial, too artiﬁcial, too (as we would put it today) industrial. He lived at a time when common people were being displaced from the countryside to live in polluted towns and cities, forced to give up peasant agriculture and cottage crafts for unhealthy labour in mines and factories. Even the wealthy, although spared the squalid living conditions of urban labourers, were surrounded by the artiﬁciality of commerce and urban society, far removed from what Wordsworth saw as the uplifting power of nature. But the world of commerce and industry, despised as it was by Romantics such as Wordsworth, clearly involved a world-view and value system of its own, and if we can judge by the staggering growth of industrial society, these must have been highly inﬂuential in modern Western thought. Because people who espoused industrialization tended to create factories rather than poetry and paintings, their world-view has attracted less scholarly attention. But because (I will argue) the world-view of Industrialism ultimately provided an alternative understanding of what constitutes a satisfactory life for animals as well as for humans, let us try to piece together its key features. The Industrial world-view did not look back to a Golden Age of innocence and harmony with nature as the Romantics did, but instead looked forward to a Golden Age when science, technology and commerce would bring about a better 29For

example: Schweitzer, A. Reverence for Life. Republished 1971 as pages 25–32 in Albert Schweitzer: Reverence for Life (P. Seymour, editor). Hallmark Editions, Kansas City. 30Wordsworth, W. 1807. ‘The world is too much with us …’ Republished 2002 in Selected Poetry of William Wordsworth (M. Van Doren, editor). Modern Library, New York.

Dfraser_C003.indd 53

6/27/2008 12:42:07 PM

54

Animal Welfare in Context

life. Nature was not an ideal state to be returned to, but an imperfect state to be controlled and improved for the betterment of all. To adherents of the Industrial world-view, any return from a life of technology to one of nature would simply be regressive. A key element in the Industrial world-view was a belief in ‘progress’. This, as historian Sydney Pollard described it in The Idea of Progress, involved ‘the assumption that a pattern of change exists in the history of mankind, … that it consists of irreversible changes in one general direction only, and that this direction is toward improvement’.31 Thus, however much the Romantics might condemn the Industrial Revolution, there could be, and should be, no going back to pre-industrial life. Initially, science played a key role in the belief in progress. Science was the ﬁrst area where change seemed inevitably to involve improvement. As Pollard noted, science ‘could move only one way, forward, for each generation started with the best of the last generation, and thus was bound to add to it’.32 By about 1700, Pollard argues, the belief that knowledge is constantly improving through science came to replace the earlier belief that the greatest knowledge had been achieved by the classic civilizations of the past. During the 1700s, the idea of progress spread far beyond the narrow world of science. By the end of the century, in the words of Pollard, ﬁrm convictions had been expressed about the inevitability of progress in wealth, in civilization, in social organization, in art and literature, even in human nature and biological make-up.33

And although this optimism was to be tested by periods of revolution, war and economic depression, a belief in progress has remained one of the inﬂuential ideas of modern times. However, for many adherents of the Industrial world-view, the idea of progress was more than a way to describe the course of history; it was also seen as a law of the social world much as gravity is a law of the physical world. Thus, to protest against ‘progress’ (as the Luddites did when they smashed the machinery that deprived craftsmen of their livelihood) was often seen as ludicrous – perhaps on a par with railing against the changing of the seasons or the rising of the sun. Within the Industrial world-view, to ‘stand in the way of progress’ is not merely unwise but ridiculous. A second key element of the Industrial world-view was the value attached to productivity. The philosopher/economist Adam Smith (1723–1790) introduced The Wealth of Nations by noting that the labour conducted by the citizens of a 31Pollard, S. 1968. The Idea of Progress: History and Society. Penguin Books, Harmondsworth,

UK. The quotation is on page 9. 32Pollard, 1968, page 20. 33Pollard, 1968, page 31.

Dfraser_C003.indd 54

6/27/2008 12:42:07 PM

A Good Life for Animals

55

nation is the source of all the goods that supply the nation with the ‘necessaries and conveniences of life’.34 Whether the workforce produces the goods directly, or whether it produces goods for exchange with other countries, it is the productivity of the workforce that dictates how well a nation’s citizens will be supplied with goods. Thus, if the workers of a nation can increase the amount of goods they produce, then the citizens of the nation will be that much better off. And Smith saw the factory system – including the use of automated machinery and the specialization of the workforce so that each person could develop a high level of skill in a speciﬁc task – as a primary means of increasing a nation’s productivity and hence the quality of life of its citizens. Of course, the industrialization that Smith seemed to welcome was not without certain costs. The greater productivity of the factories made it impossible for many traditional crafts-persons to earn a living. The result was extensive poverty and massive disruption of rural communities. This led to a demographic shift whereby rural people moved to the cities and provided the large workforce required for the new industrialized methods of manufacturing. These developments touched off a vigorous debate between the critics and the defenders of the industrial system. Critics claimed that industrial manufacturing led to miserable living conditions for the workers it exploited. Moreover, working at machines was seen as unnatural and unhealthy. In preparation for the 1833 Factory Act which limited the hours of work that could be imposed on children, a parliamentary committee received testimony that children’s bodies were being deformed by the unnatural strain imposed on immature bones and joints. More acute problems were described to Parliament by one social reformer: [D]rowsy and exhausted, the poor creatures fall too often among the machinery, which is not in many instances sufﬁciently sheathed, when their muscles are lacerated, their bones broken, or their limbs torn off.35

Perhaps even worse, critics claimed, by doing repetitive work with machines, workers lost their human nature and became machine-like themselves. Thus, critics called for legal reforms to mitigate the harm caused by industry. But the defenders of industry had arguments of their own. Far from being an unnatural development, the automated systems represented a form of progress as manufacturing evolved from laborious hand work to the complete automation that would eventually make labour unnecessary. Instead of imposing unnatural stress on workers, the machinery liberated workers from ‘the tedium of handicraft

34Smith,

A. 1776. The Wealth of Nations, Book 1, Chapter 5. Republished 1904, Dent and Sons, London. The quotation is from passage 9. 35Quoted by Bizup, J. 2003. Manufacturing Culture: Vindications of Early Victorian Industry. University of Virginia Press, Charlottesville. The quotation is on page 19.

Dfraser_C003.indd 55

6/27/2008 12:42:08 PM

56

Animal Welfare in Context

occupations’.36 In fact, the efﬁcient production that could be achieved through automation was claimed to be proof enough that the factories were well suited to human workers. Moreover, industry supporters claimed, it is in the factory owner’s own interests to ensure that workers are well treated because maximum productivity could not otherwise be achieved. Using such arguments, pro-industry voices insisted that regulation was unnecessary. Despite these arguments, many countries did opt for a gradual program of legislated reform. Indeed, the same decades that saw the ﬁrst British animal protection laws also saw a much more extensive program of reforms to protect workers in cities and factories. The values of Industrialism, although fully consistent with the use and often the exploitation of workers, were sometimes linked to a genuine concern over their welfare. A striking example was Sir Titus Salt (1803–1876) who, instead of retiring as a wealthy textile magnate in the 1850s, decided to create a model industrial community. He built a gigantic mill in a pleasant countryside area, equipped it with state-of-the-art safe-guards for worker safety together with pollution-reducing technology to keep the local air healthy, and surrounded it with housing that provided his workers with a substantially healthier and more pleasant environment than was available in the nearby cities. He even equipped the community with a school, a hospital, a bathhouse, a church and an educational centre, and he initiated a programme of subsidized health insurance for his employees. Salt was one of a number of enlightened industrialists, along with Robert Owen (1771–1858), who created model industrial communities in Scotland and the United States, and the Cadbury family who created extensive social housing and other programmes at their cocoa factory near Birmingham as well as supporting the cause of animal protection.37 These admirable attempts to create a good quality of human life were no doubt motivated partly by benevolence, but there was also a practical belief that a healthy, happy worker is a productive worker. Robert Owen developed that argument by urging his fellow industrialists to pay attention not only to the ‘inanimate machines’ in their factories but also to their ‘vital machines’ – the workers themselves – and he suggested that investment in these vital machines would return a handsome proﬁt (Box 3.3). A belief that life can be made better by the rational application of science and technology, a belief that productivity and hence prosperity can be increased by improving on the processes of unaided nature, and a belief that such ‘progress’ is 36Bizup,

2003, pages 19–20.

37See Reynolds, J. 1983. The Great Paternalist: Titus Salt and the Growth of Nineteenth Century

Bradford. Maurice Temple Smith, London; Post, A. 1974. Popular Freethought in America, 1825–1850. Octagon Books, New York; and an anonymous article called ‘The Development of Bourneville: The Factory and Village in a Garden’. Available at: http://www.cadbury.com.au/ history/04_bourneville.pdf, accessed Nov. 10, 2005.

Dfraser_C003.indd 56

6/27/2008 12:42:08 PM

A Good Life for Animals

57

Box 3.3 A passage by industrialist and social reformer Robert Owen (1771–1858) exhorting his fellow manufacturers to improve the lives and working conditions of workers (the ‘vital machines’ of factories) in order to promote both greater proﬁt and the improvement of society. If, then, due care as to the state of your inanimate machines can produce such beneficial results, what may not be expected if you devote equal attention to your vital machines, which are far more wonderfully constructed? … From experience which cannot deceive me, I venture to assure you, that your time and money so applied, if directed by a true knowledge of the subject, would return you, not five, ten, or fifteen per cent for your capital so expended, but often fifty, and in many cases a hundred per cent. Indeed, after experience of the beneficial effects from due care and attention to the mechanical implements, it became easy to a reflecting mind to conclude at once, that at least equal advantages would arise from the application of similar care and attention to the living instruments. And when it was perceived that inanimate mechanism was greatly improved by being made firm and substantial; that it was the essence of economy to keep it neat, clean, regularly supplied with the best substance to prevent unnecessary friction, and by proper provision for the purpose to preserve it in good repair, it was natural to conclude that the more delicate, complex, living mechanism would be equally improved by being trained to strength and activity and that it would also prove true economy to keep it neat and clean; to treat it with kindness, that its mental movements might not experience too much irritating friction; to endeavour by every means to make it more perfect; to supply it regularly with a sufficient quantity of wholesome food and other necessaries of life, that the body might be preserved in good working condition, and prevented from being out of repair, or falling prematurely to decay. Here, then, is an object which truly deserves your attention; and, instead of devoting all your faculties to invent improved inanimate mechanism, let your thoughts be, at least in part, directed to discover how to combine the more excellent materials of body and mind which, by a well-devised experiment, will be found capable of progressive improvement. Robert Owen, 1813–1816. From ‘Address prefixed to Third Essay’ in A New View on Society. Republished 1991 as pages 4–7 in Robert Owen: A New View of Society and Other Writings (G. Claeys, editor). Penguin Classics, London.

both good and inevitable: these I suggest (tentatively, because we need historians to correct and complete my very preliminary analysis) were hallmarks of the world-view that developed among those who embraced industrialization. But how did this come to inﬂuence perceptions of what constitutes a satisfactory life for animals?

Dfraser_C003.indd 57

6/27/2008 12:42:08 PM

58

Animal Welfare in Context

Despite the profound effects of industrialization on so much of society over the past three centuries, quasi-industrial methods of animal production did not develop until much more recently. Industrial-style production had certain precursors in the 1700s such as labour-saving, automated systems for feeding animals,38 and the 1800s occasionally saw barns designed with a precursor of modern gestation stalls for sows.39 Hence, the ideas were present much earlier, but industrial-style production of animals did not become a widespread practice until after the Second World War. Over the decades that followed, however, the industrialized countries saw most production of pigs and poultry, and some raising of cattle, shift from a mainly agrarian model to indoor, conﬁnement operations, with a large degree of automation for such tasks as feeding and manure removal. As this occurred, it sparked a debate that was remarkably similar to the one that arose over industrialized manufacturing in the 1700s and 1800s. On one side were critics of the new systems whose thinking bore the stamp of the Romantics. These critics emphasized the unnaturalness of the new systems. First, they claimed, animals are conﬁned for the whole of their lives in artiﬁcial indoor environments where they are deprived of sunshine, fresh air and all that is natural. Second, critics claimed, these systems ignore the very nature of the animals themselves by preventing them from carrying out their normal behaviour and by treating them as (almost turning them into) machines. And like the earlier Romantic critics of the Industrial Revolution, the agricultural critics put special emphasis on the emotional impact of the new systems. Critics claimed, for example, that animals in such systems ‘suffer from birth to death’, are ‘literally driven mad’ and ‘experience the same mental anguish that would drive many humans to suicide’.40 Similarly, the arguments used to defend the new systems of animal production were remarkably like those that had been used much earlier by the defenders of industrial manufacturing, and they provide us with a glimpse of how animals are seen through the lens of an industrial world-view. Defenders of industrialized production methods emphasized the good functioning and high productivity of the new systems. Far from creating unhealthy conditions, they claimed, the new systems achieve a high level of health, growth and survival by such means as controlling temperatures, preventing disease and excluding predators. As for animals suffering, surely animals would not grow and produce as rapidly as they do if they were indeed suffering. And instead of the new systems being unnatural (whatever ‘unnatural’ might mean for domestic animals hundreds of generations removed from the wild) surely the use of science and technology to conquer the 38Malcolmson, R.W. and Mastoris, S. 1998. The English Pig: A History. Hambledon Press, London. 39Baxter, S. 1984. Intensive Pig Production: Environmental Management and Design. Granada, London. 40Summarized by Fraser, 2001. Original sources, cited in the bibliography, are Sequoia, 1990; Robbins, 1987; and Penman, 1996.

Dfraser_C003.indd 58

6/27/2008 12:42:08 PM

A Good Life for Animals

59

vicissitudes of nature is a form of progress that leads inevitably to a better quality of life for animals than they could ever expect in a state of nature. Historians have not fully explored why industrialized forms of animal production began when they did, but we can point to some likely causes.41 One was social policy: faced with the food shortages that many countries experienced during and after the war years, governments looked to the expansion and increased efﬁciency of animal production as a way of ensuring a more reliable supply of food. Another likely cause was a scarcity of labour as agriculture found itself competing with other expanding sectors of the economy and could not retain the workforce needed for the older, agrarian production methods. At the same time, the use of refrigeration and the expansion of road-based transportation allowed animals and animal products to be sold into ever larger markets, thus bringing producers into competition with thousands of other producers. Farm proﬁts fell, and the cost-efﬁciency that could be achieved by the industrialized systems became almost essential for economic survival. Hence, many animal producers, although perhaps espousing the moral ideas of pastoralism and agrarianism, were more or less required to adopt economies of scale and increasingly industrial methods of production. But I would argue that when the change to more industrialized methods occurred, at least some of the men and women who adopted the new practices were sufﬁciently inﬂuenced by the world-view of Industrialism that they welcomed the changes as a form of progress. For such people it must have seemed modern and progressive to use technology and automation to replace manual labour, to specialize in one type of production so as to conduct it with greater skill, and to use the advantages of nutritional and physiological knowledge together with vaccines, antibiotics and other developments in science to increase productivity. To those who adopted the new methods willingly – and to the scientists and veterinarians who shared their enthusiasm – critics who called for a return to a more agrarian type of farming and more ‘natural’ living conditions for animals simply appeared to be standing in the way of progress. PASTORALISM, AGRARIANISM, ROMANTICISM, AND INDUSTRIALISM. In describing these world-views and their implications for animals, I do not mean to imply that the four are mutually exclusive. For example, modern Agrarian thinking reﬂects some of the values of Romanticism as applied to rural life. And although we can classify modes of thought, we cannot necessarily classify people. A dairy farmer, for example, may hold a pastoralist ethic of care as a personal credo, while also valuing rural family life, looking upon animals as sentient beings with emotional lives, and viewing vaccines as a form of progress. My claim is not that people fall into different pigeon-holes, but that different people are inﬂuenced to different degrees by the four modes of thought and thus have different ideas about what constitutes a good life for animals. 41Fraser, D. 2005. Animal Welfare and the Intensiﬁcation of Animal Production: An Alternative Interpretation. FAO Readings in Ethics No. 2, Food and Agriculture Organization of the United Nations, Rome.

Dfraser_C003.indd 59

6/27/2008 12:42:08 PM

60

Animal Welfare in Context

Inﬂuenced by the Pastoralist world-view captured in the Bible, people regarded animals as wards entrusted to humans for care, but at the same time as legitimate possessions which people are entitled to use in appropriate ways. Within this framework, for animals to have a satisfactory life, the key requirement is for their human owners to be sufﬁciently conscientious in providing the care that animals need. The central ethical issue from this viewpoint is not animal welfare but animal care. Under the inﬂuence of Agrarianism, animals are not so much wards as fellow actors in the age-old drama of rural life. A good life for animals, as for the human participants in agrarian systems, is not a life of ease or pleasure, or a life free from all hardship, but a life that is wholesome because it is lived in harmony with nature and the cycles of rural living. Through the lens of Romanticism people viewed animals as fellow beings, capable (like humans) of suffering, and all too often degraded by the constraints and artiﬁciality of modern human society. For animals, as for people, a good life is a free life lived close to nature. And with the Romantic emphasis on emotion ahead of rationality, a good life would inevitably be marked by pleasure and the avoidance of suffering inﬂicted by technology and other human creations. Seen through the lens of Industrialism, animals are cast in a role roughly analogous to that of workers in efﬁcient production systems. To pay attention to the welfare of animals in such systems is the right thing to do for practical reasons as well as ethical ones. Indeed, a healthy animal whose needs are well met will be a productive animal. And the way to make animals more healthy and productive is to be found not in returning to the vicissitudes and inefﬁciencies of nature but by the rational application of science and technology. When we speak of cultural differences in attitudes toward animals, we do not necessarily mean the far-ﬂung differences between the Ojibway and the Jains or between ancient Greece and the English Enlightenment. We also see important cultural differences within contemporary Western thought. Thus, when we apply science to issues of animal welfare we can expect to ﬁnd different people and different groups – animal protectionists, animal producers, veterinarians, organic farmers – approaching the issues with different beliefs about the nature of animals and about what constitutes a good life for them. Moreover, as we will see, the competing world-views also provided different ideas about the kind of data scientists should collect in the study of animal welfare.

Dfraser_C003.indd 60

6/27/2008 12:42:08 PM

A Science of Animal Welfare?

4

The modern application of science to moral concerns about animals came about through a remarkable combination of events which began largely in the 1960s and gave animal welfare science some of its particular ﬂavour. In 1964, the same year when Sam Burich ﬁred a harpoon into the hapless Moby Doll, a determined British vegetarian dropped a bomb on industrialized animal production. The bomb was a book called Animal Machines in which a hitherto unknown author named Ruth Harrison (Figure 4.1) shocked her readers with descriptions of the living conditions of chickens, pigs and veal calves in the conﬁnement production systems that were then becoming common.1 The book was soon followed by many others including Hans Ruesch’s Slaughter of the Innocent about the use of animals in research, Jack Olsen’s Slaughter the Animals, Poison the Earth about the extermination of wild animals in America, and Peter Singer’s Animal Liberation which alleged appalling treatment of farm and laboratory animals alike.2 The books were both symptoms and catalysts of a growing concern over human treatment of animals. Attitudes towards animals, as we have seen, had been evolving for perhaps three centuries in many countries of the West. In the late 1800s, and in the years leading to the First World War, the United Kingdom passed legislation to regulate the use of animals in research, to limit the use of ponies in coal mines, to prevent the police from delivering stray dogs for experimentation, to restrict the export of live horses for slaughter, and many other topics.3 During the twentieth century, however, for a period corresponding roughly to the two World Wars and the Great Depression, Western society seemed preoccupied with human survival and the basics of life, and animal issues were largely consigned to a back burner. 1Harrison,

R. 1964. Animal Machines. Vincent Stuart Ltd., London. H. 1978. Slaughter of the Innocent. Bantam Books, New York; Olsen, J. 1971. Slaughter the Animals, Poison the Earth. Simon & Schuster, New York; Singer, P. 1990. Animal Liberation, 2nd edition (1st edition 1975). Avon Books, New York. 3Turner, 1964. 2Ruesch,

61

Dfraser_C004.indd 61

6/27/2008 12:42:20 PM

62

Animal Welfare in Context

(a)

(b)

Figure 4.1 (a) Animal advocate Ruth Harrison whose book Animal Machines stimulated concern over the welfare of farm animals in the 1960s. (b) Scientist William Thorpe whose 1965 essay ‘The assessment of pain and distress in animals’ proposed how science could be used to assess and improve the welfare of animals. The photograph of Mrs. Harrison was kindly provided by her children; that of Dr. Thorpe was provided through the courtesy of the Royal Society, London, © The Godfrey Argent Studio.

But in the 1950s and 1960s, as security and afﬂuence returned in the industrialized nations, society began to pay increasing attention to issues that went beyond survival and personal prosperity. Rachel Carson (who wrote a foreword to Animal Machines) was already a household name for her book Silent Spring about the damage to wildlife and the environment caused by the use of pesticides in crop production.4 Civic or ‘non-governmental’ organizations arose as a force that focused social and political attention on civil liberties, community health, gender equality and other issues. In this climate, public attention to animal issues returned as well, and it eventually reached a level that arguably equaled or exceeded what had been seen in the United Kingdom a hundred years earlier. Perhaps because Animal Machines was the ﬁrst book in many years on the treatment of farm animals, perhaps because it was serialized in a major British newspaper, and perhaps because of its simple, from-the-heart sincerity, the book had an enormous impact and led to intense debate about the appropriateness 4Carson,

Dfraser_C004.indd 62

R. 1962. Silent Spring. Houghton Mifﬂin, Boston.

6/27/2008 12:42:21 PM

A Science of Animal Welfare?

63

of conﬁnement production systems for animals. It also brought together several themes which came to be repeated over and over in the extensive literature and media coverage of farm animal production that developed during the rest of the century. The ﬁrst and foremost of these themes was animal suffering. Because the animal protection movement had traditionally focused on eliminating cruelty to animals, deliberate acts of cruelty had long been illegal. The suffering that Mrs. Harrison alleged was from a different source. It arose from the production systems themselves – systems that had been designed to achieve the typical industrial goals of productivity, efﬁciency and proﬁt. To a human population that had tended to equate animal suffering with deliberate cruelty, it came as a shock to be told that animal suffering on a scale never before imagined was occurring throughout the countryside simply by the pursuit of normal business objectives. A second theme was the unnaturalness of the new production systems. Hens, instead of scratching for seeds in the barnyard, were being kept in tiers of cages where the birds never saw the sunshine, could barely move, and where the cage ﬂoors were so unsuitable as to cut into the hens’ feet. Even more unnatural were the methods described by Mrs. Harrison for raising veal calves. As one commentator summarized Mrs. Harrison’s account: Unmothered, fed artiﬁcially, warmed artiﬁcially, ‘protected’ from sunlight … stuffed with antibiotics and sedatives, the units in these factories have no contact with the Nature of which they are a part … .5

A third theme was the change in farming from an agrarian to an industrial activity. By and large, the British public had an image of farming being done by gentle country folk with a few cows, pigs and chickens that were treated almost as members of the family. People were shocked by the idea that quasi-industrial enterprises were now making this bucolic picture a thing of the past. The new term ‘factory farming’ evoked cultural memories from the previous century of down-trodden workers exploited at the hands of ruthless industrial masters. Mrs. Harrison’s title seemed to go even further by suggesting that animals were being treated not merely as oppressed workers but as the actual machinery of the industrial system. Finally, Mrs. Harrison raised doubts about the safety of eating food produced by conﬁnement methods. Surely, she implied, if animals are raised under such unnatural conditions and lead such miserable lives, the meat and other products they produce cannot be wholesome food for human consumption. Today, in a world where the public has grown sceptical of alarming messages, it is hard to imagine the impact that Mrs. Harrison’s book had. The British government, responding to a barrage of public concern, felt forced to act, and it did so 5Turner,

Dfraser_C004.indd 63

1964, page 314.

6/27/2008 12:42:24 PM

64

Animal Welfare in Context

in true governmental fashion by appointing a committee – speciﬁcally a committee charged with investigating ‘the welfare of animals kept under intensive livestock husbandry systems’. In view of what the committee described as ‘the extent of public disquiet’ on the subject, they worked with a sense of urgency.6 The committee included a number of distinguished agriculturalists but also several scientists who, although appreciating the ethical nature of the issues, saw an important role for scientiﬁc research to play in their resolution. The chairman, Professor F. W. Rogers Brambell, was a distinguished immunologist. Another member, Professor Tom Ewer, was an inﬂuential veterinary scientist and educator who was so impressed by the importance of scientiﬁc research that he pressed for veterinary students to have experience of research as a routine part of their training. But perhaps most inﬂuential was Dr. William Thorpe (Figure 4.1), a gentlemanly Quaker paciﬁst known for his scientiﬁc studies of animal behaviour, especially bird song. Dr. Thorpe, in an appendix to the committee’s report, wrote a thoughtful essay proposing how one might use the tools of science to recognize pain and distress in animals. Thorpe focused not just on the obvious concern over suffering caused by disease and injury, but also on less tangible forms of suffering. He noted, for instance, that migratory birds in captivity may ﬂutter all night in their cages during the normal season of migration because the environment does not allow the birds to perform this important element of their natural behaviour, and he suggested that this would give rise to ‘prolonged and intense emotional disturbances’.7 He proposed various means of researching the welfare of animals including physiological indicators of stress, behavioural indicators of pain and discomfort, studies of motivation that is thwarted in conﬁnement, studies of the intelligence and cognitive powers of animals, studies of animals’ capacity to develop a learned fear of humans, and studies of the preferences that animals show for different environments. It was a vision that involved several ﬁelds of science, yet Thorpe’s home ﬁeld of animal behaviour was especially prominent. Dr. Thorpe’s inﬂuence was obvious in the committee’s conclusions. The committee made many speciﬁc recommendations about animal husbandry methods, and it outlined what it considered to be the basic, minimal freedoms that animals should enjoy. But most importantly it called for research be done, in veterinary medicine, stress physiology, animal science, and particularly in the ﬁeld of animal behaviour. Such research, the committee imagined, would lead to greater effectiveness of the newer types of animal production systems, much as Robert Owen had suggested that attention to the needs of workers would improve the efﬁciency of factories. 6Brambell, F.W.R. (chairman). 1965. Report of the Technical Committee to enquire into the Welfare of Animals kept under Intensive Livestock Husbandry Systems. Her Majesty’s Stationery Ofﬁce (HMSO), London. The quotations are on the title page and page 2. 7Thorpe, W.H. 1965. The assessment of pain and distress in animals. Appendix III in Brambell, 1965. The quotation is on page 73.

Dfraser_C004.indd 64

6/27/2008 12:42:24 PM

A Science of Animal Welfare?

65

More importantly, however, such research would lead to an understanding of how the welfare of the animals themselves is inﬂuenced by different production methods. These recommendations, combined with the government’s sense of urgency in meeting public concerns, led to a sudden release of funding and other support for a type of science that had never really existed before Animal Machines: the scientiﬁc study of animal welfare. TODAY, AFTER NEARLY HALF a century of discussion and research on animal welfare, it seems only natural that the ‘welfare’ of animals should have been the central issue. In fact, however, even that development was a product of the times. In the mid- to late twentieth century, as people tried to express their ethical concerns over the use and treatment of animals, they seized on a number of ethical ideas. One group of ideas included rights, freedoms, liberty, and ending the oppression of the disadvantaged. In modern times, the ideas can be traced back to the origins of democracy and the revolutions of the 1700s in France and the United States. By the 1960s, the concepts of rights and liberation were playing important roles in establishing greater equality among races and in reforming the status of women. In such a cultural environment, it was almost inevitable that these concepts would be used in discussion about our treatment of animals. In some cases, people simply used the language of rights to state fairly commonplace claims, for example that cows should have the ‘right’ to graze on pasture in the summer. In other cases, however, people used a rights or liberation framework to advocate a programme of radical reform that involved abolishing all use of animals for human purposes. As philosopher Tom Regan put it in a famous essay called The Case for Animal Rights, ‘The fundamental wrong is the system that allows us to view animals as our resources, here for us – to be eaten, or surgically manipulated, or exploited for sport or money’.8 The issue, within the framework Regan had created, was not one of making rearing methods ‘more humane’, but rather ‘the total abolition of the use of animals in science; the total dissolution of commercial animal agriculture; the total elimination of commercial and sport hunting and trapping’.9 However there were other moral ideas that ultimately exerted a more widespread inﬂuence on thinking about the treatment of animals. The term ‘welfare’ has been in use for centuries as a broad term commonly deﬁned as ‘good fortune, health, happiness, prosperity’.10 Like rights, the concept of welfare also had a particular resonance in the late twentieth century. As the ravages of the Second World

8Regan, T. 1985. The case for animal rights. Pages 13–26 in In Defense of Animals (P. Singer, editor). Harper & Row, New York. The quotation is on page 14. 9Regan, 1985, pages 13–14. 10Random House Dictionary of the English Language, College Edition. Random House, New York, 1968. In his historic A Dictionary of the English Language, published in 1755, Samuel Johnson used the words ‘happiness; success; prosperity’ in his deﬁnition of welfare.

Dfraser_C004.indd 65

6/27/2008 12:42:24 PM

66

Animal Welfare in Context

War died down, and as people in the more prosperous countries found themselves increasingly free from basic wants and threats to personal security, more attention was paid to issues that fell under the general headings of ‘welfare’, ‘well-being’ and ‘quality of life’. Philosophers analysed these concepts; psychologists devised research methods to assess them; medical and mental health workers saw the promotion of well-being (as opposed to merely curing illness) as a goal of their professions. In this climate, some of the concerns about ill effects on animals caused by farming, biomedical research, sport and other activities came naturally to focus on the ‘welfare’ of animals, or (literally) how well they are faring. Welfare also had a second and more recent meaning. In the English social reforms of the 1800s, one of the central concerns was to protect the needy and vulnerable from being exploited or ignored. Children should not be forced, either by their parents or by economic necessity, into damaging labour in factories. Women should not be submitted to the back-breaking work of underground mining. Limits were placed on the hours that employers could require even from their adult male workers. At the same time, charitable programmes, followed by government programmes, were created to provide housing, food and medical services for those who needed them. These various programmes and reforms came to be referred to as ‘welfare’. Their focus was not on liberating people or recognizing legal rights, but simply on making better provision for vulnerable members of society. And as people perceived animals being used in animal production, biomedical research and other ways, a natural concern arose that vulnerable animals should be protected from undue exploitation. Thus, although there were competing moral frameworks – like Regan’s theory of animal rights – many of the concerns drew on less radical moral ideas about the quality of life of animals and protection of the vulnerable. In trying to capture these concerns in the language of the twentieth century, the ‘welfare’ of animals must have seemed a natural expression. But, of course, there was no established meaning of the term ‘welfare’ when applied to animals. Hence, the furor surrounding Animal Machines also stimulated a remarkable debate on an issue that had never really been settled: What actually is animal welfare? What would constitute a good life for animals? On the surface it seemed obvious – surely what was needed was to correct the terrible situations that Ruth Harrison, Peter Singer and other critics described. But, as the debate continued and the ideas moved beyond the initial stage of calling attention to the issues, and into careful analysis and comparison of alternative options, it became clear that different people had somewhat different ideas about what constitutes a good life for animals. One concern was obvious. Mrs. Harrison had claimed that the modern farming methods deprived animals of ‘all pleasure in life’ (see Box 4.1 for the relevant quotations in more detail). When Astrid Lindgren, the famous author of the Pippi Longstocking stories, entered the debate by writing letters to the major Swedish newspaper accusing the government of failing to protect farm

Dfraser_C004.indd 66

6/27/2008 12:42:24 PM

A Science of Animal Welfare?

67

Box 4.1 Sample quotations by social commentators, ethicists and animal care professionals writing on issues related to animal welfare. Sources are listed in the bibliography. 1 How far have we the right to take our domination of the animal world? Have we the right to rob them of all pleasure in life simply to make more money more quickly out of their carcasses? Have we the right to treat living creatures solely as food converting machines? Ruth Harrison (1964) 2 … legislation is necessary to guarantee dairy cows‘ grazing rights in the summer months. That way, it might even be possible to guarantee that young animals, both bulls and heifers, get a little summertime happiness, at least a temporary reprieve from the floors of barns and the crowded spaces where the poor animals are stored until they die. Let them see the sun just once, get away from the murderous roar of the fans. Let them get to breathe fresh air for once, instead of manure gas. Astrid Lindgren (1985) 3 We could wish that domestic or captive animals should be allowed to express in all respects essential behavioural and physiological possibilities rooted in their ancestral species. The adaptive mechanisms which are available, however, within limits, cope with some of the restrictive, even detrimental features of a domestic or captive existence. The welfare of managed animals is dependent upon the degree to which they can adapt without suffering to the environments provided by man. Edward Carpenter and committee (1980) 4 Animals can feel pain. As we saw earlier, there can be no moral justification for regarding the pain (or pleasure) that animals feel as less important than the same amount of pain (or pleasure) felt by humans. Peter Singer (1990) 5 In principle we disapprove of a degree of confinement of an animal which necessarily frustrates most of the major activities which make up its natural behaviour and we do not consider such confinement or restraint permissible over a long period unless the other advantages thereby conferred upon the animal are likely to be very substantial. An animal should at least have sufficient freedom of movement to be able without difficulty, to turn round, groom itself, get up, lie down and stretch its limbs. F.W.R. Brambell and committee (1965) 6 There is every indication that social valuing of and moral concern for animals will not stop at controlling negative experiences such as pain, fear, distress, boredom etc. It Continued

Dfraser_C004.indd 67

6/27/2008 12:42:24 PM

68

Animal Welfare in Context

Box 4.1

(Continued)

is likely that the emerging social ethic for animals… will demand from scientists data relevant to a much increased concept of welfare. Not only will welfare mean control of pain and suffering, it will also entail nurturing and fulfilment of the animals’ natures… Bernard E. Rollin (1993) 7 The basic moral intuition behind the [capabilities] approach concerns the dignity of a form of life that possesses both deep needs and abilities; its basic goal is to address the need for a rich plurality of life activities. With Aristotle and Marx, the approach has insisted that there is waste and tragedy when a living creature has the innate, or ‘basic’, capability for some functions that are evaluated as important and good, but never gets the opportunity to perform those functions. Martha C. Nussbaum (2004) 8 Good health is the birthright of every animal that we rear, whether intensively or otherwise. If it becomes diseased we have failed in our duty to the animal and subjected it to a degree of suffering that cannot be readily estimated. David Sainsbury (1986) 9 … my experience has been that more problems of animal welfare are to be found in the extensive – the open range – the old-fashioned methods and that by-andlarge the standard of welfare among animals kept in the so called ‘intensive’ systems is higher. On balance I feel that the animal is better cared for; it is certainly much freer from disease and attack by its mates; it receives much better attention from the attendants, is sure of shelter and bedding and a reasonable amount of good food and water. George B. Taylor (1972) 10 … well kept animals are healthier, and healthy animals are more productive and provide higher quality products. Farmers and ranchers who fail to provide good care for their animals therefore jeopardize their own livelihoods. Alan Herscovici (1996)

animals, she took up much the same theme, urging that farm animals ‘get a little summertime happiness’ during their lives. A committee convened by the Reverend Edward Carpenter under the auspices of the Church of England wrote that the welfare of animals depends on ‘the degree to which they can adapt without suffering’ to the environments in which they are kept. And Peter Singer, taking up the Utilitarian ethics of Jeremy Bentham with its emphasis on pain, suffering and happiness, claimed that we should attach as much importance to ‘the pain (or pleasure) that animals feel’ as we would to the same states in human beings.

Dfraser_C004.indd 68

6/27/2008 12:42:24 PM

A Science of Animal Welfare?

69

Pleasure, happiness, suffering, pain – terms such as these deﬁne one of the key concerns that arose in the debate about animal welfare, but curiously it is difﬁcult to ﬁnd just the right English word to capture these states. They are sometimes called ‘feelings’, but that term seems too insubstantial for states as signiﬁcant as pain and suffering. They are sometimes called ‘emotions’, but emotions (by most deﬁnitions) do not include certain important states such as hunger and thirst. They are sometimes called ‘subjective states’ meaning states that are experienced by the ‘subject’ who is having the experience, but the term is ambiguous because it does not specify that some pleasant or unpleasant quality is the deﬁning element. Perhaps the most accurate, if somewhat technical, term is ‘affective states’, a term used nearly a century ago by physiologist Walter Cannon to refer to emotions and other feelings that are experienced as either pleasant or unpleasant rather than hedonically neutral.11 Whatever term we use, we can see in these proposals an idea that resonates with the Romantic emphasis on emotion and with Bentham’s equating of good with pleasure and evil with pain. Thus, for animals to have a good life – for their welfare to be satisfactory – they should experience a minimum of suffering (in the sense of severe or prolonged negative affect) and they should be allowed the normal pleasures of life. But in the various views that were expressed over animal welfare we can also identify a second key concern. This was a call for animals to have more natural and less artiﬁcial lives. The Brambell Committee expressed disapproval ‘of a degree of conﬁnement of an animal which necessarily frustrates most of the major activities which make up its natural behaviour …’ In earthier language, Astrid Lindgren protested against the unnaturalness of conﬁnement environments for farm animals: ‘Let them see the sun just once … . Let them get to breathe fresh air for once, instead of manure gas’. Bernard Rollin, a philosopher who saw the animal welfare concerns of society extending well beyond the pain and suffering emphasized by Peter Singer, proposed: ‘Not only will welfare mean control of pain and suffering, it will also entail nurturing and fulﬁllment of the animals’ natures …’ In a similar vein, philosopher Martha Nussbaum, expanding on a conception of human well-being that emphasizes the freedom to exercise basic human capabilities, proposed that if an animal has the innate capability to perform certain important functions, then ‘there is waste and tragedy’ if the animal is prevented from doing so. In these various proposals, we hear echoes of a second element of the Romantic view of life: the idea that a good life must be a natural life, a life that is not distorted by the artiﬁcial, a life that is lived in harmony with nature and in a way that is true to one’s own nature. This idea, when applied to animals, leads in two directions. The simpler of these is Lindgren’s call for animals to

11Cannon, W.B. 1929. Bodily Changes in Pain, Hunger, Fear and Rage, 2nd edition. Appleton-Century Co, New York. The quotation is on page 4.

Dfraser_C004.indd 69

6/27/2008 12:42:25 PM

70

Animal Welfare in Context

have access to natural elements – sunshine, fresh air – in their environment. At a more subtle level, we also see a call for animals to be able to perform their ‘natural behaviour’ (Brambell), to express ‘essential behavioural and physiological possibilities’ (Carpenter), to fulﬁl their ‘natures’ (Rollin), or to exercise their innate ‘capabilities’ (Nussbaum). Harrison, Lindgren, Singer and most of the others quoted above were critics and visionaries rather than farmers and practicing veterinarians, but these last groups brought a different viewpoint to the debate. They had seen the older farming methods not through the rose-coloured lenses of Pippi Longstocking but in their harsher realities: lambs dying of cold on snowy hillsides, hens killed by foxes, sheep suffering from foot-rot in muddy ﬁelds, malnutrition arising when dietary needs were not adequately met by traditional diets. ‘Good health’ claimed the eminent veterinary educator David Sainsbury, ‘is the birthright of every animal that we rear …’ Veterinarian George Taylor, in an article called ‘One man’s philosophy of welfare’, spoke in favour of the new conﬁnement systems because, he claimed, the animal is ‘much freer from disease and attack by its mates; it receives much better attention from the attendants, is sure of shelter and bedding and a reasonable amount of good food and water’. Related to these views were ones expressed by agricultural producers and their spokespersons. For example, Alan Herscovici, writing on behalf of an agricultural organization, claimed that good animal care is essential for animal health, productivity and product quality. Thus, he claimed, ‘farmers and ranchers who fail to provide good care for their animals therefore jeopardize their own livelihoods’. In these last quotations, we see something quite different from nature-valuing Romanticism. Life for many animals, these critics seemed to claim, would likely be brutal and short were it not for heated barns, vaccines and scientiﬁcally formulated diets. The Golden Age for animals lies not in going back to traditional farming where animals live close to nature, but in applying the tools of science and technology to their housing, feeding and health care. And the evidence that their lives are better in these modern systems is not to be found in speculation about whether animals are ‘happy’ or living ‘natural’ lives but in objective, evidence that they are healthy and productive. Here was a counter-Romantic view of animal welfare, more aligned with the value system of Industrialism, but combined in some cases with a pastoralist sense that by caring appropriately for animals we beneﬁt both them and ourselves. In summary, then, as people formulated and debated various proposals about what constitutes a satisfactory life for animals in human care, three main concerns emerged: (1) that animals should feel well by being spared negative affect (pain, fear, hunger etc.) as much as possible, and by experiencing positive affect in the form of contentment and normal pleasures; (2) that animals should be able to lead reasonably natural lives by being able to perform important types of normal behaviour and by having some natural elements in their environment such as fresh air and the ability to socialize with other animals in normal ways; and (3) that

Dfraser_C004.indd 70

6/27/2008 12:42:25 PM

A Science of Animal Welfare?

71

animals should function well in the sense of good health, normal growth and development, and normal functioning of the body.12 These three issues do not necessarily exhaust the possible concerns that people may have about the welfare of animals. Another concern, which might be captured by the word ‘autonomy’ is that animals should be free to choose their own actions independently of human interference. Yet another, suggested to me by philosopher Mickey Gjerris, is that the ‘integrity’ of animals not be sacriﬁced, for example by genetic engineering. However, I would argue that the three broad issues raised above capture the great majority of the concerns that arise over the quality of life of animals. The three ways of conceptualizing animal welfare are not, of course, mutually exclusive. At least some of the commentators, although emphasizing different concerns, must have assumed that the three would go hand in hand. Lindgren, for example, clearly assumed that animals will be happier if they are allowed to live more naturally, and Sainsbury clearly saw suffering as a natural consequence of disease. And in many situations the three ways of conceptualizing animal welfare do seem to agree. If a pig is allowed to wallow in mud on a hot day, this should be good for its welfare by all three criteria: because it will feel cooler, because it can perform its natural behaviour, and because the ill effects of heat on the animal’s health will be reduced. But when people applied the different ways of conceptualizing animal welfare to speciﬁc questions of how animals should be kept, the simple picture sometimes broke down. A large-scale egg producer may feel that the most important elements of animal welfare are basic health and functioning as reﬂected by a high rate of survival, high levels of production and low incidence of disease. For such a person, the best welfare for the birds might seem to occur in a hygienic, high-health building where hens are kept in cages well separated from predators, from unpredictable weather, and from the pathogens and parasites that may be carried in soil and manure. An organic egg producer, in contrast, may feel that for the birds to have a good life, the most important requirement is that they are free to live in fresh air and sunlight with ample space to move and socialize. For such a person, a freerange system is far better for animal welfare than any cage unit, even if parasites and predators are not as well controlled. An animal protectionist might attach particular importance to affective states and not be overly concerned whether hens are 12Here

and previously (Fraser et al., 1997) I have based my argument on an informal content analysis of samples of text, combined with years of personal involvement in discussions of animal welfare. Qualitative research on people’s views of animal welfare has recently begun to bear out similar conclusions. For example, a study of Danish citizens’ views on the welfare of pigs identiﬁed much the same elements, including concern over ‘physical harm’, ‘avoidance of suffering’, and that animals should be able to lead ‘a natural life’ (Lassen et al., 2006, pages 226 and 228). In a Dutch study livestock producers tended to equate animal welfare with physical health, whereas consumers tended to emphasize ‘freedom to move and freedom to fulﬁll natural desires’, without which an animal was believed to lead a ‘miserable life’ (te Velde et al., 2002, page 210).

Dfraser_C004.indd 71

6/27/2008 12:42:25 PM

72

Animal Welfare in Context

indoors or outdoors, so long as states such as fear, pain and hunger are minimized. Thus, the different beliefs about what constitutes a good life for animals can lead to contrasting views on how animals should be housed and managed. These disagreements do not necessarily involve disagreements about facts. The conﬁnement producer and the organic producer may agree about the level of ammonia in the air of a barn, the rate of mortality in a ﬂock, the types of disease present and other factual matters. Their disagreement is over values – over what is more important or less important, more desirable or less desirable. The organic producer (in our example above) values access to the outdoors more than separation from manure-born pathogens, while the intensive producer does the opposite. But of course, animal welfare is not simply in the eye of the human beholder. The animals themselves will have, for example, preferences and desires; they will have certain types of behaviour that come naturally to the species; they will have different health status in different environments. Surely animal welfare science, by the use of objective, scientiﬁc methods for determining what is good for animals in their own terms, could resolve the disagreements that emerged in the social debate. FOR SCIENCE TO ADDRESS questions of animal welfare, the ﬁrst task was presumably to replace the conﬂicting value-based ideas about animal welfare with a single ‘scientiﬁc’ deﬁnition of the term, much as scientists establish standard deﬁnitions of terms like viscosity and atomic weight. Once the term is properly deﬁned, the deﬁnition would lead logically to certain means of identifying and measuring animal welfare. Thus many scientists, in addition to their empirical work, engaged in a remarkable out-pouring of papers proposing ‘scientiﬁc deﬁnitions’ of animal welfare. First, some semantic conventions needed to be established. Given the different lexical meanings of the term, Donald Broom, the world’s ﬁrst holder of an academic chair in animal welfare, proposed that ‘animal welfare’ be used to refer to the animal’s state, not to external beneﬁts given to the animal, and that welfare be considered as a scale running from good to bad, not just the positive end of the scale.13 As for the content of the deﬁnition, some scientists proposed very broad deﬁnitions that could conceivably cover the full gamut of issues that were present in the public debate. Broom himself deﬁned an animal’s welfare as its ‘state as regards its attempts to cope with its environment’.14 Veterinarian and animal behaviourist Barry Hughes deﬁned animal welfare as ‘a state of complete mental

13Broom, D.M. 1991. Animal welfare: concepts and measurement. Journal of Animal Science 69: 4167–4175. 14Broom, 1991, page 4168.

Dfraser_C004.indd 72

6/27/2008 12:42:25 PM

A Science of Animal Welfare?

73

and physical health, where the animal is in harmony with its environment’.15 Such deﬁnitions identiﬁed the area of discourse, but otherwise were very broad, rather like the dictionary deﬁnitions (noted earlier) that equated welfare with health, happiness and prosperity. Therefore, many scientists tried to give more precise deﬁnitions. Marian Dawkins and Ian Duncan, two of the most inﬂuential pioneers of animal welfare research, put particular emphasis on the affective states of animals. Like William Thorpe they had been trained in animal behaviour, and they saw behaviour, together with physiological and other measures, as providing a scientiﬁc means of understanding the affective states that animals experience. In fact, when Dawkins wrote her classic book about the scientiﬁc study of animal welfare she boldly entitled it Animal Suffering: The Science of Animal Welfare,16 thus underlining her belief that science can be used to help us understand affect in other species. ‘To be concerned about animal welfare’ she wrote elsewhere ‘is to be concerned with the subjective feelings of animals, particularly the unpleasant subjective feelings of suffering and pain’ (see Box 4.2 for quotations in more detail).17 Ian Duncan agreed: ‘… neither health nor lack of stress nor ﬁtness is necessary and/or sufﬁcient to conclude that an animal has good welfare’ he wrote ‘Welfare is dependent on what animals feel’. Other scientists, however, put the emphasis elsewhere. In some of his writing Donald Broom drew on the biological concept of ‘ﬁtness’ to expand on his idea of the animal ‘coping’ with its environment. Fitness, within evolutionary theory, refers to the ability of an animal to leave offspring or other near relatives in future generations. For most vertebrate animals, this requires that the individual survive to an age when it can participate in reproduction and then produce (or help close relatives to raise) viable offspring. Fitness in this sense is a measure of success within evolutionary theory, and Broom proposed it as a useful criterion for animal welfare. ‘Poor welfare’ he proposed, ‘occurs in situations in which … there is reduced ﬁtness or clear indications that ﬁtness will be reduced’.18 Gary Moberg, a physiologist, linked animal welfare to the theory of stress biology. During much of the twentieth century, biologists had been developing 15Hughes,

B.O. 1976. Behaviour as an index of welfare. Proceedings of the 5th European Poultry Conference, Malta, pages 1005–1012. The quotation is on page 1005. 16Dawkins, M.S. 1980. Animal Suffering: The Science of Animal Welfare. Chapman & Hall, London. 17In other articles, Dawkins has explicitly included good health as a key component of animal welfare. See, for example, Dawkins, M.S. 2004. Using behaviour to assess animal welfare. Animal Welfare 13: S3–S7. 18This is taken from Broom, 1991. In other passages Broom has maintained that ﬁtness is not a fully adequate criterion of welfare. Note also that within evolutionary theory, ﬁtness is commonly used as a property of genes, not individuals. However, biologists often speak of the ﬁtness of individuals, and it is in this sense that the term has been applied in animal welfare. Barnard and Hurst, 1996, have outlined some of the confusion that has arisen on this point.

Dfraser_C004.indd 73

6/27/2008 12:42:25 PM

74

Animal Welfare in Context

Box 4.2 Sample quotations indicating different conceptions of animal welfare proposed by animal welfare researchers. Sources are listed in the bibliography. 1 To be concerned about animal welfare is to be concerned with the subjective feelings of animals, particularly the unpleasant subjective feelings of suffering and pain. Marian Stamp Dawkins (1988) 2 … welfare is not simply health, lack of stress or fitness. There will usually be a close relationship between welfare and each of these. However, there will also be enough exceptions to preclude equating welfare with any of them. Thus, neither health nor lack of stress nor fitness is necessary and/or sufficient to conclude that an animal has good welfare. Welfare is dependent on what animals feel. Ian J.H. Duncan (1993) 3 Poor welfare occurs in situations in which … there is reduced fitness or clear indications that fitness will be reduced. Donald M. Broom (1991) 4 … stress is not a rare occurrence but a normal part of life. Stress becomes a threat to well-being only when the stressor, because of its frequency or its magnitude, results in a change in the animal’s biological function such that the animal enters the prepathological state with the ensuing vulnerability to pathology. I propose, therefore, that a prepathological state be used as an indicator of stress and risk to the animal’s well-being. Gary P. Moberg (1985) 5 Recently, scientists have suggested that if an animal perceives that it feels poorly (as measured primarily by behaviour) then the animal is said to be in a poor state of welfare. I dismiss this view as simplistic and inappropriate. I suggest that an animal is in a poor state of welfare only when physiological systems are disturbed to the point that survival or reproduction are impaired. John J. McGlone (1993) 6 The more adequately the organism’s needs are satisfied, the longer it may be expected to live.… Longevity may then serve as an indirect indicator of quality of life. It combines cumulative influences of harm and benefit experienced by the animal throughout its life. It is a broadly based variable, which in effect, integrates the satisfaction of various important and less important needs and avoids potentially serious inaccuracies in human interpretation of animal needs and animal desires. J.F. Hurnik (1993)

Dfraser_C004.indd 74

6/27/2008 12:42:25 PM

A Science of Animal Welfare?

Box 4.2

75

(Continued)

7 If we believe in evolution … then in order to avoid suffering, it is necessary over a period of time for the animal to perform all the behaviours in its repertoire because it is all functional … Marthe Kiley-Worthington (1989) 8 Although ethology, the study of animal behaviour, has made considerable progress towards understanding what animal welfare is and how animals must be kept to have a high quality of life, in many systems major elements of natural behaviour cannot be performed, leading to welfare problems. A major challenge for organic husbandry is … to develop more innovative systems in which all aspects of natural behaviour are taken into account. Beyond avoiding suffering, such systems make positive experiences an important part of the animal’s life. It is not important to ‘prove’ that a sow needs to root or a hen needs to dust bathe, because these are natural behaviours belonging to the animal’s species-specific nature. Susanne Waiblinger et al. (2004) 9 The organism’s priority is to maximize reproductive success by efficient selfexpenditure. Good welfare management policies should therefore strive to maintain natural or acclimatized … strategies of self-expenditure. Reliance on traditional measures of stress response to assess coping is inappropriate, since any apparently deleterious response can be evaluated only in the context of what is understood about the organism’s adaptive strategy of self-expenditure and the impact of current circumstances on it. Christopher J. Barnard and Jane L. Hurst (1996)

theories of how the body responds to a wide range of challenges or ‘stressors’ (Chapter 6). These ‘stress responses’ presumably evolved because they normally help the animal to deal with challenges. If very prolonged and extreme, however, the same responses can lead to pathological changes such as injury and disease; but before they do so, they often trigger less drastic changes such as reduced immune competence and suppression of the hormones that are basic to growth and reproduction. Moberg termed such changes ‘pre-pathological’ because they leave the animal vulnerable to disease or less able to function normally. Thus, drawing on the theory of stress biology, Moberg equated poor animal welfare with ‘suffering from stress’, and he proposed that ‘the most appropriate indicator of stress is the appearance of a pre-pathological state’. Animal scientist (i.e., an agricultural scientist specializing in animal production) John McGlone called on a different body of science in his deﬁnition of animal welfare. Animal scientists have traditionally worked to improve agricultural productivity by reducing death losses and increasing animals’ rates of growth and

Dfraser_C004.indd 75

6/27/2008 12:42:25 PM

76

Animal Welfare in Context

reproduction. McGlone tied his conception of animal welfare to these agricultural measures by proposing that ‘an animal is in a poor state of welfare only when physiological systems are disturbed to the point that survival or reproduction are impaired’. Frank Hurnik, another animal scientist, proposed a somewhat different conception of animal welfare using the traditional agricultural measure of survival rate. Hurnik argued that an animal’s quality of life is directly related to the satisfaction of the many needs that are important for survival, health and comfort in that order of importance, and that the more adequately these needs are satisﬁed, the longer the animal may be expected to live. Hence, Hurnik argued that the longevity of animals integrates the various needs over the animals’ lifespan, and can thus serve as an indicator of animal welfare. In these last four cases, we see scientists linking animal welfare to various bodies of scientiﬁc thought and methodology – to evolutionary theory for Broom, to stress biology for Moberg, and to traditional agricultural measures for McGlone and Hurnik. Yet all of these scientists in their own way tied animal welfare not to affective states, as Dawkins and Duncan had done, but to some aspect of the basic health and functioning of animals. As yet another alternative, some scientists put particular emphasis on keeping animals under more natural conditions, especially by allowing them to perform their natural behaviour. Animal behaviourist Marthe Kiley-Worthington did her early research on the vocalizations of farm animals, and she developed a descriptive list of the various types of calls given by pigs and other species.19 Sensitive to how the different elements of behaviour serve different functions, Kiley-Worthington proposed that for an animal not to suffer, it should be allowed to perform ‘all the behaviours in its repertoire because it is all functional’ in the animal’s life. In a similar vein, Susanne Waiblinger and other scientists involved in organic animal agriculture noted that in many animal production systems, ‘major elements of natural behaviour cannot be performed, leading to welfare problems’. They called for more innovative production systems ‘in which all aspects of natural behaviour are taken into account’. ‘Beyond avoiding suffering’, they claimed, ‘such systems make positive experiences an important part of the animal’s life’. In these last two quotations, the references to suffering tend to put the scientists into the camp of Dawkins, Duncan and others who emphasize affective states, but the primary goal of Kiley-Worthington and Waiblinger was not to avoid suffering by whatever means – they were not, for example, suggesting the use of antibiotics and analgesics to reduce animal suffering – but to accommodate natural behaviour on the assumption that this is a necessary condition for animals to have a good life. The idea of respecting the nature of animals took a very different form in a thoughtful reinterpretation by behavioural biologists Christopher Barnard and 19Kiley, M. 1972. The vocalizations of ungulates, their causation and function. Zeitschrift für Tierpsychologie 31: 171–222.

Dfraser_C004.indd 76

6/27/2008 12:42:25 PM

A Science of Animal Welfare?

77

Jane Hurst. They linked their analysis to evolutionary theory by arguing that animals have been designed by natural selection not simply to preserve themselves by avoiding stress and hardship, but to ‘expend’ themselves efﬁciently so as to achieve maximum reproductive success. Spawning Pink Salmon (Oncorhynchus gorbuscha), for example, have been shaped by evolution not to have long, healthy lives but to destroy themselves in the course of procreation. Using such evolutionary thinking, Barnard and Hurst argued that any scientiﬁc conception of animal welfare must be based not on longevity, low levels of stress, or other criteria that might seem intuitively reasonable to members of the long-lived human species, but on allowing animals to live and expend themselves in the manner for which they were designed by natural selection. Like the different conceptions of animal welfare in the social debate summarized earlier, the deﬁnitions of animal welfare proposed by scientists involve substantial overlap, but are sufﬁciently different that pursuit of any one does not guarantee a high level of welfare as judged by the other criteria. A high level of ﬁtness, or a lack of pre-pathological states, does not necessarily mean that the animal is content or that it can perform its natural behaviour. Similarly, the ability to perform natural behaviour is no guarantee that the animal will not suffer from fear or that it will be in good health and reproduce well. Thus, once again we see different interpretations of animal welfare which could lead to disagreements over how animals should be kept. Moreover, although the scientists related animal welfare to scientiﬁc concepts like ﬁtness and stress, to traditional scientiﬁc measures like survival and growth rate, and to newer scientiﬁc methods designed to reveal affective states, their deﬁnitions were nonetheless underlain by value-based positions about what is most important for animals to have a good life. In fact, the conceptions of animal welfare proposed by scientists were remarkably similar to the different types of concerns expressed in the public debate. Dawkins’ emphasis on the ‘subjective feelings of animals’ and Duncan’s insistence that ‘welfare is dependent on what animals feel’ essentially follow the concern for the affective states of animals that we saw in Harrison, Singer and other critics. The attempts by other scientists to deﬁne welfare in terms of ﬁtness, growth, reproduction and risk of pathology had much in common with the views expressed by veterinarians and animal producers which centred on basic health and the functioning of the body. And the views of Kiley-Worthington, Waiblinger, and Barnard and Hurst had much in common with those expressed by Astrid Lindgren, Edward Carpenter and others that for animals to have a good life they must be kept in ways that are reasonably natural for the species. Thus, instead of the science replacing value-based positions with a single scientiﬁc deﬁnition of animal welfare to which all scientists agreed, value-based positions – about what is most important or most desirable for animals – actually underlie the different deﬁnitions of animal welfare proposed by the scientists. And when the scientists then went back to the laboratory to study animal welfare scientiﬁcally, these same value-based positions were clear in the types of data that

Dfraser_C004.indd 77

6/27/2008 12:42:26 PM

78

Animal Welfare in Context

they collected: some studied rates of survival, reproduction and illness; others studied performance of natural behaviour; and others grappled with the task of understanding the affective experiences of animals. As we will see towards the end of this book, drawing conclusions about animal welfare often involves balancing these different types of evidence to some degree, and the question of how to do this constitutes one of the cutting-edge problems of the ﬁeld. Before confronting this issue, however, let us leave the broad, historical context in which the ﬁeld arose, and turn instead to the speciﬁc scientiﬁc approaches that have been used to understand animal welfare.

Dfraser_C004.indd 78

6/27/2008 12:42:26 PM

Part II Studying Animal Welfare

Dfraser_C005.indd 79

6/27/2008 12:42:40 PM

Introduction

If Professor Brambell’s committee provided an agenda for animal welfare science, it also sowed the seeds of confusion. The committee, as we have seen, proposed a wide-ranging scientiﬁc attack on the problem of understanding animal welfare. It was to involve veterinary medicine, biology and animal science. It was to include studies of the emotional states of animals, their ability to develop learned fear of humans, their preferences for different living conditions, their physiological responses to ‘stress’, and the motivations that are thwarted in conﬁnement. It was a diverse programme – what we today would call a multidisciplinary programme – and the diversity created challenges. One challenge was to reconcile the information from different ﬁelds of science. These (as we will see) include pathology, epidemiology, physiology, animal behaviour and others. However, scientists working in different ﬁelds are known for having different vocabulary, distinct modes of thought, different criteria of proof. Involving so many ﬁelds in a single scientiﬁc pursuit was bound to create problems of communication and understanding. A second challenge came from the various theories that scientists invoked as they tried to apply science to animal welfare. Some scientists, as we have seen, looked to biological theories of ‘stress’. According to one such theory, animals have a general set of responses that are triggered by all forms of adversity such as cold, disease, hunger and social tension. Inspired by this idea, some animal welfare scientists saw the measurement of the body’s ‘stress responses’ as the key to understanding animal welfare. Other scientists looked to various theories of motivation that had been developed to explain the behaviour of animals. If good welfare depends on animals being able to do the things they are highly motivated to do, then measuring the motivations of animals must be a key to understanding and improving their welfare. Yet others looked to the theory of natural selection. If animals have been shaped by natural selection to achieve high success in reproduction, then measures of reproduction would seem particularly good indicators of how well animals are faring; and where reproduction cannot be assessed directly, correlates such as longevity might serve as suitable substitutes. Here again, with such diverse theories and sources of information involved, how could we achieve a coherent account of animal welfare? 80

Dfraser_C005.indd 80

6/27/2008 12:42:40 PM

Introduction

81

A third and overlapping challenge came from the different conceptions of animal welfare that we encountered in the previous chapter. For some scientists, a good life for animals must ﬁrst and foremost be a healthy life, and those scientists looked to measures of health as the basis for assessing and improving animal welfare. For others, a good life must be a natural life, and they looked to studies of the natural behaviour of animals to assess their welfare, and tried to improve animal welfare by making living conditions more like those that the animals would encounter in the wild. For yet other scientists a good life is a hedonically pleasant life in which positive states like comfort and contentment outweigh negative states like pain and distress. For these scientists, the scientiﬁc study and improvement of animal welfare must deal primarily with understanding the affective states of animals. All this diversity – of discipline, of theory, and of value-based positions about what is important for animals to have a good life – resulted in a wide-ranging set of methods and approaches that have been used to assess and improve animal welfare, and also to a degree of confusion as different scientists applied different measures and came up with sometimes contradictory conclusions. However, the entire enterprise was also coloured by a particular view of science that exerted a profound inﬂuence during the twentieth century. On the surface, William Thorpe’s proposal that we use science to understand the pain, distress and cognitive abilities of animals seemed to build naturally on the efforts of Charles Darwin to understand emotions, and later of Romanes and Yerkes to understand the mental processes of animals. However, between Yerkes and Thorpe there had been a half century when many scientists considered that it is simply impossible to use science to help us understand the mental and affective states of animals.1 This view has been ably described by philosopher Bernard Rollin in his book The Unheeded Cry.2 A key player (as Rollin notes) was the American psychologist John Watson whose book Behaviorism, published in the 1920s, gave rise to the school of psychology that bears that name.3 Watson proposed that psychologists should not attempt to study inner, mental (‘subjective’) experience in either humans or other species, but rather that they should limit their research to observable behaviour, speciﬁcally by discovering what responses a person (or animal) gives to various stimuli. To the generation of ‘behaviourists’ that followed, if we had a thorough knowledge of the laws whereby certain stimuli lead to certain responses – laws perhaps analogous to those that apply to the movement of objects in Newtonian physics – then we would be able to predict and control behaviour without any need to venture into the murky and uncertain realm of mental processes. 1Parts of this discussion are based on Fraser, D. 1999. Animal ethics and animal welfare science: Bridging the two cultures (The D.G.M. Wood-Gush Memorial Lecture). Applied Animal Behaviour Science 65: 171–189. 2Rollin, B.E. 1990. The Unheeded Cry. Oxford University Press, Oxford. 3Watson, J.B. 1924. Behaviorism. Republished 1970, W.W. Norton & Company, New York.

Dfraser_C005.indd 81

6/27/2008 12:42:40 PM

82

Studying Animal Welfare

A similar conclusion about subjective experience was reached by many scientists working within the tradition of ‘ethology’ or the study of the behaviour of animals in their natural context. The inﬂuential ethologist Niko Tinbergen advocated that research on animal behaviour should try to answer four questions: how is the behaviour caused? how did it evolve? how does it develop as the animal matures? and what functions does the behaviour serve in the animal’s life? This was a much broader programme than Watson had laid out; but Tinbergen joined Watson in proposing that the ‘subjective’ experience of animals be excluded from scientiﬁc consideration. ‘Because subjective phenomena cannot be observed objectively in animals’, he wrote, ‘it is idle either to claim or to deny their existence’.4 The ideas of Watson and Tinbergen were not, of course, universally accepted. Some scientists continued to insist that the subjective experiences of animals are fully amenable to scientiﬁc study and crucial to understanding behaviour. For example, the eminent British ethologist Julian Huxley insisted that through the objective study of behaviour, ‘we can deduce the bird’s emotions with much more probability of accuracy than we can possibly have about their nervous processes’.5 Similarly, Dutch psychologist J.A. Bierens de Haan regarded Tinbergen’s attempt to understand behaviour without reference to subjective experience as a temporary vogue that would abate and be replaced by meaningful understanding of animal psychology. Other scientists simply carried on writing unapologetically about the affective states of animals. For example, psychologist Harry Harlow insisted that his research on maternal separation in monkeys was about ‘love’ in animals,6 and psychologist P.T. Young proposed a theory of motivation for both humans and animals based on the principle that ‘affective processes regulate and direct behaviour according to the principle of maximizing the positive and minimizing the negative’.7 Nonetheless, the view of science promoted by Watson and Tinbergen was sufﬁciently prominent during the middle decades of the twentieth century that scientiﬁc attempts to understand the subjective states of animals were relatively scarce. As a result, after the bold start by Darwin, Romanes, Yerkes and others, nearly half a century passed when remarkably little progress was made on these subjects. When animal welfare scientists began trying to understand states such as 4Tinbergen, N. 1951. The Study of Instinct. Clarendon Press, Oxford. The quotation is on page 4. 5The debate is well described by Burkhardt, R.W. Jr. 1997. The founders of ethology and the problem of animal subjective experience. Pages 1–13 in Animal Consciousness and Animal Ethics (M. Dol, S. Kasanmoentalib, S. Lijmbach, E. Rivas and R. van den Bos, editors). Van Gorcum, Assen, The Netherlands. The quotation from Huxley is on page 8. 6Blum, D. 2002. Love at Goon Park: Harry Harlow and the Science of Affection. Perseus Publishing, Cambridge, USA. 7Young, P.T. 1959. The role of affective processes in learning and motivation. Psychological Review 66: 104–125. The quotation is on page 117. For a very readable review and other examples see Toates, F. 1986. Motivational Systems. Cambridge University Press, Cambridge.

Dfraser_C005.indd 82

6/27/2008 12:42:40 PM

Introduction

83

pain and distress in animals, the science of the previous half-century thus provided surprisingly little in the way of relevant theory and methods to build on. Moreover, debates over how to study animal welfare were coloured by a lingering disagreement between those scientists who believed (with Watson and Tinbergen) that the affective states of animals fall outside the scope of science and that alternative measures (health, longevity) must be used instead, versus those who considered that the scientiﬁc study of affect is not only possible but essential for understanding animal welfare. IN THIS PART OF THE BOOK, we turn to the methods that have been used in the scientiﬁc study of animal welfare. Some of the methods (especially in Chapter 5) focus on the basic health and biological functioning of animals, others (Chapter 8) on affective states, and others (Chapter 9) on understanding and accommodating the ‘natural’ lives of animals. The remaining chapters of the section deal with speciﬁc topics and methods that have become prominent in the study of animal welfare: ‘stress’ in Chapter 6, abnormal behaviour in Chapter 7, and the preferences and motivations of animals in Chapter 10. In all these chapters I have focused not just on the most recent thinking and debates, but have also tried to sketch a little of the historical development of the ideas – how concepts evolved, how methods changed, how certain types of confusion arose. Throughout these chapters I have tried to give a critical analysis that brings out both the strengths and the limitations of the various methods, and the logic by which we link scientiﬁc ﬁndings to an understanding of animal welfare. If, at the end, readers are left wondering how the different types of information can be combined to create a coherent account of animal welfare, then they will be in the right mind-set to proceed to the ﬁnal part of the book.

Dfraser_C005.indd 83

6/27/2008 12:42:40 PM

Disease, Injury and Production

5

Long before the phrase ‘animal welfare science’ had been coined, veterinarians, agriculturalists, zoo biologists and others had been studying, treating and preventing problems of animal health, growth and reproduction. Inasmuch as concern over animal welfare includes concern over the basic health and functioning of animals, this body of veterinary and biological research quickly became one of the fundamental elements of animal welfare science. In time we will confront the philosophical debate that arose about how welfare is actually related to concepts such as health, growth and reproduction. Let us begin, however, by reviewing some of the different methods and approaches that have been used, and the kind of insights they have provided. IN MANY CASES VETERINARIANS on the front lines of animal health protection ﬁnd themselves confronted with sick or dead animals and need to work out what caused the condition and how to treat or prevent it. As a veterinary pathologist dealing with farm animals, Ernest Sanford was troubled by a number of cases of dead sows arriving for post mortem examination in seemingly good condition but with the stomach and spleen fatally twisted. In looking into the cases he found that many were from farms that practiced an unusual method of feeding their animals. Pregnant sows have prodigious appetites, likely as a result of many decades when pigs were bred for rapid growth and high food intake. With the concentrated diets normally used on pig farms, pregnant sows need to be restricted to only a fraction of the volume of food they would eat by choice, so that they do not become obese. For sows kept on pasture, producers will sometimes restrict their total intake by delivering an abundant amount of food every second day. This way the animals can eat their ﬁll without having to compete for a limited food supply, and on alternate days the sows eat hay or other roughage, and they forage for what food they can ﬁnd by rooting in the soil. However, alternate-day feeding was also adopted by some conﬁnement pig producers who kept their sows indoors in individual gestation stalls where the animals had no opportunity to forage when hungry. 84

Dfraser_C005.indd 84

6/27/2008 12:42:40 PM

Disease, Injury and Production

85

Under these conditions the animals became intensely excited at feeding time. As the staff began to roll the feed carts through the barn, dispensing food stall by stall, the animals that were last to be fed could work themselves into a frenzy of trumpeting and vigorous movement while waiting for their food to arrive, and they then bolted down the food that would last them for the next two days. Sanford found that the problem of twisted stomachs and spleens was most common on conﬁnement units that used alternate day feeding, and that it occurred especially on weekends – times when there was likely to be reduced staff and perhaps only one person feeding hundreds of sows who thus had to wait longer than usual to be fed. Putting these observations together Sanford concluded that the problem was due to some combination of over-excitement, excessive hunger and rapid eating, caused by the alternate-day feeding system and exacerbated by weekend work schedules.1 Improvements could be made by doing away with alternate-day feeding for sows in conﬁnement, and by using an automatic feeding system whereby food is dropped simultaneously to all sows at the throw of a switch so that the animals, although still hungry, would not work themselves into a frenzy at feeding time. Sanford’s ﬁndings played an important role in animal welfare standards that recommended against the use of alternate-day feeding for sows unless roughage is available.2 Torsion of the stomach and spleen is an uncommon condition whose likely causes had to be pieced together by a combination of knowledge and ingenuity on the part of the pathologist. In other cases, however, health problems are common enough to be studied experimentally. When egg producers began keeping laying hens in ‘battery cages’ (i.e., groups of cages joined together in batteries – Figure 5.1), they encountered a range of basic health problems including injuries and feather damage. Swedish scientist Ragnar Tauson and co-workers used an experimental approach to understand the problems. They housed birds in a wide variety of commercially available cages and recorded the incidence of various health problems as well as practical aspects of egg production in the different environments.3 They found that many birds developed signiﬁcant foot lesions if kept in cages with steeply sloped ﬂoors constructed of poorly galvanized wire mesh. The steep slope was designed to make eggs roll out of the cage onto a collection belt, but it also appeared to make the birds grip the ﬂoor strongly for secure footing. The problem was largely eliminated if the ﬂoor had only a gentle slope and especially if the wire mesh was coated in a plastic material that covered any areas of sharp metal. The gentler slopes also reduced the chance of eggs being cracked as they rolled across the cage ﬂoor. 1Sanford, S.E., Waters, E.H. and Josephson, G.K.A. 1984. Gastrosplenic torsions in sows. Canadian Veterinary Journal 25: 364. 2For example: Connor, M.L. (chairperson). 1993. Recommended Code of Practice for the Care and Handling of Farm Animals: Pigs. Agriculture and Agri-Food Canada, Ottawa. 3Tauson, R. 1998. Health and production in improved cage designs. Poultry Science 77: 1820–1827.

Dfraser_C005.indd 85

6/27/2008 12:42:40 PM

86

Studying Animal Welfare

Figure 5.1 ‘Battery cages’ (i.e., groups of cages joined together in batteries) for laying hens. Photo courtesy of Prof. I.J.H. Duncan.

Similarly, many birds developed lesions on the neck when feeding from deep, sharp-lipped troughs that were installed too high for comfortable access. The problem was largely eliminated by a shallower food trough located more conveniently for the birds. A third problem was that many birds developed overgrown claws. In a more natural environment the ground is sufﬁciently abrasive to wear down the hens’ claws simply in the course of walking and scratching for food, but this did not happen on the ﬂoor of the cages. The problem was solved by installing a strip of abrasive material, like sandpaper, on the rear part of the feed trough. A fourth problem was loss of feathers caused by a mixture of abrasion and pecking by other birds. Replacing the wire-mesh side walls of the cages with solid partitions led to noticeable improvements in feather cover and also reduced the birds’ food requirements because poorly feathered birds eat more to stay warm. These various ﬁndings were the result of a long series of experiments, initially comparing different commercially available cages and then working through one environmental factor after another to identify cause and effect, and ultimately to ﬁnd practical solutions. The results were very inﬂuential with cage manufacturers and formed the basis of regulations on cage design initially in Sweden and subsequently in the European Union. Many of the new designs also improved proﬁts for the producers by eliminating unnecessary injuries and other losses.

Dfraser_C005.indd 86

6/27/2008 12:42:41 PM

Disease, Injury and Production

87

For many health issues, the number of possible causal factors is simply too great for an experimental approach to yield a satisfactory understanding of the problem. In these cases, epidemiological surveys have often produced valuable insights. The pioneer of veterinary epidemiology in animal welfare research is the Swedish veterinarian Ingvar Ekesbo, who started in the 1960s to compile health statistics on farm animals in relation to how they were housed and managed. At the time, dairy production was in a state of change as dairy farmers moved away from traditional ‘tie-stalls’ in which cows spent much of their time tethered, in favour of various ‘loose housing’ systems. Some of these had open areas where the cattle could rest; others had individual cubicles which cows could enter at will, plus an open area for feeding. The ﬂooring of the new units ranged from bare concrete to deep bedding of straw or sawdust, and in some cases the cows were kept permanently indoors with no time on pasture. The industry was in a state of ﬂux, and producers were switching to new systems with little relevant research to guide them. Ekesbo’s ﬁrst major study involved a 5-year examination of health records painstakingly compiled and analysed from over 700 dairy herds.4 The study produced a wealth of information about the relationship between the environment and disease. For example ‘milk fever’, a metabolic disorder caused by a severe drop in blood calcium level just after calving, was relatively common in all three of the major housing systems (Figure 5.2); and regardless of the type of barn, the disease occurred slightly more often in the months when the cows were on pasture rather than indoors. In this case, keeping cows indoors may have allowed farmers to exert better control over the animals’ mineral intake. Trampled teats followed the opposite pattern (Figure 5.2). This problem rarely occurred while cows were on pasture, and when the cows were indoors the type of environment was of major importance. Speciﬁcally, cows in loose-housing systems had relatively few trampled teats, but the problem was more common in tie stalls, especially if the stalls were not bedded. Mastitis (inﬂammation of the udder) followed a different pattern again. It was slightly more common in tie stalls than in loose housing, but it was about twice as common on farms where the cattle had to lie on hard surfaces (usually concrete) compared to farms where the resting area was bedded in straw or sawdust (Figure 5.3). The results of Ekesbo’s study provided invaluable guidance to the dairy industry as it underwent a widespread transformation from tie-stalls to less restrictive forms of housing. Working in the 1960s, Ekesbo was limited to the types of comparisons he could make by hand calculations, but today’s epidemiologists have access to powerful ‘multivariate’ methods of analysis that can identify relationships between variables in large data sets after adjusting statistically for other factors. An example comes from the study of hyperthyroidism (excessive activity of the thyroid gland) in cats. Hyperthyroidism commonly involves loss of weight, increased appetite and 4Ekesbo, I. 1966. Disease incidence in tied and loose housed dairy cattle and causes. Acta Agriculturae Scandinavica 15 (Suppl): 1–74.

Dfraser_C005.indd 87

6/27/2008 12:42:43 PM

While housed

Milk fever

Loose housing Tied bedded Tied unbedded While on pasture Loose housing Tied bedded Tied unbedded While housed Loose housing

Trampled teats

Tied bedded Tied unbedded While on pasture Loose housing Tied bedded Tied unbedded 0

2

4

6 8 10 Per cent of cows

12

14

Figure 5.2 Percentage of parturitions involving milk fever (above) and percentage of cows with trampled teats in one recording year (below) on Swedish farms divided according to the housing system in place: loose housing, tie stalls with bedding, and tie stalls without bedding. Results are shown for periods of the year when the animals were housed indoors and when they were kept outdoors on pasture. Results are averages of the two most common dairy breeds in Sweden. Data are from Tables 26 and 53a of Ekesbo, 1966.

Loose housing

Mastitis

Bedded Unbedded Tied Bedded Unbedded 0

4

8 12 16 Per cent of cows

20

24

Figure 5.3 Percentage of cows with mastitis in one recording year for the two most common dairy breeds in Sweden, shown for cows on farms with loose housing (above) or tie-stalls (below), either with or without bedding. Data are from Table 44a of Ekesbo, 1966.

Dfraser_C005.indd 88

6/27/2008 12:42:43 PM

Disease, Injury and Production

89

excessive activity, sometimes combined with excessive drinking and urination, vomiting, diarrhoea and heart failure. The disease was ﬁrst described in 1979, but rapidly became very common. A team of veterinarians at Purdue University in the United States carried out an analysis of hospital records in order to identify risk factors. They ﬁrst compiled nearly a quarter million records of hospital visits involving cats at nine teaching hospitals from 1978 to 1997. In the ﬁrst ﬁve years, hyperthyroidism was seen in less than 0.1% of hospital visits, but by the mid-1990s the condition was diagnosed in nearly 3% of such visits. The data showed that the average age of cats brought to the hospital had increased during the 20 years of the study, owing to a general trend for cats to live longer, but statistical analysis showed that the increase in the disease was far more rapid than could be explained by age alone. Believing that some change in the cats’ diet might be involved, the team next assembled records of 642 older cats brought to their teaching hospital and tested for hyperthyroidism, and they surveyed the owners for information on what the animals were fed. Statistical analysis comparing cats with and without hyperthyroidism showed, as expected, that the most pronounced risk factor had to do with diet. Compared to cats receiving only dry food, cats fed partly or largely on canned food were about three times more likely to develop hyperthyroidism, and the risk increased further for cats fed on brands of food that were packaged exclusively in ‘pop-top’ containers (which can be opened simply by pulling away the metal lid). The team noted that containers used for wet cat-food include a range of lacquers, resins and other chemicals, some of which are known to disrupt endocrine functions in the body and which can migrate into food that is high in oil or fat.5 Sometimes it is possible to combine the analytical power of epidemiology and the precision of a planned experiment in the same study. Many animal welfare standards used for ‘broiler’ chickens (i.e., chickens bred for meat rather than egg production) put particular emphasis on the amount of ﬂoor space available per bird. Marian Dawkins and co-workers did an extensive study in the United Kingdom involving 2.7 million birds raised by 10 chicken production companies that agreed to house birds at a wide range of stocking densities, from a crowded 46 kilograms of birds per square metre, down to the most spacious commercial standard of 30.6 Dawkins and co-workers then measured a large number of environmental variables, as well as basic health and functioning variables such as lameness, survival, growth and aggression. The planned experimental comparison showed that different levels of space allowance did inﬂuence the health and functioning of the birds as expected. Notably, 5Edinboro, C.H., Scott-Moncrieff, J.C., Janovitz, E., Thacker, H.L. and Glickman, L.T. 2004. Epidemiologic study of relationships between consumption of commercial canned food and risk of hyperthyroidism in cats. Journal of the American Veterinary Medical Association 224: 879–886. 6Dawkins, M.S., Donnelly, C.A. and Jones, T.A. 2004. Chicken welfare is inﬂuenced more by housing conditions than by stocking density. Nature 427: 342–344.

Dfraser_C005.indd 89

6/27/2008 12:42:43 PM

90

Studying Animal Welfare

the percentage of birds showing some sign of lameness increased from 19% under the most spacious conditions to 39% at the most crowded, and the growth rate of the birds fell by about 5% over the same range. However, the really striking differences were not between different space allowances but between different companies. Total deaths ranged from 1.4% of birds in the best case to 14% in the worst. Leg health was even more variable: in one company lame birds were relatively rare, whereas in another most birds showed some impairment of gait. Multivariate analysis showed that two of the key environmental variables were the temperature and the humidity in the barns. In particular, the percentage of birds dying was correlated with high temperature and high humidity in the ﬁnal weeks of the growing period. Another key variable was the number of staff involved: houses that were tended consistently by the same one or two workers had fewer deaths than those where a larger number of staff were involved. The proportion of birds with sound legs was also correlated with low light levels, presumably because the birds ate less voraciously in dim light and their body weight did not increase too rapidly for their skeletal development. The authors concluded that space allowance is important, but that in devising standards to improve the basic health and functioning of chickens, emphasis should also be placed on other key variables. Each of the above examples has dealt with a speciﬁc research programme involving an individual scientist or team: Sanford’s sleuthing to piece together the causes of twisted stomachs in sows, Ekesbo’s analysis of the records of 700 dairy herds, the series of experiments by Tauson’s team on cage design, and so on. In other cases, however, an area of research has been tackled by a large number of researchers working independently in different locations and without any effort to standardize their methods. When this happens, the different studies often produce a complex picture which, on the surface, does not lead to any simple conclusion. In such cases, is it possible to combine information from different sources in some systematic manner? The issue arose in the debate over whether injections of bovine growth hormone, used in some countries to increase milk yield, have any harmful effect on the health and welfare of cows. In the 1970s basic research on lactation showed that injections of growth hormone cause lactating cows to direct more nutrients to milk production, thus increasing milk yield appreciably.7 On this basis an agricultural pharmaceutical company developed a method to produce bovine growth hormone (or BST for Bovine Somatotrophin) from genetically modiﬁed bacteria, and the company promoted the product as a way of increasing milk yield on dairy farms. When the company applied for permission to market the product in Canada, an expert panel was appointed by the Canadian Veterinary Medical Association to determine, among other things, whether BST has any ill effects on the animals’ health and welfare. 7Bauman, D.E. 1992. Bovine somatotropin: Review of an emerging animal technology. Journal of Dairy Science 75: 3432–3451.

Dfraser_C005.indd 90

6/27/2008 12:42:43 PM

Disease, Injury and Production

91

The panel noted that there were a great many relevant studies conducted on different herds under a wide range of conditions, sometimes with contradictory results. To combine the different sources of information the committee carried out a ‘meta-analysis’ of results from the different studies. This is a statistical procedure that estimates the size of an effect over all the studies considered; it also tests whether this overall effect was at more than chance level, and whether the effect differed signiﬁcantly between studies. The meta-analysis indicated that the effect of BST injections on cow health did vary substantially from study to study. Overall, however, BST was associated with an increase of about 25% in the risk of clinical mastitis and about 50% in the risk of lameness (Figure 5.4). Both of these effects were likely due not to any direct effect of BST itself, but to the higher metabolic demand that it created in augmenting milk production.8 One of the most important factors affecting the health and functioning of animals is their genetic make-up. My own introduction to welfare problems of genetic origin came in 1970 when my colleague Martin Waddell brought me a cage of crippled mice. The mice had a hereditary form of muscular dystrophy that had been discovered some years before. The animals had such weak and underdeveloped muscles that they were incapable of mating, but they continued to be produced (through a delicate process of artiﬁcial insemination) because they were considered valuable animal models for research on muscular dystrophy in humans. Waddell was concerned because the coats of the dystrophic mice had become thin and patchy. He recognized that their ability to groom themselves might well be deﬁcient, but if this was the problem, was there a way to help the mice maintain better coat condition? By looking at the mice under a low-power microscope, we could see that they were infested with tiny fur mites (Myobia musculi) attached to their skin. Normal mice control the mites by grooming their coats with their teeth. If the dystrophic mice could not groom themselves, we wondered whether other mice could provide this service for them. We set up a simple experiment involving normal and dystrophic mice housed either alone or two to a cage. After a few weeks we looked at their coats and did our best to estimate the number of mites they were carrying. We found that if normal mice were kept singly, most of the body was clear of mites, but they developed a dense build-up on the head and shoulders – the parts of the body that they could not reach to groom themselves. In contrast, normal mice housed in pairs were virtually free from mites, even on the head and shoulders. Evidently, mice rely on their cagemates to groom those parts of the body. When a 8Canadian Veterinary Medical Association (CVMA). 1998. Report of the Canadian Veterinary Medical Association Expert Panel on rbST. Health Canada, Ottawa. Available at: http://www.hcsc. gc.ca/english/protection/rbst/animals/index.htm, accessed July 2005; Dohoo, I.R., DesCôteaux, L., Leslie, K., Fredeen, A., Shewfelt, W., Preston, A. and Dowling, P. 2003. A meta-analysis review of the effects of recombinant bovine somatotrophin. 2. Effects on animal health, reproductive performance, and culling. Canadian Journal of Veterinary Research 67: 252–264.

Dfraser_C005.indd 91

6/27/2008 12:42:44 PM

92

Studying Animal Welfare

1 2 3 4 5 6 7 8 9 10 11 Combined .33

.5

.75 1 2 Risk ratio – lameness

4

6

Figure 5.4 A meta-analysis of the effects of bovine growth hormone (BST) injections on the risk of lameness in dairy cows. Each line represents the results of one set of data (of 11 in total) that compared the incidence of lameness in cows treated with BST and untreated control cows. If treated cows had the same incidence of lameness as the control group, the risk ratio would be 1. The locations of the rectangles show the risk ratio in the study; the area of the rectangle shows the weight assigned to the study in the overall analysis, based especially on the number of cows tested. The length of the line shows the 95% confidence interval of the estimate (truncated in some cases to fit on the page). The dashed vertical line indicates the overall risk ratio based on all studies, and the diamond indicates the 95% confidence interval of the estimate. From Dohoo et al., 2003. Figure kindly provided by Dr. Dohoo and reproduced with the permission of the Canadian Veterinary Medical Association.

normal mouse and a dystrophic mouse were housed together, the normal mouse ended up looking like singly housed animals: they had no mites on the parts of the body that they could groom for themselves, but they developed a dense growth of mites (several hundred per square centimetre of skin) on the head and shoulders. This told us that the dystrophic cage-mate was incapable of performing the service of social grooming. However, dystrophic mice that were housed with a normal mouse had excellent coats. Presumably, a normal cagemate was willing to groom the entire body of mice that could not groom themselves. In this case, there was nothing we could do to prevent the misery caused by the muscular dystrophy, but as long as we housed the crippled mice with normal cage-mates, they could at least have the beneﬁt of adequate grooming.9 Although animals with inherited disorders have had a long history in medical research, the recent ability of geneticists to tamper deliberately with the genome has 9Fraser, D. and Waddell, M.S. 1974. The importance of social and self-grooming for the control of ectoparasitic mites on normal and dystrophic laboratory mice. Laboratory Practice 23: 58–59.

Dfraser_C005.indd 92

6/27/2008 12:42:44 PM

Disease, Injury and Production

93

led to an explosion of mice suffering from genetic diseases. In 2007, for example, a group of scientists found a way to produce mice that would develop rheumatoid arthritis.10 They began by identifying a human gene that increases the risk that the person will develop the disease. Then, applying the techniques of genetic engineering to mice, they were able to replace the comparable mouse gene with the defective human gene. When these mice were then injected with collagen, the main protein in cartilage, they developed an immune response that attacked the cartilage of their own bodies and produced a condition that closely resembled human rheumatoid arthritis. Such animals, and the many thousands of others that have been deliberately created to have disease conditions, pose a new generation of animal welfare problems. Preventing the diseases is out of the question, because the diseases were created deliberately for the purpose of research. The main approaches for improving the animals’ welfare are to identify mitigating treatments (which may then be useful for treating the corresponding diseases in humans) and to make thoughtful decisions about when to euthanize such animals so that their suffering can at least be terminated as promptly as the research allows. Domestic dogs present a wide range of genetic diseases, some of which are, at least in principle, more amenable to improvement. Dogs have been bred for thousands of years to serve purposes such as guarding, herding, transporting goods, hunting by sight, and hunting by scent. More recently some of the physical traits that arose from this early genetic selection have been further accentuated by breeding ‘show dogs’ to conform to written standards for the breed. As this artiﬁcial selection replaced natural selection, many genetically based disorders became very common. Some of these arose because dogs were deliberately bred to have extreme physical characteristics such as large heads (which complicate the birth process) and ﬂat noses (which create difﬁculty in breathing). In other cases, selecting for certain body shape and size seems to increase the risk of physical disorders. One of the most common is hip dysplasia in which the ball and socket surfaces of the hip joint are not held tightly together, and movement in the joint can damage the cartilage or cause unnatural bone growth that leads to pain and chronic arthritis. Veterinary scientist Elizabeth LaFond and co-workers analysed over 10 000 cases of hip dysplasia diagnosed out of 300 000 records of dogs brought to veterinary teaching hospitals in the United States. Compared to mixed-breed dogs, purebred dogs of several larger breeds (Bernese Mountain dog, German Shepherd, Newfoundland, Saint Bernard and others) were ﬁve to seven times more likely to be diagnosed with this condition.11 10Taneja, V., Behrens, M., Mangalam, A., Grifﬁths, M.M., Luthra, H.S. and David, C.S. 2007. New humanized HLA-DR4-transgenic mice that mimic the sex bias of rheumatoid arthritis. Arthritis & Rheumatism 56: 69–78. I am grateful to Elisabeth Ormandy for helpful discussion. 11LaFond, E., Breur, G.J. and Austin, C.C. 2002. Breed susceptibility for developmental orthopedic diseases in dogs. Journal of the American Animal Hospital Association 38: 467–477.

Dfraser_C005.indd 93

6/27/2008 12:42:44 PM

94

Studying Animal Welfare

In other cases, purebred dogs are bred from such a small gene pool that they have a high incidence of harmful recessive genetic conditions. For example, ‘collie eye anomaly’ results from a recessive gene on chromosome 37, which was likely present but rare in ancestral dogs that gave rise to the modern herding breeds.12 The condition, which leads to abnormalities in the structure of the eye and can impair vision in severe cases, is now extremely common in several herding breeds. For example, the Companion Animal Welfare Council in the United Kingdom reported that 72% of Shelties and 64% of Rough Collies in that country were affected.13 This may have occurred through an accidental concentration of the harmful gene when the breeds were established, combined with breeding from too small a number of animals. In principle, problems like hip dysplasia and collie eye anomaly could be greatly reduced through an organized and science-based effort by breeders. Animal welfare scientist Paul McGreevy and geneticist Frank Nicholas proposed several changes to make this possible. They proposed that breed associations and other groups review the breed standards to ensure that people are not encouraged to breed for traits that cause or contribute to animal welfare problems. Breeders could also use modern scientiﬁc methods, such as markers that identify carriers of harmful genes, to eliminate such genes from breeding programmes. Breed associations could also change the rules for registering purebred dogs to increase the genetic diversity of the breeds. Perhaps most fundamentally, breed associations could identify animal welfare as one of their goals, and give precedence to animal welfare in their actions and decisions.14 Genetic selection of food-producing animals has led to some similar problems. For example, artiﬁcial selection of dairy cattle for high milk production has led to a dramatic increase in yield, but this is often accompanied by a higher incidence of mastitis, lameness and reproductive failure. As a result, most dairy cows in North America are culled for poor health or reproductive failure by the end of their second or third lactation, after only a small fraction of their potential productive lives. By modifying the criteria, so that selection is based on factors such as lifespan and resistance to disease as well as milk production, it should be possible to achieve gains in yield without the same harmful effects on the animals’ welfare. In fact, given the economic cost of veterinary treatment, such an approach will not necessarily result in less proﬁt, even if average milk production per cow is not quite as great.15 12Parker,

H.G. et al. (10 authors). 2007. Breed relationships facilitate ﬁne-mapping studies: A 7.8-kb deletion cosegregates with Collie eye anomaly across multiple dog breeds. Genome Research 17: 1562–1571. 13Companion Animal Working Group 2006. Breeding and Welfare in Companion Animals. Companion Animal Welfare Council, Sidmouth, UK. 14McGreevy, P.D. and Nicholas, F.W. 1999. Some practical solutions to welfare problems in dog breeding. Animal Welfare 8: 329–341. 15Sandøe, P., Nielsen, B.L., Christensen, L.G. and Sørensen, P. 1999. Staying good while playing god – the ethics of breeding farm animals. Animal Welfare 8: 313–328.

Dfraser_C005.indd 94

6/27/2008 12:42:44 PM

Disease, Injury and Production

95

One of the most important positive contributions of science to animal welfare is to develop and test practical measures to prevent naturally occurring animal diseases. Of all the diseases that humans share with other species, rabies must surely rank as one of the most terrible. We have recognized since the time of Louis Pasteur that rabies is caused by a pathogen (now known to be a virus) which travels slowly from the site of infection to the central nervous system where it causes inﬂammation of the brain leading to dementia, paralysis and death. India has an estimated 20 million stray dogs, and rabies is a signiﬁcant problem. Mainly as a result of dog bites, India sustains roughly 20 000 human rabies fatalities per year.16 Research on the control of feral animal populations has now identiﬁed how the control of rabies could be achieved. Simply killing or removing stray dogs turns out not to be an effective option because stray dogs are capable of a high rate of reproduction and can move easily from one area to another; hence, animals that are removed by trapping or shooting are rapidly replaced by other dogs. However, if the dogs are trapped, vaccinated, neutered, and then released back into their home area, the population remains stable but rabies is eliminated. If this is done in a systematic way, sector by sector throughout a city, the disease can be brought under control.17 If this procedure could be implemented on a large scale, the result would be that a huge number of dogs, and a very signiﬁcant number of people, would be spared one of the worst deaths imaginable. These are just a few of a vast number of examples that illustrate the role of the animal health sciences in assessing and improving animal welfare. I chose the examples to illustrate some of the different approaches – pathology, epidemiology, experimental approaches, meta-analyses – and some of the various areas of application: diagnosing health problems, improving housing and management systems, understanding factors that increase the risk of disease, developing animal welfare standards, and mitigating problems caused by extreme breeding practices. The examples also illustrate the diverse types of disease conditions that occur in domestic animals. Some, like rabies, are infectious diseases that are presumably ancient in origin, and it is a tribute to modern veterinary medicine that so many of these animal diseases are reasonably well controlled, at least in the more prosperous countries. But we also see emerging diseases, and diseases whose incidence and severity have been greatly increased by modern conditions. Hyperthyroidism in cats appears for the most part to be a new disease, perhaps made worse by modern methods of packaging cat food. The torsion of the stomach and spleen

16Sudarshan,

M.K. et al., 2007. Assessing the burden of human rabies in India: Results of a national multi-center epidemiological survey. International Journal of Infectious Diseases 11: 29–35. 17World Health Organization (WHO). 2004. WHO Expert Consultation on Rabies. First Report. Technical Report Series, WHO, No. 931, Geneva; World Society for the Protection of Animals (WSPA). 1994. Stray Dog Control. WSPA, London. I am grateful to Drs. S.A. Rahman and Carla Forte Maiolino Molento for helpful discussion of these issues.

Dfraser_C005.indd 95

6/27/2008 12:42:44 PM

96

Studying Animal Welfare

studied by Ernest Sanford was mainly a product of modern breeding and feeding systems for pigs. The lameness studied by Marian Dawkins occurred partly because meat-producing chickens have been bred for such rapid muscle growth that their skeletal development sometimes fails to keep pace with the increase in body weight. The milk fever studied by Ingvar Ekesbo was very likely exacerbated by high milk production leading to excessive demand for body calcium at the time of calving. Given that so many common disorders are caused or exacerbated by high levels of production in animals, let us now consider how the ‘productivity’ of animals is related to their welfare. IN THE CASE OF farm animals, there is a traditional argument linking animal welfare to the productivity of animals and the proﬁtability of the farm. According to this argument, if farm animals did not have a good quality of life, they would simply fail to thrive. Because they would not grow and produce well, the owner would not make a satisfactory proﬁt and would go out of business. Therefore (according to this argument), a high level of agricultural productivity, and a satisfactory proﬁt from the enterprise, are enough to ensure that the animals’ welfare is good. We will return to this argument below, but ﬁrst let us look at how measures of productivity have been used in assessing animal welfare. When large numbers of farm animals are conﬁned in enclosed buildings, one of the concerns is a build-up of harmful gasses in the air. The most obvious of these is ammonia which comes mainly from the manure that accumulates either in the building or in a manure pit below the animals’ living area. When people step into a barn from outside, they may ﬁnd ammonia mildly irritating to the eyes and nose at concentrations of 10 parts per million (ppm), but they seem to adapt after some exposure. A common human health standard is that workers should not be required to remain in air with more than 25 ppm ammonia for an eight-hour shift. But how much is too much for animals who live in the environment continuously? One approach to setting standards has been to compare simple measures of productivity. A group of agricultural scientists in the United States raised broiler chicks in controlled-atmosphere chambers where the level of ammonia was maintained at about 0, 25, 50 or 75 ppm. The birds were kept in the chamber for their ﬁrst four weeks and then had three more weeks in clean air before being slaughtered. The data showed no obvious effects of ammonia at 25 ppm compared to 0, but at 50 and 75 ppm there was a clear decline in weight gain and an increase in the percentage that died (Table 5.1). The authors compared their data to the results of similar studies done 20 years earlier. They concluded that even though genetic selection for productivity had resulted in tremendous increases in growth rate in the previous two decades, the birds’ sensitivity to ammonia remained essentially unchanged.18 18Miles, D.M., Branton, S.L. and Lott, B.D. 2004. Atmospheric ammonia is detrimental to the performance of modern commercial broilers. Poultry Science 83: 1650–1654.

Dfraser_C005.indd 96

6/27/2008 12:42:44 PM

Disease, Injury and Production

97

Table 5.1 Measures of agricultural productivity of broiler chicks kept at four levels of ammonia in the air. Ammonia Body weight Depression in Deaths in concentration (ppm) at 4 weeks (g) body weight (%) Feed/gain 7 weeks (%) 0 25 50 75

1421 1395 1178 1128

— 2 17 21

1.53 1.52 1.62 1.62

5.8 2.8 10.6 13.9

Notes: Chicks were kept for 4 weeks at different levels of ammonia, and were then raised to 7 weeks in clean air. Data represent the 4 weeks, except that deaths were calculated for the full 7 weeks. ‘Depression in body weight’ is the per cent reduction in 4-week weight compared to birds kept at 0 ppm ammonia. Feed/gain represents the number of grams of feed consumed for each gram of live body weight gain. Data are from Miles et al. (2004).

A similar approach has been used to recommend minimum space allowances for laying hens. At the end of the twentieth century most hens in the industrialized countries were kept in cages that contained small groups typically consisting of three to ten birds. After the use of cages had become widespread, agricultural scientists did literally dozens of studies to compare the productivity of hens with different amounts and conﬁgurations of space. Thirty such studies were summarized by United States poultry scientists Albert Adams and James Craig. They divided space allowances into three categories: high density (averaging 310 cm2 per bird), medium density (387 cm2) and low density (516 cm2). Increasing the crowding from low to medium density resulted in the mortality rate rising from 7.7% to 10.5% during a laying cycle (roughly one year), and the birds laid 7.8 fewer eggs on average. Increasing the density further from medium to high resulted in the mortality rate rising to 15.8% and an additional reduction of 16.6 eggs laid.19 These ﬁndings, made in the 1980s, did not result in the egg industry in the United States adopting animal welfare standards at that time. Twenty years later, however, United Egg Producers – an association of egg producers in the United States – commissioned a new review of research by an expert committee. The committee, largely agreeing with the earlier study, concluded that below about 465 cm2 per bird in the cage (with a range of about 430–550 cm2 depending on the size of the birds and the shape of the cages) egg production falls and mortality rate increases. The committee also found that access to the feeder is a critical factor. They noted that because of competition among hens, ‘lower-ranking hens are unable to consume sufﬁcient feed or are unable to consume feed at the time they are most highly motivated to do so, if the feeder space is inadequate’. Thus they recommended that birds should have about 465 cm2 of ﬂoor space and 10 cm 19Adams, A.W. and Craig, J.V. 1985. Effect of crowding and cage shape on productivity and proﬁtability of caged layers: A survey. Poultry Science 64: 238–242.

Dfraser_C005.indd 97

6/27/2008 12:42:44 PM

98

Studying Animal Welfare

of trough space per bird so that harmful effects on production and survival would be avoided. These recommendations then formed the basis of a system of animal welfare standards endorsed by the producer association.20 Studies have also used the rate of reproduction of animals to indicate environmental problems. As noted earlier, sows on many farms are conﬁned throughout most of pregnancy in gestation stalls that allow the animals to be fed individually without ﬁghting over food, but are too small for them to walk, socialize normally, or even turn around. A similar effect is sometimes achieved by tethering the animals by a short chain attached to a strap around the neck or torso. To compare these housing methods, John McGlone and co-workers21 kept sows during two pregnancies and lactations either in stalls or in tether systems. The tethered sows, in addition to showing various differences in behaviour, produced smaller litters than the sows in stalls. On this basis McGlone and co-workers recommended against the use of the tether system for both welfare and economic grounds, and this advice has been widely adopted in industry recommendations.22 Similar approaches have been used to some extent in non-domestic species. Small cats such as servals (Felis serval), ocelot (F. pardalis) and caracals (F. caracal) are often kept in zoos, but their breeding success is inconsistent for reasons that could reﬂect deﬁciencies in diet, environment, handling, or other factors. Zoo biologist Jill Mellen collected data on 154 small cats of 20 species from eight zoos. Some of the cats bred successfully, producing up to two litters per year, while others did not breed at all. Statistical analysis showed no evidence of differences in breeding success arising from some of the obvious housing and management factors such as the size of the enclosure, the presence of additional den sites, or a varied diet. However, the amount of contact with other cats proved to be important: Mellen concluded that for these mainly solitary species, the best reproduction could be achieved by housing the animals singly but with periodic breeding opportunities, whereas keeping three or more cats permanently together was the least natural and least successful option. The behaviour of the human caretakers was also important: breeding was more successful in zoos where caretakers spent considerable time with the cats, talking to the animals and soliciting interaction. Mellen suggested that for successful reproduction ‘cats must learn to “feel comfortable” with the inevitable presence of the human caretakers’, and she suggested that the policy of some zoos of minimizing interaction of caretakers with the cats is misguided.23 20Bell,

D., Chase, B., Douglass, A., Hester, P., Mench, J., Newberry, R., Shea-Moore, M., Stanker, L., Swanson, J. and Armstrong, J. 2004. UEP uses scientiﬁc approach in its establishment of welfare guidelines. Feedstuffs 76: 1–9. The quotation is on page 3. 21McGlone, J.J., Salak-Johnson, J.L., Nicholson, R.I. and Hicks, T. 1994. Evaluation of crates and girth tethers for sows: Reproductive performance, immunity, behavior and ergonomic measures. Applied Animal Behaviour Science 39: 297–311. 22For example: Connor, 1993. 23Mellen, J.D. 1991. Factors inﬂuencing reproductive success in small captive exotic felids (Felis spp.): A multiple regression analysis. Zoo Biology 10: 95–110. The quotation is on page 106.

Dfraser_C005.indd 98

6/27/2008 12:42:45 PM

Disease, Injury and Production

99

Table 5.2 Correlation coefficients, based on farm averages, between the frequency of negative behaviour by human caretakers (slaps, pushes, hits and tail-twists) and (1) yearly milk yield of the cow, (2) two measures of cow behaviour in an approach test, and (3) average levels of cortisol in the milk. Measures used in the approach test were the time taken by cows to approach within one metre of a seated person, and the percentage of cows that did not approach within one metre of a seated person in three minutes. Data are based on Tables 4 and 5 of Hemsworth et al., 2000. Variable Yearly milk yield Approach test Time taken to approach within 1 m Percentage of cows that did not approach within 1 m Cortisol in milk P

r 0.36 0.30 0.33 0.34

 0.01.

The effect of human caretakers on the basic biological functioning of animals has been emphasized in extensive research by Australian animal scientist Paul Hemsworth and his co-workers. In one of many studies, Hemsworth’s team used 66 commercial dairy herds to monitor the behaviour of the human staff, fear reactions by the cattle, and the animals’ milk yield. Statistical analysis showed that milk yield, which ranged from 3447 to 7141 litres per cow per year, was signiﬁcantly lower on those farms where the workers showed a large amount of negative behaviour (slaps, pushes, hits and tail-twists) when moving the animals (Table 5.2). The cows’ level of fear of humans was measured in a standardized ‘approach test’ in which animals were moved brieﬂy to an enclosure where they encountered a seated person. The average time taken by the cows to approach within one metre of the person, and percentage of animals that did not approach within one metre of the person, were taken as indicators of fear. Negative behaviour by the workers was correlated with higher levels of fear in the cows, and with higher levels of the stress-related hormone cortisol in the milk. Based on this and many other studies, Hemsworth and co-workers proposed that a high level of fear of humans is an important factor in the welfare and productivity of many farm animals in situations where close contact with humans is inevitable.24

24Hemsworth,

P.H., Coleman, G.J., Barnett, J.L. and Borg, S. 2000. Relationships between human–animal interactions and productivity of commercial dairy cows. Journal of Animal Science 78: 2821–2831; Hemsworth, P.H. and Coleman, G.J. 1998. Human–Livestock Interactions: The Stockperson and the Productivity and Welfare of Intensively-Farmed Animals. CAB International, Oxford.

Dfraser_C005.indd 99

6/27/2008 12:42:45 PM

100

Studying Animal Welfare

In these various examples, reductions in growth, reproduction and survival reﬂect clear problems with normal functioning of the body, and there seems little question that animal welfare is reduced. In such cases, improving the basic functioning of animals by providing them with fresher air, more space, better access to feed, a more natural social grouping or less aversive handling must surely indicate some improvement in their welfare. But how far can we go in relying on such measures as indicators of animal welfare? This is a debate that has gone on for four decades as we can see in some of the contrasting historical positions quoted in Box 5.1. The Brambell Committee reported hearing many submissions claiming that the continued productivity of farm animals is clear evidence of good animal welfare; and when Ingvar Ekesbo began his work on the then-new methods of housing cows he encountered much the same view. Both Brambell and Ekesbo saw the arguments as over-simpliﬁed for reasons we will explore below, yet the claims continue to be made by some industry ﬁgures and were heard from some scientists well into the 1980s. For example, in a 1981 publication the Council for Agricultural Science and Technology in the United States, although acknowledging ‘some degree of conﬂict or trade-off’, claimed that the pursuit of proﬁt by animal producers leads ‘automatically’ to improved welfare (Box 5.1).25 As we have seen above, the argument linking productivity and animal welfare has merit in many contexts. There are abundant examples of animals living with inadequate access to water or pasture, animals lacking treatment for disease and injury, animals losing body condition during arduous journeys, and so on. Under all these circumstances, improvements in animal welfare would clearly be accompanied by improved productivity. In modern production, however, there are many complicating factors that cause a degree of uncoupling between welfare and productivity. One complicating factor, as we have seen, is genetic selection for production traits. Modern hens have been bred so strongly for egg production that they will mobilize calcium from their bones to create egg shell. This can lead to signiﬁcant weakness in the leg bones and a high frequency of broken bones when the birds are removed from their cages for slaughter.26 Genetic selection of beef cattle for very large muscles has produced certain breeds whose carcasses have high commercial value, but these breeds are prone to a variety of animal welfare problems: compared to other breeds they tend to have more difﬁculty at calving and poorer calf survival, and some animals react to heat stress with an excessive build-up of lactic acid which leads to rigidity of the muscles, sometimes to the point of paralysis.27 We have already encountered disease conditions that are made worse when very high levels of productivity put a strain on normal metabolic processes. In these cases, increases in productivity achieved by breeding animals for production traits seem to have occurred partly by pushing the body to extremes, at some risk to 25Some

years later, the same organization took up a more nuanced view. See CAST, 1997. T.G. and Wilkins, L.J. 1998. The problem of broken bones during the handling of laying hens – a review. Poultry Science 77: 1798–1802. 27Gregory, N.G. 1998. Animal Welfare and Meat Science. CABI Publishing, Wallingford, UK. 26Knowles,

Dfraser_C005.indd 100

6/27/2008 12:42:45 PM

Disease, Injury and Production

101

Box 5.1 Contrasting viewpoints on the relationship between productivity and animal welfare. Many witnesses have represented to us that the growth rate of an animal for meat or the egg production of a laying hen are the only reliable objective measures of their welfare. It is claimed that suffering of any kind is reflected by a corresponding fall in productivity. The argument is that in the absence of any scientific method of evaluating whether an animal is suffering, its continued productivity should be taken as decisive evidence that it is not. This is an oversimplified and incomplete view and we reject it. It is true that a satisfactory growth or egg production rate is a reliable guide to the welfare of the animal in certain respects – for example that it is being well fed – but it is inadequate in other respects. F.W.R. Brambell (chairman) (1965, pages 10–11) There is a widespread belief that the effect of the … environment can be judged simply from the animals’ production. This method is both deceptive and inexact. The state of health, like production, is regulated by many different factors, many of which affect both. Knowledge of these factors is however too incomplete to allow definite conclusions on the state of health to be drawn from production data. Ingvar Ekesbo (1966, page 7) The goal of maximum profitability pursued by animal producers (and others) leads automatically to improved welfare of both animals and humans. In the competitive free-enterprise system, maximum profitability cannot be achieved without careful attention to the welfare of animals. And the welfare of human consumers of animal products is increased by the greater quantities and lower prices of the products available for purchase. Nonetheless, one must recognize the existence of some degree of conflict or trade-off between animal welfare and human welfare because the combination of conditions that leads to the maximum profitability of an animal-production operation involving many animals is not necessarily the same as the combination of conditions that leads to the maximum welfare of the animals individually. Council for Agricultural Science and Technology (CAST) (1981, page 1) … it is in the best interests of the producer to treat his animals as well as possible to get the greatest economic return and, therefore, there really isn’t any basic conflict between the ethics and the economics of poultry production. Arnold S. Rosenwald (1981, page 578) Assessment of well-being in agricultural animals cannot include an estimate of any mental suffering because we still are unable to do such an evaluation. But such an assessment must take into account all possible physiologic, immunologic, behavioral, anatomic, and agricultural-performance indicators of stress and distress. Meanwhile, health, reproductive, and productive traits continue to be the most reliable farm-level indicators of fit between agricultural animals and their environments. Stanley E. Curtis (1987, page 373)

Dfraser_C005.indd 101

6/27/2008 12:42:45 PM

102

Studying Animal Welfare

normal health. Further gains in productivity achieved in this way are likely to be at the expense of animal welfare, not in support of it.28 Similar complications arise from the use of hormones and other interventions to push productivity to very high levels. The Canadian veterinary panel, cited above, noted that BST clearly increased milk yield, but this was associated with an increased risk of mastitis and lameness because the increased metabolic activity can put further strain on an already strained metabolism. Even antibiotics, which have been of enormous beneﬁt for the health of both humans and animals, have proven to be a two-edged sword in the case of farm animal welfare. When used therapeutically, of course, antibiotics can cure such painful conditions as mastitis, obviously to the beneﬁt of the animal’s welfare. Even when used preventively, the net result may be strongly positive for animal welfare. For example, when piglets are weaned by being removed suddenly from the mother, they often experience signiﬁcant digestive problems and may develop serious diarrhoea and weight loss lasting several days. If they are given feed medicated with antibiotics, these problems can be greatly reduced.29 But is this truly a welfare advantage, or does it merely allow a stressful practice to be perpetuated? For example, beef calves are sometimes removed abruptly from the cow before their normal weaning age, transported to auctions, mixed with unfamiliar calves, and then transported again to a feedlot. Presumably this is highly stressful for the animals, and it can cause digestive and respiratory problems lasting for days. However, with the use of antibiotics to prevent these undesirable outcomes, such practices can still be commercially advantageous. Perhaps most importantly, productivity should not be confused with proﬁt. Proﬁt requires a certain level of productivity, but proﬁt can also be increased by limiting input costs. Reducing space allowance, staff time, bedding, veterinary care and other amenities can help to reduce costs; and even if these cut-backs reduce productivity to some extent, the net result may still be greater proﬁt overall. A striking example was provided by the study of Adams and Craig, reported above, whose analysis of space allowances for hens also included calculations of the proﬁtability associated with the different levels of crowding under market conditions in the United States. The analysis showed that especially if egg prices are high and feed costs are low, then proﬁt can often be increased by adding extra birds to a facility so that crowding is more severe, even though the death rate is increased and the birds’ individual rate of lay declines.30 28Rauw, W.M., Kanis, E., Noordhuizen-Stassen, E.N. and Grommers, F.J. 1998. Undesirable side effects of selection for high production efﬁciency in farm animals: A review. Livestock Production Science 56: 15–33. Other useful sources on the genetic aspects of animal welfare are found in Van Zutphen, L.F.M. and Bedford, P.G.C. (editors). 1999. Genetics and Animal Welfare. Animal Welfare (Special Issue) 8: 309–438. 29For example: Kyriakis, S.C., Bourtzi-Hatzopoulou, E., Alexopoulos, C., Kritas, S.K., Polyzopoulou, Z., Lekkas, S. and Gardey, L. 2002. Field evaluation of the effect of in-feed doxycycline for the control of ileitis in weaned piglets. Journal of Veterinary Medicine, Series B 49: 1439–1450. 30Adams and Craig, 1985.

Dfraser_C005.indd 102

6/27/2008 12:42:45 PM

Disease, Injury and Production

103

It is perhaps an indication of the success of modern animal production that animals are often functioning not at the lower end of their productivity range, where productivity is likely to be increased by solving basic problems of health and nutrition, but at the upper end where increases in productivity are no longer likely to improve animal welfare. At these levels of productivity, the relationship between welfare, productivity and proﬁt is too complex for any simple inferences to be drawn. IN THIS CHAPTER WE have looked at some of the most basic aspects of the health and biological functioning of animals and what they tell us about animal welfare. On the surface, diseases, injuries and reductions in growth and reproduction would seem to be straightforward indicators of welfare problems, and often they are. But we have also seen that caution is needed in drawing conclusions about animal welfare even from such seemingly straightforward evidence. Now we will turn to cases where the link between bodily changes and animal welfare is far from straightforward.

Dfraser_C005.indd 103

6/27/2008 12:42:45 PM

‘Stress’

6

When research on animal welfare began, many scientists tried to link the concept of welfare to some existing body of scientiﬁc theory. A prime candidate was the biological theory of ‘stress’. The word ‘stress’ refers (in everyday speech) to ‘physical, mental or emotional strain or tension’.1 Hence, demonstrating that an animal is in a state of ‘stress’ would seem to provide clear evidence of a welfare problem, and conversely the absence of ‘stress’ might be taken as an indicator that no major welfare problem exists. Moreover, inﬂuential physiologists had theorized that the body has a set of characteristic changes in the nervous and endocrine systems that occur in response to all manner of ‘stress’. Could it be, then, that by measuring these physiological changes, we could in effect ‘measure’ animal welfare? IN THE EARLY 1900s, the American physiologist Walter Cannon proposed that a wide range of emotional states bring about a general activation of the body which prepares the individual for exertion.2 Cannon had been impressed by the inﬂuence of affective states on the process of digestion. The classic studies of Ivan Pavlov and others had shown that the production of saliva and other digestive ﬂuids is not caused simply by the presence of food in the mouth and stomach, but often begins when an animal ﬁrst detects food or even a neutral stimulus (such as a bell) that has been paired with food in the past. Cannon concluded that the digestive processes are not unconscious reactions but rather ‘psychic’ phenomena which depend in part on ‘pleasurable excitation’ caused by the anticipation and enjoyment of eating. But whereas these pleasurable feelings help to trigger the processes of digestion, Cannon noted that negative emotions can bring the same processes to an end.

1Random

House Dictionary of the English Language, College Edition. Random House, New York, 1968. 2Cannon, W.B. 1929. Bodily Changes in Pain, Hunger, Fear and Rage, 2nd edition (1st edition 1915). Appleton-Century Co, New York. The phrase ‘pleasurable excitation’ appears on page 10.

104

Dfraser_C006.indd 104

6/27/2008 12:42:57 PM

‘Stress’

105

He reported one experiment in which the arrival of food led to copious secretion of digestive ﬂuids by a dog, but when a cat was brought into the laboratory to ‘infuriate’ the dog, all production of digestive juices ceased, and did not resume for many minutes after the cat had been removed and the dog had been paciﬁed. Such observations led Cannon to focus his attention on the sympathetic (literally, ‘with feeling’) nervous system. Unlike the nerves that enervate the voluntary muscles of the arms and legs, and which are long extensions of nerve cells originating within the central nervous system, nerves exiting the central nervous system in the thoracic and lumbar regions end in ganglia located close to the spinal cord. These nerves activate other nerves which originate in the ganglia and travel diffusely throughout much of the body. This set of nerves constitutes the sympathetic nervous system, and it causes a wide range of effects including an increase in heart rate, increased blood ﬂow to the skeletal muscles, improved contraction of fatigued muscles, and (as we have seen) reduction or cessation of the digestive processes (Figure 6.1). Complementing the sympathetic nervous system is the action of the medulla (or central portion) of the adrenal gland. The adrenal medulla is itself composed of modiﬁed nerve cells. Whereas the nerves of the sympathetic nervous system release the chemical norepinephrin (‘noradrenalin’), the cells of the adrenal medulla secrete the slightly modiﬁed chemical epinephrin (‘adrenalin’) which is released into the blood stream and travels throughout the body causing many of the same changes. Thus the sudden appearance of the cat or other stimuli that evoke the strong emotions of rage or fear would trigger two physiological responses with similar effects: activation of the sympathetic nervous system resulting in the release of norepinephrin in many parts of the body, and activation of the adrenal medulla resulting in release of epinephrin into the blood stream. In Cannon’s day this generalized response was sometimes termed the ‘ﬁght or ﬂight’ response; in drier modern terminology it is sometimes called the ‘SAM’ response, named for the sympathetic nervous system and adrenal medulla.3 Although Cannon used the term ‘stress’ in reporting his work on the SAM system, and although this system is correctly regarded as a key physiological response to emergencies and other demands on the body, the term ‘stress’ has come to be closely associated with the work of the Austrian-Canadian physiologist Hans Selye. As a medical student, Selye had been impressed by the common features that occur in many different illnesses. He noted that although the different disease pathogens have their speciﬁc effects (which were the focus of medical attention because they could be used to diagnose the illness), many illnesses also produce a set of very similar symptoms: diffuse aches and pains, fatigue, loss of appetite and a general sense of being unwell. These Selye later called ‘the syndrome of just being sick’.4 3Sapolsky, R.M. 1992. Neuroendocrinology of the stress-response. Pages 287–324 in Behavioral

Endocrinology (J.B. Becker, S.M. Breedlove and D. Crews, editors). MIT Press, Cambridge. 4Selye, H. 1956. The Stress of Life. McGraw-Hill Book Co, New York. The quotation is on page 16.

Dfraser_C006.indd 105

6/27/2008 12:42:58 PM

106

Studying Animal Welfare

CRH Hypothalamus Pituitary gland

TH AC

Adrenal cortex Adrenal medulla

>

Sympathetic ganglia > > >

> > >

Thoracic region

Epinephrin Glucocorticoids

Norepinephrin

> > > > > > > > > >

Lumbar region

Figure 6.1 Schematic drawing of the SAM and HPA response systems. In the SAM system (for sympathetic nervous system and adrenal medulla), nerves exiting the central nervous system in the thoracic and lumbar regions end in the sympathetic ganglia. There they activate nerves of the sympathetic nervous system which are widely distributed throughout the body and release the neurotransmitter norepinephrin. In addition, nerves travel directly from the central nervous system to the adrenal medulla where they stimulate release of epinephrin into the blood. In the HPA system (for hypothalamus–pituitary–adrenal cortex), the hypothalamus (in the brain) releases the hormone CRH (corticotrophin releasing hormone), which causes the pituitary gland to release the hormone ACTH (adrenocorticotrophic hormone) into the blood. This in turn causes the adrenal cortex to release glucocorticoids into the blood. Drawing by Vicky Earle, University of British Columbia.

Soon after he began his long career in medical research, Selye discovered that injections of various toxins or impure preparations of hormones, in addition to whatever speciﬁc effects they might have on the body, reliably produce three additional, general reactions: ulcers of the stomach and upper intestine, shrinking of the thymus and lymph nodes, and enlargement of the cortex (the outer portion)

Dfraser_C006.indd 106

6/27/2008 12:42:58 PM

‘Stress’

107

of the adrenal glands. By the late 1940s, Selye’s research on these non-speciﬁc reactions led him to propose that in addition to the SAM system which Cannon had studied, the body has a second major response system by which it reacts to virtually all types of challenges. In this system, as we now understand it, the hypothalamus (in the brain) releases a hormone called CRH (for corticotrophin releasing hormone), which causes the anterior pituitary gland to release a second hormone called ACTH (for adrenocorticotrophic hormone) into the blood. This in turn causes the cortex of the adrenal gland to increase its output of steroid hormones called glucocorticoids (Figure 6.1).5 The system described by Selye came to be called the HPA system for the hypothalamus, pituitary and adrenal cortex.6 In reporting his work, however, Selye adopted the term ‘stress’ which he deﬁned as ‘the nonspeciﬁc response of the body to any demand made upon it’,7 and Selye’s highly inﬂuential writing strongly emphasized the HPA response system as playing a central role in ‘stress’. Activation of the HPA system has a wide range of effects which, for the most part, appear to prepare the body for sustained exertion. An increased secretion of glucocorticoids, together with other related responses to challenge, reduce the secretion of insulin and thus help to prevent glucose in the blood from being stored in the form of glycogen and fat. At the same time, they help to create glucose from stored glycogen and thus make energy available for immediate use. In addition, activation of the HPA system reduces the secretion of sexual hormones and the sensitivity of the ovaries and testes to the inﬂuence of these hormones. Activation of the HPA system also involves lowered secretion of growth hormone and other hormones related to growth. The net result is that storage of energy is reduced, liberation of energy is increased, and the long-term processes of reproduction and growth are held back in favour of making bodily resources available for use. In addition, activation of the HPA system has a number of effects that reduce the activity of the immune system.8 Selye’s work stimulated an entire industry of scientiﬁc studies of ‘stress’, much of it focused on the HPA system. Since that time our understanding of ‘stress’ has undergone many developments. One development has been an appreciation of the vast complexity of the body’s responses to challenge. The ‘stress system’, as described by George Chrousos in a 1997 5The generic term ‘corticosteroids’ is used for hormones produced by the cortex of the adrenal gland and having a steroid structure, as do testosterone and oestrogen. The corticosteroids most involved in the HPA response are ‘glucocorticoids’, so called because they inﬂuence the metabolism of glucose and related nutrients. Cortisol and corticosterone are the most typical glucocorticoids in mammals and birds, and the levels of these hormones in blood plasma are the most commonly used measures of activity of the HPA system. 6Sapolsky, 1992. 7Selye, H. 1974. Stress without Distress. McClelland and Stewart, Toronto. The quotation is on page 27. 8For a very readable introduction to this area see Sapolsky, 1992.

Dfraser_C006.indd 107

6/27/2008 12:42:59 PM

108

Studying Animal Welfare

memorial lecture for Selye, includes multiple sites and pathways within the brain and a wide range of neurotransmitters and other chemicals produced by the brain, the pituitary gland, the sympathetic nervous system, the adrenal gland and other tissues. Moreover, the stress system has complex interactions with virtually all major endocrine systems of the body; it has a wide range of effects on the cardiovascular system, the respiratory system, the digestive system, the immune system, and the reproductive system; and (Chrousos argues) it inﬂuences an impressive array of diseases and other conditions including depression, hypertension, anorexia, obesity, osteoporosis, dwarﬁsm, reproductive failure and auto-immune diseases.9 A second development has been a growing understanding of how injury and infection trigger and inﬂuence responses of the HPA system. Beginning in the 1970s, immunologists discovered that certain cells produced by the immune system at a site of infection release a group of protein molecules called ‘cytokines’, which help to activate a general immune response. Cytokines are carried by the blood to lymphatic tissue and bone marrow where they stimulate the production of white blood cells, and thus promote inﬂammation and ﬁght infection. At the same time, the inﬂammation-enhancing cytokines trigger responses in the brain, which result in production of corticotrophin releasing hormone and thus activate the HPA component of the stress system. The cytokines have another important effect. Acting within the brain, certain cytokines produce the diverse group of symptoms – headache, fever, muscular aches, fatigue and loss of appetite – which characterize a wide range of illnesses. Thus, the ‘syndrome of just being sick’, which Selye had observed as a student, presumably results from the action of cytokines produced by the body in response to a wide range of diseases and infections. This syndrome is now regarded not simply as weakness and debility caused by disease, but as an evolved and adaptive affective state (‘malaise’), which motivates a diseased individual to rest, keep warm, and otherwise act in ways that are beneﬁcial in ﬁghting infection.10 A third key development has been an understanding of the role of affective processes in activating the HPA system. In a lengthy series of experiments, endocrinologist John Mason exposed animals, especially rhesus monkeys (Macaca mulatta), to a wide range of unpleasant experiences including heat, cold, restraint and forced exercise, all of which, by Selye’s logic, could be expected to produce a non-speciﬁc stress response. Mason found that whether he suddenly increased the temperature in a monkey’s cage from 20 to 30C, or suddenly decreased it to 10C, the animal in both cases showed a pronounced HPA response as evidenced by an increase in a glucocorticoid metabolite in the urine. Mason reasoned, however, that this could have been due not to the challenge to the body posed by the 9Chrousos, G.P. 1998. Stressors, stress and neuroendocrine integration of the adaptive response. The 1997 Hans Selye Memorial Lecture. Annals of the New York Academy of Sciences 851: 311–335. 10Dantzer, R. 2001. Cytokine-induced sickness behavior: Where do we stand? Brain, Behavior, and Immunity 15: 7–24.

Dfraser_C006.indd 108

6/27/2008 12:42:59 PM

‘Stress’

109

temperature itself, but to the sudden and unexpected nature of the change. When Mason took care to change the temperature very gradually over many days, then HPA and SAM activity increased in the cold (as might be expected because of the increased energy demand for thermoregulation) but declined in the heat.11 Figure 6.2, taken from a paper by French physiologists Robert Dantzer and Pierre Mormède, nicely illustrates the principles of the model that Mason proposed. These ﬁndings have important implications for Selye’s view that activation of the HPA system and related responses constitute a non-speciﬁc reaction to challenge. Selye had written, It is difﬁcult to see how such essentially different things as cold, heat, drugs, hormones, sorrow and joy could provide an identical biochemical reaction in the body. Nevertheless, this is the case; it can now be demonstrated, by highly objective quantitative biochemical determinations, that certain reactions are totally non-speciﬁc, and common to all types of exposure.12

Not so, argued Mason. Instead of different types of challenge producing the same non-speciﬁc activation of the HPA system, they actually produce quite distinct endocrine responses. It is only the animal’s short-term psychological reaction to unexpected change that causes different types of adversity to have the same initial effect. Mason readily acknowledged that certain physical stimuli clearly result in an HPA response; he cited cold, hypoxia and haemorrhage, and we could add many types of illness and injury. He argued, however, that in many cases the seemingly non-speciﬁc response to diverse challenges is due to a temporary, emotional reaction to the sudden change that such treatments often entail.13 Finally, a further body of research has shown that the degree to which an animal can control or predict an adverse situation plays a key role in determining the magnitude of the physiological response. In a grim but classic series of experiments performed around 1970, psychologist Jay Weiss restrained pairs of rats in an apparatus that delivered electric shocks to their tails, and he used ulceration of the stomach (one of Selye’s classic indicators) as a measure of the animals’ stress response. In one study, two rats received shocks at the same time, but one of the rats also received a warning signal to indicate that a shock was imminent. The animal that received unsignalled shock developed substantial ulceration of the stomach, whereas the rat that received the warning signal had very little. In another study, both rats received a

11Mason,

J.W. 1974. Speciﬁcity in the organization of neuroendocrine response proﬁles. Pages 68–80 in Frontiers in Neurology and Neuroscience Research 1974 (P. Seemand and G.M. Brown, editors). University of Toronto Press, Toronto. 12Selye, 1974, page 29. 13Mason, J.W. 1975a. A historical view of the stress ﬁeld. Part 1. Journal of Human Stress 1: 6–12; Mason, J.W. 1975b. A historical view of the stress ﬁeld. Part 2. Journal of Human Stress 1: 22–36.

Dfraser_C006.indd 109

6/27/2008 12:42:59 PM

110

Studying Animal Welfare

High temperature

Plasma cortisol level

Psychological response Specific response Baseline

Acute exposure Gradual exposure

Hours

Days

Time

Low temperature

Plasma cortisol level

Psychological response Acute exposure Gradual exposure Baseline

Specific response

Hours

Days

Time

Figure 6.2 A model proposed by Robert Dantzer and Pierre Mormède to explain changes in plasma cortisol levels when animals are exposed to high or low temperatures in two different ways. With gradual exposure, cortisol levels decline when environmental temperature is high (above) and increase when it is low (below). A sudden change to either hot or cold temperatures is followed by a sudden rise in plasma cortisol because of the animals’ ‘emotional reaction’ to the rapid change. The graphs (based on results for cattle) have been redrawn after Dantzer and Mormède (1983), with the kind permission of Journal of Animal Science.

warning signal before the shock was delivered, but one of the two could prevent the shock (to both animals) by turning a wheel. Although both rats received exactly the same shocks, the ‘helpless’ rat developed more ulceration of the stomach than the rat that had some means of control. In such situations, Weiss concluded, ‘the psychological factors were the main cause of stomach ulcers and other disorders’.14 14Weiss, J.M. 1972. Psychological factors in stress and disease. Scientiﬁc American 226: 104–113. The quotation is on page 104.

Dfraser_C006.indd 110

6/27/2008 12:42:59 PM

‘Stress’

111

HOW HAVE THE ‘STRESS’ responses of the body been used in the study of animal welfare? Let us start by surveying some examples before attempting a critical analysis of the link between ‘stress’ and animal welfare. Recall from the Introduction to this book that a long-standing concern over the humaneness of hunting red deer with hounds and horses led the United Kingdom’s National Trust (which owns land where such hunts were carried out) to commission a study of hunting. In conducting the studies, zoologists Patrick Bateson and Elizabeth Bradshaw noted that whereas wolves (the deer’s natural predator) normally kill deer by ambush followed by a short chase, hunts conducted by mounted hunters with hounds can last for many hours. During this time, the deer can often be seen alternating between bursts of activity and periods of resting until the animal ﬁnally stops running and is killed by gunshot at close range. The key question for Bateson and Bradshaw was whether the level of distress and exertion associated with hunting causes the deer to suffer more than they would if their numbers were controlled by other means such as stalking the animals and shooting them with a riﬂe. With the cooperation of hunt organizers, the researchers observed hunts and collected blood samples from more than 60 freshly killed animals. As comparison groups they also collected blood from farmed red deer that had been killed suddenly by riﬂe shot, and from a sample of injured deer that were destroyed for humane reasons after being hit by cars, entangled in fences, or injured in other ways. The blood samples from deer killed after being chased by hunters and hounds showed a wide range of changes suggesting extreme exertion. The plasma of hunted animals became stained with haem, presumably from the rupture of red blood cells. In lengthy hunts, muscle enzymes in the blood reached very high levels likely because of rupturing of muscle cells. One muscle enzyme (creatine kinase) increased by more than ten-fold, and in some animals it reached levels that would be taken to indicate muscle pathology in race horses. Blood sugar fell to low levels and glycogen in the muscles became depleted – changes indicating that the animals had exhausted the energy available to the muscles and were incapable of further exertion. This was consistent with the behavioural observations of animals stopping to rest, even when in obvious danger. In addition to all this, levels of cortisol – the major glucocorticoid in deer and many other mammals – provided a particularly dramatic picture. The laboratory assay was capable of detecting cortisol down to 1 nanogram per millilitre (ng ml1) of blood plasma.15 Most of the farmed deer that were killed suddenly by riﬂe shot had cortisol levels below the detection limit of the test. The hunted animals showed a median of 71 ng ml1. This was perhaps 100 times higher than the non-hunted deer, and was substantially higher than the 44 ng ml1 for the injured deer that were killed for humane reasons.16 Based on such evidence, Bateson reported to as 2.7 nanomoles per litre of blood. I have converted all the values into ng ml1. P. and Bradshaw, E.L. 1997. Physiological effects of hunting red deer (Cervus elaphas). Proceedings of the Royal Society, London, B 264: 1707–1714; Bradshaw, E.L. and Bateson, P. 2000. Welfare implication of culling red deer (Cervus elaphus). Animal Welfare 9: 3–24. 15Reported 16Bateson,

Dfraser_C006.indd 111

6/27/2008 12:43:00 PM

112

Studying Animal Welfare

the National Trust that ‘lengthy hunts with hounds impose extreme stress on red deer and are likely to cause them great suffering’.17 In this example, the extremely large changes in cortisol indicated an unusually strong activation of the HPA system, but the precise causes of this activation remained unclear. Given that emotional states are important triggers of HPA activity, it seems very plausible that affective states – presumably fear, perhaps combined with feelings of exhaustion – played a large role. However, severe exercise by itself can produce an HPA response, and damage to muscle tissues might also have contributed. Hence, it is difﬁcult to be certain about the relative importance of psychological and other factors. The authors concluded simply that the evidence provided ‘a strong indicator of great physiological and psychological stress’.18 Moreover, the fact that most of the measures were more extreme for hunted deer than for injured deer that were killed for humane reasons made it difﬁcult to dispute that hunting imposed a signiﬁcant welfare cost. In contrast to the deer study, activation of the HPA system has also been used under more controlled conditions as an indicator of the distress caused by procedures that are presumed to be painful. Many dairy calves have their horn-buds removed so that the horn will not develop and become a threat to the safety of other animals and farm workers. The ‘scoop’ method of dehorning is done with two joined, semi-circular blades which close around the growing horn and cut it away, together with adjacent skin and some underlying bone. In some countries, such dehorning is considered a veterinary procedure and must be done with local anaesthesia. Elsewhere, however, dehorning of calves is normally done by non-veterinary staff without any form of pain management. A research team in New Zealand, wanting to test workable methods of pain reduction, used plasma cortisol levels to quantify the stress response associated with the pain of dehorning.19 They monitored plasma cortisol levels for three groups of calves: those that were dehorned without any form of pain management, those that received local anaesthetic to freeze the area beforehand, and a sham control group that were handled in a similar manner but not actually dehorned. The dehorning was quickly followed by a large increase in plasma cortisol if the procedure was done without local anaesthetic, but this cortisol response was largely prevented in calves that received local anaesthetic to freeze the area (Figure 6.3). This ﬁnding, combined with the lack of a cortisol response in the sham controls that were handled in the same way but not dehorned, made a very plausible case that the cortisol response in this situation served as an indirect indicator of the pain caused by the procedure. 17Bateson, P. 1997. The Behavioural and Physiological Effects of Culling Red Deer. Report to the Council of the National Trust, unpublished. The quotation is on page (Roman numeral) x. 18Bateson and Bradshaw, 1997, page 1712. 19Petrie, N.J., Mellor, D.J., Stafford, K.J., Bruce, R.A. and Ward, R.N. 1996. Cortisol response of calves to two methods of disbudding used with or without local anaesthetic. New Zealand Veterinary Journal 44: 9–14. I am using ‘dehorning’ as a generic term to include removal of horns or horn-buds.

Dfraser_C006.indd 112

6/27/2008 12:43:00 PM

113

Change in cortisol concentration (ng ml−1)

‘Stress’

40

30

20

10

0 −10

−2 −1

0

1

2

3 4 5 Time (h)

6

7

8

9

10

Figure 6.3 Change in plasma cortisol levels in calves after surgical dehorning. Results are shown for three groups of calves: calves dehorned without local anaesthetic (black squares), calves dehorned after receiving local anaesthetic (open squares), and a sham control group which were handled in a similar manner but not dehorned or anaesthetized (open circles). This figure has been redrawn from Petrie et al. (1996) Cortisol response of calves to two methods of disbudding used with or without local anaesthetic. New Zealand Veterinary Journal 44: 9–14. Copyright Elsevier.

However, a closer examination of Figure 6.3 reveals a second and more subtle result. About two hours after the dehorning, at the time when the freezing was presumably wearing off, the animals that had received the local anaesthetic began to show a slow, steady increase in cortisol lasting for several hours; and three to four hours after the procedure, blood cortisol for these animals was signiﬁcantly higher than for calves that received no local anaesthesia. In fact, the total production of cortisol during the experiment, estimated by calculating the area under the curves in Figure 6.3, was actually higher (in this experiment) for the calves that received local anaesthetic. What does this indicate? Presumably, once the anaesthesia had dissipated, the animals that had received it began to experience pain from the inﬂamed and injured area, and the onset of the pain could have activated the HPA response. For the animals that did not receive local anaesthetic, the initial release of cortisol may have helped to suppress the inﬂammation, and the natural pain-suppressing agents that are often secreted in conjunction with an HPA response may have helped to mitigate the sensation of pain. Hence, two hours after the procedure these animals might have been in less pain than those that received the local anaesthetic. Or alternatively perhaps the pain was just as severe for the untreated animals, but because it was diminishing rather than increasing, it did not cause as great an emotional response. The experiment nicely demonstrated that effective pain management for surgical dehorning

Dfraser_C006.indd 113

6/27/2008 12:43:00 PM

114

Studying Animal Welfare

requires an additional treatment such as a long-acting analgesic and anti-inﬂammatory drug to manage the pain as the local anaesthetic wears off.20 However, the example also shows the care that is needed in drawing conclusions about animal welfare from measures of HPA system activation. If we were to simply equate HPA activity with poor welfare, might we have to conclude that the freezing did not improve, or even worsened, the welfare of the calves that received it? Hunting and dehorning are examples of short-term inﬂictions on animals, but activation of the HPA system has also been used to study welfare problems caused by longer-term housing and management methods. Clouded leopards (Neofelis nebulosa) are widely kept in zoos, partly because their wild populations are dwindling and captive breeding may play an important role in the future conservation of the species. In captivity, however, these cats often perform seemingly abnormal behaviour such as excessive pacing, and some inﬂict injury on themselves by chewing their tails or plucking their fur. Believing that these behavioural problems may reﬂect poor welfare among clouded leopards, zoo biologist Nadja Wielebnowski and co-workers wanted to test whether the undesirable behaviour was correlated with increased HPA system activity, and to ﬁnd what housing and husbandry variables might contribute to the problems.21 As a non-invasive measure of HPA activity the team used the concentrations of glucocorticoid metabolites in samples of faeces collected from 72 clouded leopards housed in 12 different zoos. For each animal they also collected information on the size and location of the animal’s enclosure, how the animals were handled, and the keepers’ assessments of the animals’ behaviour. The faecal concentrations of glucocorticoid metabolites showed the expected relationship with behavioural indicators of agitation in the leopards. The concentrations were positively correlated with the keepers’ ratings of how often the animals seemed ‘tense’ and how much they paced, hid and slept. Moreover, about half of the cats were reported to damage themselves by fur-plucking and tailchewing, and these animals had higher average concentrations of glucocorticoid metabolites than the others. The team then used the metabolite concentrations to compare different aspects of how the animals were housed and managed. There were three particularly interesting results. First, there were signiﬁcant correlations between concentrations of faecal glucocorticoid metabolites and the location of the animals’ housing. Leopards that were on public display had higher concentrations than those that were not exposed to zoo visitors, and leopards housed within sight of larger predators 20Stafford, K.J. and Mellor, D.J. 2005. Dehorning and disbudding distress and its alleviation in calves. The Veterinary Journal 169: 337–349. 21Wielebnowski, N.C., Fletchall, N., Carlstead, K., Busso, J.M. and Brown, J.L. 2002. Noninvasive assessment of adrenal activity associated with husbandry and behavioral factors in the North American Clouded Leopard population. Zoo Biology 21: 77–98.

Dfraser_C006.indd 114

6/27/2008 12:43:00 PM

‘Stress’

115

which might pose a danger for them (lions, tigers) showed higher concentrations than those that were not. Second, the physical properties of the enclosure proved to be important. The ﬂoor space available to the leopards, which ranged from 100 to nearly 700 m2, showed no relationship with concentrations of glucocorticoid metabolites, but enclosure height, ranging from about two to six metres, was strongly related. Speciﬁcally, high enclosures, where cats could climb and perch well above ground level (a natural defensive behaviour for the species), were associated with lower glucocorticoid metabolites than low enclosures where climbing was limited. Finally, the keepers played an important role. The different leopards had between two and seven regular keepers. Statistical analysis showed that the more different keepers were involved, the higher the concentrations of glucocorticoid metabolites. However, the largest effect had to do with the amount of contact with the keepers. The primary keepers spent from 7 to 32 hours per week interacting with the leopards, and where individual contact time was high, concentrations of glucocorticoid metabolites were relatively low. In this study, activation of the HPA system (as reﬂected by faecal glucocorticoid metabolites) seemed to be a plausible indicator of the welfare of the animals, as levels were related to pacing, hiding, self-destructive behaviour and to the keepers’ ratings of how often the animals seemed ‘tense’. The results also identiﬁed features that may help calm the animals. These included keeping the cats in enclosures that allow them to perform their natural behaviour of climbing and perching well above ground; sheltering the cats from the public and from other predators that might pose a danger to them; and ensuring that the cats are handled by relatively few keepers who spend considerable time with the animals. A second example of long-term activation of the HPA system comes from the extensive work of Australian physiologist John Barnett and colleagues on HPA activation and other responses of pregnant sows to different housing systems. As we have seen, many sows are conﬁned throughout most of pregnancy, either in individual gestation stalls or by the use of tethers. There are also various forms of loose housing for groups of sows; these give the animals more freedom of movement, but also permit aggression whose severity depends on group size, the degree of competition over food, and other factors. In a series of studies Barnett and co-workers housed sows in individual stalls, tethers and various forms of loose housing, and they monitored HPA system activation by determining levels of ‘free’ corticosteroids22 in samples of blood. In one 22Of the cortisol and related compounds (generically, ‘corticosteroids’) circulating in the blood, a certain percentage, which varies over time and between species, is bound to a protein, and the remainder is unbound or ‘free’. The corticosteroids are thought to be biologically active only when in the free state. There is a debate over whether animal welfare research should measure levels of free, biologically active corticosteroids (Barnett’s view) or whether total corticosteroids provide a more comprehensive indicator of activation of the HPA system (Mormède et al., 2007) .

Dfraser_C006.indd 115

6/27/2008 12:43:00 PM

Free corticosteroid

3.0

Glucose 4.0

2.5

3.0

2.0 1.5

2.0

1.0 1.0

0.5

0.0 Paddock

Group

Stalls

Tethers

Paddock

Group

Stalls

Tethers

0.0

Mean concentration (mmol l −1) glucose

Studying Animal Welfare

Mean concentration (ng ml−1) corticosteroid

116

Figure 6.4 Results of blood samples taken from pregnant gilts (i.e., sows in their first pregnancy) in four housing treatments: neck tethers, individual gestation stalls, a group pen with six animals (all of these treatments were indoors with a concrete floor), and an outdoor paddock. Results show plasma levels of free corticosteroid and glucose. Data are from Tables 3 and 5 of Barnett et al. (1985).

study, blood samples collected 18 and 46 days after the animals had been moved into the various housing systems showed signiﬁcantly higher levels of free corticosteroids in animals kept in tethers compared to all other systems (Figure 6.4). This was accompanied by higher levels of blood glucose in these same animals, a ﬁnding that could have been due to corticosteroids making energy available for immediate use. Based on this and other evidence, the team concluded that the tether system causes ‘a chronic stress response and a signiﬁcant metabolic cost’.23 However, later observations by the same group suggested that the greater HPA response in the tethers was actually due to a speciﬁc design feature of the system they had studied, rather than to tethering in general. In the tether system used for this research, the short partitions that separated the heads of the sows from their neighbours were relatively open and allowed substantial aggression to occur between adjacent animals. In a follow-up study, Barnett’s group included an additional treatment consisting of the tether system with the partitions covered so as to prevent aggressive behaviour through the partitions. In this modiﬁed system there was virtually no aggression, and the level of plasma corticosteroids was no higher than in other housing systems. The authors concluded that the reduced

23Barnett,

J.L., Winﬁeld, C.G., Cronin, G.M., Hemsworth, P.H. and Dewar, A.M. 1985. The effect of individual and group housing on behavioural and physiological responses related to the welfare of pregnant pigs. Applied Animal Behaviour Science 14: 149–161. The quotation is on page 149.

Dfraser_C006.indd 116

6/27/2008 12:43:01 PM

‘Stress’

117

welfare they had seen resulted not from tethering itself but from the continual aggression that occurred in the original design.24 A third example of an HPA response to housing conditions comes from veterinary researchers Jan Ladewig and Diedrich Smidt, who carried out a detailed study of cortisol secretion in 16 young bulls.25 The bulls had been loose-housed in large pens on deep straw bedding until they were 14 months of age. Then, eight of the animals were moved from the relative comfort of the home pen to a set of stanchions formerly used for tying dairy cows. There the animals were tethered so that their movements were limited to a small area with unbedded concrete ﬂooring. Blood samples were collected by very gradual, automated withdrawal of blood throughout the day, yielding a sample every 20 minutes. The frequent sampling revealed the remarkable ﬂuctuations in blood cortisol that occur over the course of a day. For much of the time, levels hovered around 2–3 ng ml1. Then about ten times per day there was a secretory episode when levels rose to roughly 10 ng ml1. There was also a daily pattern, with low levels typically occurring in the evening before the lights were turned out, and higher levels in the early morning before the lights came on. When the bulls were ﬁrst moved to the unfamiliar and presumably uncomfortable stanchions, the secretion of cortisol increased many fold for some bulls, especially in the ﬁrst three days. Thereafter, however, the pattern and level of cortisol secretion returned to normal, and remained similar to the control bulls when the animals were retested after spending more than a month tied in the stanchions. Nonetheless, the bulls’ behaviour continued to indicate that they found the environment uncomfortable. In particular, the animals had fewer bouts of lying – a result that Ladewig and Smidt took to indicate that changing position was uncomfortable on the bare concrete. In this case, the higher levels and altered pattern of cortisol secretion appeared to be a transient effect due to the novelty of the presumably unwelcome change of environment, but did not reﬂect the (presumed) on-going discomfort of being tied for a month on a concrete ﬂoor. As a ﬁnal example, psychologist Michael Hennessy and co-workers used activation of the HPA system to monitor ‘stress’ among dogs in an animal shelter. The shelter received dogs relinquished by their owners together with strays and dogs seized by shelter staff. To assess ‘stress’ in the dogs in shelters, Hennessy’s team began with a simple study in which they determined plasma cortisol levels in blood samples collected from dogs on various days after admission. As a comparison group, the researchers collected samples from their own dogs and from dogs of friends, taken in the home situation where the dogs were presumably 24Barnett, J.L., Hemsworth, P.H. and Winﬁeld, C.G. 1987. The effects of design of individual stalls on the social behaviour and physiological responses related to the welfare of pregnant pigs. Applied Animal Behaviour Science 18: 133–142. 25Ladewig, J. and Smidt, D. 1989. Behavior, episodic secretion of cortisol, and adrenocortical reactivity in bulls subjected to tethering. Hormones and Behavior 23: 344–360.

Dfraser_C006.indd 117

6/27/2008 12:43:01 PM

Studying Animal Welfare

Plasma cortisol (ng ml−1)

118

40 30 20 10 0 1

2

3

4

5

Days in shelter

6–9

10+

Dogs in home

Figure 6.5 Levels (mean and standard error) of plasma cortisol in blood samples collected from dogs in an animal shelter, shown in relation to how long the dog had been in the shelter. Cortisol levels were highest for dogs studied on their first three days in the shelter. Dogs that had been in the shelter for 10 days or more had cortisol levels similar to those of a comparison group of dogs in their homes. Adapted from Figure 1 of Hennessy et al. (1997). Copyright Elsevier 1997.

comfortable and relaxed. The results (Figure 6.5) showed that plasma cortisol levels averaged about 25 ng ml1 for dogs sampled on their ﬁrst three days in the shelter. Thereafter, cortisol levels were generally lower. For dogs sampled after 10 days or more in the shelter, levels were about 10 ng ml1, similar to the levels seen in dogs in their homes.26 These data then provided the basis for further research to test whether petting by humans could help dogs adapt to the shelter environment. Petting (the research showed) does help to prevent the very temporary increase in cortisol caused by such short-term disturbance as taking a blood sample, but it does not mitigate the stress response occasioned by putting a dog in the shelter.27 IN CHOOSING THESE VARIOUS examples, I have (for sake of simplicity) selected studies that looked at activation of the HPA system rather than the SAM system, and that reported levels of cortisol or its metabolites as a principal measure of ‘stress’. There are, of course, many other approaches to quantifying the stress responses of the body. These include many measures of immune competence (because increased activation of the HPA system often suppresses the immune system), measures of changes in the action of glucocorticoid receptors in

26Hennessy, M.B., Davis, H.N., Williams, M. T., Mellott, C. and Douglas, C.W. 1997. Plasma cortisol levels of dogs at a county animal shelter. Physiology & Behavior 62: 485–490. 27Hennessy, M.B., Williams, M.T., Miller, D.D., Douglas, C.W. and Voith, V.L. 1998. Inﬂuence of male and female petters on plasma cortisol and behaviour: Can human interaction reduce the stress of dogs in a public animal shelter? Applied Animal Behaviour Science 61: 63–77.

Dfraser_C006.indd 118

6/27/2008 12:43:01 PM

‘Stress’

119

the body, and measures of the ability of the adrenal cortex to mount a strong glucocorticoid response to injections of the hormone ACTH. However, these few examples provide a basis for discussing the relevance of stress responses to animal welfare, without going into the technical complexities of what is now a huge scientiﬁc literature. The dehorned calves and hunted deer provide examples of a stress response to short-term disturbance. Given that affective states are powerful drivers of stress responses, it seems very plausible that the spike of cortisol secretion seen in the calves immediately after dehorning was due to pain caused by the procedure. The cortisol response then formed the basis for recommending that calves need both local anaesthetic and an analgesic in order to fully manage the pain. The deer, driven to exhaustion in a literally life-or-death situation, were presumably in a state of fear which, combined with extreme exertion and tissue damage, produced a massive physiological stress response; and because the response was more extreme than that seen even in injured deer that were killed for humanitarian reasons, it provided policy makers with the evidence they needed to justify banning this form of hunting. In both these cases we can think of stress responses as representing activation of bodily resources in response to some emergency registered through an affective state such as pain or fear, and perhaps through neural and chemical signals from bodily damage. The same logic – that the stress response represents an activation of the body for a real or potential emergency – may also apply to the longer-term studies. For clouded leopards in captivity, it seems plausible that constant exposure to lions, tigers, and unfamiliar zoo visitors could be a continuing cause of fear or anxiety. For these animals a natural means of self-protection is to climb to the relative safety of a tree branch, and the low cages that made this behaviour impossible may have contributed further to the animals’ sense of unease. Similarly, Barnett’s sows in the unmodiﬁed tether system were subjected to continual aggression from neighbours. Under less conﬁned conditions, the smaller or more timid animals would respond to such aggression by moving away, but the housing system forced them to remain close beside the aggressors. Thus, in both the leopards and the sows we can imagine that natural selection fashioned the animals to treat such situations as long-lasting emergencies meriting a continued mobilization of the body’s resources. Hence, it makes sense that these adverse conditions were accompanied by activation of the HPA system. Ladewig’s bulls and Hennessy’s dogs showed a very different pattern, with high cortisol secretion only during their ﬁrst days in the new surroundings. Again, the initial exposure to a strange and inhospitable environment could well have triggered an emotional state with a corresponding activation of the HPA system. But once the new situation had become familiar and routine, it seems plausible that the animals would no longer respond to it as one that required emergency mobilization of resources. Dogs kept long-term in the kennel may have been lonely and understimulated; the bulls tethered for a month on a bare concrete ﬂoor may

Dfraser_C006.indd 119

6/27/2008 12:43:01 PM

120

Studying Animal Welfare

have been chronically uncomfortable; but there was no emergency that required the HPA system to remain highly active.28 Clearly, activation of the HPA and SAM systems can provide valuable insights that arguably could not have been achieved in other ways, but we need to think carefully about how these biochemical changes relate to animal welfare. Were we simply to equate ‘stress’ responses with poor welfare (and, conversely, a lack of response with good welfare), we would be led to some counter-intuitive conclusions. We would recognize, of course, the welfare problems of the deer and the leopards, but we would need to conclude that the welfare of dogs kept long-term in rescue shelters is as good as that of pet dogs in their homes, and that the welfare of bulls tethered on an uncomfortable ﬂoor is as good as that of bulls loose-housed on deep straw. We might even have to conclude that local anaesthesia reduced the welfare of calves undergoing painful dehorning because total cortisol secretion was somewhat higher for calves that received anaesthesia than for those that did not. If, on the other hand, we see ‘stress’ responses as indicating an activation of the body in situations that the animal registers as requiring some active response, then we would conclude that stress responses denote certain kinds of welfare problems but not others. This line of reasoning also makes sense of another issue that arises in using HPA and SAM responses to assess animal welfare. We know from other work that these systems are activated not only by unpleasant situations that are likely to cause fear and pain, but by situations that are entirely natural for the animals or indeed by situations that animals may well seek out. For example, cortisol secretion is increased by exercise and by presumably pleasurable activities such as mating, nursing and exploration.29 All this ﬁts with the view of the SAM and HPA responses as making resources available for current or anticipated demands on the body, but not with the view that they necessarily denote welfare problems. The studies we have discussed also raise a quantitative question. For the dehorned calves, the immediate spike in blood cortisol levels represented a roughly 6-fold increase over baseline levels. The deer killed after being chased for hours had cortisol levels perhaps 100 times those of undisturbed animals that were killed instantly by riﬂe shot. In contrast, the sows exposed to continual aggression showed free corticosteroid levels about 1.5 times greater than sows in other treatments, and on their ﬁrst day in the rescue shelter Hennessy’s dogs had cortisol levels about 2.5 times greater than dogs in their own homes. These last two differences, although 28In a detailed review, Pierre Mormède and co-workers, 2007, propose that although longterm exposure to unpleasant situations may not affect the concentration of glucocorticoids in plasma, it may still lead to other changes in the HPA system, such as an increased responsiveness to short-term stressors. If this is correct, then it may be possible to use measures of the HPA system to capture problems such as long-term housing of dogs in shelters, but the measures would need to go beyond simple levels of glucocorticoids in the blood. 29Rushen, J. and de Passillé, A.M.B. 1992. The scientiﬁc assessment of the impact of housing on animal welfare: A critical review. Canadian Journal of Animal Science 72: 721–743.

Dfraser_C006.indd 120

6/27/2008 12:43:01 PM

‘Stress’

121

reliable results, are actually smaller than the changes that Ladewig and Smidt observed as normal ﬂuctuations over the course of a day as the body gears up for activity and gears down for rest. Given that changes in the HPA and SAM systems represent adaptive adjustments to varying demands on the body, is there a way to separate normal, adaptive ﬂuctuations from changes that are signiﬁcant for assessing the animal’s welfare? One proposal has been to identify a threshold value that separates normal adjustments from those that indicate a risk to animal welfare. Most notably, John Barnett and Paul Hemsworth reviewed research by their group and others, and concluded that if the concentration of free corticosteroids in blood plasma shows a sustained increase of 40% or more, then the animal is likely to incur some ‘detrimental consequences’. They recognized two types of detrimental consequences: (1) physiological consequences including reduced immune competence and changes in plasma levels of glucose and other nutrients, and (2) production-related consequences involving ‘growth rate, sexual behaviour or pregnancy rate’. Using this logic, they proposed that ‘a sustained elevation of 40% in free corticosteroid concentration’ constitutes evidence of a risk to welfare.30 This proposal has met with opposition from several scientists. Jeffrey Rushen questioned whether the evidence really supported a 40% increase as a threshold value. He argued that there have been many studies where crowding and other conditions led to detrimental consequences in variables such as rates of growth and survival without corresponding differences in plasma corticosteroid levels.31 Michael Mendl questioned Barnett and Hemsworth’s use of ‘detrimental consequences’ as determinants of animal welfare. In many cases, he argued, people think of welfare being reduced in cases where the animal is having difﬁculty adapting to its circumstances even though there might be no detrimental consequences in terms of growth, immune competence, or the other variables that Barnett and Hemsworth considered relevant to animal welfare.32 A second and more general proposal is that of physiologist Gary Moberg. As we saw in Chapter 4, Moberg noted that activation of the HPA and SAM systems is often a natural and adaptive response of the body, but if sufﬁciently intense or prolonged it can lead to pathological changes such as disease, failure of reproduction, or outbreaks of harmful behaviour. However, such breakdowns of normal biological functioning are often preceded by a ‘prepathological’ state; 30Barnett,

J.L., and Hemsworth, P.H. 1990. The validity of physiological and behavioural measures of animal welfare. Applied Animal Behaviour Science 25: 177–187. The quotations are from page 182. 31Rushen, J. 1991. Problems associated with the interpretation of physiological data in the assessment of animal welfare. Applied Animal Behaviour Science 28: 381–386. 32Mendl, M. 1991. Some problems with the concept of a cut-off point for determining when an animal’s welfare is at risk. Applied Animal Behaviour Science 31: 139–146. For other critiques of the use of physiological data to draw conclusions on animal welfare see Dantzer and Mormède (1983a), Rushen and de Passillé (1992), and Mormède et al. (2007).

Dfraser_C006.indd 121

6/27/2008 12:43:01 PM

122

Studying Animal Welfare

for example, sustained high activation of the HPA system can lead to reduced immune competence before any disease is observed, and to altered secretion of reproductive hormones before there is any reproductive failure. Moberg proposed (Box 4.2, p. 74) that a stress response ‘becomes a threat to well-being only when the stressor … results in a change in the animal’s biological function such that the animal enters the pre-pathological state with the ensuing vulnerability to pathology’. Thus, he proposed that ‘pre-pathological’ states be used to indicate the point where the body’s ‘stress’ responses constitute a risk to animal welfare.33 Moberg’s proposal avoids the problem of having to ﬁnd and defend a threshold value for any one variable, and thus it avoids some of the criticisms directed at the proposal of Barnett and Hemsworth. Note, however, that both proposals, in addition to linking the concept of welfare to ideas and measures derived from stress biology, involve a philosophical stance on what does and does not count as an animal welfare problem. For Barnett and Hemsworth, activation of HPA system can be said to indicate a risk to welfare only when it reaches a level that incurs a metabolic cost or otherwise impairs the functioning of the body. For Moberg the criterion is a ‘change in the animal’s biological function’ which creates ‘vulnerability to pathology’. These criteria would cover certain welfare problems where there is a threat to health, growth and reproduction; hence, welfare problems would be acknowledged in the case of the hunted deer, the agitated leopards, and the sows that were exposed to continual aggression. By these criteria, however, no welfare problem would be recognized in various other situations where people might well consider that an animal’s quality of life is reduced, for example in cases of long-term discomfort (the tethered bulls), boredom and loneliness (the shelter dogs), and acute pain (the dehorned calves), unless these were accompanied by some metabolic cost, change in health status, or other threat to basic health and functioning. THE CONFUSION OVER THE relationship of ‘stress’ to animal welfare has much to do with the controversial decision made by Hans Selye to use the term ‘stress’ to refer to ‘the nonspeciﬁc response of the body’ to any form of demand.34 As we have seen, ‘stress’ is a common English word that refers to ‘physical, mental or emotional strain or tension’,35 and it clearly implies something undesirable. When later work made it apparent that the HPA system is not speciﬁc to unpleasant states, but is invoked in both pleasant and unpleasant situations where some activation of the body is needed, Selye continued to use the word ‘stress’ for the general response, but he took to speaking of ‘eustress’ (literally, ‘good stress’) in cases 33Moberg, G.P. 1985. Biological response to stress: Key to assessment of animal well-being? Pages 27–49 in Animal Stress (G.P. Moberg, editor). American Physiological Society, Bethesda, USA. 34Selye, 1974, page 27. 35Random House Dictionary of the English Language, 1968.

Dfraser_C006.indd 122

6/27/2008 12:43:02 PM

‘Stress’

123

where the situation is pleasant and beneﬁcial, and ‘distress’ where it is not, and he wrote about ﬁnding an ‘optimal level’ of stress, between too little and too much. By this usage, the everyday connotation of stress as something undesirable had essentially disappeared, and ‘stress’, in effect, had acquired two meanings: in everyday speech it meant an undesirable state of strain or tension, but in Selye’s science it was a term for activation of the HPA system and related responses. Selye was such an articulate, proliﬁc and energetic popularizer of his scientiﬁc ideas (a biographer described him as a ‘travelling salesman of stress’36) that a generation of scientists virtually equated ‘stress’ with activation of the HPA system. This created a natural pitfall for animal welfare scientists. They wanted to tie the concept of animal welfare to some body of established science, and the idea that one could assess ‘stress’ through activation of the HPA system seemed a god-send. If indeed activation of the HPA system represents (as Selye put it) ‘an identical biochemical reaction of the body’ which is ‘totally non-speciﬁc, and common to all types of exposure’, then surely quantifying HPA activity provides an objective way to identify all types of animal welfare problems and perhaps even to measure their severity. And by extension, the absence of an HPA response might be taken to indicate that no animal welfare problem exists.37 But if, as now seems more plausible, the HPA and SAM systems are best seen as ways that the body prepares for certain increased demands, and if these responses are triggered by some forms of adversity but not all, then measures of the HPA and SAM systems fall far short of being a royal road to understanding animal welfare. How, then, can they contribute? First, as Walter Cannon recognized a century ago, the bodily responses provide windows on certain affective states of animals. In particular, the sensitivity of the HPA and SAM systems to emotions such as fear and pain can provide a useful means of monitoring these states. Thus, for example, the biochemical measurements of HPA activity and related measures played a key role in supporting recommendations for abolishing hunting deer with dogs, and for recommending the use of more effective pain management for dehorning calves. Second, a very strong or prolonged activation of these systems can indicate some actual or potential impairment of basic health and functioning. Following Gary Moberg’s logic, instead of waiting for a pathological state to develop, we could detect problems of basic health and functioning in their early stages by identifying reduced immune competence or reduced secretion of gonadal hormones, before disease or reproductive failure occurs. 36Yanacopoulo, A. 1992. Hans Selye, ou, La Cathédrale du Stress. Le Jour, Montreal. The quotation reads ‘commis-voyageur du stress’ and is on page 281. 37The confusion is ably described by Dawkins, M.S. 1998. Evolution and animal welfare. Quarterly Review of Biology 73: 305–328. The broader confusion over the term ‘stress’ is nicely interpreted by Toates, F. 1995. Stress: Conceptual and Biological Aspects. John Wiley & Sons, Chichester.

Dfraser_C006.indd 123

6/27/2008 12:43:02 PM

124

Studying Animal Welfare

And third, knowledge of the HPA and SAM systems contributes to basic understanding of certain animal welfare problems. When Paul Hemsworth and co-workers noted that negative handling of farm animals often leads to lower productivity, an understanding of the HPA system provided a plausible mechanism to explain the result. Speciﬁcally, chronic fear of people could lead to activation of the HPA system which in turn could suppress the hormones that promote growth, reproduction and milk yield. In Chapter 8, we will return to the use of endocrine measures in the study of affective states, but ﬁrst let us deal with one other area where scientists have struggled to understand how a group of complex bodily changes relate to animal welfare.

Dfraser_C006.indd 124

6/27/2008 12:43:02 PM

Abnormal Behaviour

7

In 1944, psychiatrist David Levy published a classic article on behavioural disturbances associated with mental illness, especially the excessive activity and repetitive movements that are sometimes seen in emotionally disturbed children. Levy also described similar behaviour in animals. He cited repetitive weaving movements of horses in stables and head movements by caged birds, and he proposed that these behavioural abnormalities in animals reﬂect the same kinds of emotional disturbances that he saw in his young patients.1 Levy was one of several researchers at that time who perceived abnormal behaviour of animals as an indication of underlying psychological abnormalities; and as concern grew over the welfare of captive animals, researchers paid increasing attention to abnormal behaviour. Andrew Fraser, one of the early advocates of including animal behaviour in the training of veterinarians, published observations on abnormal somnolence, aggression and other aberrant behaviour in farm animals.2 Veterinarian and animal behaviourist Michael W. Fox compiled one of the earliest books on applied animal behaviour and it dealt speciﬁcally with behavioural abnormalities.3 German zoo biologists Heini Hediger and Monika Meyer-Holzapfel published numerous accounts of abnormal movement patterns among wild animals in captivity.4 Yet for all this long history of attention to abnormal behaviour, the actual interpretation of such behaviour has proven to be one of the most troublesome

1Levy, D.M. 1944. On the problem of movement restraint: Tics, stereotyped movements, hyperactivity. American Journal of Orthopsychiatry 14: 644–671. 2Fraser, A.F. 1959. Displacement activities in domestic animals. British Veterinary Journal 115: 195–200. 3Fox, M.W. (editor). 1968. Abnormal Behavior in Animals. W.B. Saunders Company, Philadelphia. 4Hediger, H. 1950. Wild Animals in Captivity (G. Sircom, translator). Butterworths Scientiﬁc Publications, London; Meyer-Holzapfel, M. 1968. Abnormal behavior in zoo animals. Pages 476–503 in Abnormal Behavior in Animals (M.W. Fox, editor). W.B. Saunders Company, Philadelphia.

125

Dfraser_C007.indd 125

6/27/2008 4:57:54 PM

126

Studying Animal Welfare

Figure 7.1 Self-biting by a rhesus macaque. Photo by Dr. Joe Erwin. Reproduced with the kind permission of Dr. Erwin and the publisher from: Erwin, J., Mitchell, G. and Maple, T. Abnormal behavior in non-isolate-reared rhesus monkeys. Psychological Reports, 1973; 33: 515–523. Copyright Psychological Reports 1973. Photo kindly supplied by Dr. Erwin with the cooperation of Dr. Viktor Reinhardt.

topics in animal welfare science. If we see sows chewing the bars of their pens, or mice biting the whiskers off their cagemates, or an otter swimming in repetitive circles around a small pool in a zoo, what can we conclude about the animals’ welfare? Is the behaviour a form of pathology signalling a breakdown of normal health and functioning? Is it a coping mechanism that allows the animal to feel less distressed in a restrictive environment? Is it merely a habit that the animal has formed, perhaps on a par with people drumming their ﬁngers? Or is it something else entirely? THE TERM ‘ABNORMAL’ IS perhaps most easily applied in cases where animals actually cause injury to themselves. A disturbing example is self-biting by macaque monkeys housed in individual cages in laboratories (Figure 7.1). In reviewing the literature on this behaviour, animal welfare scientists Viktor Reinhardt and Matt Rossell noted that self-biting is generally seen in animals showing ‘a high level of emotional arousal’ and that it occurs in ‘emotionally disturbing situations’

Dfraser_C007.indd 126

6/27/2008 4:57:55 PM

Abnormal Behaviour

127

over which the animal has no control.5 Self-biting may occur, for example, when monkeys are transferred to a different cage or when they are approached by staff who appear to cause the animals fear. However, given that macaques are highly social animals and that the behaviour is generally associated with solitary housing, the scientists concluded that ‘the principal stressor is obviously the absence of a companion’. The behaviour does not appear to respond to standard methods of enriching the environment such as puzzle feeders (which require the animals to work for food), although it may be reduced by drugs that calm the animals. The most effective treatment, however, is to house the animals with social companions. Reinhardt and Rossell estimated that there were 15 000 macaque monkeys kept in solitary housing in the United States when they wrote their article. Based on various reports, they calculated that roughly 10% of these animals would bite themselves severely enough to cause injury, and another 10% would do less severe self-biting which would not cause obvious injuries and would go largely unnoticed. They questioned the scientiﬁc usefulness of these 3000 animals which, in their view, show evidence of serious emotional disturbance. Self-injury by macaques seems an obvious example of abnormal behaviour because we cannot imagine that natural selection shaped animals to harm themselves in this way. Much the same applies to certain forms of social behaviour. For example, when young calves are separated from their mothers at only a day or two of age, they show a strong motivation to suck on almost any available object, especially during the ﬁrst few minutes after they have been fed milk. Some calves develop the habit of sucking on the ears, navel or prepuce of other calves. The sucking can be so persistent that it wears away the hair or skin in the affected area, and in exceptional cases the behaviour may involve consumption of urine.6 Similarly, when sheep are conﬁned in pens where they cannot graze, a few animals may develop the habit of denuding their pen-mates by clipping their wool in a way that resembles grazing. The behaviour is relatively harmless for the recipient unless the weather is cold and the denuded animal suffers from a lack of insulation. However, very young lambs, around the age when they would normally begin eating vegetation, sometimes develop the habit of ingesting wool from the bodies of their mothers. With little other solid material passing through the digestive tract, the wool may form into balls that remain in the stomach, and these may pose a serious health problem by blocking the passage of food.7

5Reinhardt, V. and Rossell, M. 2001. Self-biting in caged macaques: Cause, effect, and treatment. Journal of Applied Animal Welfare Science 4: 285–294. The quotations are from pages 288, 286 and 290 respectively. 6Stephens, D.B. 1982. A review of some behavioral and physiological studies which are relevant to the welfare of young calves. Pages 47–67 in Welfare and Husbandry of Calves (J.P. Signoret, editor). Martinus Nijhoff, The Hague. 7Sambraus, H.H. 1985. Mouth-based anomalous syndromes. Pages 391–422 in Ethology of Farm Animals (A.F. Fraser, editor). Elsevier Science Publishers, Amsterdam.

Dfraser_C007.indd 127

6/27/2008 4:57:56 PM

128

Studying Animal Welfare

‘Tail-biting’ in pigs is perhaps the most notorious form of abnormal behaviour directed to another animal. If pigs are kept in conﬁned conditions where there are no natural materials for them to root and chew, many of them will chew on the tails, ears and other body parts of their pen-mates. In some cases, this initially innocuous behaviour leads to an injury of the tail, and this can have two further effects. First, pigs are strongly attracted to chew on objects that have begun to disintegrate: a length of rope attracts a certain level of chewing by pigs, but once it becomes frayed it attracts far more.8 Second, the taste of dried blood and scabby tissue seems to appeal to pigs, especially if their diet is deﬁcient in salt or protein.9 Consequently, an injured tail can attract so much attention that it is quickly reduced to a bleeding stump. In extreme cases the victim may lose signiﬁcant amounts of blood, and the injury may become the entry point for bacteria that infect the spinal cord and lead to paralysis and death.10 These various activities are seen as abnormal because we assume that natural selection did not favour animals that suck urine, ingest wool or chew tails. Nonetheless, the actual movement patterns of sucking, grazing, and chewing are part of the animals’ normal repertoire of behaviour; speciﬁcally, each of these actions is involved in obtaining and ingesting food, and under natural conditions the animals might spend substantial amounts of time performing these actions. Animal behaviourist Georgia Mason refers to the normal movements as the ‘source behaviour’ for the abnormality.11 In cross-sucking, wool-clipping and tail-biting we see a normal source behaviour ‘redirected’ from its typical target to some different and seemingly inappropriate target, generally in environments where the normal targets for the behaviour are lacking. Related to redirected behaviour is an abnormality described by the eminent ethologist Konrad Lorenz. In a classic paper, Lorenz described how a young starling, which Lorenz had reared in his home, went through the exact movements of catching ﬂying insects even though no insects were present: With head and eyes the bird made a motion as though following a ﬂying insect with its gaze; its posture tautened; it took off, snapped, returned to its perch, and with its bill performed the sideways lashing, tossing motions with which many insectivorous birds slay their prey against whatever they happen to be sitting upon. Then the starling swallowed several times, whereupon its closely laid plumage loosened up somewhat, and there often ensued a quivering reﬂex, exactly as it does after real satiation.12 8Feddes, J.J.R. and Fraser, D. 1994. Non-nutritive chewing by pigs: Implications for tail-biting and behavioral enrichment. Transactions of the American Society of Agricultural Engineers 37: 947–950. 9Fraser, D., Bernon, D.E. and Ball, R.O. 1991. Enhanced attraction to blood by pigs with inadequate dietary protein supplementation. Canadian Journal of Animal Science 71: 611–619. 10Sambraus, 1985. 11Mason, G.J. 1991 Stereotypies: A critical review. Animal Behaviour 41: 1015–1037. 12Lorenz, K. 1937. Über die Bildung des Instinktbegriffes. Die Naturwissenschaften 25: 289–300, 307–318, 324–331. Republished 1957 as: The nature of instinct: The conception of instinctive behavior. Pages 129–175 in Instinctive Behavior: The Development of a Modern Concept (C.H. Schiller, editor and translator). International Universities Press, New York.

Dfraser_C007.indd 128

6/27/2008 4:57:56 PM

Abnormal Behaviour

129

This type of behaviour has been called ‘vacuum activity’ because it occurs in the absence of any obvious stimulus or target. One vacuum activity that has been used in the study of animal welfare is ‘tonguerolling’ by calves. Calves that are raised for ‘white’ veal are generally fed a milk-like diet from birth until they are slaughtered at about four months of age. The calves are prevented from consuming roughage such as grass or hay partly because the iron contained in such plant-based food would cause their muscles to assume a normal reddish colour instead of the pale colour that purchasers of this product demand. The diet, however, is unnatural because calves would normally start to forage and ruminate from about two weeks of age. When limited to a milky diet, some calves will spend hours per day in what appears to be ‘vacuum grazing’. They extend the tongue out of the mouth and curl it to the side in what appears to be the action that cattle use to grasp a sward of grass and pull it into the mouth, but the calves do this simply in the air, without the tongue contacting any physical object. Another seemingly anomalous type of behaviour has been called ‘displacement activity’. During the 1930s and 1940s, a number of ethologists reported observing that when animals are strongly motivated to act in certain ways (ﬂeeing, attacking, mating), they sometimes interrupt an on-going sequence of behaviour and switch to some other activity that seems quite irrelevant in the context. Examples were described by Niko Tinbergen.13 He noted, for example, that ‘two skylarks engaged in furious combat [may] suddenly peck at the ground as if they were feeding’, or birds on the point of mating may suddenly begin to preen themselves. Tinbergen adopted the term ‘displacement activities’ because the behaviour appeared to be displaced from one behavioural system into another. He noted that such behaviour often occurs when animals are simultaneously motivated to act in two different ways, for example to both attack and ﬂee from an intruder, or if they are strongly motivated to perform an action that they cannot carry out. In the assessment of animal welfare, displacement activities are sometimes used as evidence that an animal is motivated to perform a type of behaviour that the environment does not permit. An example, which we will explore in more detail in the next chapter, is Ian Duncan’s demonstration that when hungry hens are trained to eat from a particular food dispenser and then ﬁnd the dispenser blocked, they often begin to pace and preen themselves vigorously. Duncan interpreted these actions as displacement behaviour, and he then used similar pacing and preening as evidence of frustration in other situations.14 More recently, psychiatrist/primatologist Alfonso Troisi has proposed that displacement activities can be used as non-invasive measures of ‘stress’ in primates. He noted that various non-human primates perform self-directed activities such as grooming and 13Tinbergen,

N. 1952. ‘Derived’ activities; their causation, biological signiﬁcance, origin, and emancipation during evolution. Quarterly Review of Biology 27: 1–32. The quotation is on page 6. 14Duncan, I.J.H. 1970. Frustration in the fowl. Pages 15–31 in Aspects of Poultry Behaviour (B.M. Freeman and R.F. Gordon, editors). British Poultry Science Ltd., Edinburgh.

Dfraser_C007.indd 129

6/27/2008 4:57:56 PM

130

Studying Animal Welfare

scratching in situations that are presumed to involve anxiety and uncertainty, and that these behaviours are increased by anxiogenic (anxiety-producing) drugs and reduced by anxiolytic (anxiety-reducing) drugs. In humans, he noted that similar self-directed behaviour, together with aimless manipulation of objects (chewing pens, twisting rings), can be used as indicators of ‘stressful stimuli and may reﬂect an emotional condition of negative affect’.15 Well after the early ethologists had described displacement behaviour, experimental psychologists identiﬁed a seemingly related class of activities which they termed ‘adjunctive behaviour’. In 1960, psychologist John Falk was studying hungry rats that had been trained to press a lever for a small food pellet in an experimental chamber. In Falk’s original experiment, once a rat had received a pellet, it was obliged to wait an average of one minute before another press of the lever would be rewarded. The rats developed the habit of drinking water during these intervals, but their consumption far exceeded what Falk expected. Many consumed three to four times their normal daily water intake during a three-hour session in the apparatus, and some drank nearly half of their body weight in water during this time.16 Later research showed that many other types of behaviour could be induced in a similar way. If a running wheel is present, many rats will run energetically in the wheel during the intervals between rewards; pigeons in a similar situation will peck aggressively at a restrained pigeon; monkeys will consume wood chips picked up from the litter on the ﬂoor of the chamber. Falk referred to such behaviour as ‘adjunctive’ because the behaviour occurs as an adjunct to another activity that the animal is strongly motivated to perform.17 Adjunctive behaviour has also been used as evidence of animal welfare problems. As we noted earlier, pregnant sows are typically fed only a fraction of the amount of food they would consume by choice, and they remain hungry for virtually the whole day. If a water dispenser is available, some sows will drink two or three times their normal daily intake, and under winter conditions, warming this amount of cold water to body temperature, only to discharge it as dilute urine, involves an appreciable caloric cost. However, if such sows are given a bulky high-ﬁbre food (which under typical circumstances would result in an increase in water intake), they spend much longer eating, and the excessive drinking largely disappears.18

15Troisi,

A. 2002. Displacement activities as a behavioral measure of stress in nonhuman primates and human subjects. Stress 5: 47–54. The quotation is from page 52. 16Falk, J.L. 1961. Production of polydipsia in normal rats by an intermittent food schedule. Science 133: 195–196. 17Falk, J.L. 1971. The nature and determinants of adjunctive behavior. Physiology & Behavior 6: 577–588. 18Robert, S., Matte, J.J., Farmer, C., Girard, C.L. and Martineau, G.P. 1993. High-ﬁbre diets for sows: Effects on stereotypies and adjunctive drinking. Applied Animal Behaviour Science 37: 297–309.

Dfraser_C007.indd 130

6/27/2008 4:57:56 PM

Abnormal Behaviour

131

In this case, much of the sows’ water intake appeared to be adjunctive drinking that was not linked to thirst. In the case of displacement and adjunctive activities, behaviour is performed in an abnormal context and in response to a seemingly unrelated motivation. However, abnormal behaviour can also involve a simple increase in the amount or intensity of an activity. In the 1920s, the British anatomist Solly Zuckerman studied a colony of Hamadryas baboons (Papio hamadryas) in the London Zoo.19 The colony was created by the introduction of 100 baboons into a rocky enclosure known as Monkey Hill. The intention had been to use only males, but six females were included by accident. This proved to be a fateful error because with so many mature males competing for so few females, the level of aggression in the colony became horriﬁc. Of the original 100 animals, 27 died during the ﬁrst six months mostly from injuries, and after two years the number had been reduced to 56. At that point, the disastrous decision was made to add 30 additional females to the colony. Zuckerman described the result: The new arrivals stirred the Hill into great excitement, and all the old males tried to secure females, ﬁfteen of whom were killed in the ﬁghts that occurred between the 27th of July and the end of August. These ﬁghts are deﬁnitely sexual in nature. The males ﬁght for the females, who are usually fatally injured in the mêlée which rages around them … . The injuries inﬂicted were of all degrees of severity. Limb-bones, ribs, and even the skull have been fractured. Wounds have sometimes penetrated the chest or abdomen, and many animals showed extensive lacerations in the ano-genital region.

Decades later, ﬁeld research on free-living baboons by a new generation of primatologists such as Jeanne Altmann, Barbara Smuts and Thelma Rowell have produced a very different and more peaceful picture of baboon social behaviour. Fatal wounding during aggression among males certainly occurs, and mauling of females does play a role in baboon life.20 However, the aggression described by Zuckerman was clearly at an abnormal level, presumably because of the highly unnatural gender ratio and the fact that the enclosure made escape impossible. SELF-INJURY, REDIRECTED BEHAVIOUR, VACUUM ACTIVITIES, displacement and adjunctive behaviour, and abnormal levels of behaviour – each of these has played a role in animal welfare science. However, what has been called ‘stereotyped behaviour’ has been used in the assessment of animal welfare probably as much as the other categories combined. It merits careful analysis both because of its importance, and because of the amount of confusion it has caused. 19Zuckerman,

S. 1932. The Social Life of Monkeys and Apes. Kegan Paul, Trench, Trubner & Co, London. The quotation is from pages 219 and 220. 20Altmann, J. 1980. Baboon Mothers and Infants (Republished 2001). University of Chicago Press, Chicago.

Dfraser_C007.indd 131

6/27/2008 4:57:56 PM

132

Studying Animal Welfare

In a critical review of the topic, Georgia Mason noted that stereotyped behaviour is commonly deﬁned as behaviour that is ‘repetitive, invariant and has no obvious goal or function’.21 Mason went on to describe other common features of stereotyped behaviour: that the animal may continue performing the behaviour much longer than would normally be expected, that stereotyped behaviour may be performed predictably in the same location or at the same time of day, and that the animal may appear to have difﬁculty stopping. Stereotyped activities (or ‘stereotypies’) can be derived from various types of source behaviour. Some of the most common involve the mouth. These may begin as a redirected form of sucking, eating or foraging but, through constant repetition, come to be performed in a repetitive, largely unvarying pattern. Animal behaviourist Gregory Cronin provided a detailed description of the stereotyped behaviour developed by sows after they were moved into narrow gestation stalls equipped with a short length of chain that the animals could toy with.22 For one sow (Figure 7.2), he described ﬁve behavioural elements which occurred in an almost unvarying sequence. The sow would begin by picking up the chain and chewing it to the top. Then the sow would turn its head to the left, bringing the chain to the right-hand side of the mouth. Then the sow would pull the snout backward, turn the head to the right, and ﬂick the chain over the snout. Next the sow would chew the chain for about two seconds. Then the sow would either return to the previous element, or else lower its head, drop the chain, and then begin the sequence again. Different sows had different sequences of actions, and some would perform the behaviour over and over for many hours each day. Cronin also described how sows developed these repetitious sequences of behaviour. ‘With time’, he noted, ‘there was a gradual organization of random behaviour into sequences, a reduction in the number of different environmentdirected acts performed by individual sows … and a corresponding increase in the proportion of activity that was spent on particular acts’. In a sample of nine sows that he studied closely from the time they were put into the stalls, Cronin noted that it took a median of 16 days, with a range of 8–71 days, before a stereotyped sequence became the sow’s most commonly performed action. A second group of stereotypies are derived not from actions of the mouth but from general locomotion. Working with zoo animals in the 1930s, Monika Meyer-Holzapfel provided some classic descriptions of such behaviour, including repetitive walking in circular or ﬁgure-eight patterns (Figure 7.3). In one of her examples, a dingo in the Amsterdam Zoo was separated from its pack and

21Mason,

1991, page 1015.

22Cronin, G.M. 1985. The Development and Signiﬁcance of Abnormal Stereotyped Behaviours

in Tethered Sows. Doctoral thesis, Agricultural University of Wageningen, Wageningen, The Netherlands. The quotation is on pages 29 and 32.

Dfraser_C007.indd 132

6/27/2008 4:57:56 PM

Abnormal Behaviour

133

Figure 7.2 Five elements that made up a stereotyped sequence of behaviour performed by one sow (Sow 23) as described by Cronin (1985). Arrows and corresponding numbers show the percentage of cases in which each element was followed by each other element. Most of the elements lasted for one to two seconds. Redrawn from Cronin (1985) with kind permission of Dr. Gregory Cronin.

placed alone in an adjacent enclosure where it could still see its group members through a partition. At the beginning, the dingo ran back and forth along the separating trellis, keeping quite close to it; when it turned around at each end of its path, the turns were quite small. In the following days, these paths were paced farther and farther from the trellis. Thus the path gradually took the form of a ﬁgure eight.

Meyer-Holzapfel suggested that the emergence of the repetitive ﬁgure-eight pattern reﬂected a ‘gradual calming down of the animal as attempts to pass through the partition fence become less intense’.23 Intuitively, it seems reasonable to suggest that the performance of stereotyped behaviour represents some insult to the animal’s welfare, and indeed this has been a widespread assumption. In fact, however, the links between stereotyped behaviour 23Meyer-Holzapfel,

Dfraser_C007.indd 133

1968, pages 482 and 483.

6/27/2008 4:57:57 PM

134

Studying Animal Welfare

A

B

C

Figure 7.3 The development of stereotyped route-tracing by a dingo in the Amsterdam Zoo described by Monica Meyer-Holzapfel. The dashed line represents a fence separating the dingo from its social group. When first separated, the dingo ran back and forth, close to the fence, making tight turns at each end (pattern A). Over time, the author stated, ‘the animal becomes less intensely aroused to escape’ (page 483), and the path adhered less closely to the fence (pattern B). Eventually the dingo was running repeatedly in a rough figure-of-eight pattern (pattern C). Redrawn from Figure 25-3 of Meyer-Holzapfel (1968) with permission of Elsevier.

and welfare have been more often assumed than clearly articulated. To make the links more explicit, let us consider four lines of evidence. One is the fact that stimulant drugs that cause mental disturbance in humans often trigger stereotyped behaviour in animals. The 1960s saw a rash of pharmacological studies showing that administration of amphetamines triggered the spontaneous appearance of stereotyped behaviour in a wide range of animals. For example, two Danish scientists studied the behaviour of 25 guinea-pigs injected with amphetamine. Stereotyped behaviour generally began 30–60 minutes after the injection and lasted for about three hours. Two of the animals showed continuous, vertical tossing movements of the head. Twenty-three bit continuously at the wires of the cages, at the edge of the food bowl, or on the skin of their cage mates. In four animals this biting was either preceded or followed by a period of head-tossing. While the animals were engaged in this bizarre behaviour, their normal behaviour (walking, grooming, etc.) was suspended, and did not resume until the reaction to the amphetamine had worn off. The authors also noted that similar dosage rates of amphetamines in humans tend to induce mental disturbance and, in many patients, the continual repetition of apparently purposeless actions.24 A second line of evidence shows that stereotypies may be linked to physiological stress responses. In the study of clouded leopards in 12 zoos described earlier, keepers were asked to rate various features of the animals’ behaviour. Leopards that 24Randrup, A. and Munkvad, I. 1967. Stereotyped activities produced by amphetamine in several animal species and man. Psychopharmacologia (Berlin) 11: 300–310.

Dfraser_C007.indd 134

6/27/2008 4:57:57 PM

Abnormal Behaviour

135

had high concentrations of faecal corticosteroid metabolites, indicating chronic activation of the hypothalamus–pituitary–adrenal-cortex (HPA) system, were described as performing more stereotyped pacing in their cages, and were more likely to have a history of self-inﬂicted injury.25 In a dramatic demonstration linking the HPA system to abnormal behaviour, animal scientist Janeen Salak-Johnson and co-workers injected corticotrophin releasing hormone (CRH), which activates the HPA response system, directly into the ventricles of the brain in pigs. They reported three main behavioural reactions: hyperactivity (constant movement), ‘freezing’ (a behaviour which they took as an indicator of fear), and continuous rooting, chewing and rubbing movements of the snout and mouth. They proposed that the use of CRH for producing hyperactivity and stereotyped behaviours ‘may represent a new model for the study of mania or obsessive–compulsive behaviors’ in humans.26 As a third line of evidence, it has sometimes been possible to link stereotyped behaviour to a speciﬁc affective state. The pregnant sows in gestation stalls again provide an interesting example. Various researchers have proposed that this type of stereotyped behaviour develops from exploratory motivation, or from attempts to escape from a conﬁned space, or that the behaviour helps animals to ‘cope’ with an aversive environment. These interpretations need not be mutually exclusive, but they tend to overlook the possible role of hunger. As noted earlier, pregnant sows are typically fed a restricted diet to prevent them from gaining too much weight, and research has shown that sows in stalls are much more likely to perform stereotyped behaviour when food-restricted in this way.27 But if that is true, then is the stereotyped behaviour mainly a reﬂection of hunger or of the restriction of movement in the narrow stalls? Veterinary scientist Claudia Terlouw and co-workers carried out the critical experiment.28 They housed some sows individually in narrow stalls where the animals were tethered by a chain, and other sows in group pens equipped with chains hanging from the walls. In each housing treatment, half of the sows were fed a restricted diet similar to normal commercial feeding levels, while the other half received nearly twice as much. Regardless of the housing treatment, sows on the limited diet were seen chewing, biting, and rooting the chain for about 8% of the time that they were observed, whereas the well fed sows spent only about 1% of the time in such behaviour. Thus these elements of stereotyped behaviour appeared to be related more speciﬁcally to hunger than to other motivational states such as frustration at being kept in a narrow stall. Of course, both the stalls and the 25Wielebnowski

et al., 2002. J.L., Anderson, D.L. and McGlone, J.J. 2004. Differential dose effects of central CRF and effects of CRF astressin on pig behavior. Physiology & Behavior 83: 143–150. The quotation is on page 150. 27Appleby, M.C. and Lawrence, A.B. 1987. Food restriction as a cause of stereotypic behaviour in tethered gilts. Animal Production 45: 103–110. 28Terlouw, E.M.C., Lawrence, A.B. and Illius, A.W. 1991. Inﬂuences of feeding level and physical restriction on development of stereotypies in sows. Animal Behaviour 42: 981–991. 26Salak-Johnson,

Dfraser_C007.indd 135

6/27/2008 4:57:57 PM

136

Studying Animal Welfare

group pens represented very restrictive environments. In more complex or natural environments, the animals might respond to hunger by rooting in the soil and other natural food-ﬁnding activities. Hence, these particular stereotypies might best be regarded as indicators of hunger that are likely to occur when animals cannot respond to hunger in more natural ways. A ﬁnal line of evidence shows that stereotyped behaviour may involve basic brain pathology. The basal ganglia are a collection of structures located deep in the brain and consisting of the nuclei of nerve cells that are involved in the control of bodily movement. Some disturbances to the basal ganglia result in involuntary actions like tics and the uncontrollable movements that are seen in disorders such as Huntington’s disease.29 In animal studies, when brain lesions and administration of amphetamine induce stereotyped behaviour, these effects often occur by disrupting the functioning of the basal ganglia. Thus stereotyped behaviour can result from the malfunctioning of the basal ganglia; but when we see stereotyped behaviour developing in caged animals in the absence of any pharmacological treatment or disease, does it actually represent the same kind of neural pathology? Animal welfare scientists Joseph Garner and Georgia Mason proposed an ingenious and non-invasive way to test this idea.30 They noted that in experimental gambling tasks, human autistics or schizophrenics with disfunctioning of the basal ganglia are likely to persist in performing a previously learned response even when it is not rewarded. Garner and Mason tested whether the same type of persistent repetition of learned behaviour would be seen in animals that perform high levels of stereotyped behaviour. The animals they studied were bank voles (Clethrionomys glareolus) that performed stereotyped gnawing on the bars of their cages in very different amounts, ranging from one animal that showed little of this behaviour to another that spent 28% of its active time gnawing on the bars. For the study, each animal was trained to move from its home cage into a simple T-maze where it could then return to the home cage by either a passageway to the right or one to the left. In the ﬁrst part of the experiment, the voles received a sugar-water reward in one of the passageways but not the other. The high-stereotyping and low-stereotyping animals took a similar number of trials to learn to use the side where the reward was available. However, when the reward was no longer provided, the low-stereotyping animals soon reverted to using the two routes about equally, while the high-stereotyping animals persisted in using the previously rewarded route. The difference was dramatic: the animal that was most free from stereotypies took only 26 trials to stop repeating the previously correct response in the maze; the vole that performed the most bar-gnawing took 244 trials. 29Albin,

R.L., Young, A.B. and Penney, J.B. 1989. The functional anatomy of basal ganglia disorders. Trends in Neurosciences 12: 366–375. 30Garner, J.P. and Mason, G.J. 2002. Evidence for a relationship between cage stereotypies and behavioural disinhibition in laboratory rodents. Behavioural Brain Research 136: 83–92.

Dfraser_C007.indd 136

6/27/2008 4:57:58 PM

Abnormal Behaviour

137

Using these and several other lines of evidence, Garner and Mason showed that the amount of stereotyped behaviour shown by the voles was closely correlated with various measures typical of basal ganglia dysfunction. They concluded that laboratory animals showing persistent stereotyped behaviour may well have the kind of brain dysfunction commonly seen in human patients with autism, schizophrenia and degenerative brain disorders, and they questioned the validity of research data collected from such animals. Garner and Mason’s conclusion could also help explain the remarkable persistence of some stereotyped behaviour. For example, Meyer-Holzapfel noted that some circus animals that developed repetitive rocking or weaving movements while restrained in small cages would continue these actions while standing continually in one place even after they had been moved to large enclosures which permitted much more freedom.31 It is possible that the behaviour had simply become so habitual that it continued outside the environment where it had ﬁrst developed. Alternatively, if repetitive behaviour reﬂects basic pathology of the nervous system, then it may continue after the original cause of the pathology is no longer present. To summarize, these lines of evidence illustrate various ways that stereotyped behaviour may be related to impaired welfare. Stereotyped behaviour can involve pathology of the basal ganglia or abnormal activation of the HPA stress-response system, and thus represent a breakdown in basic health and functioning. Stereotyped behaviour can indicate unpleasant affective states – in some cases chronic hunger, in other cases agitation induced either by amphetamines or by a distressing environment. And stereotyped behaviour can prevent animals from performing their natural behaviour or, as in the case of the hungry sows, it may indicate that the environment does not allow the animals to respond to some challenge in their normal way. DESPITE THESE POSITIVE EXAMPLES, the use of stereotyped behaviour in the assessment of animal welfare has created a great deal of confusion. The ﬁrst problem has been one of deﬁnition. The traditional deﬁnition of stereotyped behaviour – behaviour which is repetitive, unvarying and without obvious goal or function – can encompass a wide range of potentially unrelated phenomena. A great deal of normal behaviour, such as walking and chewing for example, involves repeating the same action over and over in a way that varies as little as most stereotyped movements. Moreover, to judge that behaviour is without obvious function may simply mean that its function is not yet known. If a baby sucks on a soother, or if a person chews gum or goes for a walk every day along the same route, would these activities qualify as stereotyped behaviour and hence (according to the common assumption) be used as evidence of poor welfare? Let us examine three examples to show how broadly the term ‘stereotyped’ has been applied. 31Meyer-Holzapfel,

Dfraser_C007.indd 137

1968, page 485–487.

6/27/2008 4:57:58 PM

138

Studying Animal Welfare

In one inﬂuential study (to which we will return below) physiologists Robert Dantzer and Pierre Mormède studied pigs using a method modelled after John Falk’s procedure for inducing adjunctive drinking.32 Dantzer and Mormède kept hungry pigs in a chamber where a small amount of food was delivered every four minutes. The chamber also contained a short length of chain, and the pigs developed the habit of biting and pulling on the chain during the interval between food deliveries. Every time this occurred, the movement of the chain closed a switch that allowed the action to be recorded automatically. With repeated sessions the pigs came to average 300 pulls on the chain in a 40-minute session. In reporting the study, Dantzer and Mormède referred to the chain-pulling as ‘stereotyped’, presumably because the behaviour was repeated frequently and was without obvious function, although there was no reason to believe that it involved unvarying patterns of movement. As a second example, poultry behaviourist John Savoury and co-workers explored why ‘broiler breeders’ – the breeding birds that produce the chicks that are raised for meat – spend so much time pecking at objects in their environment. Genetic selection for rapid growth in these birds has resulted in chicks with huge appetites, and when these birds are raised to adulthood for breeding purposes their food intake has to be severely restricted so that they do not become obese. In one study, Savoury and co-workers observed broiler breeders fed at a typical level of feed restriction compared to similar birds that had constant access to feed. They found that birds on the restricted diet spent a great deal of time (averaging 14–49% of observation time) pecking at the walls of their pens, whereas birds with constant access to feed were not seen doing this behaviour at all.33 This suggested that chronic hunger is a key part of the motivation. However, a second study showed that in the hours after they have been fed, the birds pecked much more if they had received a sizeable meal of 40 grams than after a very small meal of only 10 grams.34 Thus, despite the link to hunger, some form of feedback after a sizeable meal seemed to maintain rather than reduce the motivation to peck. In reporting this work, Savoury and co-workers referred to the pecking they observed as ‘stereotyped’, presumably because it was unexpectedly persistent and directed at objects other than food. Finally, in a study of digging by young Mongolian gerbils (Meriones unguiculatus), Christoph Wiedenmayer observed that caged gerbils sometimes dig in the bedding in the middle of the cage, and at other times perform more sustained bouts of digging around the periphery. He noted that when the gerbils dug in the middle of the cage, most bouts lasted for less than 6 seconds and none lasted as long

32Dantzer,

R. and Mormède, P. 1983b. De-arousal properties of stereotyped behaviour: Evidence from pituitary–adrenal correlates in pigs. Applied Animal Ethology 10: 233–244. 33Savoury, C.J., Seawright, E. and Watson, A. 1992. Stereotyped behaviour in broiler breeders in relation to husbandry and opioid receptor blockade. Applied Animal Behaviour Science 32: 349–360. 34Savoury, C.J. and Mann, J.S. 1999. Stereotyped pecking after feeding by restricted-fed fowls is inﬂuenced by meal size. Applied Animal Behaviour Science 62: 209–217.

Dfraser_C007.indd 138

6/27/2008 4:57:58 PM

Abnormal Behaviour

139

as 12 seconds. In contrast, digging around the edge of the cage often continued for longer. Wiedenmayer found that if young gerbils were kept in a cage where they had access to an artiﬁcial burrow system, then they did less of the short bouts of digging and did not do sustained digging at all. In reporting the work, Wiedenmayer called the digging stereotyped if a bout lasted longer than 12 seconds, and non-stereotyped if the duration was less than 12 seconds, arguing that the criterion of 12 seconds separated ‘locally ﬁxed, long-lasting’ digging from other digging.35 These three examples produced interesting ﬁndings, but they also show that the term ‘stereotyped’ has been used to cover a range of behaviour which may be related only loosely if at all. Recall that the behaviour described by Cronin and Meyer-Holzapfel involved idiosyncratic sequences of actions which, after a gradual development, ﬁt the traditional deﬁnition of stereotyped behaviour in that they were repeated for long periods of time, were notably unvarying in form, and were without obvious function. However, in the above three examples we see the term being used for various types of puzzling behaviour including what appear to be adjunctive behaviour (chain-pulling by pigs), redirected behaviour (pecking by broiler breeders), and behaviour performed in abnormally long bouts (prolonged digging by gerbils). Whether these represent the same kind of phenomena studied by Cronin and Meyer-Holzapfel is far from clear. Based on such observations, both Georgia Mason and laboratory animal ethologist Hanno Würbel have proposed that the rather broad, traditional deﬁnition of stereotyped behaviour needs to be revised. Würbel suggests that the key features of stereotyped behaviour are not that it is repetitive, unvarying and functionless but rather ‘the dynamic, developmental changes’ whereby stereotypies gradually ‘increase in frequency and duration while becoming more and more ﬁxed in form and orientation’.36 This description would include Cronin’s sows and Meyer-Holzapfel’s dingo, but would likely exclude many types of behaviour that have been termed ‘stereotyped’. Mason suggests that scientists develop a classiﬁcation of allegedly stereotyped behaviour based more speciﬁcally on its causes. In particular, she suggests distinguishing between those stereotypies that occur in response to ‘motivational frustration, fear or physical discomfort’ and that can be eliminated by solving the underlying problem, versus those that result from some pathology of the central nervous system and may be more difﬁcult to eliminate once they are established.37 35Wiedenmayer,

C. 1997. Causation of the ontogenetic development of stereotypic digging in gerbils. Animal Behaviour 53: 461–470. The quotation is on page 464. 36Würbel, H. 2007. The motivational basis of caged rodents’ stereotypies. Pages 86–120 in Stereotypic Animal Behaviour: Fundamentals and Applications to Welfare, 2nd edition (G. Mason and J. Rushen, editors). CABI, Wallingford, UK. The quotation is on page 88. 37Mason, G. 2007. Stereotypic behaviour in captive animals: Fundamentals, and implications for welfare and beyond. Pages 325–356 in Stereotypic Animal Behaviour: Fundamentals and Applications to Welfare, 2nd edition (G. Mason and J. Rushen, editors). CABI, Wallingford, UK. The quotation is on page 348.

Dfraser_C007.indd 139

6/27/2008 4:57:58 PM

140

Studying Animal Welfare

30 25

% Time

20 15 10 5 0 12

14

16 Age (weeks)

18

20

Figure 7.4 The percentage of observation time spent in ‘tongue-rolling’ by veal calves fed a milk-like diet with no roughage. Calves that did not develop ulceration of the stomach spent increasing amounts of time performing tongue-rolling from the age of 12–20 weeks (open circles). Calves that did develop stomach ulcers spent much less time in this behaviour (filled diamonds). The figure has been redrawn from Figure 4 of Wiepkema et al. (1987) with permission of Elsevier.

A second complication in interpreting stereotyped behaviour is that some such activities turn out to be far from functionless. In the 1980s, Dutch ethologist Piet Wiepkema and co-workers monitored tongue-rolling and other abnormal behaviour in veal calves, and then examined the stomach of each calf when the animals were slaughtered at 22 weeks of age.38 Some of the calves had visible stomach ulcers or scars that represented ulcers from the past, whereas other calves showed no signs of ulceration. When Wiepkema and co-workers compiled the behavioural records, they found that the calves whose stomachs remained free of ulcers had developed persistent tongue-rolling. By about 20 weeks of age, these ulcer-free animals were seen tongue-rolling for more than 20% of the time when they were observed. In contrast, calves whose stomachs showed ulcers or ulcer scars had done relatively little tongue-rolling while alive (Figure 7.4). An obvious hypothesis is that the tongue-rolling had somehow prevented ulcers in the calves

38Wiepkema,

P., van Hellemond, K., Roessingh, P. and Romberg, H. 1987. Behaviour and abomasal damage in individual veal calves. Applied Animal Behaviour Science 18: 257–268. The stomach in this case was the abomasum – the chamber that corresponds to the stomach in human beings. Because the calves were fed a milk-like diet, the rumen was not involved in digestion.

Dfraser_C007.indd 140

6/27/2008 4:57:58 PM

Abnormal Behaviour

141

that had learned to do this behaviour. But how could the performance of such odd behaviour contribute to the health of the digestive system? A hint comes from the work of animal behaviourist Anne Marie de Passillé and co-workers. Noting that young calves have a strong motivation to suck on objects in their environment, de Passillé provided calves with a simple rubber teat attached to the wall of the stall. The calves developed the habit of sucking on the teat (without obtaining milk, of course) during the minutes after they had drunk a normal milk meal from a bucket. De Passillé found that if calves sucked on the dry teat after drinking milk, they showed evidence of greater secretion of the hormones insulin and cholecystokinin which regulate the digestion of food and the uptake of nutrients.39 Thus, the simple action of sucking contributed to the physiological processes of digestion. In the case of Wiepkema’s calves, perhaps the action of tongue-rolling was itself sufﬁcient to inﬂuence the digestive processes – for example by stimulating the production of saliva which helps to buffer the acidity of the stomach – in a way that helped to prevent stomach ulcers. Something similar may occur during a form of abnormal behaviour that is often seen in horses. ‘Crib-biting’ is a common abnormal activity in which the horse grasps an object with its incisors and contracts its lower neck muscles in a way that draws the larynx back and pulls air into the oesophagus. For decades this behaviour has been attributed to ‘boredom’, but recent work by Christine Nicol and co-workers at the University of Bristol suggests that there may be a link with the functioning of the digestive system.40 Nicol and co-workers ﬁrst noted that crib-biting is much more common when horses are fed a diet that is high in grain and low in roughage. Such diets can be eaten quickly, without the long periods of foraging and chewing that would be common for horses kept on pasture. Nicol also noted that horses need saliva to buffer the acidity of the stomach, but they secrete saliva only when chewing. Could it be that by crib-biting, horses are able to increase their secretion of saliva and thus reduce excess stomach acidity? In one study, Nicol and co-workers used endoscopy to compare the stomachs of 19 horses that had begun to perform crib-biting and 16 that had not. The cribbiters had more inﬂammation and ulceration of the stomach, as well as higher levels of acidity in the faeces, all of which suggested poorer buffering of acid in the stomach. The horses were then put on either an antacid diet or a control diet. After 14 weeks, a second endoscopy showed that horses on the antacid diet showed greater improvement in the condition of the stomach, and they tended to have a greater reduction in the time spent crib-biting than horses on the control diet. 39Passillé, A.M.B. de, Christopherson, R. and Rushen, J. 1993. Nonnutritive sucking by the calf and postprandial secretion of insulin, CCK, and gastrin. Physiology & Behavior 54: 1069–1073. 40Nicol, C.J., Davidson, H.P.D., Harris, P.A., Waters, A.J. and Wilson, A.D. 2002. Study of crib-biting and gastric inﬂammation and ulceration in young horses. Veterinary Record 151: 658–662.

Dfraser_C007.indd 141

6/27/2008 4:57:58 PM

142

Studying Animal Welfare

The results were far from conclusive, but they were consistent with the hypothesis that crib-biting develops in response to an unhealthy acidity of the stomach, and that the abnormal behaviour may increase the ﬂow of saliva and thus reduce stomach acidity. However, whereas tongue-rolling had apparently succeeded in preventing ulcers in Wiepkema‘s calves, the crib-biting was not enough to prevent or remedy inﬂammation and ulceration of the stomach of the horses. Abnormal behaviour may have other types of beneﬁcial consequences. When Dantzer and Mormède induced a high level of chain-pulling in pigs by allowing the animals only intermittent access to food, they also tested the pigs for plasma levels of corticosteroids as an indicator of HPA activity. They found that corticosteroid levels declined from about 20 ng ml1 at the start of the session to about 10 ng ml1 at the end if the chain was present, but levels remained high on the one day when the chain was removed from the chamber. In this case it appeared that performance of seemingly abnormal behaviour helped to reduce HPA activity, perhaps by reducing the emotional arousal created by having to wait for the arrival of small amounts of food.41 Somewhat analogous ﬁndings have been made with autistic children. According to one theory, childhood autism commonly involves pathology of the amygdala (two clusters of nerve cells in the brain that help to process emotional reactions) leading to disturbances of the sympathetic nervous system. Using electrical conductance of the palms of the hands (‘sweaty palms’) as a measure of sympathetic nervous system arousal, cognitive scientist William Hirstein and co-workers found that in many autistic children, the performance of repetitive actions is accompanied by a lowering of the excessive sympathetic nervous system arousal that is often seen in autistic people. They concluded that through activities such as repetitive movements, autistic children are attempting to control a sympathetic nervous system which ‘fails to govern itself and seems to require certain behaviours on their part for its regulation’.42 Even the oral stereotypies discussed above may have similar effects. The evidence remains modest, but one study found that horses that had formed the habit of crib-biting had lower heart rate when they were crib-biting than during other types of behaviour.43 Another study found that during tongue-rolling, calves showed a small but repeatable decline in heart rate, and the authors concluded that the behaviour may help to reduce activity of the sympathetic nervous system.44 41Dantzer

and Mormède, 1983b. W., Iversen, P. and Ramachandran, V.S. 2001. Autonomic responses of autistic children to people and objects. Proceedings of the Royal Society, London, B 268: 1883–1888. The quotation is from page 1886. 43Minero, M., Canali, E., Ferrante, V., Verga, M. and Odberg, F.O. 1999. Heart rate and behavioural responses of crib-biting horses to two acute stressors. Veterinary Record 145: 430–433. 44Seo, T., Sato, S., Kosaka, K., Sakamoto, N. and Tokumoto, K. 1998. Tongue-playing and heart rate in calves. Applied Animal Behaviour Science 58: 179–182. 42Hirstein,

Dfraser_C007.indd 142

6/27/2008 4:57:58 PM

Abnormal Behaviour

143

Thus, even those stereotypies that are thought to inﬂuence digestive processes may also have a more general effect of reducing arousal. These and many similar results leave us with a conundrum. On the one hand, some stereotyped and other abnormal behaviour can reﬂect pathology, negative affective states, and environments that do not accommodate important types of natural behaviour. In these respects, the abnormal behaviour seems clearly linked to reduced welfare. On the other hand, stereotyped and other abnormal behaviour can also have positive effects which, to some extent, help to mitigate certain harms. What, then, can we conclude about animal welfare from measures of stereotyped and other abnormal behaviour? The ﬁeld is indebted to Georgia Mason and Naomi Latham for proposing a synthesis of this confusing literature.45 They began with a meta-analysis of several hundred scientiﬁc articles about behaviour that the original authors had described as stereotyped. Mason and Latham found 153 articles reporting that the performance of stereotyped behaviour was related in some way to indicators of poor welfare. In some of these articles, for example, the stereotyped behaviour was correlated with higher stress responses or with poor health, or it occurred more in highly restrictive environments than in environments that were thought to promote good welfare. However, there were also 133 articles in which stereotyped behaviour had no such relationship with indicators of poor welfare, and in more than half of these articles, the stereotyped behaviour was actually correlated with indicators of good welfare. Wiepkema’s study, showing better stomach health in calves that performed stereotyped behaviour, would be an example. Mason and Latham then classiﬁed the papers into two types: those that compared different environments, and those that compared different individuals within the same environment. In studies that compared different environments – such as cage designs that led to high levels of stereotyped behaviour and those that did not – greater performance of stereotyped behaviour was associated with poorer welfare in a large majority of cases. In studies that compared different animals in the same environment, Mason and Latham found the opposite relationship: individuals that performed stereotypies often seemed to have better welfare than other animals living under the same conditions but with less or no stereotyped behaviour. A tentative conclusion would be that environments that induce stereotyped behaviour are likely to create welfare problems, but in those environments there is a reasonable chance that the animals that perform stereotypies are in some way better off than those that do not. How to make sense of this? Mason and Latham summarized two ways in which stereotyped behaviour might actually improve welfare for animals kept in difﬁcult circumstances. First, simple repetition of an action might have a calming effect,

45Mason, G.J. and Latham, N. 2004. Can’t stop, won’t stop: Is stereotypy a reliable animal welfare indicator? Animal Welfare 13: S57–S69.

Dfraser_C007.indd 143

6/27/2008 4:57:58 PM

144

Studying Animal Welfare

analogous to humans rocking, pacing or using a mantra to meditate. Second, they suggested, stereotyped behaviour could be a kind of ‘do-it-yourself enrichment’. Wiepkema’s calves had no roughage to act as the object of foraging behaviour, but those animals that learned to do the tongue-rolling action without roughage were able to prevent at least some of the harmful consequences of being unable to forage. In such situations, abnormal behaviour may be close enough to the original source behaviour to have at least some of its beneﬁcial effects, but how successfully the abnormal behaviour substitutes for the real thing would need to be determined case by case. In situations where abnormal behaviour helps to mitigate the harmful consequences of an unsuitable diet or environment, preventing the behaviour without solving the underlying cause may simply make matters worse. Some horse-owners, for example, go to lengths to prevent crib-biting by ﬁtting horses with a ‘cribbing collar’ which restricts the animal’s ability to arch its neck, or by electrifying the surfaces that the horses might use for cribbing, or even by a surgical operation in which certain muscles and nerves are altered so that the horse cannot perform the movements involved in cribbing.46 Such approaches could create the worst option of all: the conditions that cause the abnormal behaviour continue, but what little the animal can do to ameliorate the situation is prevented. WHAT CAN WE CONCLUDE about animal welfare from abnormal behaviour? First, we need to recognize that ‘abnormal behaviour’ is a catch-all phrase that includes a number of seemingly different phenomena. Moreover, the labels we use for different types of abnormal behaviour – displacement activities, adjunctive behaviour, redirected behaviour and so on – did not arise from a systematic attempt to classify behavioural abnormalities; rather, they appeared at different times as different scientists invented terms for the seemingly anomalous behaviour they encountered. Thus, instead of a well-thought-out classiﬁcation, we have a dog’s breakfast of categories, often without clear dividing lines between them. And in the case of stereotyped behaviour, at least, the label is commonly attached to different types of behaviour that may have little in common. For these reasons, we cannot expect that different types of abnormal behaviour will map neatly onto speciﬁc kinds of animal welfare problems. To make matters more complex, in some cases the performance of abnormal behaviour can have beneﬁcial effects, at least in the sense of helping to mitigate problems. Redirected or vacuum behaviour may have enough in common with its original source behaviour to produce some of the same beneﬁts, such as stimulating the secretion of digestive ﬂuids or hormones. More generally, the performance of a seemingly abnormal action may have effects on the neuroendocrine system such as

46McGreevy, P. and Nicol, C. 1998. Physiological and behavioral consequences associated with short-term prevention of crib-biting in horses. Physiology & Behavior 65: 15–23.

Dfraser_C007.indd 144

6/27/2008 4:57:59 PM

Abnormal Behaviour

145

reducing arousal or down-regulating physiological stress responses. Hence, animals that perform abnormal behaviour may sometimes be better off than animals kept under similar circumstances whose behaviour appears more normal. All that said, there are many ways in which behavioural abnormalities and animal welfare may be related. Abnormal behaviour may be a direct cause of injury to the animals themselves or to others. Abnormal behaviour can indicate that the environment is missing some important element that is needed for normal functioning of the animal. It may also reﬂect some negative affective state, such as frustration at not being able to act in a way that the animal is strongly motivated to do. Stereotyped and perhaps other repetitious actions may reﬂect an abnormal activation of the nervous system in the short term (as from amphetamines or distressing circumstances), or an underlying pathology of the brain. Hence, where animal housing or management conditions induce stereotyped or other abnormal behaviour, it is reasonable to suppose that the situation may be deﬁcient in some way and that the causes and effects warrant investigation.

Dfraser_C007.indd 145

6/27/2008 4:57:59 PM

Affective States

8

For many people, concern about animal welfare consists mainly – even solely – of concern about the affective states of animals. Recall Ruth Harrison complaining that modern production methods rob animals of ‘all pleasure in life’, and Peter Singer’s proposal that ‘the pain (or pleasure) that animals feel’ should concern us as much as similar states in human beings. To put such concerns into practice, many countries have created legislation designed to give animals in human care some degree of protection against what is termed ‘pain’, ‘distress’ and ‘suffering’. Even people who are concerned primarily with the basic health of animals, or with the ability of animals to lead natural lives, will often acknowledge that the animals’ own experience of the situation – its emotions, its feelings – constitute an important element of animal welfare. In responding to these concerns, animal welfare scientists have developed various methods designed to identify and quantify certain affective states in animals. Because such states cannot be observed directly, but must be inferred from other evidence, we need to be clear on the logic that is used to draw conclusions about affective states from empirical observations. The logic is not always explicit in the literature, so let us be as clear as possible on the different ways that conclusions about affective states are reached, and the strengths and limitations of the different approaches.1 IN SOME CASES, ANIMALS are thought to have speciﬁc signals that communicate particular affective states to other animals, and we can ‘listen in’ on this communication.2 For example, if an unweaned piglet becomes separated from its mother, its behaviour follows a typical sequence. The piglet begins by walking slowly and giving

1For

a valuable review of how behaviour is inﬂuenced by affect combined with cognition, see: Toates, F.M. 2004. Cognition, motivation, emotion and action: A dynamic and vulnerable interdependence. Applied Animal Behaviour Science 86: 173–204. 2I am grateful to D.M. Weary for his important contribution to the thinking in this section.

146

Dfraser_C008.indd 146

6/27/2008 12:43:35 PM

Affective States

147

Frequency (kHz)

8

0 0

1

2 Time (s)

3

4

Figure 8.1 Vocalizations of a piglet removed from its mother and litter-mates. The sonogram shows nine calls given over four seconds. The first are low-pitched grunts. Later calls rise to high-pitched squeals. Originally published as Figure 15.6 in Fraser and Weary (2005). Reproduced with kind permission of Dr. Dan Weary. Copyright Blackwell Publishing, 2005.

short, quiet grunts separated by a second or two of silence. The pattern of calling is so characteristic that an experienced pig keeper will immediately recognize such calls as coming from a lost piglet. If the piglet does not ﬁnd the mother, its movements gradually take on a more agitated appearance, the calling becomes louder and more frequent, and the calls change in nature. Instead of short grunts, the call will often rise at the end to a high-pitched squeal (Figure 8.1), and in time the squeals replace the grunts completely. Experiments have shown that newly separated piglets give more calls, especially more of the loud, high-pitched calls, if they have not been fed recently or if they are in a cool environment; both of these are conditions that presumably increase their need to be re-united with the mother. Moreover, by playing the calls to sows through a hidden speaker, we can see that sows approach the speakers, and they respond more vigorously to calls given by piglets in conditions of greater need.3 As a working hypothesis, it is easy to imagine how affect could play a central role in causing this behaviour. It seems plausible, for example, to suggest that piglets have been shaped by natural selection to experience an unpleasant emotional state – let’s call it ‘separation distress’ – when separated from the mother under conditions that would threaten their survival in the wild. Let us hypothesize further that this unpleasant state stimulates the agitated movement and the distinctive calling, that variation in calling (between individuals and over time) is due to variation in the strength of the affective state, and that the pattern of sound attracts the 3Literature on this subject is summarized by Weary, D.M., Ross, S. and Fraser, D. 1997. Vocalizations of isolated piglets: A reliable indicator of piglet need directed towards the sow. Applied Animal Behaviour Science 53: 249–257.

Dfraser_C008.indd 147

6/27/2008 12:43:35 PM

148

Studying Animal Welfare

mother and is likely to lead to a reunion. If this hypothesis is correct, then the calls serve as a signal that communicates separation distress to the mother, and we can use the calls to identify (and to a degree quantify) this affective state.4 It is possible, of course, that the hypothesis about separation distress is incorrect. Perhaps natural selection simply wired the nervous system of piglets so that they respond reﬂexively and unconsciously to certain stimuli (perhaps the absence of the familiar smell of the mother) with the calls and other behaviour. Nature provides many examples of organisms responding in adaptive ways in situations where we might not expect an affective state to be involved: sunﬂowers orienting their heads toward the sun, woodlice scurrying away when exposed to light. What is our rationale for suggesting that the calls of the piglet are mediated by an affective state rather than occurring as an unconscious response? One element of the logic is that the calls are accompanied by other responses. The separated piglet also dashes about in a seemingly agitated manner, and it urinates and defaecates far more often than we would expect under normal circumstances. Moreover, the different responses are correlated: piglets that give the greatest number of high-pitched calls are generally the most active and defaecate the most often; and older piglets, which are less dependant on the mother, show less of all these responses than young ones.5 To explain these ﬁndings as unconscious, reﬂex responses, we would need a group of hypotheses, one to explain why the absence of the mother stimulates rapid movement, another to explain why it causes increased defaecation, another to explain why the more rapid movements tend to accompany louder calls, and so on. In contrast, the affect-based hypothesis allows us to postulate one response to separation – a feeling of distress – which then leads to a characteristic set of responses, only some of which (the calling and perhaps the rapid movement) serve any obvious adaptive function by helping to reunite the piglet with the mother. The affective hypothesis also makes further predictions we could test, for example that the shrieking piglet would also show physiological signs of upset such as the increase in heart rate and production of adrenal hormones that are commonly seen during emotional arousal. In nature, we would expect that many affective states will not have corresponding signals. This is because signals are expected to evolve only in those cases where the signals would provide a ﬁtness advantage – that is, where animals that signal a need would, on average, leave more copies of their genes than animals that do not. For this to happen there must be a recipient who will detect the signal and then act in a way that assists the sender. Separation distress is a perfect example because the mother has presumably been shaped by natural selection to protect her offspring, and the signal allows her to do this more effectively. We might expect, then, that signals 4Weary,

D.M. and Fraser, D. 1995. Signalling need: Costly signals and animal welfare assessment. Applied Animal Behaviour Science 44: 159–169. 5Fraser, D. 1975. Vocalizations of isolated piglets. I. Sources of variation and relationships among measures. Applied Animal Ethology 1: 387–394.

Dfraser_C008.indd 148

6/27/2008 12:43:35 PM

Affective States

149

of separation distress would be widespread among species of animals where one or both parents provide care for their young, and this seems to be the case. Ducklings, chicks, and other birds that follow the mother give characteristic peeping sounds if they become separated; and on the farm no sound is more distressing than the bawling of a calf that has bonded with its mother and is then separated from her. Alarm calls provide another example where we might expect evolution to favour the signalling of affect. In this case there is a signaller who notices and signals some potential source of danger such as an approaching predator, and there are recipients who can detect the signal and then avoid the danger. As long as the recipients – or a reasonable proportion of them – are genetically related to the signaller, then signalling is likely to protect the kin or even the offspring of the signaller, and the basic conditions are met for signalling to evolve. And again, in species where families or extended families tend to remain together, signals of alarm appear to be widespread. One example comes from vervet monkeys (Cercopithecus aethiops) which give three distinct alarm signals, one for an eagle, one for a leopard and one for a snake. Monkeys hearing the calls tend to take appropriate action: after the leopard signal they run up a nearby tree, but after the eagle signal they drop down from the treetops into the branches below. But do the calls simply alert the monkeys to an approaching danger and they then see the predator themselves and act accordingly? In a famous study described in their book How Monkeys See the World, primatologists Dorothy Cheney and Robert Seyfarth played the different calls from a speaker when no predator was nearby, and they found that the monkeys acted appropriately based on the sound alone. In this case, in addition to its affective component (‘Be afraid!’), the signal also appears to have a referential component (‘Be afraid of something overhead!’).6 Hunger in dependent offspring is a third situation in which we might expect signalling of unpleasant affect. Here again the basic conditions for evolution of a signalling system seem to be met: a signaller that needs assistance, and genetically related recipients (the parents) that can detect the signal and improve their own reproductive success by responding with a supply of food. The most spectacular examples are the ‘begging calls’ of nest-reared birds which are totally dependent on the parents for their food. The effectiveness of begging calls was illustrated in a classic experiment by the Finnish ornithologist Lars von Haartman in the 1950s. He studied food provisioning by pied ﬂycatchers (Ficedula hypoleuca) that were raising broods in specially designed nest-boxes in which von Haartman could temporarily conceal a second brood of hungry nestlings. Exposed to a double dose of begging calls, from both their own brood and the concealed brood, the parents increased their rate of bringing food to the nest, to the beneﬁt of the fortunate offspring.7 6Cheney

and Seyfarth, 1990. a modern version of this classic experiment see Ottosson, U., Bäckman, J. and Smith, H.G. 1997. Begging affects parental effort in the pied ﬂycatcher, Ficedula hypoleuca. Behavioral Ecology and Sociobiology 41: 381–384. 7For

Dfraser_C008.indd 149

6/27/2008 12:43:35 PM

150

Studying Animal Welfare

Begging calls, however, bring us to a core problem in interpreting signals: what is there to prevent animals from giving exaggerated signals – signals that are not really a reﬂection of their needs? If by giving extra begging calls, the young can make the parents work themselves to the bone bringing food, why not do this and live off the fat of the land? If signals can be ‘dishonest’ in this sense, then perhaps they cannot be used as true indicators of the strength of an affective state. A solution has been worked out in the theory of ‘honest signalling’, which identiﬁes the conditions that would have to be met for evolution to favour signalling that corresponds to the needs of the signaller. The key criterion is that producing the signal should incur some ﬁtness cost to the signaller. The separated piglet that dashes about and deafens the world with its calls, and the nest of begging birds calling so distinctly that any passing cat can tell there are young birds in the tree, seem likely to attract predators as well as parents. We would not expect evolution to favour gratuitous (‘dishonest’) signalling where such risks are involved unless there was real need as well. Thus, when we want to use signals to infer affective states, we need to ensure that the conditions for honest signalling are met.8 Separation distress, alarm and hunger are three cases where, under the right social circumstances, we might expect evolution to favour signals of affective states, but there are many other cases where we would not expect affective states to have corresponding signals. For example, if wildebeest were to give some signal of pain from an injured leg or to call out from the pain of calving, there is little chance that a related animal would come to their aid, but more chance that the signal would be detected by potential predators. Hence we might expect that species which, in a wild state, are subject to predation and are unlikely to receive relevant social support might show little outward sign of states such as pain; or alternatively, could it be that they have evolved a tendency not to experience pain at all? We will return below to how we may be able to separate these two possibilities, but ﬁrst let us deal with other indicators of affective states. IN CASES WHERE ANIMALS do not signal an affective state, the state may still be accompanied by some characteristic behaviour or other change. For example, if a person holds a chicken (or animal of many other species) ﬁrmly on a ﬂat surface for a few seconds and then releases it, the chicken is likely to remain immobile for some seconds or minutes in a reaction commonly called ‘tonic immobility’ or ‘death-feigning’. Tonic immobility has commonly been viewed as an indicator of fear, partly because it occurs in situations that might cause fear, partly because procedures intended to calm the birds (tranquillizers, habituation) reduce the 8Weary

Dfraser_C008.indd 150

and Fraser, 1995.

6/27/2008 12:43:35 PM

Affective States

151

Figure 8.2 A mechanical ‘chicken harvester’ which moves through a barn of chickens, gathers them between rotating rubber fingers, and moves them onto a conveyor belt that deposits them in shipping cages. Reproduced with permission of Anglia Autoflow Ltd, through the kind assistance of Barry Landymore.

response, and partly because tonic immobility persists longer if chickens have just been exposed to loud noises or other events intended to create fear.9 An interesting application of this method occurred in a debate about how best to catch chickens. On many chicken farms, crews of people are employed to catch the birds for shipping when they reach market weight. Typically each person enters the barn, grabs several birds by the legs, carries them upside-down, drops them in the shipping crates, and repeats this until the barn has been emptied. A newer alternative is the mechanical ‘chicken harvester’ – a large machine that moves slowly through the building, gathers up birds in a set of rotating rubber ﬁngers, and moves them onto a conveyor belt which then transfers them to the crates (Figure 8.2). When mechanical harvesters ﬁrst appeared, there was concern that they would cause fear and panic in the birds. However, in testing the equipment, Ian Duncan and co-workers found that immediately after being 9Gallup, G. 1981. Hypnosis. Pages 294–297 in The Oxford Companion to Animal Behaviour (D. McFarland, editor). Oxford University Press, Oxford; Jones, R.B. 1986. The tonic immobility reaction of the domestic fowl: A review. World’s Poultry Science Journal 42: 82–96.

Dfraser_C008.indd 151

6/27/2008 12:43:35 PM

152

Studying Animal Welfare

caught, birds captured by machine remained in tonic immobility for signiﬁcantly less time than birds that had been captured by hand. The researchers also recorded the birds’ heart rates as a general indicator of activity of the sympathetic nervous system. Although heart rate increased sharply during catching by both methods, it returned to normal more quickly for birds that had been captured by machine. Thus, both lines of evidence suggested that machine catching caused less fear than manual catching.10 ‘Vigilance’ is another possible indicator of fear. When we observe wild birds at a feeder or antelope on open grassland, the animals generally alternate between feeding (with the head lowered) and standing with the head raised in a way that allows them to scan the environment. This ‘vigilance’ is increased, and feeding time is correspondingly decreased, if the apparent risk of predation is high. There is also a degree of safety in numbers: animals in large groups, where there are many eyes and ears to detect danger, tend to spend less time in vigilance and more time feeding.11 A Canadian research team developed a test of vigilance as a way to quantify fear in dairy cattle. They designed a feeder structured so that a cow had to raise her head from the feeder in order to look around. Cows were then offered a highly palatable type of food in the feeder, and an observer recorded the time they spent with the head raised. When ﬁrst exposed to the novel feeder the cows were vigilant during much of the sessions, but this behaviour declined as the cows became accustomed to the environment over repeated trials. When a dog was present, vigilance increased greatly. In another experiment, cows were exposed for three weeks to one person who handled them roughly and a second person who handled them gently. When the cows were then observed at the experimental feeder, they spent more time vigilant if the rough handler was present than in the presence of either the gentle handler or an unfamiliar person. The team saw the vigilance test as an improvement over ‘approach tests’ for measuring fear of humans. Recall how Paul Hemsworth tested animals for fear of humans by bringing them individually into an enclosure that contained an unfamiliar person, and then monitoring how quickly the animal approached the person and how much time it spent in proximity to the person. Animals that were slow to approach, and that spent little time near the person, were said to show higher levels of fear of humans. The Canadian team noted that in such tests the animal’s behaviour could be inﬂuenced by its level of curiosity or by whether it had learned to associate people with food. Moreover, in an approach test there is little or no disadvantage to an animal that remains distant from the person. The team argued 10Duncan, I.J.H., Slee, G., Kettlewell, P., Berry, P. and Carlisle, A.J. 1986. Comparison of the stressfulness of harvesting broiler chickens by machine and by hand. British Poultry Science 27: 109–114. 11Elgar, M.A. 1989. Predator vigilance and group size in mammals and birds: A critical review of the empirical evidence. Biological Reviews 64: 13–33.

Dfraser_C008.indd 152

6/27/2008 12:43:41 PM

Affective States

153

that in the vigilance test, fear is less likely to be confounded with other factors, and by making the animal choose between feeding and vigilance, some cost was created so that the behaviour ‘should be more closely related to the degree of the animal’s fear’.12 In these examples, the working hypothesis is that a type of behaviour, such as tonic immobility or vigilance, evolved to accompany a particular affective state (in this case fear) not as a signal that communicates the state to other animals, but because the behaviour confers a ﬁtness advantage in some other way. It is easy to see why vigilance would be an adaptive response to fear, but what about tonic immobility? One possible explanation is that tonic immobility evolved as a response to being captured by a predator, because predators may simply store an immobile prey animal (who can later escape) without making further attempts to kill it. If this is true, then this fortuitous situation leaves us with a plausible tool for assessing fear, or at least the type of fear induced by being captured by a predator. EVEN IF THERE IS no speciﬁc behaviour that accompanies an affective state, it may still be possible to use general changes in either physiology or behaviour as indirect evidence that affect is present. We have already encountered the use of physiological indicators such as increased heart rate and activation of the SAM and HPA systems as indices of fear and pain. These physiological changes are not, of course, speciﬁc to any one affective state, but where there is reason to believe that a state such as pain or fear is likely to be present, then activation of these general responses may help to quantify it. An instructive example comes from studies of tail-docking of lambs. Lambs often have their tails docked because long tails can become matted with faeces and provide a breeding area for parasitic insects. The docking can be done surgically with a knife or docking tool, or by applying a tight elastic ring which cuts off the circulation and causes the tail tissue to die. In one study, behavioural observations showed that docking with elastic rings gave rise to a strong response: in the hour after application of the ring, the lambs became very restless, lying down and standing up repeatedly, and lying on their sides in a manner that is not usually seen when lambs rest normally. Behavioural changes were less noticeable when the tail was simply cut off with a knife, and on this basis the observers had concluded that ring docking, although bloodless, actually causes ‘more acute pain’ to the lambs than surgical docking.13

12Welp, T., Rushen, J., Kramer, D.L., Festa-Bianchet, M. and Passillé, A.M.B. de, 2004. Vigilance as a measure of fear in dairy cattle. Applied Animal Behaviour Science 87: 1–13. The quotation is on page 2. 13Molony, V., Kent, J.E. and Robertson, I.S. 1993. Behavioural responses of lambs of three ages in the ﬁrst three hours after three methods of castration and tail docking. Research in Veterinary Science 55: 236–245. The quotation is on page 244.

Dfraser_C008.indd 153

6/27/2008 12:43:41 PM

Studying Animal Welfare

Cortisol (ng ml−1)

154

Figure 8.3 Plasma cortisol concentration (mean) in lambs whose tails were docked either surgically by knife, or by the application of a tight rubber ring which cut off blood circulation and caused the tail tissue to die. Redrawn based on data from Lester et al. (1996). Reproduced with kind permission of the New Zealand Veterinary Association.

Animal welfare scientist David Mellor and co-workers repeated this study but using blood cortisol levels as well as behaviour to monitor the lambs’ reactions. They found the expected period of restless movement and abnormal lying after the elastic rings were applied. The surgically docked lambs also showed behavioural abnormalities including walking with stiff movements. The team argued that because the behavioural changes caused by the two procedures were different in kind, they could not be used to say which procedure was more distressing. However, cortisol responses showed a distinct difference between the treatments. Both groups had a sharp rise in plasma cortisol in the hour after the procedure, but the peak was higher with surgical docking and it remained high throughout the four hours that the lambs were monitored, whereas cortisol levels returned to normal soon after the ring procedure (Figure 8.3). On this basis the team argued that the surgical treatment ‘caused greater and more protracted distress’.14 Like general physiological changes, general behavioural responses have also been used to quantify short-term pain or fear. Animal scientist Karen SchwartzkopfGenswein and co-workers used such an approach to test the painfulness of 14Lester, S.J., Mellor, D.J., Holmes, R.J., Ward, R.N. and Stafford, K.J. 1996. Behavioural and cortisol responses of lambs to castration and tailing using different methods. New Zealand Veterinary Journal 44: 45–54. The quotation is on page 45.

Dfraser_C008.indd 154

6/27/2008 12:43:41 PM

Affective States

155

different methods of branding beef cattle for identiﬁcation. A traditional method of identifying cattle in western North America is to use a hot branding iron to sear the calf’s skin and create a distinctive pattern of scar tissue. A more modern alternative is to put the branding iron in liquid nitrogen and use it to freeze the skin. It is generally assumed that freeze branding is less painful than hot-iron branding, but is this true? To test this, Schwartzkopf-Genswein and co-workers restrained cattle in a holding chute and monitored their movements while the cattle received either hot-iron branding or freeze branding. In addition to video-recording the animals’ behaviour, the team used load cells and strain gauges to register the amount of force the animals applied to the equipment in the sudden movements they made when the branding irons were applied. Cattle branded by the hot iron applied greater force and longer duration of force against the restraining chute; they were also more likely to ﬂick their tails, kick, fall and vocalize than the animals that received freeze branding. The freeze branding did produce deﬁnite behavioural changes, but these were less pronounced. The team concluded that both methods cause pain but that hot-iron branding is more painful than freeze branding.15 In another interesting approach, Ian Duncan and David Wood-Gush looked for signs that could be used to identify frustration in domestic hens. In the hour before laying an egg, domestic hens will normally enter a secluded nest box if one is available. When housed in standard ‘battery’ cages, no nest box is present and the hens must lay their eggs on the ﬂoor of the cage. Hens in this situation become noticeably more active shortly before laying, but is this just normal pre-laying activity or does it indicate some form of frustration at not being able to ﬁnd a suitable nest site? To answer this question, Duncan and Wood-Gush deliberately caused frustration in hens in order to see whether frustration leads to any characteristic signs. They trained hens to expect a large meal in a certain container at a certain time of day, and then observed the hens when the food was covered with a transparent panel that prevented them from eating. They noted that hens responded to frustration with characteristic behaviour including stereotyped back-and-forward pacing, ‘displacement’ preening, and increased aggression toward less dominant birds. Duncan and Wood-Gush then looked for these indicators of frustration among birds in cages and found that the birds did perform the characteristic pacing and displacement preening just before the time of lay. They concluded that despite many generations in captivity, the hens retained a strong motivation to lay in a nest box (a conclusion that later research has conﬁrmed in other ways) and that preventing this behaviour was causing frustration.16 The approach gave convincing results 15Schwartzkopf-Genswein, K.S., Stookey, J.M. and Welford, R. 1997. Behavior of cattle during hot-iron and freeze branding and the effects on subsequent handling ease. Journal of Animal Science 75: 2064–2072. 16The work is summarized in Duncan, I.J.H. 1987. The welfare of farm animals: An ethological approach. Science Progress, Oxford 71: 317–326.

Dfraser_C008.indd 155

6/27/2008 12:43:42 PM

156

Studying Animal Welfare

in this case, but there is a risk that a given affective state might be manifested in different ways in different situations. A fearful rat, for example, might ﬂee, freeze, or bite depending on the circumstances. When an animal experiences a state such as fear or pain, its physiology and behaviour presumably undergo an immensely complex set of changes involving many aspects of the nervous system, the endocrine system, posture, bodily movements and so on. When we make simple measurements – such as the force exerted against a strain gauge, the duration of tonic immobility, or levels of blood cortisol – we capture only a small fraction of the changes that are occurring. An alternative approach is to sacriﬁce quantitative precision in favour of using a wider range of information. An example is a system developed by animal behaviourists Mirjam Kessler and Dennis Turner to monitor how quickly cats adapt to a boarding cattery.17 When cats are ﬁrst brought into a boarding facility or rescue shelter they generally show signs of fearfulness that may go on for days. They may assume a tense posture, retreat to the back of the cage, hide under any available bedding, and hiss and scratch if someone tries to touch them. Cat behaviourist Sandra McCune had described postures and movements of cats corresponding to what she viewed as seven levels of adaptation to a new environment. Kessler and Turner began their study by identifying cats that clearly ﬁtted into one of McCune’s categories, and they made further observations in order to capture other features of the animals’ posture and behaviour. The results formed the basis of the classiﬁcation system shown (in simpliﬁed form) in Table 8.1. The system uses a large number of features including posture, the position of the tail and ears, and the degree of opening of the eyes. Trained observers show a high but not perfect level of agreement in assessing cats. Disagreements can occur, for example, when the different items in the scoring system do not occur together in the expected way. For instance, how should a cat be rated if the tail is fully extended (level 1) but the eyes are opened in a normal manner (level 3)? Kessler and Turner named their system the cat ‘stress’ score although the system seems designed more speciﬁcally to assess the degree to which cats have adapted to an unfamiliar environment, rather than the much larger class of challenges (cold, heat, injury, disease) that are commonly assumed under the term ‘stress’. In some cases, statistical analysis has played an important role in identifying the items to be included in comprehensive scoring systems. British pain researchers John Roughan and Paul Flecknell studied 57 laboratory rats that had undergone abdominal surgery after receiving various doses of analgesics or saline as a control. Once the rats recovered from the general anaesthetic, Roughan and Flecknell watched them for 10 minutes and scored occurrences of eight types of behaviour. Although behaviour was extremely variable, analysis showed that there were four 17Kessler, M.R. and Turner, D.C. 1997. Stress and adaptation of cats (Felis silvestris catus) housed singly, in pairs and in groups in boarding catteries. Animal Welfare 6: 243–254.

Dfraser_C008.indd 156

6/27/2008 12:43:42 PM

Dfraser_C008.indd 157

Table 8.1 A simplified version of the ‘Cat-Stress-Score’ of Kessler and Turner (1997). For simplicity, this version shows only levels 1, 3, 5 and 7, and it omits certain items that can be scored only when cats are inactive.

Score

Body

Legs and tail

Head

Eyes

Ears and whiskers

Activity and vocalization

1 Fully relaxed

Laid out on side or back, belly exposed, normal breathing

Legs fully extended, tail extended or loosely wrapped

Laid on the surface with chin upwards or on the surface

Closed or half closed, may be blinking slowly, pupils normal

Normal: ears half back and whiskers lateral

Sleeping or resting, no vocalization

3 Weakly tense

Laid ventrally or sitting, belly not exposed, normal breathing

Legs bent, tail on the body or curved backward, tail may be twitching

Head above the body, some movement

Opened normally, pupils normal

Ears half back or erected to front or back and forward on head; whiskers lateral or forward

Resting, awake or actively exploring, meowing or quiet

5 Fearful, stiff

Laid ventrally or sitting, belly not exposed, normal or fast breathing

Legs bent, tail close to the body

On the plane of the body, little or no movement

Widely opened with pupils dilated

Ears partly flattened, whiskers lateral, forward or back

Alert, may be trying to escape; plaintive meow, yowling, growling or quiet

7 Terrorized

Crouched directly on top of paws, shaking, belly not exposed, fast breathing

Legs bent, tail close to the body

Lower than the body, motionless

Fully opened with pupils fully dilated

Ears fully flattened and back on head, whiskers back

Motionless and alert; plaintive meow, yowling, growling or quiet

6/27/2008 12:43:42 PM

158

Studying Animal Welfare

features – arching the back, abnormal gait, writhing and staggering or falling – which were signiﬁcantly more common in the rats that received little or no analgesic to manage the pain. On this basis, Roughan and Flecknell devised a composite scoring system based on these four behavioural traits, and suggested that it could be used as a simple way to identify rats that had not received adequate analgesia.18 An even more comprehensive and unstructured approach, known as ‘free choice proﬁling’ has been pioneered by animal welfare scientist Françoise Wemelsfelder. Wemelsfelder asked observers to watch the behaviour of a pig for several minutes in an open pen where the animal could explore the environment and interact with a seated person. The observers were then asked to describe the ‘expressive qualities’ of the pig’s behaviour. Without prompting, the observers selected terms such as ‘conﬁdent’, ‘playful’, ‘bold’, ‘anxious’ and ‘excitable’. The observers were subsequently asked to score various animals on a linear scale (running from minimum to maximum) corresponding to each of the descriptive terms they had selected. Adapting a statistical method originally used in the sensory evaluation of food, Wemelsfelder analysed the ratings of many observers to identify the major axes corresponding to the quantitative evaluations. The ﬁrst axis turned out to be correlated with terms such as ‘conﬁdent’ and ‘playful’ at the one end, versus ‘timid’ and ‘wary’ at the other. The second axis was correlated with terms such as ‘excitable’ and ‘alert’ at the one end versus ‘relaxed’ and ‘calm’ at the other. The results showed that the observers were very consistent when asked to rate the same animals on different occasions, and that there was a high level of agreement between observers.19 This approach differs from others in important respects. First, the observations are not limited to pre-selected behaviours such as vocalizing and approaching; rather, the observer watches the entire performance including subtle expressive elements such as the suddenness of bodily movements that might be impossible to quantify. Hence, although the observations are unstructured they have the potential to take much more detail into account. Second, the approach does not assume that speciﬁc elements of behaviour indicate pre-determined affective states; rather, the observers use their own intuitive judgement to select terms that seem (to them) to capture the expressive quality of the behaviour. The method takes advantage of the tremendous human capacity to recognize patterns in complex situations, but it has limitations. As Wemelsfelder and co-workers noted, chimpanzees sometimes adopt a facial expression that appears to be a broad grin. The behaviour is actually a threat display, but people, being accustomed to the human smile, are likely to misinterpret it as a friendly expression, 18Roughan, J.V. and Flecknell, P.A. 2003. Evaluation of a short duration behaviour-based post-operative pain scoring system in rats. European Journal of Pain 7: 397–406. 19Wemelsfelder, F., Hunter, T.E.A., Mendl, M.T. and Lawrence, A.B. 2001. Assessing the ‘whole animal’: A free choice proﬁling approach. Animal Behaviour 62: 209–220. The quotations appear on pages 211 and 217.

Dfraser_C008.indd 158

6/27/2008 12:43:42 PM

Affective States

159

at least when shown only a static photograph. And because the human smile is so universally recognized, different observers may well show a high level of agreement in this mistaken judgement. Hence, strong agreement by itself is no indication that a judgement is correct. If we assume, however, that such cases are the exception rather than the rule, then states such as playfulness, wariness and excitability may be captured more accurately and consistently by this type of overall evaluation than by more speciﬁc measures that lend themselves to precise quantiﬁcation. The use of the approach to date has tended to focus on temperament traits (conﬁdence, timidity), but it would seem a small step to use the method to quantify affective states such as pain, fear and contentment. THE ABOVE EXAMPLES HAVE focused mostly on negative states. Scientiﬁc study of positive states in animals raises additional complexities, but there have been some promising beginnings. The following few examples are intended to tempt scientists into considering what is, to date, an underdeveloped area of research. In a few cases, there may be actual signals of positive affect, analogous to signals of alarm or separation distress. Adolescent rats produce ‘chirps’ at a frequency of about 50 kHz (well beyond the range of human hearing) during play and when tickled by human handlers. Psychologists Jeffrey Burgdorf and Jaak Panksepp have suggested that these calls are analogous to human laughter, and they carried out simple studies that lend weight to the idea. In series of training sessions, a human handler tickled certain rats and gently touched other rats without tickling them. The rats that were tickled approached the person’s hand much more quickly than rats that were not, and they approached the hand more and more quickly as training progressed. These results suggested that the rats that were tickled found the experience pleasant. The tickled rats also produced more chirps during the tickling, and the rats that approached most quickly also vocalized most often. In addition, rats that were individually housed were more likely to show these effects than those that were socially housed, suggesting that tickling was more pleasurable for rats that could not engage in normal social contact with other rats.20 But why should animals evolve a signalling system to indicate that they are having fun? The evolutionary function of alarm calls or begging signals seems fairly obvious, but how would signals of positive affect contribute to survival and reproductive success? Part of the answer may hinge on the signiﬁcance of play. Play is widespread among animals, especially mammals, but its adaptive signiﬁcance is far from clear. Play may allow young animals to practise movements such as hunting and ﬁghting that will be used later in life for survival and reproduction. It may also involve stimulating the body in ways that promote neural development: the rough-and-tumble play that young children enjoy helps to stimulate the development of basic skills 20Burgdorf, J. and Panksepp, J. 2001. Tickling induces reward in adolescent rats. Physiology & Behavior 72: 25–38.

Dfraser_C008.indd 159

6/27/2008 12:43:42 PM

160

Studying Animal Welfare

such as righting and balance. Marek Špinka and co-workers have offered a more speciﬁc hypothesis. They suggested that in play, young animals deliberately put themselves at a disadvantage, and that this constitutes a kind of rehearsal for situations later in life when they will encounter similar disadvantages in, for example, ﬁghting or escaping from predators.21 If play is beneﬁcial, and if at least some play requires others to participate, then signals of enjoyment – the rat’s ‘laugh’, the ‘play face’ of chimpanzees and the ‘play bow’ of domestic dogs – could have evolved as a way to stimulate others to begin or prolong a bout of play. But signals of positive affect may do more than encourage play. If it is advantageous for animals to signal alarm when danger is present, perhaps it is also advantageous to have an on-going signal that keeps family and group members informed that no cause for alarm has been detected. The ‘singing’ of hens and the snufﬂy sounds that pigs make when rooting (‘contentedly’) in a ﬁeld, may be such signals. Mammalogist R.F. Ewer suggested that the purring of kittens could be a kind of ‘all’s-well’ signal that facilitates positive social interaction including nursing.22 Thinking in evolutionary terms may also help us identify the kinds of situations where positive affect is likely to occur.23 We can think of behaviour as serving two broad goals: to prevent harm in ‘need situations’ where there is some tangible threat to survival or reproduction; and to take advantage of ‘opportunity situations’ where there is a chance to promote long-term health and reproduction through certain actions. Need situations arise, for example, when an individual is becoming dehydrated and needs to drink in order to survive, when dependent offspring are in danger and require protection, or when an injury has occurred and the individual needs to rest the injured area so that healing can occur. Opportunity situations include the opportunity to groom, explore, play, or reinforce social bonds (activities that we assume have beneﬁts for the animal in the long term) at times when there are no pressing matters to attend to, and hence when there is little cost to engaging in these non-urgent activities. A little reﬂection on human experience suggests that need situations commonly involve negative affective states such as thirst, fear and pain, whereas the motivation in opportunity situations is based more on the pleasure inherent in the activity. Thus, we stop walking on an injured foot to relieve pain rather than to induce pleasure, but we play a game because we enjoy it, not because of an unpleasant state – analogous to pain or thirst – that spurs us to do so. Certain types of behaviour can, of course, occur in either need or opportunity situations. Thus we may feel driven to eat by a gnawing feeling of hunger when we need nutrients, but we may also eat a good-tasting desert at the end of a meal even though we are 21Špinka, M., Newberry, R.C. and Bekoff, M. 2001. Mammalian play: Training for the unexpected. Quarterly Review of Biology 76: 141–168. 22Ewer, R.F. 1973. The Carnivores. Weidenfeld and Nicolson, London, pages 336–338. 23This argument is developed in Fraser, D. and Duncan, I.J.H. 1998. ‘Pleasures’, ‘pains’ and animal welfare: Toward a natural history of affect. Animal Welfare 7: 383–396.

Dfraser_C008.indd 160

6/27/2008 12:43:43 PM

Affective States

161

not experiencing hunger. A neural analysis of such situations by physiologist Kent Berridge is consistent with the hypothesis that positive and negative affect are distinct processes. Berridge has shown that animals have two distinct types of motivation for food (he calls them ‘wanting’ or appetite, versus hedonic ‘liking’) and that they involve different neurophysiological mechanisms.24 This view of motivation suggests that positive affect is likely to accompany those behavioural activities that contribute to long-term ﬁtness but do not seem to meet any near-term need or threat. Depending on the species and life history, such pleasurable behaviour is likely to include play, grooming, hoarding, exploring, renewing territorial markings, and certain types of social behaviour. Hence, accommodating these types of behaviour (by enriching the environment, for example) could contribute to an animal’s welfare by promoting positive affect. Neuroendocrine approaches also show promise of shedding light on positive affect. Swedish physiologist Kerstin Uvnäs-Moberg and others have proposed a distinct neuroendocrine basis of positive social behaviour.25 For half a century, the hormone oxytocin has been recognized as the agent that causes contractions of smooth muscle in the mammary glands (thus triggering the ejection of milk) and in the uterus (where it causes contractions during the birth process). Only more recently were the psychological effects of oxytocin also recognized. In an early demonstration of this effect, a research team studied ewes that had recently given birth and were in the process of licking their newborn lambs and forming a lasting attachment. When the team brought a foreign lamb, the ewes generally butted it away. If, however, the experimenters manually distended the birth canal a second time, the ewes then licked and accepted the foreign lamb as if it were their own. The explanation appears to be that distension of the birth canal during parturition results in a large release of oxytocin which, acting in the brain, creates a period when the ewe will respond positively to a lamb and bond with it, and that even artiﬁcial distension of the birth canal can trigger a similar reaction.26 Uvnäs-Moberg and others have extended such ﬁndings by proposing a broad role for oxytocin in positive social behaviour. According to this hypothesis, oxytocin is commonly released during pleasurable tactile contact which may happen during nursing, mating, or other behaviour, and may be induced by pleasant stroking, massage or even acupuncture. Moreover, oxytocin has powerful ‘antistress’ properties. It reduces activity of the SAM and HPA systems, thus resulting in lower cortisol levels, reduced heart rate and blood pressure, and increased digestion of 24Berridge, K.C. 1996. Food reward: Brain substrates of wanting and liking. Neuroscience and Biobehavioral Reviews 20: 1–25; Berridge, K.C. 1999. Pleasure, pain, desire, and dread: Hidden core processes of emotion. Pages 525–557 in Well-Being: The Foundations of Hedonic Psychology (D. Kahneman, E. Diener and N. Schwarz, editors). Russell Sage Foundation, New York. 25Uvnäs-Moberg, K. 1998. Oxytocin may mediate the beneﬁts of positive social interaction and emotions. Psychoneuroendocrinology 23: 819–835. 26Keverne, E.B., Levy, F., Poindron, P. and Lindsay, D.R. 1983. Vaginal stimulation: An important determinant of maternal bonding in sheep. Science 219: 81–83.

Dfraser_C008.indd 161

6/27/2008 12:43:43 PM

162

Studying Animal Welfare

food. Uvnäs-Moberg suggests that oxytocin may play an important role in social bonding and the health-promoting effects of positive social contact. Thus she proposes that oxytocin is a key element of a generalized response system that is roughly the opposite to the ﬁght-or-ﬂight system made famous by Walter Cannon. Jaak Panksepp, whose work on rat ‘laughter’ was noted above, has also amassed a large volume of evidence of neural systems in the brain which control behaviour that appears to involve emotional states. Based on stimulation of speciﬁc regions of the brain combined with other evidence mainly from rats, Panksepp proposed seven ‘executive networks’, each involving an emotional component. Some of these systems – notably those whose emotional element consists of fear, aggressive motivation and separation distress – involve seemingly unpleasant states and promote escape and avoidance. Other systems – those underlying social play, exploration, maternal care and sexual behaviour – appear to involve pleasant states, and activation of these neural systems can serve as a reward. If Panksepp’s hypothesis is correct, then (in rats at least) the performance of these last four types of behaviour should be accompanied by positive affect.27 MANY OF THE METHODS described so far, such as the use of signals and general ‘stress’ responses, make the tacit assumption that when animals show evidence of a state such as pain or fear, they are indeed experiencing the state rather than merely responding in an automatic and unconscious way. Intuitively it may seem obvious that this is true; it seems implausible to suggest, for example, that the shrieking piglet or the death-feigning chicken is acting unconsciously without experiencing an affective state. Nonetheless, the evidence falls short of actually proving that affect is involved. Let us now turn to some examples that would be very difﬁcult to explain without invoking actual experience of affect. As one approach, many studies have used the method of allowing animals to selfadminister pharmaceutical products, especially analgesics to reduce pain, as a way of identifying whether the animals are experiencing an affective state. Pain researcher Francis Colpaert and co-workers have done many studies in which rats had the opportunity to administer analgesics to themselves. In one case they gave arthritic and non-arthritic rats a choice of drinking from two water bottles, one of which contained sweetened water and the other a dilute but unpalatable solution of fentanyl, an opiate analgesic. Healthy rats consumed very little of the fentanyl, but arthritic rats consumed substantial amounts, and the time course of self-administration corresponded with changes in the severity of the arthritis. To rule out the possibility that self-administration was simply due to an addiction to the opiate properties of the 27I have based this brief discussion on: Knierim, U., Carter, C.S., Fraser, D., Gartner, K., Lutgendorf, S.K., Mineka, S., Panksepp, J. and Sachser, N. 2001. Good welfare: Improving quality of life. Pages 79–100 in Coping with Challenge: Welfare in Animals including Humans (D.M. Broom, editor). Dahlem Workshop Report 87, Dahlem University Press, Berlin. For more detailed evidence see: Panksepp, J. 1998. Affective Neuroscience: The Foundations of Human and Animal Emotions. Oxford University Press, New York.

Dfraser_C008.indd 162

6/27/2008 12:43:43 PM

Affective States

163

drug, the researchers also studied the rats after treating them with a corticosteroid that would be expected to reduce the severity of the arthritis. The corticosteroid proved effective in reducing the amount of fentanyl that the rats consumed; hence, the researchers concluded that the consumption of fentanyl was due to pain from the arthritis, not to addiction. In human medicine, self-administration of pain-killers is regarded as an objective measure of the severity of chronic pain. Colpaert and colleagues concluded that self-administration of fentanyl provides the same kind of objective indicator of chronic pain in rats.28 This line of research gives us a way to answer a question posed earlier in the chapter: if a wildebeest with an injured leg shows no outward sign of pain, how can we tell whether it is simply very good at hiding pain (presumably because limping wildebeest tended to attract predators during the evolution of the species), or whether it does not actually feel pain at all? Following Colpaert’s logic we could propose that if the wildebeest walks normally but voluntarily consumes bad-tasting analgesics, then we would conclude that it is experiencing pain even though its outward behaviour provides little indication. Scientiﬁc attention has also focused on the experience of anxiety in animals. Japanese animal behaviourist Yasushi Kiyokawa and co-workers found that a chemical released by male rats during periods of stress appears to serve as an ‘alarm pheromone’ that might create anxiety in other rats. They initially tested the effect of the pheromone on ‘stress-induced hyperthermia’. This refers to the increase in body temperature that commonly occurs in the hour or so after a stressful event, presumably because of activation of the sympathetic nervous system. Kiyokawa and colleagues had found that rats exposed to the pheromone showed a greater stressinduced hyperthermia than control rats, but it was not clear whether this reﬂected increased anxiety in the animals or merely an unconscious physiological response. To answer this question the team placed rats individually in a large and brightly lit enclosure, and allowed them to explore it for ﬁve minutes. Then a small ‘hiding box’ was placed inside the enclosure (Figure 8.4). The rats could then hide in the box, peer out from the entrance hole (a behaviour the researchers called ‘risk assessment’), or they could continue to explore and groom themselves in the open. During the tests, ﬁlter papers impregnated with either the pheromone or plain water were hung nearby. Rats tested in the presence of the pheromone spent more time concealed in the box and peering out the entrance, less time in the open, and performed virtually no grooming. Because the pheromone induced seemingly cautious behaviour as well as higher body temperature, Kiyokawa and co-workers concluded that it was not simply causing an automatic activation of the sympathetic nervous system, but rather produced an anxiety-like state in the rats exposed to it.29 28Colpaert, F.C., Tarayre, J.P., Alliaga, M., Slot, L.A.B., Attal, N. and Koek, W. 2001. Opiate self-administration as a measure of chronic nociceptive pain in arthritic rats. Pain 91: 33–45. 29Kiyokawa, Y., Shimozuru, M., Kikusui, T., Takeuchi, Y. and More, Y. 2006. Alarm pheromone increases defensive and risk assessment behaviors in male rats. Physiology & Behavior 87: 383–387.

Dfraser_C008.indd 163

6/27/2008 12:43:43 PM

164

Studying Animal Welfare

Figure 8.4 Schematic diagram of the apparatus used by Yasushi Kiyokawa and co-workers to test the effects of ‘alarm pheromone’ on the behaviour of rats. While in the apparatus, rats could hide in the box, peer out from the opening, or remain in the exposed area. Either water or alarm pheromone was applied to the filter papers before the test. From Kiyokawa et al. (2006). Reproduced with kind permission of Dr. Kiyokawa and Elsevier, © Elsevier, 2006.

In another interesting ﬁnding, animal welfare scientist Emma Harding and co-workers looked for evidence of an affective state in rats analogous to depression in humans. It is known from clinical psychology that depressed people tend to put a pessimistic interpretation on ambiguous events. Wishing to test for a similar phenomenon in animals, Harding and co-workers trained rats to distinguish between two tones. When the ‘positive’ tone sounded, pressing a lever resulted in the rat receiving a food pellet, but when the ‘negative’ tone sounded, pressing the same lever resulted in 30 seconds of unpleasant noise. The rats learned to press in response to the positive tone and avoid pressing in response to the negative tone. The rats were then given tones intermediate between the two, to see whether they would interpret the ambiguous sounds as positive or negative. Before this test, some of the rats were exposed to several days of unpredictable events in their home cages – the sudden appearance of a strange rat in their cage, lights coming on and off at irregular times, the cage suddenly tilting to one side. Such unpredictability has been shown by other research to induce signs of a mild, depression-like state. Rats that were exposed to this type of unpredictability tended to interpret the ambiguous tones more pessimistically, pressing the lever less quickly and less often than those that had been spared the unpredictable events. Harding and co-workers

Dfraser_C008.indd 164

6/27/2008 12:43:43 PM

Affective States

165

concluded that such a method could be used to detect ‘an enhanced expectation of positive events – a correlate of happy mood in humans’.30 Finally, neurophysiologist Michael Gentle devised a simple but intriguing approach to test whether animals have conscious experience of pain. If a person accidentally touches a burning object, a simple reﬂex in the nervous system may cause an immediate movement away, even before any sensation of pain occurs. The person can then report when and if a conscious experience of pain develops. Since animals cannot report directly on their experiences, how can we be sure, when we see some change in their behaviour, that they are actually experiencing pain instead of acting by reﬂex? Gentle noted that the human experience of pain can be mitigated by redirecting the patient’s attention elsewhere through such methods as relaxation training, hypnosis and other therapies. He reasoned that if an animal’s reaction to some harmful event is simply an unconscious reaction, then shifting the animal’s attention should not inﬂuence the response. If, however, the animal experiences pain, then redirecting its attention should reduce the signs of pain as it does in the case of human patients. To test this idea, Gentle injected sodium urate crystals into one leg joint of domestic chickens. This produced a temporary arthritis-like condition and led to changes in behaviour consistent with mild pain lasting about three hours. Birds that were kept in their barren home cage during this time would avoid putting weight on the affected leg, would sit or stand on the good leg, and if encouraged to walk they would do so with a limp. However, these changes in behaviour were greatly reduced or eliminated by placing the bird into a pen where the ﬂoor was strewn with wood shavings that attracted the hen’s attention, and especially if a second, unfamiliar chicken was present to further distract the injected bird. Much the same occurred if the injection was given shortly before the hen was due to lay an egg – a time when hens are strongly focused on ﬁnding a nesting place. As Gentle described it, During this period, the birds’ attention seemed to be fully occupied with ﬁnding a suitable nest site, and they seemed to be totally unaware of pain coming from the ankle. Following oviposition [egg laying], the birds would brieﬂy feed and would then show pain-related behaviour, either one-legged standing or sitting.

Gentle concluded that because the birds’ reaction to the injection was modiﬁed by shifting its attention elsewhere, the reaction cannot have been an unconscious adjustment of behaviour that happened automatically. Instead, it must have been mediated by awareness. Gentle argued that this simple result ‘has far-reaching consequences for our understanding of pain in birds’ and ‘provides evidence for consciousness’.31 30Harding,

E.J., Paul, E.S. and Mendl, M. 2004. Animal behaviour: Cognitive bias and affective state. Nature 427: 312. 31Gentle, M.J. 2001. Attentional shifts alter pain perception in the chicken. Animal Welfare 10: S187–S194. The quotations are from pages S187 and S192 respectively.

Dfraser_C008.indd 165

6/27/2008 12:43:44 PM

166

Studying Animal Welfare

EARLIER WE DISCUSSED THE view of John Watson, Niko Tinbergen and others that it is inappropriate, and perhaps futile, to use science to try to understand the affective states of animals. Let us now return brieﬂy to those views and how they ﬁt with the methods and ideas described in this chapter.32 As philosopher Bernard Rollin has pointed out, the exclusion of affect from behavioural psychology and ethology reﬂected the inﬂuence of Positivism, the school of thought associated with the ‘Positive Philosophy’ of the French sociologist Auguste Comte (1798–1857).33 Comte, working in an emerging science during the nineteenth century, was trying to deﬁne a clear demarcation between science and such non-scientiﬁc ﬁelds as metaphysics and theology. Part of his solution was to maintain that science deals only with the material world, not with immaterial souls or hypothetical entities.34 Positivism had widespread effects on science generally, and it inﬂuenced the place of affect in science in perhaps three ways. First, with its emphasis on tangible, observable phenomena, Positivism held that we should not postulate unobservable processes to explain observable ones. Under this inﬂuence, some physicists in the early twentieth century considered it unscientiﬁc to invoke entities such as atoms and electrons, which could not be observed directly, to explain physical and chemical processes.35 Under this same inﬂuence, the emotions, feelings and other mental states of animals were often banished as explanatory concepts. Second, Positivism held that processes that cannot be observed, even if they may occur in the physical world, fall outside the realm of science. On this basis, whole areas of investigation – including evolution and the origin of the universe – were deemed by some Positivists to be unsuitable for scientiﬁc enquiry, and the same thinking was applied to the subjective experience of animals. Finally, Positivism held that the sciences are built upon each other in a hierarchical manner.36 Thus, Comte considered that sociology rests on biology, which rests on chemistry, which rests on physics. According to this view, the behaviour of animals was seen as ultimately explainable in terms of underlying physiological processes. When the physiological processes cause the movements of animals that we call behaviour, perhaps the animal does experience some agreeable or disagreeable feeling, but the subjective experience is merely a by-product – an 32Parts

of this discussion are based on Fraser, D. 1999. Animal ethics and animal welfare science: Bridging the two cultures (The D.G.M. Wood-Gush Memorial Lecture). Applied Animal Behaviour Science 65: 171–189. 33Rollin, B.E. 1990. The Unheeded Cry. Oxford University Press, Oxford. 34Kolakowski, L. 1968. The Alienation of Reason: A History of Positivist Thought (N. Guterman, translator). Doubleday, New York. 35Feigl, H. 1995. Positivism and Logical Empiricism. Pages 630–636 in The New Encyclopedia Britannica, 15th edition, Volume 25. Encyclopedia Britannica Inc., Chicago. 36Lenzer, G. (editor). 1975. Auguste Comte and Positivism: The Essential Writings. University of Chicago Press, Chicago.

Dfraser_C008.indd 166

6/27/2008 12:43:44 PM

Affective States

167

‘epiphenomenon’ – which accompanies but does not in any way cause or explain the occurrence of the behaviour. Charles Darwin’s famous contemporary Thomas Huxley provided a classic example of this view: The consciousness of brutes would appear to be related to the mechanism of their body simply as a collateral product of its working, and to be completely without any power of modifying that working as the steam-whistle which accompanies the work of a locomotive engine is without inﬂuence on its machinery.37

According to this view, the subjective experiences of animals are actually irrelevant to science because they play no role in causing events in the physical world. In summary, then, under the inﬂuence of Positivist thinking, the subjective experiences of animals were seen as inadmissable as explanatory concepts, as not amenable to scientiﬁc study, and as playing no role in the causation of behaviour. The Positivist view of science, and the constraints it placed on the study of animal behaviour, have been criticized by both scientists and philosophers in arguments that I will not try to review here.38 Sufﬁce it to say that from today’s vantage point, it seems rather odd to suggest that trying to understand affective and other mental states is ‘unscientiﬁc’. As Jeffrey Rushen has pointed out, there is nothing ‘unscientiﬁc’ in hypothesizing processes that cannot be observed directly.39 No one has ever witnessed the evolution of a new genus of animals, nor watched the Big Bang, nor dissected a quark, yet it is accepted practice for scientists to postulate the existence and properties of such processes and entities, even though they cannot be observed directly, in order to make sense of available data and to build theories which generate testable predictions. But how exactly could the postulating of affective states help to provide more satisfactory explanations and generate more correct predictions than explanations that avoid reference to affective and other mental states? Some insight is provided by psychologist Ronald Baenninger, who used a description of the behaviour of a hungry dog to comment on the reluctance of psychologists to postulate expectation in animals: Positive or negative reinforcement may be adequate to explain a dog’s movement to a place where food … has previously occurred. But what of the responses that were not even present during the acquisition trials? Why does the dog perk up its ears, whimper,

37Campbell, N. 2004. What was Huxley’s epiphenomenalism? Biology and Philosophy 16: 357–375. The passage quoted appears on page 361. 38For example, see Grifﬁn, 1976, 1984 and 1992; Rollin, 1990; Dawkins, 1990; and Midgley, 1983. 39Rushen, J. 1985. The scientiﬁc status of animal consciousness. Applied Animal Behaviour Science 13: 387–390.

Dfraser_C008.indd 167

6/27/2008 12:43:44 PM

168

Studying Animal Welfare

turn its head and eyes toward the learned location, and prepare to spring toward it? These responses appear only on subsequent trials, and persist into extinction trials.40

If we wanted to explain the dog’s behaviour purely in terms of stimulus–response links acquired because of food reward, we could (Baenninger suggested) postulate a number of stimulus–response chains of acquisition for each of these movements, but it would be simpler and less cumbersome to postulate a single mediating mental event called expectation, which causes dogs to attend to certain stimuli and act in certain ways. We see in this example the importance of behavioural detail. If we conﬁne scientiﬁc investigation to fairly abstract and quantitative data, such as the average speed with which dogs run to a food source after different periods of deprivation, we may be able to express the ﬁndings in a mathematical relationship, and we may gain nothing by postulating that dogs feel hunger and expect food. But if we try to account for qualitative, narrative data – for why a particular dog responds in a particular, detailed manner – then postulating subjective states such as hunger and expectation may be essential to achieve workable explanations. What links Baenninger’s example to those we saw earlier in Romanes’ narrative about the blind elephant, Yerkes’ accounts of Congo, and Goodall’s description of the death of Flint, is attention to context, behavioural detail, and unique individual responses. Here, perhaps, we see the inﬂuence of a type of thought that has come to be called ‘Feminist’. Feminist thought is often described as narrative and contextual more than theoretical, as particular rather than abstract, as relational more than rational, and as empathetic rather than objectifying. In these terms, we can see some of the hallmarks of Feminist thought in the approach to science shown in the various examples from Romanes to Goodall. Thus, if science moved away from considering the experiential lives of animals at a time when Positivist thought was prevalent, today, when Feminist thought is prevalent, we are seeing that trend gradually reversed, and a respectable literature on the study of affect is emerging. Marian Dawkins’ 1980 book Animal Suffering has now been joined by other scientiﬁc titles such as Well-Being: The Foundations of Hedonic Psychology41 and Panksepp’s Affective Neuroscience: The Foundations of Human and Animal Emotions.42 Inasmuch as this work will lead to affect-based theories of behaviour with greater explanatory power than competing theories, it may usher in a new scientiﬁc understanding of animals.

40Baenninger, R. 1990. Consciousness and comparative psychology. Pages 249–269 in Reﬂections on the Principles of Psychology: William James after a Century (M.G. Johnson and T.B. Henley, editors). Erlbaum, Hillsdale, USA. The quotation is on page 257. 41Kahneman, D., Diener, E. and Schwarz, N. (editors). 1999. Well-Being: The Foundations of Hedonic Psychology. Russell Sage Foundation, New York. 42Panksepp, 1998.

Dfraser_C008.indd 168

6/27/2008 12:43:44 PM

‘Natural’ Living

9

In the previous chapters we have encountered many examples of animal welfare problems created by keeping animals under highly unnatural conditions: macaque monkeys injuring themselves when housed in individual cages, calves developing ulcers when kept on a milk-like diet long after the age when they would normally graze and ruminate, conﬁned sows developing torsion of the stomach and spleen after being fed a two-day ration while in a frenzy of hunger and excitement. Given such blatant problems, it is no surprise that many people see ‘natural’ living conditions as important for animal welfare. But when scientists tried to incorporate this thinking into their efforts to assess and improve animal welfare, they quickly encountered a range of issues. Some were conceptual: for example, how should we deﬁne ‘natural’ conditions for animals, especially for domestic animals that have lived for thousands of years under human care? Other issues were practical: how, for example, can we actually make rearing conditions more ‘natural’ in barns, kennels, and laboratories? Another issue is philosophical: is naturalness only instrumentally important for animal welfare inasmuch as it helps to avoid problems of ill health or negative affect, or is naturalness inherently important as a fundamental element of animal welfare? Let us start with a thought-provoking example. BY 1980, WHEN CONFINEMENT systems for farm animals had become widespread in the industrialized countries, two scientists based in Edinburgh decided to try a radically different approach to raising animals. David Wood-Gush had already established himself as one of the pioneers of animal welfare science. As early as the 1950s he had been doing basic research on the behaviour of chickens, modelling his work to a degree on the studies of Konrad Lorenz whose work had inspired a generation of young scientists to study ‘ethology’ or the natural behaviour of animals. When concern over farm animal welfare erupted in the 1960s, Wood-Gush had already created an ethology section at the Poultry Research Centre in Edinburgh and he expanded the group and reoriented their work toward issues of bird welfare. By 1980 he had moved to a position at the University of Edinburgh where he launched a string of young scientists into behavioural studies of the farm animals. 169

Dfraser_C009.indd 169

6/27/2008 12:43:55 PM

170

Studying Animal Welfare

Alex Stolba was one of these: a Swiss scientist whose short career – he died suddenly near the age of 40 – was marked by startling originality. He was aware of research underway to try to improve the welfare of animals in conﬁnement systems by improving pen designs or giving animals somewhat more living space, but he dismissed these as ‘reductionist’ methods,1 preferring instead an approach that would take the full nature of the pig into account. Wood-Gush already had a solution in mind. While working with chickens, he had decided to study how these highly domesticated birds would behave if kept in a state more like the habitat of their wild ancestors. He had gained permission to release domestic chickens on Holy Isle – a tiny, uninhabited island off the west coast of Scotland. His intention was to study the birds in a ‘natural’ environment (albeit not the tropical jungle in which the species had evolved) and thus gain a fuller understanding of the birds’ behavioural capacities. He found that the birds showed much of the behaviour of their wild ancestors – roosting in trees, ﬁnding secluded areas to lay eggs, protecting and brooding their young, and stimulating the young to feed by pecking and scratching at food sources.2 Unfortunately, the experiment came to a premature end because this ‘natural’ environment included predators and harsh weather. With Stolba favouring the same type of approach, the two scientists did a similar but more successful study on pigs. They began by turning pigs loose in a hilly, wooded area in the Pentland Hills close to Edinburgh. Together with doctoral student Ruth Newberry and other co-workers, Stolba and Wood-Gush logged hundreds of hours observing the animals’ behaviour. The pigs, despite being domesticated through hundreds of generations of artiﬁcial selection, proved to retain virtually all the behavioural repertoire of their wild ancestor, the European wild boar.3 The pigs lived in small groups of several sows and their offspring. They spent hours per day rooting in the soil with the remarkable snout which serves both as a sturdy shovel and a delicate sense organ. They exercised their powerful neck muscles by levering against fallen logs and rocks. They used dunging areas well removed from their resting areas. And when a sow was about to give birth, she would move away from the social group and build an elaborate nest in a secluded, partly hidden area where she would remain with the new litter for several days before rejoining the social group. Of course all this behaviour could only be carried out if there were certain features in the environment – Stolba and Wood-Gush called them ‘key stimuli’ – such as a rooting substrate, nesting material and other objects. 1Stolba

expressed this opinion to me during a personal conversation in 1984. D.G.M., Duncan, I.J.H. and Savoury, C.J. 1978. Observations on the social behaviour of domestic fowl in the wild. Biology of Behaviour 3: 193–205. 3Stolba, A. and Wood-Gush, D.G.M. 1984. The identiﬁcation of behavioural key features and their incorporation into a housing design for pigs. Annales de Recherches Vétérinaires 15: 287–298. 2Wood-Gush,

Dfraser_C009.indd 170

6/27/2008 12:43:55 PM

‘Natural’ Living

171

Front gate

1m

Manure area

Threshold

Levering bar

Rooting area (peat) Rail Feed hoppers (for growing pigs)

Water trough

Activity area

Rubbing post Sow feeding stalls

Straw rack

Covered nest area

Rail Piglet area Rear passage

Figure 9.1 The floor plan of one version of the Edinburgh Family Pen with a manure area at the front, a rooting area with peat plus levering bars that the animals can use to exercise their neck muscles, an activity area with rubbing posts and straw racks, a covered nest area where sows can be confined for farrowing, and a heated piglet area protected by rails that prevent larger pigs from entering. Sows can be fed individually in the sow feeding stalls. Rails near the centre of the pen are used to exclude sows from feed hoppers intended for growing pigs. Redrawn after Kerr et al. (1988).

Based on these observations, Stolba and Wood-Gush began designing a complex pen that would allow pigs in captivity to behave much as they behaved in the wooded hillside (Figure 9.1). The pen contained a manure area at the front, well separated from where the pigs rested. Next was a rooting area with peat moss, and two heavy wooden ‘levering bars’ which the pigs could use to exercise their neck muscles. Then came an activity area with a rubbing post and straw rack. At the rear were several feeding stalls, divided so that sows could eat in relative peace, adjacent to a covered nest area where sows could be isolated to build nests and give birth. Each such pen housed the kind of social group seen in the ﬁeld work, generally four sows and their litters of similar-aged offspring which remained together at least until the sows had weaned the young onto solid food. Stolba and Wood-Gush built several of these ‘Edinburgh Family Pens’ and together with other researchers they monitored the animals’ behaviour, health,

Dfraser_C009.indd 171

6/27/2008 12:43:55 PM

172

Studying Animal Welfare

survival and growth. They found that the animals’ health was good, and economic performance was roughly similar to that seen in viable commercial units at that time. They concluded that the welfare of pigs in the Family Pens was ‘much improved especially in comparison to commercial intensive units’, mainly because the animals showed ‘the normal range and frequency of behaviour patterns and no abnormal behaviour’.4 However, critics of the approach were not convinced by Stolba and Wood-Gush’s conclusion. In particular, the neonatal death rate was high by commercial standards,5 and given that most piglet deaths are due to either injury (typically from the sow crushing the piglets through clumsy movements) or exposure (especially of small piglets in cold weather), the high death rate arguably involved animal welfare problems of the ﬁrst magnitude. The Edinburgh Family Pen never caught on commercially for a variety of reasons. These likely include the slightly higher labour requirements, the greater difﬁculty of observing the animals and intervening with veterinary treatments, and the lack of control over breeding and weaning. But perhaps most important was a philosophical difference between the promoters of the system, who saw it as a major welfare advance because of the naturalness it provided, and more typical pig producers who worked to produce clean, healthy animals of uniform size in buildings that could be sanitized and isolated from pathogens. For them, improving animal welfare did not mean providing ‘natural’ conditions, but creating a hygenic environment, minimizing deaths, promoting rapid growth, and preventing the introduction of disease. As one unimpressed American visitor put it after seeing the Family Pen demonstration unit, ‘Most miserable pigs I ever saw’.6 THE FAMILY PEN WAS a colourful episode in the long history of the belief that in order for animals to have a good life, they must lead the kind of life that is ‘natural’ for the species. Behind this intuitively plausible belief there is a certain logic, but also a need for critical analysis. Without question, as we have seen, unnatural living conditions have created major animal welfare problems by any criterion of welfare, and some of these problems could presumably be solved or reduced if the animals were kept under more ‘natural’ conditions. But when we try to create, or even deﬁne, natural conditions or a natural environment for a species, we quickly run into difﬁculty. The boreal forest is presumably the natural environment for moose. In Canada, this 4Kerr,

S.G.C., Wood-Gush, D.G.M., Moser, H. and Whittemore, C.T. 1988. Enrichment of the production environment and the enhancement of welfare through the use of the Edinburgh Family Pen System of Pig Production. Research and Development in Agriculture 5: 171–186. The quotations are from pages 184 and 171 respectively. 5A thoughtful analysis is provided by Edwards, S.A. 1995. Designing systems to meet behavioral needs: The Family Pen system for pigs. Pages 115–125 in: Animal Behavior and the Design of Livestock and Poultry Systems. Northeast Regional Agricultural Engineering Service, Ithaca, USA. 6I am grateful to Dr. Stan Curtis for relating this incident to me in 1982.

Dfraser_C009.indd 172

6/27/2008 12:43:55 PM

‘Natural’ Living

173

forest is characterized by cold winters, vast stretches of coniferous trees, abundant lakes and swamps, mineral-poor soils, and wolves. If we try to create a natural environment for captive moose, which of these features need to be incorporated? The question is especially complicated in the case of domestic animals which have been bred for thousands of years to thrive in circumstances that are presumably very different from those in which the ancestral species evolved. Chickens, for example, are believed to be descended from one or more species of Asian junglefowl (Gallus), birds that live mainly in habitats dominated by trees, shrubs and bamboo.7 For thousands of years, however, domestic chickens have been bred to survive and reproduce around human habitations, in small coops and cages, and in human homes. Since about 1950 some meat strains have been bred to live at ground level in ﬂocks of many thousands in large barns; other strains have been bred to lay (but not incubate) a huge number of eggs while living in small groups in cages. Thus the ‘natural environment’ of the domestic chicken is, to some extent, a meaningless term. The closest candidate might be an Asian jungle consisting of dense vegetation and overhead cover, but today few people – including those who raise chickens under free-range and other supposedly ‘natural’ conditions – provide anything like such an environment. Indeed, traditional agricultural environments, natural as they may seem for farm animals, may actually not correspond well to the animals’ needs. Veterinary scientist John Webster, who has devoted his long and distinguished career to promoting practical approaches to animal welfare, provided the sobering example of sheep wintered on poorly drained pasture near a housing development. Natural and bucolic as the scene may appear to the casual passer-by, the animals are likely to be malnourished by the winter pasture, suffering from cold because of the muddy conditions and lack of shelter, in pain from foot rot, and periodically frightened by uncontrolled dogs from nearby houses.8 Even without the complexities created by domestication, there are major problems in equating animal welfare with natural environments. After working closely with moose, I ﬁnd it difﬁcult not to feel sympathy for what these animals endure: temperatures ranging from  40 to 40 degrees, clouds of insects so dense that the animals seek refuge in water where they are attacked by leeches instead, legs so long that they can easily be broken by a false move in dense brush, and frequent loss of their young to predators. To make matters worse, moose entered North America on the land bridge from Eurasia perhaps 100 000–150 000 years ago, a period long enough for the species to have spread across the continent and to have evolved into four distinct types,9 but too short for them to acquire one 7Wood-Gush, D.G.M. 1971. The Behaviour of the Domestic Fowl. Heinemann Educational Books, London. 8Webster, J. 1994. Animal Welfare: A Cool Eye towards Eden. Blackwell Science, Oxford. The passage is on page 168. 9Peterson, R.L. 1955. North American Moose. University of Toronto Press, Toronto.

Dfraser_C009.indd 173

6/27/2008 12:43:56 PM

174

Studying Animal Welfare

Figure 9.2 An emaciated moose in springtime. The animal survived the winter even though its coat showed a pattern of damage typical of animals heavily infested with ticks. Photograph courtesy of Hank Hristienko.

key adaptation. Winter ticks, Dermacentor albipictus, are found in small numbers on cattle, deer and other species in North America, but they infest moose by the tens of thousands. In response to a heavy load of ticks, moose rub their skin on trees to the point of wearing away the hair that would protect them from the cold of winter.10 It is a truly pathetic sight to see emaciated, tick-infested moose dying of exposure in punishingly cold weather (Figure 9.2). For those who think that providing a natural environment is the answer to animal welfare problems, a little acquaintance with the beleagured Monarch of the North would be cause for reﬂection. In summary, although there are many examples of animal welfare problems caused by artiﬁcial conditions, looking to ‘natural’ environments as a way to safeguard animal welfare raises serious difﬁculties. For a given species it may not be possible to identify a ‘natural’ environment nor, for practical reasons, to recreate it; and the problem is further compounded for animals with a long history of domestication. However, even if we could identify and replicate a ‘natural’ environment, animals may still face serious animal welfare problems as judged on the basis of their health and affective states, even if they are spared the welfare problems caused by artiﬁcial conditions. 10Samuel, W.M. and Welch, D.A. 1991. Winter ticks on moose and other ungulates: Factors inﬂuencing their population size. Alces 27: 169–182.

Dfraser_C009.indd 174

6/27/2008 12:43:56 PM

‘Natural’ Living

175

BUT THE APPROACH USED by Stolba and Wood-Gush was subtly different: they were not trying to replicate the animals’ natural environment but to accommodate the animals’ natural behaviour. Even in an artiﬁcial environment like the Edinburgh Family Pen, if the animals behave as they would in the wild, would this not be enough to ensure a high level of animal welfare? Indeed, could we not use the behaviour of animals in the wild as a yardstick to assess the welfare of animals in captivity? Such an approach was tested by zoo biologist Jake Veasey and co-workers who asked whether the welfare of giraffes in four different zoos could be assessed by comparing the behaviour of the captive animals to that of free-living giraffes. To do this, Veasey and co-workers conducted dawn-to-dusk observations to determine the ‘time budgets’ of the animals – summaries of how they apportioned their time among different activities – in the various zoos and in an area of natural giraffe habitat in Africa.11 Collecting suitable baseline data in the wild proved to be a challenge. When Veasey and colleagues followed giraffes on foot or observed them from vehicles, the giraffes spent a good deal of their time watching the researchers (warily, one assumes) and sometimes walking or galloping away. When the observers positioned themselves at a watering hole, the giraffes (not surprisingly) spent a good deal of the time drinking. Observations seemed more representative when made from a high vantage point overlooking a plain, but at that distance ﬁne aspects of behaviour could not be distinguished. Thus, just for seemingly simple, technical reasons, establishing ‘natural’ behaviour proved remarkably difﬁcult. A more profound complication arose when they tried to use the differences to draw conclusions about animal welfare. Certainly the time budgets were very different between free-living giraffes and those in the zoos. Most noticeable was a large difference in the amount of time spent in feeding. Feeding occupied most of the day for the wild animals but only a small portion of the day for the giraffes in zoos who were, of course, provided with food. But did the lower feeding time mean that the zoo animals were somehow deprived of the opportunity to forage, or was it simply that the wild giraffes had difﬁculty ﬁnding enough to eat? With little of the day spent feeding, the zoo animals spent correspondingly more time just standing, and those with a comfortable grassy surface in the enclosure spent some time lying down – a behaviour that was not seen during the day for the wild giraffes. But should lying down be viewed as ‘abnormal’ and hence a sign of impaired welfare, or did the zoo animals merely have sufﬁcient leisure and freedom from fear that they could indulge in a luxury that was not feasible for their wild cousins? Deeper reﬂection on the results reveals further layers of complexity. If the wild giraffes had to spend long hours ﬁnding food, perhaps they did have a harder life 11Veasey, J.S., Waran, N.K. and Young, R.J. 1996. On comparing the behaviour of zoo housed animals with wild conspeciﬁcs as a welfare indicator, using the giraffe (Giraffa camelopardalis) as a model. Animal Welfare 5: 139–153. The quotation is on page 139.

Dfraser_C009.indd 175

6/27/2008 12:43:58 PM

176

Studying Animal Welfare

from day to day but ended up more ﬁt and healthy than well-fed zoo animals that ﬁlled their time with after-dinner siestas. Which group had better welfare? Or is it simply impossible to combine welfare advantages of one sort, such as ease of ﬁnding food, with disadvantages of another sort, such as decreased physical ﬁtness? Is it, in effect, an error to view animal welfare as a single trait where the components can be added and subtracted in some common currency? We will return to this question later. For now, let us just note that although Veasey and co-workers began with the idea that the wild environment is often assumed to be ‘a blueprint for optimal welfare’, and although there were deﬁnite differences between behaviour in the wild and in captivity, there was no simple way to translate these differences into conclusions about animal welfare. And, of course, some of the same issues that arise over natural environments also pertain to natural behaviour. Just as natural environments contain elements that would seem detrimental to animal welfare – things like extreme weather and predators – likewise natural behaviour includes elements such as shivering, panting and ﬂeeing that the animal would presumably never wish to perform. And with domestic animals, natural behaviour can be hard to deﬁne, although the Edinburgh pigs and Holy Isle chickens clearly suggested that domestication, although changing quantitative thresholds such as ﬂight distances, has left much of the behaviour of the wild ancestor largely intact. WE CAN AVOID SOME of these complications if we interpret naturalness not in terms of a natural environment, nor speciﬁcally in terms of natural behaviour, but in terms of the ‘adaptations’ that animals possess. Adaptations (as the term is used by evolutionary biologists) are features that the species has evolved as means of meeting challenges and exploiting opportunities, and thus managing to survive and reproduce successfully. We can think of adaptations as falling into several types. The most obvious adaptations – those that give species their characteristic physical features – are anatomical. It is generally believed that the moose’s long legs are an adaptation to deep snow, and thus allow the species to exploit vast areas of forest where snow depth would make life impossible for smaller members of the deer family. The hollow hairs of the caribou (Rangifer tarandus) provide both insulation and buoyancy which help the animals to endure the cold of arctic winters and to cross fast-moving rivers on migration between seasonal habitats. Other adaptations are physiological. For example, most mammalian species are capable of synthesizing vitamin C and thus, unlike humans, can live on diets that include no fresh fruit or other sources of this vitamin. Humans and other primates, together with fruit bats and guinea-pigs, lack this adaptation and need vitamin C in their diet.12 12Pauling, L. 1976. Vitamin C, the Common Cold, and the Flu. W.H. Freeman and Company, San Francisco.

Dfraser_C009.indd 176

6/27/2008 12:43:58 PM

‘Natural’ Living

177

A third group of adaptations are behavioural, but here we must draw a distinction. Some behavioural adaptations are more or less automatic reactions to speciﬁc stimuli. A person whose core body temperature drops by two or three degrees is likely to begin shivering uncontrollably. This rapid contraction of muscles helps to generate heat and is presumably an adaptation that could help people to survive cold temperatures. However, people do not decide to begin shivering; rather, shivering simply begins and people are unable to stop. Weeping and sneezing are other common types of behaviour which may simply ‘happen’ even when people make a conscious effort to prevent the behaviour. The shivering person is conscious of feeling cold and of shivering, just as the weeping person feels sad and the sneezing person feels a tickling sensation in the nose; but the behaviour itself is largely if not totally outside the person’s conscious control. In all these cases – shivering, weeping, sneezing – the behaviour involves a relatively simple set of actions that vary little from one performance to another or from individual to individual. Konrad Lorenz’s term for such behaviour, translated into English, was ‘ﬁxed action pattern’. For some simple animals, perhaps all of life’s challenges are met with such ﬁxed units of behaviour occurring automatically in response to certain stimuli. For example, the ticks that harrass moose hatch from eggs in the soil. The larvae (as far as we understand) have been ‘programmed’ by their genes to move upward, against gravity, until they reach the soil surface and then continue upward on nearby vegetation. There they cling until chemicals from a passing moose – carbon dioxide, butyric acid and others – cause them to grasp the animal’s hair and burrow inward until they can attach their mouthparts to the skin.13 Thus the tick (we presume) goes through a set of preprogrammed movements in response to speciﬁc stimuli, and it thereby obtains all the nutrition it will need for its lifetime. Food-ﬁnding for chimpanzees could hardly be more different. Chimpanzees learn to eat many kinds of fruit, and they return to speciﬁc fruit trees each year at the time when the fruit is ripe. They have learned to ‘ﬁsh’ for termites by inserting a twig into a termite mound and then pulling it out and eating the termites that cling to it. They raid bees’ nests for honey, and some groups of chimpanzees work cooperatively to capture and kill pigs, monkeys and other animals.14 In the case of chimpanzees, obtaining food is not a matter of ﬁxed action patterns. Instead, the animals’ manner of obtaining food varies greatly from one situation to another, and from one individual to another. This ﬂexibility allows chimpanzees to exploit various habitats and food sources which, as far as we can imagine, could not be exploited through a set of ﬁxed movements and simple stimulus–response connections. What, then, is the ‘adaptation’ that makes this ﬂexible behaviour possible? Part of the answer (or so we hypothesize) is that chimpanzees have a motivational 13Sonenshine, 14Goodall,

Dfraser_C009.indd 177

D.E. 1993. Biology of Ticks, Volume 2. Oxford University Press, New York. 1971.

6/27/2008 12:43:58 PM

178

Studying Animal Welfare

system that involves experiential states. If we use human experience as a model to suggest hypotheses for chimpanzees, we might expect these experiential states to include an unpleasant state (‘hunger’) when the animal is badly in need of food, states of pleasure when it eats food with certain properties that would typically correspond to its needs, and a state of satiety which replaces the feeling of hunger when the animal has eaten enough. Because these states are experienced as pleasant and unpleasant, they serve as an incentive to look for food when it is needed and to learn to ﬁnd food. The adaptation, then, is not a set of ﬁxed action patterns programmed genetically to occur in response to speciﬁc stimuli, but (in part) the capacity to experience certain affective states that motivate behaviour. Of course, a state of motivation would be of limited use unless the animal also possessed ‘cognitive’ adaptations that allow certain types of perception and learning. A classic example of adaptive learning is the ‘imprinting’ of precocial species of birds studied by the early ethologists. Precocial birds, such as young ducks, geese and chickens, hatch in a relatively mature state and are capable of leaving the nest and following the mother almost immediately. But instead of the young birds having a genetically ﬁxed image of the mother, evolution appears to have equipped them with a propensity to rapidly learn the visual characteristics of a nearby large object, especially if it moves and makes sounds, and to form a strong attachment to this stimulus and follow it. Normally, of course, it is the mother that the young birds ﬁrst see and follow. However, if the birds are hatched artiﬁcially and raised by a foster mother of another species, they will follow the foster mother devotedly, and later in life will direct sexual behaviour not toward their own species but toward hers.15 Imprinting is one type of learning that forms a particular adaptation in certain species, but the capacity for learning in general is clearly an important adaptation for many animals. In some cases, such as the learning shown by chimpanzees to ﬁnd different food sources, the learning appears to be highly variable and to involve insight and social transmission of behaviour. In other cases, the capacity to learn is very speciﬁc. Rats, which eat a wide range of foods and have survived during many generations of human attempts to poison them, appear to have a highly developed ability to learn to avoid harmful food. When presented with a novel food, rats will typically eat a very small amount. If they become sick in the next few hours, this single experience is enough to instil a strong and long-lasting avoidance of any foods with the same ﬂavour. In contrast, rats seem to ﬁnd it impossible to learn to avoid poisoned food on the basis of a distinctive colour or if it is consistently offered in a speciﬁc location. In this case, the adaptation consists of a capacity for extremely efﬁcient learning of a very speciﬁc association.16 15Manning,

A. 1972. An Introduction to Animal Behaviour, Second Edition. Edward Arnold, London. 16Rozin, P. and Kalat, J.W. 1971. Speciﬁc hungers and poison avoidance as adaptive specializations of learning. Psychological Review 78: 459–486.

Dfraser_C009.indd 178

6/27/2008 12:43:58 PM

‘Natural’ Living

179

In these few examples I have divided adaptations into categories – anatomical, physiological, behavioural, affective and cognitive – that will be familiar to scientiﬁcally trained readers. However, nature does not always conform to our tidy categories; what we may see as separate types of adaptations often ﬁt together to give a species some characteristic ability. The sheep’s four-chambered stomach (anatomical), its ability to digest ﬁbrous plant matter (physiological), its efﬁcient methods of ingesting plants (behavioural), its (presumed) capacity to experience hunger (affective), and its ability to learn to recognize beneﬁcial plants and avoid harmful ones (cognitive) – these ﬁt together to allow sheep to thrive by eating the vast supply of grassy and other ﬁbrous vegetation which would not support animals lacking such adaptations. One further comment about adaptations: When we try to keep animals in ways that suit their adaptations, one of the challenges is to understand the range of conditions under which animals are adapted to live. Koalas (Phascolarctos cinereus) are adapted to eating only the leaves of eucalyptus trees and may fail to thrive on other diets. In contrast, domestic pigs seem adapted to eat a very wide range of foods. The common domestic animals, which we keep on farms or in homes, are generally species that can thrive under a wide range of conditions. Domestic cattle, for example, can grow thick hair and thrive outdoors under cold winter conditions, but can also stand hot summer weather if they have adequate access to shade and water. In contrast, one of the objections to raising certain ‘exotic pets’ is that the species may be adapted to a narrow range of diets and environmental conditions which typical pet-owners cannot reproduce.17 IF WE KEEP ANIMALS in ways that do not fully match their adaptations, what kinds of animal welfare problems may arise? In some cases there may be no welfare problems at all. A moose kept in a large zoo enclosure where it does not need its long legs to survive, a rat with sufﬁcient access to fresh fruit that it does not need to synthesize vitamin C, a person who never experiences temperatures cold enough to shiver uncontrollably – all these seem unlikely to suffer any ill health or suffering as a result of an environment that does not require them to use these adaptations. In these examples, the adaptations in question appear to be ways that the animal deals with some form of adversity, and barring an unforeseen complication (such as cold temperatures being needed to kill parasites), removing the need to use these adaptations may, by at least some criteria, improve the quality of life of the animal rather than decrease it. In other cases, however, we encounter problems of basic health and functioning if animals are kept in ways that fail to match their adaptations. Cattle provide an example. Although cattle are adapted to thrive on a diet of grasses and forbs, they can also digest richer foods such as grain, and in many cases they prefer grain 17Schuppli, C.A. and Fraser, D. 2000. A framework for assessing the suitability of different species as companion animals. Animal Welfare 9: 359–372.

Dfraser_C009.indd 179

6/27/2008 12:43:58 PM

180

Studying Animal Welfare

to coarse forages. However, diets that include too much grain can leave cattle vulnerable to health problems such as excessive acidity in the rumen. More extreme diets can have even more drastic effects. When veterinarian Frank Loew visited Cuba in the 1970s to assist the ﬂedgling intensive cattle industry, he saw animals suffering from polioencephalomalacia or necrosis of brain tissue accompanied by loss of coordination. Cuba has a large sugar industry which produces huge amounts of molasses as a by-product, and this was being fed to cattle as the main source of dietary energy. For an animal adapted to digest the complex carbohydrates of grasses, a diet of sugar by-product was so unnatural as to cause pathological changes to the brain.18 A second type of welfare concern relates to affective states that motivate behaviour. From our own experience we understand how it feels to experience strong hunger or thirst at times when we are unable to eat or drink. Similar concerns arise if a calf is strongly motivated to suck, or a hen is strongly motivated to ﬁnd a nest site, or a caged bird is strongly motivated to migrate, in situations where the corresponding behaviour is impossible. Lacking a more speciﬁc term, we often speak generically of ‘frustration’ occurring in such situations. We will discuss this topic in more detail in Chapter 10. For the moment, let us just note it as another type of welfare concern that can arise when animals are unable to live in the manner to which they are adapted. A third type of welfare concern is even more difﬁcult to label. Heini Hediger recognized the problem decades ago through his experience of animals in zoos. He noted that if an animal has its food provided, and if it never needs to escape from predators, then it might seem to have an ideal life, but the situation can actually create, as he put it, ‘a vacuum – an occupational blank’ because the animal cannot do anything for itself and does not experience ‘contingencies’ between its actions and some meaningful outcome.19 Hediger himself provided a telling example. He had been visiting a colony of chimpanzees in an American laboratory. The animals’ physical needs were well provided for, but some of the older animals were, in Hediger’s opinion, ‘obviously bored’. However, the chimpanzees had learned one way to exert some control over their surroundings. They would surreptitiously ﬁll their mouths with a large quantity of water and then drench any passing human. Hediger, having been warned about this behaviour, watched carefully before he passed the cage of a particular chimpanzee. The animal appeared distracted; it was looking in another direction and appeared to be playing with its toes, and it showed no bulging of its cheeks 18Mella, C.M., Perez-Oliva, O. and Loew, F.M. 1976. Induction of bovine polioencephalomalacia with a feeding system based on molasses and urea. Canadian Journal of Comparative Medicine 40: 104–110. 19Hediger, H. 1955. Studies of the Psychology and Behaviour of Captive Animals in Zoos and Circuses (G. Sircom, translator). Butterworths Scientiﬁc Publications, London. The quotations are from pages 133–134.

Dfraser_C009.indd 180

6/27/2008 12:43:58 PM

‘Natural’ Living

181

that would suggest a mouthful of water. Thinking that he had judged the situation correctly, Hediger walked past the cage. At the precise moment when Hediger was in the most vulnerable location, the chimpanzee whirled about and soaked him. Hediger commented that he did not mind the prank: I was nevertheless glad for the old ape’s sake, that his joke had come off so well. Who knows how much good this well-prepared incident did to him? It was no doubt a very welcome break in his life of boredom.

In this example, Hediger used the term ‘boredom’, but the term needs to be interpreted carefully. As Hediger seemed to realize, dispelling ‘boredom’ for animals is not simply a matter of providing a more complex environment, but of providing possibilities for the performance of behaviour and active interaction with the environment.20 Thus, for example, a television might help to relieve boredom in pigs mainly because they would chew energetically on the electrical cord. Finally, the inability to interact normally with the environment, both physical and social, can have detrimental results for the basic development of young animals. The American psychologist Harry F. Harlow is both famous and infamous for his studies showing that if young monkeys are raised artiﬁcially, without normal contact with a mother and play-mates, they develop bizarre and dysfunctional social behaviour. Harlow’s observation began by chance when he and his colleagues wanted ‘to produce and maintain a colony of sturdy, disease-free young animals’ for research purposes.21 Their procedure was to separate infant rhesus macaque monkeys from their mothers a few hours after birth, and raise them in individual cages where their feeding could be controlled and where the animals could be isolated from disease-causing pathogens. From their individual cages, the young monkeys could see and hear other monkeys but they had no opportunity for play or other physical contact. The method produced monkeys with excellent physical health, but when the animals began to mature, Harlow realized ‘that our monkeys were emotionally disturbed as well as sturdy and disease-free’: As a group they exhibit abnormalities of behavior rarely seen in animals born in the wild and brought to the laboratory as preadolescents or adolescents, even after the latter have been housed in individual cages for many years. The laboratory-born 20See

discussion in: Wemelsfelder, F. 1993. The concept of animal boredom and its relationship to stereotyped behaviour. Pages 65–95 in Stereotypic Animal Behaviour: Fundamentals and Applications to Welfare (A.B. Lawrence and J. Rushen, editors). CAB International, Wallingford, UK. 21Harlow, H.F. and Harlow, M.K. 1962. Social deprivation in monkeys. Scientiﬁc American 207: 136–146. The quotations are on pages 138 and 143. Prof. Alan Hein pointed out to me that this work was published largely in the form of descriptive rather than quantitative data. Its acceptance by other scientists, at a time when descriptive data played little role in comparative psychology, is perhaps a reﬂection of Harlow’s formidable scientiﬁc reputation.

Dfraser_C009.indd 181

6/27/2008 12:43:58 PM

182

Studying Animal Welfare

monkeys sit in their cages and stare ﬁxedly into space, circle their cages in a repetitive stereotyped manner and clasp their heads in their hands or arms and rock for long periods of time.

Intrigued by these observations, Harlow and colleagues began varying the rearing procedure and were able to identify the speciﬁc requirements for normal development. They found that a six-month period of isolation was enough to produce animals that were ‘permanently inadequate’, whereas the effects of two to three months in isolation appeared largely reversible. Moreover, 20 minutes per day of play in a complex environment with other young monkeys was enough to produce relatively normal behavioural development, even in monkeys raised with no other social contact. Play, as we have seen, may serve a number of functions in the behavioural development of young animals.22 Harlow’s work showed that in this species, exercising the capacity for play is essential for the development of functional social behaviour. The development of normal perception also appears to require that animals can interact with the environment in basic ways. In the 1960s, experimental psychologists did extensive research on how the environment affects the development of visual perception in kittens. If kittens were raised in the dark or in featureless environments with unpatterned walls, they failed to develop the ability to process visual information in normal ways. In a strange but informative experiment, psychologists Richard Held and Alan Hein raised ten pairs of kittens in the dark, but starting at about ten weeks of age the kittens had three hours per day of exposure to a circular enclosure with striped walls. One kitten of each pair could walk independently in the enclosure. This kitten wore a collar attached to a device that moved a small gondola containing a restrained sibling kitten (Figure 9.3). Thus, the walking kitten experienced the changing visual stimulus as correlated with its own actions, whereas the kitten in the gondola saw the same changes in the visual pattern, but these were not correlated with its own selfproduced movement. After some days of this exposure, the kittens that could move actively in the enclosure responded normally to an approaching hand and could distinguish between a shallow and a steep step down, whereas the siblings that had moved only passively in the striped enclosure showed no ability to respond to visual information, although they later developed seemingly normal vision when they were allowed to move freely in an illuminated environment.23 Even the fundamental integrative processes in the central nervous system may fail to develop normally if the animal cannot exercise basic adaptations. When mice are raised in the aptly named ‘shoe-box cages’, these rather timid, burrowing animals have only the simplest of environments where they cannot burrow, 22Špinka

et al., 2001. R. and Hein, A. 1963. Movement-produced stimulation in the development of visually guided behavior. Journal of Comparative and Physiological Psychology 56: 872–876. 23Held,

Dfraser_C009.indd 182

6/27/2008 12:43:58 PM

‘Natural’ Living

183

Figure 9.3 An apparatus used by psychologists Richard Held and Alan Hein to study the development of visual perception in kittens. One kitten could walk independently in the enclosure but wore a collar attached to a device that moved a small gondola containing a restrained sibling. The walking kitten thus experienced the changing visual stimulus as correlated with its own actions, whereas the kitten in the gondola saw the same changes in the visual pattern, but these were unrelated to its own self-produced movement. Reproduced from Held and Hein (1963), by kind permission of Prof. Hein.

hide, explore, store seeds, and do the other things that free-living mice might do. In such environments some mice develop the types of repetitive behaviour we encountered earlier which may reﬂect abnormal functioning of the basal ganglia of the brain. Ethologist Hanno Würbel suggests that this represents one example among others demonstrating a lack of normal brain development in animals raised in the kind of impoverished environments that are normally used for laboratory animals.24 To summarize, not allowing an animal to exercise its adaptations can have diverse effects. In some cases, especially where an adaptation allows an animal to deal with a speciﬁc problem such as deep snow or extreme cold, not requiring the animal to exercise the adaptation may pose no apparent problem for the animal and may actually appear (at least on the surface) to be beneﬁcial. In other cases, not allowing the animal to use its adaptations may lead to various animal welfare concerns. These include ill health, a set of affective states that we might collectively term ‘frustration’ and ‘boredom’, and abnormalities of social, emotional, and cognitive functioning. 24Würbel, H. 2001. Ideal homes? Housing effects on rodent brain and behaviour. Trends in Neurosciences 24: 207–211.

Dfraser_C009.indd 183

6/27/2008 12:43:58 PM

184

Studying Animal Welfare

MANY EFFORTS TO IMPROVE animal welfare, in zoos, laboratories, and to some extent on farms, are based on allowing animals to live in ways that correspond better to their adaptations. For example, we noted earlier that many animals spend a great deal of time obtaining and consuming food, and when abnormal behaviour occurs, it often involves similar behaviour (grazing, chewing, sucking) redirected in bizarre or undesirable ways. Hence, a common strategy for improving animal housing has been to devise environments where animals can carry out elements of their natural foraging and feeding behaviour. We have already seen that wild chimpanzees will create ‘ﬁshing rods’ to capture termites. Real termite ﬁshing would be difﬁcult to re-create in most zoos, but if chimpanzees are provided with long sticks and a container with a small opening, they will occupy a good deal of time ﬁshing for a tasty treat such as honey.25 Servals (long-legged African cats) hunt by leaping on their prey, sometimes ﬂushing birds from low vegetation and catching them by spectacular jumps into the air. When kept and fed in a standard zoo enclosure, servals spend much of their time inactive or pacing repetitively in the cage. In his classic book, Behavioral Enrichment in the Zoo, psychologist Hal Markowitz described how servals came alive when fed in a way that allowed them to exercise their normal hunting behaviour. To do this, Markowitz devised ‘ﬂying meatballs’ – food items attached to a rope or rod that was swung over the heads of the cats (Figure 9.4). Servals fed in this way would leap vertically twice their body length to capture the prey.26 Presumably it would be neither practical nor humane to provide servals with real African grassland and low-ﬂying birds to capture, but it was still possible to allow them to obtain food in the manner for which they are adapted. Mock termite mounds have been used in another way as a form of environmental enrichment. Hediger described how he had observed zebras in Africa using these tall, rough structures as rubbing posts to groom their coats, to the extent that the tops of termite mounds were sometimes rubbed smooth. Impressed by this observation, Hediger had a concrete replica of a termite mound constructed for the zebra enclosure in the zoo in Zurich. When the zebras saw the new structure, they immediately approached it and rubbed themselves on it as the free-living zebras had done, but so vigorously that it was soon overturned. Hediger then had a second one built, but this one was better reinforced. The mock termite mound was subsequently put to daily use by the zebras (Figure 9.5), but Hediger reported that he had to assign two attendants ‘armed with whips’ to keep the zebras away until the concrete hardened.27 25Celli, M.L., Tomonaga, M., Udono, T., Teramoto, M. and Nagano, K. 2003. Tool use task as environmental enrichment for captive chimpanzees. Applied Animal Behaviour Science 81: 171–182. 26Markowitz, H. 1982. Behavioral Enrichment in the Zoo. Van Nostrand Reinhold, New York. 27Hediger, 1955, page 18.

Dfraser_C009.indd 184

6/27/2008 12:44:00 PM

‘Natural’ Living

185

Figure 9.4 A serval leaping into the air to capture a ‘flying meatball’ in the San Diego Zoo. Originally published by Markowitz (1982); reproduced with the kind permission of Dr. Markowitz.

Providing a suitable social environment is another major avenue for improving animal welfare. Flamingos normally live in large ﬂocks where ‘safety in numbers’ presumably provides individuals with a level of protection that such large, conspicuous birds would not have if living alone or in small groups. Zoo biologist Robert Young noted that many zoos with only a small number of ﬂamingos have little success in breeding the birds. Some zoos have installed artiﬁcial models of ﬂamingos to give the birds the impression that they are in much larger groups. Others have equipped ﬂamingo enclosures with mirrors on either side of the group to create the visual impression of a large ﬂock.28 Of course, quality as well as quantity is important when the social environment is used to improve animal welfare. When domestic guinea pigs are kept in large 28Young, R.J. 2003. Environmental Enrichment for Captive Animals. Blackwell Publishing, Oxford.

Dfraser_C009.indd 185

6/27/2008 12:44:00 PM

186

Studying Animal Welfare

Figure 9.5 A zebra in the zoo in Zurich scratching itself on a concrete replica of a termite mound in 1949. Originally published in Hediger (1955). Photo generously provided by Zoo Zurich, through the kind cooperation of Dr. Robert Zingg and with permission of Elsevier.

colonies, strong bonds develop between speciﬁc males and females. Behavioural biologist Norbert Sachser reported an experiment in which male guinea-pigs were removed from their colonies to an unfamiliar enclosure for a few minutes as a mild cause of stress. The animals generally showed an increase in blood cortisol levels in the unfamiliar environment, indicating activation of the HPA system. The cortisol response was similar whether the males were alone, or accompanied by an unfamiliar female, or accompanied by a familiar female to which they were not bonded. However, if the bonded female was present, the stress response was much reduced (Figure 9.6).29 For many animals, interaction with a human keeper appears to provide an opportunity for social enrichment, and for no species is this more obvious than for the domestic dog. As animal behaviourist David Tuber noted, handling and training dogs can help to reduce their responses to stressful situations and to eliminate problematic behaviour.30 The opportunity to play with humans is so positive for many dogs that it can serve as a powerful reward when dogs are being trained. 29Sachser, N., Dürschlag, M. and Hirzel, D. 1998. Social relationships and the management of stress. Psychoneuroendocrinology 23: 891–904. 30Tuber, D.S., Miller, D.D., Caris, K.A., Halter, R., Linden, F. and Hennessy, M.B. 1999. Dogs in animal shelters: Problems, suggestions, and needed expertise. Psychological Science 10: 379–386.

Dfraser_C009.indd 186

6/27/2008 12:44:01 PM

‘Natural’ Living

187

500

Cortisol (ng ml−1)

400

300

200

100

0 Alone

Unfamiliar female Familiar female

Bonded female

Figure 9.6 Plasma cortisol levels of 10 male guinea-pigs before and after being moved into an unfamiliar enclosure. The animals were tested under four conditions: alone, with an unfamiliar female, with a familiar female, and with the female to which the male was bonded. Cortisol levels increased substantially in the first three conditions, but significantly less when the bonded female was present. Redrawn after Sachser et al. (1998), with kind permission of Elsevier. Values are mean and standard error.

In the Australian customs service, for example, trainers who teach dogs to sniff out narcotics keep a short length of rope nearby, and they reward the dogs with a game of tug-of-war when the dog shows the correct behaviour.31 In the above examples, environments have been modiﬁed in order to allow animals to exercise some adaptation, especially by performing a type of natural behaviour that would otherwise be impossible. In other cases, however, the animals’ natural manner of living is used in a heuristic way to suggest changes to quantitative variables rather than permit the performance of speciﬁc actions. If dairy cows and their calves are allowed to remain together during the calf’s ﬁrst few weeks, the cows will normally nurse their calves many times per day. On most dairy farms, however, calves are separated from their mothers within the ﬁrst day after birth, and are then fed milk from a bucket, often only twice per day because this ﬁts with the daily cycle of work on the farm. Only a modest amount of milk can be fed at each meal because a large quantity given in too short a time could overwhelm the digestive system and cause illness. Under these conditions, 31Adams, G.J. and Johnson, K.G. 1994. Sleep, work, and the effects of shift work in drug detector dogs Canis familiaris. Applied Animal Behaviour Science 41: 115–126.

Dfraser_C009.indd 187

6/27/2008 12:44:02 PM

188

Studying Animal Welfare

the calves grow much more slowly than they would if left with the cow. Although it is not feasible to leave calves with the cow on most commercial diary farms, we can still arrange artiﬁcial feeding in a way that better mimics natural feeding. If calves are fed a number of smaller meals per day, they can consume and digest much more milk, and can achieve more or less normal growth rates; and if fed by an artiﬁcial teat rather than a bucket the calves can also consume milk in a more natural manner.32 Beyond these examples of stimulating natural patterns of behaviour and respecting the natural time course of behaviour, there have also been attempts to allow animals to create contingencies whereby their own actions produce some type of outcome. Again, the innovative Hal Markowitz provided a classic example with a project to enrich an enclosure in the San Francisco Zoo housing a 16-year-old African leopard (Panthera pardus). Markowitz equipped the enclosure with four speakers that could emit a simulated bird song. When the ﬁrst speaker, installed near a high ledge, began playing the song, the leopard could approach the speaker and thus activate a motion detector. This began a programmed sequence of events. First, the initial speaker became silent and the song was taken up by a second speaker part way down a diagonal tree trunk, and then by a third near the base of the tree trunk, and then by a fourth on the far side of the enclosure. The sequence was meant to simulate a bird ﬂeeing from the leopard. If the leopard followed the sound trail to the ﬁnal speaker, it would activate a second motion detector at the far end, and this triggered the release of a piece of meat as a reward. Markowitz noted that the apparatus allowed the leopard to trigger the system in various ways. Sometimes the leopard would spring up the tree trunk toward the ﬁrst speaker and then bound down by the shortest route to retrieve the reward; at other times it adopted a leisurely pace and a different route. Once it had learned to use the apparatus, the leopard would typically obtain 10–20 rewards per day and it spent much less time pacing and hiding than before the device had been installed. As Markowitz and co-workers noted, the apparatus did not simply allow animals to carry out a single action. Rather, it ‘provides animals with broad contingencies and allows them to discover different ways to use it’.33 WHAT, THEN, IS THE relationship between ‘natural’ living and animal welfare? Without question, keeping animals under artiﬁcial conditions can lead to major animal welfare concerns. Some concerns involve disease and injury. Others involve the affective states of animals, including what we might generically term ‘boredom’ if animals cannot carry out the natural behaviour by which they would normally 32Appleby, M.C., Weary, D.M. and Chua, B. 2001. Performance and feeding behaviour of calves on ad libitum milk from artiﬁcial teats. Applied Animal Behaviour Science 74: 191–201. 33Markowitz, H., Aday, C. and Gavazzi, A. 1995. Effectiveness of acoustic ‘‘prey’’: Environmental enrichment for a captive African leopard (Panthera pardus). Zoo Biology 14: 371–379. The quotation is on page 377.

Dfraser_C009.indd 188

6/27/2008 12:44:03 PM

‘Natural’ Living

189

interact with their environment, and ‘frustration’ if they are highly motivated to perform behaviour that is impossible under their actual circumstances. Yet other welfare concerns involve an inability to use normal cognitive processes. And if young animals cannot interact in normal ways with their environment – both physical and social – they may show abnormal development of social behaviour, perceptual ability, or even of basic neural organization. It is hardly surprising, therefore, that many people consider that for animals to have good quality of life, they must be able to live in a way that is ‘natural’ for the species. Indeed, many useful strategies have emerged to improve animal welfare by allowing animals to live in ways that match their adaptations. Some of the simplest involve providing diets and physical conditions that suit the animals’ anatomy and physiology. ‘Behavioural enrichment’ tries to let animals perform natural types of behaviour, for example by allowing them to obtain food in ways that are normal for the species. And some more elaborate types of enrichment allow animals to exercise their cognitive and affective adaptations, for example by making it possible for animals to explore, play and experience contingencies linking their behaviour with its outcomes. But when we try to use ‘natural living’ as a measure of animal welfare, we encounter certain problems. ‘Natural living’ is sometimes interpreted to mean that animals should be kept in environments similar to those in which the species evolved. However, a ‘natural’ environment is often difﬁcult to identify; for practical reasons it may be impossible to create; and for domestic animals the concept is especially difﬁcult to interpret. Some people emphasize natural behaviour rather than the environment itself, but natural behaviour also raises problems of interpretation. I have argued that the least troublesome way to characterize a ‘natural’ life is that animals can live in a manner that corresponds to their adaptations. But whatever way we characterize a ‘natural’ life for animals, problems of interpretation persist. In particular, a truly ‘natural’ environment will involve features such as cold, heat, drought, predators and disease; and by the same token, animals commonly possess behavioural reactions and other adaptations that help them to deal with these hardships. If these natural adversities never arise, and if the animal never has to use the corresponding adaptations, few would argue that the animal’s welfare is impaired. This line of thought brings us back to the philosophical issue raised earlier. If people recognize that ‘natural living’ involves both advantages and disadvantages in terms of the basic health and affective experience of animals, will they continue to believe that naturalness is itself important for animal welfare? In other words, are natural living conditions only instrumentally good inasmuch as they support better health and functioning of animals, reduce negative states such as distress and frustration, and promote more enjoyment of life; or is naturalness inherently good for the quality of life of animals, independently of the other advantages and disadvantages that it brings? To some, as we have seen, animal welfare is all about the basic health and functioning of animals; to others, it is all about the balance of positive and

Dfraser_C009.indd 189

6/27/2008 12:44:03 PM

190

Studying Animal Welfare

negative affect. To people who hold these views, natural living conditions are only instrumentally valuable inasmuch as they promote animal welfare as deﬁned by other criteria. By this logic, if chickens in barns or giraffes in zoos are as healthy as those living in the wild, and if animals in artiﬁcial environments are warmer, safer and better fed than those living in nature, then there is no welfare advantage to putting the animals in a ‘natural’ environment or an environment where they will show the same behaviour as their wild counterparts. Nonetheless, the positive weighting that some people attach to naturalness reﬂects deeply held beliefs. To many organic farmers, a good life for animals must be a ‘natural’ one, even if there is a cost to some aspects of comfort and hygiene,34 much as the organic producers themselves may give up certain personal conveniences in order to live the kind of life they value. To behavioural biologists such as Barnard and Hurst (as we have seen), apparent hardships such as ‘stress’ and short life span should not be viewed as detrimental to animal welfare if these are natural for the species in question.35 The debate may never be resolved fully, but it may help to distinguish between the freedom to live naturally and the fact of living naturally. A rough analogy in human terms is the difference between allowing people the freedom to live as close to nature as they wish, versus requiring them to live in caves and forests because that is how their ancestors lived. Allowing animals to choose certain environmental features that they would encounter in nature, and choose to perform certain elements of their natural behaviour, should accommodate some of the concern for naturalness in the lives of animals, while making a better ﬁt to other criteria of animal welfare. This approach changes the question from: Is the animal in a natural environment? or Does the animal perform all its natural behaviour? to, Does the animal act as it would choose to act? This will be the topic for Chapter 10.36

34Alroe, H.F., Vaarst, M. and Kristensen, E.S. 2001. Does organic farming face distinctive livestock welfare issues? A conceptual analysis. Journal of Agricultural and Environmental Ethics 14: 275–299. 35Barnard, C.J. and Hurst, J.L. 1996. Welfare by design: The natural selection of welfare criteria. Animal Welfare 5: 405–433. 36For further discussion of natural behaviour and animal welfare, see: Špinka, M. 2006. How important is natural behaviour in animal farming systems? Applied Animal Behaviour Science 100: 117–128.

Dfraser_C009.indd 190

6/27/2008 12:44:03 PM

Preferences and Motivation

10

In the essay that he wrote for the Brambell Committee about assessing pain and distress in animals, William Thorpe raised the intriguing possibility of ‘asking’ animals what kinds of environments they prefer.1 As an example he cited a chance observation of some African buffalo which, after experiencing life both in the wild and in captivity, gave the strong impression that they preferred the protection of captivity, at least during the dark of night (Box 10.1).2 Appropriately enough, the ﬁrst attempt to resolve an animal welfare issue through research on animals’ preferences arose from one of the speciﬁc recommendations made by the Brambell Committee itself. The committee had concluded that the ﬂooring materials used for hens in cages were often unsatisfactory. The committee was particularly critical of ‘chicken wire’ ﬂooring (ﬁne-gauge wire netting) which, over time, sometimes injured the birds’ feet. Instead, the committee recommended that the cage ﬂoor should consist of a heavy-gauge rectangular metal mesh. In an attempt to obtain the hens’ own view of this recommendation, Barry Hughes and Alan Black, two members of David Wood-Gush’s growing team of poultry welfare researchers in the 1970s, tested the preferences of hens for different types of ﬂooring. They housed hens in cages consisting of two sections, each ﬂoored in a different material, and they simply observed how much time the birds spent on the different ﬂooring products when offered various choices. The hens showed no strong preferences or aversions for the different materials, but overall they tended to select the ﬁne-gauge ‘chicken wire’ which the Brambell Committee had deemed unsuitable. This preference, Hughes and Black suggested, may have occurred because the chicken wire provided more points of contact and hence ﬁrmer footing, even though it occasionally caused injuries. In addition to 1Portions of this chapter have been reworked from Fraser, D. and Matthews, L.R. 1997. Preference and motivation testing. Pages 159–173 in Animal Welfare (M.C. Appleby and B.O. Hughes, editors). CAB International, Wallingford, UK. I am grateful to Lindsay Matthews and the publishers for permission to use some of that material here. 2Thorpe, 1965, pages 73–74.

191

Dfraser_C010.indd 191

6/27/2008 12:44:15 PM

192

Studying Animal Welfare

Box 10.1 A passage from W.H. Thorpe‘s inﬂuential essay, ‘The Assessment of Pain and Distress in Animals’. Obviously domestic animals cannot be allowed to run entirely free, to breed at random, to rear their own young in the normal way of the wild, or to establish their own social groupings and hierarchies. Nor indeed if, having experience of both, they could be asked their opinion, is it probable that they would always prefer the wild. In the early part of 1964 a group of African buffalo were captured in a region of Kenya where their natural existence was no longer tolerable or possible, and were taken for release in the Nairobi National Park. They resented being captured, as any wild animal would. After capture, during the process of transport and preparation for release, they were of course kept in pens or yards much like those in which domestic cattle are kept. When the time came for their release in the new environment, they showed many signs of distaste for it. They would return toward human habitations toward nightfall and try to enter the paddocks where they had been. One even tried to walk through the french windows of the office of the Director of the Kenya National Parks. The natural assumption is that the unfamiliar National Park, reeking of lion, leopard and other dangerous and uncomfortable neighbours, must have seemed a very unfriendly place; far inferior to the luxurious though restricted quarters they had become used to inhabiting! These buffalo could be ‘asked’ because they had experienced both the wild and domestic states. Domestic animals cannot usually be ‘asked’ because they have never experienced the former; and animals, like men, doubtless prefer to keep the ills they have than to fly to others that they know not of. But there is no doubt that a well managed farm or ranch run on approximately ‘natural’ lines can provide an environment for animals which, on any estimate that we are able to make, must be in many ways preferable to the wild. W.H. Thorpe, 1965. The assessment of pain and distress in animals. Appendix III in Report of the Technical Committee in Enquire into the Welfare of Animals Kept Under Intensive Livestock Husbandry Systems, F.W.R. Brambell (chairman). Her Majesty’s Stationery Office (HMSO), London, pages 73–74.

this speciﬁc ﬁnding, Hughes and Black expressed enthusiasm for the potential of research on animals’ preferences to improve our understanding of animal welfare. ‘We feel’, they wrote in 1973, ‘that this type of experiment offers a new approach to animal welfare; objective assessment of animals’ preferences should ultimately make subjective value judgements superﬂuous’.3

3Hughes, B.O. and Black, A.J. 1973. The preference of domestic hens for different types of battery cage ﬂoor. British Poultry Science 14: 615–619. The quotation is on page 619.

Dfraser_C010.indd 192

6/27/2008 12:44:16 PM

Preferences and Motivation

193

Other early preference research reﬂected similar optimism. In 1977 Marian Dawkins published a paper entitled, ‘Do hens suffer in battery cages? Environmental preferences and welfare’. She used preference experiments to test whether hens would prefer large pens or outdoor runs ahead of battery cages. In one experiment she gave hens free access to cages and to larger pens for 12 hours, and observed the birds at intervals to see which environment they selected. The (perhaps surprising) result was that the hens spent similar amounts of time in the cages as in the large pens. She then did a series of trials which involved placing hens in a central location where they could turn in one direction to enter a battery cage or in the other direction to enter an outdoor run, and the hens were then conﬁned in the chosen location for ﬁve minutes before being tested again. With this procedure, the hens tended to select the outdoor run. While fully recognizing the preliminary nature of her results, Dawkins, too, expressed optimism for the ability of preference research to shed light on animal welfare. ‘I hope’, she wrote, ‘that one day such work will enable us to take an objective and humane judgment on life in a battery cage’.4 THE INK HAD BARELY dried on these early reports when debate broke out over what we can actually conclude from research on animals’ preferences. Initially Ian Duncan, and subsequently many others, provided a string of criticisms of preference research: that animals’ preferences may simply reﬂect what they are accustomed to, that animals may not prefer what is truly beneﬁcial for them, that a less preferred option may still be important, and so on. The result has been a lively debate about the interpretation of preference research and a substantial evolution of research methods.5 Let us follow a few of the changes in preference research and how these inﬂuence the conclusions we can draw. First, it quickly became clear that determining an overall preference for one environment ahead of another is of limited value. A hen, for example, may prefer an open area for foraging, a secluded box for laying, and a perch for resting. To really understand animals’ preferences, experiments must be designed not just to establish an ‘overall’ preference but to identify how preferences vary depending on the circumstances and the animal’s behaviour. The point was brought home to me by one of my own attempts at preference research.6 I had wanted to ﬁnd simple ways to improve environments for pigs raised in conﬁnement, and the use of bedding on the bare concrete ﬂoors of typical pig pens was an obvious option. My intention was to establish a simple procedure for testing the preferences of pigs for bedding of various types, but the ﬁrst step 4Dawkins, M.S. 1977. Do hens suffer in battery cages? Environmental preferences and welfare.

Animal Behaviour 25: 1034–1046. The quotation is on page 1045. 5Duncan, I.J.H. 1978. The interpretation of preference tests in animal behaviour. Applied Animal Ethology 4: 197–200. Other key works in the early debate include Dawkins, 1980 and 1983, and Duncan, 1992. Much of the discussion is summarized in Fraser and Matthews, 1997. 6Fraser, D. 1985. Selection of bedded and unbedded areas by pigs in relation to environmental temperature and behaviour. Applied Animal Behaviour Science 14: 117–126.

Dfraser_C010.indd 193

6/27/2008 12:44:16 PM

194

Studying Animal Welfare

Figure 10.1 Preference research to test pigs’ preferences for different substrates. Two adjoining pens were identical except that one had a bare concrete floor while the other was bedded with straw. Photo kindly provided by Agriculture and Agri-Food Canada, Ottawa.

was to establish that pigs would, indeed, show a clear preference for bedded pens ahead of identical pens without bedding. I therefore designed a simple apparatus with two adjoining pens, each equipped with food and water, the only difference being that the concrete ﬂoor was bedded with straw in one pen but not the other (Figure 10.1). I then used time-lapse video recording to gain a representative sample of where the pigs located themselves over the several days of the test. The initial results came as a surprise. Some groups of pigs did indeed show a clear preference for the straw-bedded side, but other groups seemed to choose more at random, and a few consistently spent more time in the bare concrete pen. This was not what I had expected. On more careful analysis of the pigs’ behaviour, it became clear that there were some subtleties that I had not captured in my initial thinking. The pigs spent about 15% of their time active but not feeding or drinking. Much of this activity appeared to be foraging – walking slowly with the snout at ﬂoor level in a behaviour which, if performed outdoors, would involve the animals rooting in the soil and eating insects, worms and vegetation. The majority of this behaviour occurred in the pen with straw; hence, straw was clearly acting as a stimulus and target for this behaviour. When the pigs were eating, they used the two pens about equally; this suggested that the pigs were relatively indifferent to the softness of the ﬂoor when standing in one place. When the pigs were resting, however, I sensed that

Dfraser_C010.indd 194

6/27/2008 12:44:16 PM

Preferences and Motivation

195

there was a pattern: the pigs seemed to select the straw-bedded side when the room was relatively cool but to avoid the straw when the room was warm. I should, of course, have anticipated this from the outset, because it is well known that straw serves as a form of thermal insulation. When I repeated the experiment in a room where I could maintain a constant air temperature, the pigs gave clear-cut results: at about 20C, they strongly preferred to rest on straw, but at about 26C they generally chose to lie on bare concrete ahead of straw. The lesson from this study (and in hind-sight, it should have been obvious to me from the outset) is that when animals select environments, their preferences will vary depending on what they are doing, what time of day it is, the ambient conditions, and so on, and that environmental preference research needs to be designed to identify these sources of variation and how they inﬂuence animals’ preferences. As a result, the research methods of the 1970s, which often recorded which of two environments an animal would choose in a simple test, have given way to methods such as continuous video or electronic monitoring over periods of days or weeks to record how animals allocate their time and behaviour in relation to different environmental features. A second development was to separate preference from familiarity. If we want to use preference research to draw conclusions about the general class of animals being tested (such as laying hens or albino rats), then we need to ensure that the results are not simply a reﬂection of the previous experience of the particular animals used in the study. In the Hughes and Black study, the hens had been housed previously in cages with ‘chicken wire’ ﬂoors. When the birds showed a mild preference for that type of ﬂooring ahead of others, did this reﬂect a preference typical of hens in general, or were those speciﬁc hens merely selecting what they found most familiar? Animal welfare scientist Cassandra Tucker took such thinking into account when she studied the preferences of dairy cattle for different types of bedding.7 She started by letting cows have free access for seven days to three stalls, one bedded with sand, one with sawdust and one with a geotextile ‘cow mattress’ ﬁlled with recycled rubber. The cows showed a strong preference for the sawdust stalls, and one even broke down a barrier to enter this stall when it was blocked, but Tucker knew that this result could have been inﬂuenced by the fact that the cows had previously been kept mainly on sawdust. During the next six days, therefore, Tucker blocked the different stalls in turn so that each cow was required to spend two days on each of the surfaces. Then she retested the animals and found that even with this recent exposure to all three materials, most cows still chose sawdust although two of the twelve converted to sand. Nonetheless, although a two-day exposure might be enough to overcome any initial reluctance to try a new material, cows might still retain a long-term preference 7Tucker, C.B., Weary, D.M. and Fraser, D. 2003. Effects of three types of freestall surfaces on preferences and stall usage by dairy cows. Journal of Dairy Science 86: 521–529.

Dfraser_C010.indd 195

6/27/2008 12:44:17 PM

196

Studying Animal Welfare

for the kind of environment in which they were raised. With this in mind, Tucker located a farm whose cows were normally kept in sand-bedded stalls. When these animals were put through the same study, they preferred both the familiar sand and the unfamiliar sawdust, but largely rejected the mattresses. Tucker concluded that both sand and sawdust could become well accepted by cows, at least after enough exposure. A third development has been to use preference research to identify the speciﬁc design features that inﬂuence animals’ choices. If hens choose a ﬂoor covered with wood shavings ahead of a wire ﬂoor, is this because they are choosing the softer surface, the surface that gives the most secure footing, the surface that they cannot see through, or something entirely different? This level of understanding is important if we want to generalize beyond the speciﬁc options presented in the study. My engineering colleague Peter Phillips used this thinking in a study of the features that make ramps acceptable to pigs.8 Ramps are widely used in loading pigs, and occasionally in special kinds of housing that involve more than one level, but pigs balk at walking up certain ramps for reasons that were not well understood. Phillips therefore set up a room with a small holding pen surrounded by four ramps. Pigs placed in the holding pen were free to walk up and down the ramps as they wished. Most of the animals treated the apparatus like a play-structure and tried out their skill walking and running on the ramps and exploring a kind of environment that they had never experienced. Phillips did a series of different experiments, each one comparing a different design feature: one experiment used ramps with four different slopes; another compared four levels of illumination; and so on. By determining whether the pigs showed clear preferences among the ramps in the different experiments, Phillips was able to conclude which design features made a difference to the pigs’ choices. He found that the slope was a crucial factor: the steeper the ramp, the less it was used, especially if the slope exceeded about 25. The other important factor was the spacing of ‘cleats’ – horizontal bars attached to the ramp surface to provide secure footing. Ramps were well used if the cleats were only 5–10 cm apart, but with cleats farther apart the pigs made little use of the ramp. When other factors were varied (level of illumination, ramp width, enclosure of the side walls), the pigs showed no particular preferences for one ramp ahead of another. In this way Phillips used preference research not simply to compare the speciﬁc options offered but to identify the design features that inﬂuence the animals’ preferences. These few examples illustrate three lessons that preference research has taught us. First, it is rarely very helpful to determine an overall preference for one environment ahead of another under a narrow range of experimental conditions; rather, we need to determine how different variables (temperature, behaviour, time of day and so on) inﬂuence the animals’ choices. Second, in order to generalize beyond the speciﬁc animals used in the study, we need to know how the experimental subjects’ 8Phillips, P.A., Thompson, B.K. and Fraser, D. 1988. Preference tests of ramp designs for young pigs. Canadian Journal of Animal Science 68: 41–48.

Dfraser_C010.indd 196

6/27/2008 12:44:17 PM

Preferences and Motivation

197

preferences were inﬂuenced by their previous experience. And third, rather than determining preferences among certain speciﬁc alternatives, it is more useful if we can establish the features or design variables that underlie the animals’ choices. BUT EVEN IF AN ANIMAL prefers one environment to another, this does not mean that either environment has much effect on the animal’s welfare. A preference may be a weak (like a preference for dark chocolate over milk chocolate) or strong (like a preference to live in a house rather than a dungeon), and we presume that giving animals their preferred options will be important for their welfare only if the preference is relatively strong. What we need, then, is some measure of how badly an animal wants (or wants to avoid) a particular environment or how badly it wants to perform a type of behaviour that requires a certain environmental feature. In a series of papers, Marian Dawkins developed several approaches for measuring the strength of animals’ motivation as a way of understanding animal welfare. Perhaps the simplest approach is to see whether an animal will learn to perform an ‘instrumental’ task, such as pushing a lever or running down a runway, to obtain a particular option. For example, having established that hens prefer ﬂoors with litter ahead of bare wire, Dawkins and her student Tim Beardsley tested whether hens would learn an instrumental task to gain access to a cage with litter.9 They began by allowing hens to peck on two keys, one red and one yellow, which opened the door to two cages, one with a wire ﬂoor and one with a litter-covered ﬂoor. Although the hens had chosen the litter-covered ﬂoor when a simple choice was offered, they showed no evidence of learning to peck the key that opened the cage with litter. Dawkins and Beardsley realized, however, that their choice of pecking as an instrumental task might have created a problem. Pecking is a natural part of the hen’s feeding behaviour, and hens have little difﬁculty learning to peck a key if the reward is food. However, pecking a key for access to a particular type of ﬂooring may be so artiﬁcial that hens cannot learn to make this association. (Recall the rats that quickly learned to avoid poisoned food based on its ﬂavour but not based on its colour or location.) Dawkins and Beardsley therefore modiﬁed the experiment so that hens could open the doors simply by walking toward the cages and thus breaking a photo-beam. With this method the hens learned to access the cage with the litter-covered ﬂoor. Because the hens would perform a task in order to obtain a litter-covered ﬂoor, the experiment suggested that the motivation for litter must be more than trivial. But the experiment also provided a caution: failure to learn an instrumental task might be due to technical aspects of the experiment – including any mis-matching of the instrumental task and the reward – rather than a lack of motivation.10 9Dawkins,

M.S. and Beardsley, T. 1986. Reinforcing properties of access to litter in hens. Applied Animal Behaviour Science 15: 351–364. 10A classic work on such ‘constraints on learning’ is: Hinde, R.A. and Stevenson-Hinde, J. (editors). 1973. Constraints on Learning: Limitations and Predispositions. Academic Press, London.

Dfraser_C010.indd 197

6/27/2008 12:44:17 PM

198

Studying Animal Welfare

If animals will learn to perform the same instrumental task for two or more different rewards, then it may be possible to provide a more quantitative comparison of the strength of different motivations. In particular, Dawkins proposed that we might assess the strength of a poorly understood motivation by ‘titrating’ it against a second, better understood motivation for a reward such as food. In one experiment, Dawkins tried to estimate the strength of hens’ motivation to perform ‘dust-bathing’, a natural behaviour that keeps feathers in good condition by exposing them to dust and other loose materials that help to absorb oils. Dawkins placed hens in a central area from which they could enter two cages. One cage contained litter (to permit dust-bathing) but no food, while the other contained food but no litter. After exposing the birds to the two cages in preliminary trials, Dawkins then studied which cage the hens chose after they had been without food for 0, 3 or 12 hours. When not deprived of food at all, the hens generally chose the cage with litter; when deprived of food for 12 hours, they generally chose the cage with food; but when deprived of food for 3 hours, the hens chose the two cages about equally. The results suggested that the hens’ motivation for dust bathing, under the speciﬁc conditions of the experiment, was about as strong as their motivation to eat when mildly food deprived.11 As a even more elaborate option, Dawkins proposed an approach which she described in the language of micro-economics (Figure 10.2). She noted that consumers purchase some commodities (such as bread) in roughly the same amounts even if the price increases or if people’s incomes fall. These commodities are said to have ‘inelastic demand’ and are sometimes called ‘necessities’ because consumers are willing to pay more and more of their income to obtain them. For other commodities (such as ﬁne wine) consumption falls if the price increases or if consumers’ incomes decline. Such commodities are said to have ‘elastic demand’ and may be regarded as ‘luxuries’ because consumers will forego them more readily. Applying this thinking to animals, Dawkins proposed that elasticity of demand could be used to show ‘how important different environments or commodities are to the animals themselves’.12 To apply this concept to animals, a commodity such as food can be provided in response to some work (‘price’) that the animal has to perform, and the price can then be varied experimentally to determine the ‘price elasticity’ of the demand. Alternatively, the animal can be given a limited amount of time (‘income’) to access various resources; then the amount of time can be varied to determine the ‘income elasticity’ of the demand. The assumption is that commodities that are very important to the animal should show relatively little price elasticity or income elasticity; that is, the animal should devote more and more effort or 11Dawkins,

M.S. 1983. Battery hens name their price: Consumer demand theory and the measurement of ethological “needs.” Animal Behaviour 31: 1195–1205. 12Dawkins, M.S. 1990. From an animal’s point of view: Motivation, ﬁtness, and animal welfare. Behavioural and Brain Sciences 13: 1–9 and 54–61. The quotation is from page 6.

Dfraser_C010.indd 198

6/27/2008 12:44:17 PM

Preferences and Motivation

199

Amount purchased

High

Inelastic demand

Elastic demand

Consumer’s surplus

Reservation price

Low Low

High Price

Figure 10.2 An idealized illustration of ‘demand curves’ for commodities with elastic and inelastic demand. For commodities with inelastic demand (such as bread), the amount purchased remains fairly constant over a wide range of prices. For commodities with elastic demand (like fine wine), the amount purchased falls with increasing price. The slope of the line (i.e., how steeply the line falls with increasing price) is commonly used as a quantitative measure of elasticity. Two other measures of willingness to pay are also shown. Reservation Price is the highest price paid for the commodity. In animal studies, Consumer’s surplus can be calculated as the area to the left of the demand curve, and applies to a given quantity of the commodity. Note that in economics, the axes are typically reversed, and they are often shown on a logarithmic scale.

available time to maintain a given level of reward if that reward is very important. By thus establishing the elasticity of demand (Dawkins proposed) we should be better able to judge the importance that animals attach to resources such as food, companionship, bedding and exercise. Behavioural scientist Chris Sherwin provided an example of the approach in his study of the importance that mice attach to different opportunities for physical activity. One of the criticisms of standard laboratory cages for mice is that they restrict the movement of these normally active animals to a few steps in any direction. Sherwin devised an apparatus that allowed mice to move from the home cage into three options for extra movement. One was an oval-shaped loop where the mouse could run like a horse on a racetrack; another was a complex system of tunnels; and the third was a running wheel (Figure 10.3). The mice could gain access to each option by pressing a lever to open a door. At the start of the experiment, a single press of the lever was enough to open the door. Then the ‘cost’ was increased over successive days to 10 presses to open the door, then 20, then 40, and then 80. At the lowest cost, the mice gained access to the tunnels and

Dfraser_C010.indd 199

6/27/2008 12:44:18 PM

200

Studying Animal Welfare

Tunnel system (simplified)

‘Locomotion-loop’

Running wheel

Home cage

I.R. sensors Door + switch

30 cm

15

Number of visits

Running wheel Tunnel system 10

Loop

5

0 0

20

40

60

80

Lever presses required

Figure 10.3 Top: a schematic diagram of the apparatus used by Chris Sherwin to test demand of mice for three different opportunities for locomotion: a running wheel, an oval-shaped loop, and a complex system of tunnels. Bottom: demand curves showing the average number of times per day that mice visited and used the three devices when they were required to press a lever 1, 10, 20, 40 or 80 times for each visit. Drawing and data from Sherwin (1998). Drawing courtesy of Dr. Chris Sherwin, reproduced by permission of Elsevier.

the running wheel roughly 14 times per day, but visited the loop much less. As the cost increased, the number of visits to the tunnel system declined quickly, whereas the mice continued to visit the wheel many times per day until the cost became very high (Figure 10.3). In Dawkins’ terms, demand for the wheel was more inelastic than for the other options. In addition, when the cost of entry to the wheel was high, the mice showed more activity per visit, and thus maintained much the same overall

Dfraser_C010.indd 200

6/27/2008 12:44:18 PM

Preferences and Motivation

201

level of wheel running over a wide range of cost. Sherwin concluded that mice are strongly motivated to use a running wheel, and that wheels make an appropriate form of environmental enrichment for the species.13 An alternative method of varying the cost of a resource is to create physical obstacles that animals must overcome. In a test of the hen’s motivation to enter a secluded nest box before laying an egg, animal behaviourists Jonathan Cooper and Michael Appleby allowed hens to move from their home pen into a special ‘nesting pen’ equipped with small boxes where the hens could retreat to lay. To enter the nesting pen, the birds had to pass through a gap which could be set from relatively wide (220 mm) down to very narrow (95 mm). The hens themselves measured about 120 mm wide, so they had to squeeze themselves through the gap when it was at its narrower settings. In the two hours before laying, the hens showed keen interest in the nesting pen. On days when the gap was wide, the hens entered and re-entered the nesting pen many times. On days when the gap was narrow, they made numerous unsuccessful attempts to enter, but ﬁnally squeezed themselves through the gap in time to lay in one of the boxes. Cooper and Appleby concluded that nesting sites should be considered a ‘necessity’ for the hens in the study because the birds ‘persevered with gaining access to the nest box as the cost of access was increased’.14 Demand elasticity has great intuitive appeal. If we could generate simple demand curves for various commodities such as food, litter and social contact, would the slope of these lines not provide an objective way to rate the relative importance of the commodities for a given animal species? In hind-sight and after various attempts, there appear to be many complications that stand in the way.15 Let us look at three examples. One complication is that an animal’s strength of motivation for a reward is likely to change from one time to another depending on a host of relevant variables. A rat might show relatively inelastic demand for food when it is very hungry but much more elastic demand when it is not; a hen might be very eager to roost on a perch for resting at night, but have little interest in a perch at other times. In theory we could design an elaborate experiment to show how demand elasticity varies with level of deprivation, time of day, and other factors; but this would be a large 13Sherwin, C. 1998. The use and perceived importance of three resources which provide caged laboratory mice the opportunity for extended locomotion. Applied Animal Behaviour Science 55: 353–367. 14Cooper, J.J. and Appleby, M.C. 1996. Demand for nest boxes in laying hens. Behavioural Processes 36: 171–182. The quotation is from page 180. 15A detailed critique is provided by: Kirkden, R.D., Edwards, J.S.S. and Broom, D.M. 2003. A theoretical comparison of the consumer surplus and the elasticities of demand as measures of motivational strength. Animal Behaviour 65: 157–178. Elements of the controversy were discussed earlier by Dawkins, 1988 and 1990, Dantzer, 1990, Mench and Stricklin, 1990, and by Mason et al., 1998, and commentaries that followed that article. I am grateful to Richard Kirkden for helpful discussion of the issues.

Dfraser_C010.indd 201

6/27/2008 12:44:19 PM

202

Studying Animal Welfare

undertaking, and in any case it would not give us a single value of demand elasticity to characterize the importance of the given resource for a given type of animal. A second issue is how to divide rewards into ‘parcels’ in order to determine demand elasticity. In the case of food, it seems fairly straightforward to divide the reward into discrete quantities such as food pellets, and thus be able to collect the many measurements needed to generate a demand curve. But how should we parcel out rewards such as bedding or opportunities for dust-bathing or mating? In her experiment with hens working for food or litter, Dawkins allowed hens ﬁve minutes access to the rewards each time they performed the required amount of work. But is ﬁve minutes long enough to allow meaningful dust-bathing? If human beings were given rewards such as ﬁve minutes of sleep, would the time allowed per reward not inﬂuence its perceived value? Again, we could use an elaborate experiment to determine how elasticity of demand is inﬂuenced by the manner of parcelling a reward into discrete quantities, but even with such data we would not have a single value of elasticity that would characterize the importance of the resource. Finally (for purposes of our discussion, although this does not exhaust the issues that arise in estimating demand elasticity), when demand curves are plotted over a range of prices, they often turn out not to have any single slope that characterizes the relationship between consumption and price. Food is commonly regarded as a typical necessity which should show inelastic demand. Thus, for example, if rats are required to press a lever to obtain their food, and if the number of presses required per reward is varied from day to day, we would expect the animals to adjust their rate of pressing and obtain a fairly consistent daily food intake. This turns out to be true over a considerable range of prices, but if the price is very high, the animals allow their food intake to drop to a very low level,16 possibly because they switch to a strategy of conserving bodily resources rather than expending resources to obtain food. Hence, the rats’ demand for food appears inelastic over low and medium prices and elastic at very high prices. In this and many examples, there is room to doubt whether we can ﬁnd a single value, representing the slope of a demand curve, which adequately characterizes the animal’s motivation for a given resource. Indeed, in economics it is generally recognized that a demand elasticity is only relevant for a point on a demand curve, or at most for a limited range of prices, rather than for a commodity in general. Fortunately, an animal’s willingness to pay for a resource can be measured in more straightforward ways than demand elasticity.17 The most easily understood measure is simply the highest price that the animal will pay (in economics, the ‘reservation price’) for a unit of the resource (Figure 10.2). For example, by varying the price one can simply determine the maximum price a hen will pay for a pellet of food, or for a nest box, or for a perch where she can roost at night. 16Hurst, S.R., Raslear, T.G., Shurtleff, D., Bauman, R. and Simmons, L. 1988. A cost–beneﬁt analysis of demand for food. Journal of the Experimental Analysis of Behavior 50: 419–440. 17Kirkden et al., 2003.

Dfraser_C010.indd 202

6/27/2008 12:44:19 PM

Preferences and Motivation

203

The value of these resources will ﬂuctuate from time to time depending on level of motivation and other factors, but if the hen will pay a very high price at any given time, this suggests that the resource is important to the hen. Moreover, this measure can be applied even to resources (such as mating opportunities) which do not lend themselves to being divided into parcels as would be needed to generate a demand curve. Where it is possible to generate a demand curve, the area to the left of the curve (Figure 10.2) can be used in a more general way to express the price the animal will pay for any given quantity of the resource. In the language of economics this is called the ‘consumer’s surplus’ (or ‘consumer surplus’). It represents the price that the consumer is willing to pay for a given amount of a commodity, minus the price that they are actually required to pay. Thus, if I am willing to pay ﬁve dollars for a basket of apples but need to pay only three dollars, then my consumer’s surplus – the surplus value that I as a consumer gain on the transaction – is two dollars. In animal husbandry, resources such as food and bedding are provided ‘free’ to the animals; hence, whatever price they are willing to pay for a given quantity of the resource represents their consumer’s surplus. Georgia Mason and co-workers provided an elegant example of these various measures in a study of how strongly American mink (Mustela vison) are motivated to obtain different resources. Mink are partially aquatic carnivores that are often raised in cages for their fur. In the wild, mink perform a wide range of behaviour that is impossible in captivity; for example, they swim, rest in several nest sites, survey the environment from raised perching places, and explore the burrows of potential prey animals. In Mason’s study, 16 mink in standard cages were trained to push against weighted doors for access to various resources including a tunnel, a raised platform, an alternative nest box, and a small pool of water where they could swim. An empty cage was included as an additional option that the mink would not be very motivated to enter. The animals lived in the experimental apparatus for six weeks, and had no access to the resources except by working for them. The amount of weight that the mink had to lift in order to open the doors was varied systematically between 0 and 1.25 kilograms which represented roughly the maximum amount the animals would lift. Some resources, such as the tunnel and the raised platform, were used when the price of entry was low, but not when it was high. For other resources, especially the pool of water, the mink worked harder and harder as the price increased. Mason used several measures of the animals’ willingness to pay (Table 10.1). The maximum price paid (reservation price) for access to water was 1.25 kilograms because all animals paid this amount (the highest amount charged) for this resource; in contrast, some animals did not open the door to the tunnel when the price was that high. Consumer’s surplus varied from 81 kilograms for the water to 21 for the tunnel and 9 for the empty cage. The mink showed relatively inelastic demand for the pool of water (a slope of only 0.26) and more elastic demand for the tunnel (0.73), indicating that the mink would work harder and harder to maintain access to water for swimming,

Dfraser_C010.indd 203

6/27/2008 12:44:19 PM

204

Studying Animal Welfare

Table 10.1 Three measures of mink’s willingness to pay for seven different resources. The measures were elasticity of demand, consumer’s surplus and reservation price. Data are taken from Mason et al. (2001).

Resource Water pool Alternative nest site Novel objects Raised platform Toys Tunnel Empty cage

Elasticity of demand

Consumer’s surplus (kg)

Reservation price (kg)

0.26 ± 0.04 0.41 ± 0.08 0.58 ± 0.08 0.57 ± 0.07 0.62 ± 0.05 0.73 ± 0.07 0.77 ± 0.06

81 ± 10 61 ± 6 54 ± 5 51 ± 8 24 ± 3 21 ± 2 9±1

1.25 ± 0.00 1.17 ± 0.05 1.16 ± 0.04 1.14 ± 0.06 1.06 ± 0.07 1.06 ± 0.06 0.84 ± 0.07

Notes: Elasticity of demand is expressed as the slope of the demand curve, plotted as the logarithm of the number of rewards obtained versus the logarithm of the price; larger numbers indicate a steeper slope and hence greater elasticity. Consumer’s surplus (in kg) was calculated as the area under the demand curve plotted as the price per visit versus the number of visits. Reservation price (in kg) is the greatest weight that the animals would lift for access to the resource; the maximum weight used in the experiment was 1.25 kg. All values are the mean and standard error based on 16 animals.

but tended to forego the tunnel when the price was high.18 Animal welfare scientist Richard Kirkden and co-workers have argued that under many circumstances the consumer’s surplus and reservation price are more likely to provide a valid estimate of willingness to pay than is demand elasticity.19 MEASURES OF MOTIVATION HAVE most commonly been used in cases where animals are motivated to obtain a resource that they value, but an animal’s welfare is also inﬂuenced by situations that the animal would prefer to avoid. Examples include unpleasant temperatures, noise, or management practices that cause fear or pain. We can gain a better understanding of these situations by a kind of motivation research that deals speciﬁcally with avoidance. Animal welfare scientists Lee Niel and Dan Weary used avoidance testing to study the humaneness of killing rats with carbon dioxide. Of the millions of rats used each year by scientists, the majority are killed at the end of the experiments, and exposure to lethal levels of carbon dioxide is one of the most common methods. Typically, the rats are placed in a sealed chamber and carbon dioxide is gradually introduced to replace the air. The rats lose consciousness when the gas reaches a concentration of 30–40% and then die as the concentration increases. Carbon dioxide can cause pain to the eyes and respiratory passages, presumably 18Mason, G.J., Cooper, J. and Clarebrough, C. 2001. Frustrations of fur-farmed mink. Nature 410: 35–36. 19Kirkden et al., 2003.

Dfraser_C010.indd 204

6/27/2008 12:44:19 PM

Preferences and Motivation

205

Figure 10.4 The apparatus used by Lee Niel and Dan Weary to test whether rats will avoid different concentrations of carbon dioxide in the air. Rats lived in the upper (home) cage and were trained to walk down the black tunnel to the lower (test) cage, where they received a treat of sugared breakfast cereal. On different days of the experiment, the lower cage was filled with different concentrations of carbon dioxide to identify which concentrations would deter the rats from entering the test cage and obtaining the treat. Photograph reproduced by kind permission of Joanna Makowska.

because it combines with moisture to create carbonic acid. Pain is thought to develop when the concentration reaches around 30–50%; hence, the hope is that the animals will lose consciousness before the gas becomes painful. However, Matthew Leach, David Morton and other laboratory animal specialists had found that rats will predictably vacate a chamber when carbon dioxide reaches levels well below what would be needed to make the animals unconscious.20 As a stronger test of avoidance, Niel and Weary constructed an apparatus consisting of an upper home cage where the animals lived, linked by a tunnel to a lower test cage where carbon dioxide or other gases could be introduced (Figure 10.4). They trained the rats to walk down from the home cage to the test cage once a day when a signal was given, and there they received a treat of 20 pieces of a sugared breakfast cereal that rats ﬁnd very attractive. Once the rats had become accustomed to the routine, Niel and Weary pre-ﬁlled the test cage with various concentrations of carbon dioxide. In pure air, or with carbon dioxide at 5 or 10%, the rats remained in the test cage for several minutes and ate all the treats. With 20% carbon dioxide, the rats spent only a few seconds in the test 20Leach, M.C., Bowell, V.A., Allan, T.F. and Morton, D.B. 2002. Aversion to gaseous euthanasia agents in rats and mice. Comparative Medicine 52: 249–257 with erratum on 52: 449 and 572.

Dfraser_C010.indd 205

6/27/2008 12:44:20 PM

206

Studying Animal Welfare

cage and only two of the nine animals touched the treats. These and other results indicated that rats ﬁnd carbon dioxide very aversive at concentrations well below what would be required to render them unconscious.21 For pigs, at least, the use of an ‘inert’ gas such as argon may help to solve this problem. Gas stunning of pigs is normally done by lowering the animals into a shaft pre-ﬁlled with a gas, usually carbon dioxide. Veterinary scientists Mohan Raj and Neville Gregory tested the responses of pigs when lowered into various concentrations of carbon dioxide and other gases. Carbon dioxide at 40% or higher concentrations generally produced ‘severe respiratory distress’ before the pigs lost consciousness and many pigs tried to escape from the chamber. However, if the air in the chamber was largely replaced by the inert gas argon so that oxygen was reduced to a concentration of only 2%, the animals generally lost consciousness with little respiratory distress and without attempting to escape. Because carbon dioxide is produced naturally in the body and cleared by breathing, pigs have apparently evolved a response (and likely an unpleasant affective state) when the gas approaches dangerous levels; but because argon is completely unnatural, the animals have likely not evolved any means of detecting or avoiding the gas. Hence, Raj and Gregory argued, stunning pigs by using argon to replace most of the air should be more humane than using a high concentration of carbon dioxide.22 In Niel and Weary’s experiment rats could demonstrate that they found carbon dioxide unpleasant simply by escaping from it. Where such a direct approach is not feasible, it may be possible instead to see whether animals will learn to avoid a situation after repeated exposure. Jeffrey Rushen provided a classic example in his work to evaluate the humaneness of electro-immobilization of sheep. Electroimmobilization involves passing a pulsed, low-voltage current through the body. The method has been used to immobilize animals temporarily for procedures such as shearing wool, and it had also been claimed to function as an analgesic that makes procedures such as shearing less unpleasant for the animals; but does it really prevent discomfort, or does it merely render the animals unable to move? To test these alternatives, Rushen trained sheep to move along a runway to a pen where they received mildly aversive treatments such as rough shearing, either with or without electro-immobilization. The experiment showed that over repeated trials, sheep that received the electrical treatment became more difﬁcult to move along the runway than those that received the mildly aversive treatments without electro-immobilization. Rushen concluded that electro-immobilization actually made unpleasant procedures more aversive, not less.23 21Niel,

L. and Weary, D.M. 2006. Rats avoid exposure to carbon dioxide and argon. Applied Animal Behaviour Science 107: 100–109. 22Raj, A.B.M. and Gregory, N.G. 1996. Welfare implications of the gas stunning of pigs 2. Stress of induction of anaesthesia. Animal Welfare 5: 71–78. The quotation is on page 77. 23This and related studies are reviewed in: Rushen, J. 1996. Using aversion learning techniques to assess the mental state, suffering, and welfare of farm animals. Journal of Animal Science 74: 1990–1995.

Dfraser_C010.indd 206

6/27/2008 12:44:22 PM

Preferences and Motivation

207

Experimental psychology has provided other well researched methods that use avoidance learning to test whether animals ﬁnd situations aversive. ‘Shuttle avoidance’ involves two joined compartments, and the animals (usually rats) are required to ‘shuttle’ from one to the other, typically whenever a signal sounds, in order to avoid some unpleasant stimulus such as an electric shock delivered through the ﬂoor. Ian Duncan and his student Mark Rutter attempted to use this procedure to study fear in chickens. The fear-producing stimulus was a balloon that could be inﬂated rapidly in either compartment; the hens merely had to move from one compartment to the other when a tone sounded in order to prevent the balloons from inﬂating. Instead of doing this, however, the chickens appeared to ‘freeze’ when the tone was given, and they did not learn the avoidance task.24 Duncan and Rutter concluded that shuttle avoidance is not a suitable procedure to test for fear in the type of chickens that they used. They then tried a different approach called ‘passive avoidance’. They trained chickens to peck a key for a food reward whenever a signal had been given. Once the hens had thoroughly learned the response, two tones were introduced. One tone signalled that pecking would produce food as usual; the other indicated that pecking the key would lead to a loud noise instead. With this procedure, when all that was required was the passive response of not pecking, the hens successfully learned to avoid the noise.25 In principle, shuttle avoidance and passive avoidance provide two approaches that can be used to test whether an animal ﬁnds a given situation aversive, but Duncan and Rutter’s research showed the care that is needed in selecting a valid procedure. STUDIES OF ANIMAL MOTIVATION have played a key role in debates over the troublesome concept of ‘behavioural needs’. When Alex Stolba and David Wood-Gush set out to design the better pig pen, they also spelled out their rationale for trying to accommodate natural behaviour as a means of improving animal welfare. They stated that animals have certain ‘behavioral needs’ including ‘the needs to perform the obligatory … elements of a behavior sequence’, and ‘the needs to perceive speciﬁc external key stimuli or compound features’ which are necessary for the ‘obligatory’ behaviour to be performed.26 The term ‘behavioural needs’ (or ‘ethological needs’) had already been brought into use, especially by a number of German-speaking animal welfare scientists such as Werner Bessei, Dieter Fölsch and Rose-Marie Wegner; and a formal requirement to provide for animals’ ‘ethological needs’ had been created in 1978 by the European Convention for the Protection of Animals kept for Farming Purposes.27 This was a curious development because, as Marian Dawkins and others pointed 24Rutter, S.M. and Duncan, I.J.H. 1991. Shuttle and one-way avoidance as measures of aversion in the domestic fowl. Applied Animal Behaviour Science 30: 117–124. 25Rutter, S.M. and Duncan, I.J.H. 1992. Measuring aversion in domestic fowl using passive avoidance. Applied Animal Behaviour Science 33: 53–61. 26Stolba and Wood-Gush, 1984, page 289. 27Sources are cited in Dawkins, 1983.

Dfraser_C010.indd 207

6/27/2008 12:44:22 PM

208

Studying Animal Welfare

out, the term ‘ethological needs’ had no established meaning in science and was rarely mentioned by behavioural scientists (in English at least) except when warning readers against using such nebulous terms. Hence, the introduction of the term into animal welfare policy and legislation created a dilemma for scientists: should scientists dismiss the term on the grounds that it forms no part of scientiﬁc theory, or should they accept that it had acquired a role in policy, and therefore develop criteria to determine what should count as an ethological (or behavioural) need? Dawkins, for one, suggested that animal behaviourists should engage in the issue. In 1983 she urged that, ‘Unless ethologists tackle the issue of ethological needs soon, ethology is in danger of lending the appearance of respectability to a term which has no agreed meaning, is obscure and misleading and, because of its widespread use in welfare circles, is in danger of jeopardizing ethology as a scientiﬁc discipline with a major contribution to make to animal welfare’.28 Thus began a debate over whether and how to give an operational, scientiﬁc meaning to a term which, in the view of some scientists, had been foisted on science by policy makers. The debate hinges in part on what the word ‘need’ means in everyday speech and what it is likely to imply when included in policy and legislation. When we speak of a need, there is generally a corresponding goal or function. Once nutritionists understood that vitamin C prevents scurvy, they could determine how much vitamin C is needed to achieve this goal. However, ‘need’ is used in both a weak and a strong sense. In the weak sense, we can speak of what we ‘need’ to accomplish any goal, no matter how inessential it might be. Thus, to win the Americas Cup I would ‘need’ an expensive yacht. In the strong sense, however, we use ‘need’ only if there would be serious consequences if the need is not met. In this sense of the word, I do need vitamin C but not a yacht. Need, in this strong sense, is sometimes distinguished from mere desire, and the two terms have very different moral implications. For example, parents are considered foolish if they give a child everything it desires, but they are considered evil if they deprive the child of its needs. In the case of animal welfare, then, what circumstances would justify invoking the language of ‘need’ with its strong ethical implications? If an animal acts in a certain way, or prefers an environment where it can perform a certain behaviour, can we devise criteria for claiming that this reﬂects a need (speciﬁcally a ‘behavioural need’) as opposed to merely a desire? In particular, if an animal is ostensibly healthy, does it make sense to speak of its having a ‘need’ to perform certain types of behaviour? In her treatment of this issue, Dawkins made an important distinction. A wild bird, she argued, needs food so that it will not die; it needs to make a nest so that its offspring will survive. Survival and reproduction are the ‘ultimate’ goals or functions served by feeding and nest-building, and they justify our speaking about what the bird needs to accomplish these goals. However, the bird itself may have no concept of these ultimate goals. Rather, the bird has feelings that motivate it to 28Dawkins,

Dfraser_C010.indd 208

1983, page 1196.

6/27/2008 12:44:22 PM

Preferences and Motivation

209

eat and build a nest. These feelings that motivate behaviour are, in Dawkins’ terms, the ‘proximate’ needs that correspond, in the bird’s experiential life, to the ultimate goals of surviving and reproducing. In a natural environment, Dawkins argued, proximate needs and ultimate goals would be largely aligned. If a calf failed to suck, it would starve; if a sow failed to ﬁnd and prepare a suitable nest, her piglets would die. But on a commercial farm, the calf is given a bucket of milk which it can consume by drinking and thus be nourished without ever sucking; and the sow is placed in a warm, hygienic stall where her piglets will thrive without her making a nest. The ultimate goal is met by human intervention, but the animal may be driven by the proximate need nonetheless. From the animal’s viewpoint, it ‘needs’ to suck or make a nest just as much as if these activities were actually required to accomplish their ultimate function. By this logic, then, if we want to use the language of need, we can do so by identifying proximate needs that animals remain highly motivated to pursue even when the corresponding ultimate goals are met in other ways. The argument was taken up by Barry Hughes and Ian Duncan.29 They pointed out that the idea of behavioural needs harks back to a theory of motivation advanced by Konrad Lorenz.30 Lorenz had proposed that the motivation to perform a particular action (his term for the motivation, translated into English, was ‘action-speciﬁc energy’) builds up gradually over time. Once sufﬁcient action-speciﬁc energy has accumulated, then a key stimulus in the environment triggers the behaviour. With the action-speciﬁc energy thus discharged (rather like the ﬂushing of a toilet), the motivation to perform the behaviour begins to build again. If a large amount of action-speciﬁc energy has accumulated and no suitable stimulus is present, the behaviour may occur spontaneously in the form of a ‘vacuum activity’ (Chapter 7) as illustrated by Lorenz’s pet starling that went through its natural insect-catching with no insects present. Lorenz’s model makes a plausible ﬁt to behaviour like eating: during a period without food, the motivation to eat grows gradually stronger and stronger, and then declines once a large meal has been consumed. However, Lorenz tried to apply his model to behaviour in general. In a book that was both famous and infamous, Lorenz proposed that his theory of motivation explains human violence and warfare.31 He saw aggression as a natural part of human behaviour which (according to his theory) will inevitably occur when the action-speciﬁc energy for aggression

29Hughes, B.O. and Duncan, I.J.H. 1988. The notion of ethological ‘need’, models of motivation and animal welfare. Animal Behaviour 36: 1696–1707. 30Lorenz, K. 1939. Comparative study of behavior. Report given at the Zoological Convention, Rostock, 1939. Republished 1957 as pages 239–263 in Instinctive Behavior: The Development of a Modern Concept (C.H. Schiller, editor and translator). International Universities Press, New York. 31Lorenz, K. 1966. On Aggression (M.K. Wilson, translator). Harcourt, Brace & World, New York.

Dfraser_C010.indd 209

6/27/2008 12:44:23 PM

210

Studying Animal Welfare

accumulates. The idea was denounced by many critics,32 but beyond the speciﬁc debate on aggression, Lorenz’s simple model of motivation was also rejected by many animal behaviourists who saw it as scientiﬁcally inadequate. One critic was the English ethologist Robert Hinde who made the much more ﬂexible (and plausible) proposal that the causes of behaviour include both internal and external factors. The motivation to drink, for example, is inﬂuenced by internal factors such as feedback from the receptors in the body that sense dehydration, and by external factors such as the sight of cold beer. In Hinde’s view, the relative role of the internal and external factors will vary from behaviour to behaviour and from species to species. Hence, how a given type of behaviour is motivated is an empirical question that must be determined by investigation, not assumed on the basis of a single, universal model of motivation. Lorenz’s model had an important implication for animal welfare. If all behaviour depends on a kind of ‘energy’ speciﬁc to that action, then it would seem inevitable that animals will be motivated to perform every element of their natural behaviour. If a well fed pet starling is still compelled to perform insect-catching because of the build-up of action-speciﬁc energy for insect-catching, then why would the same not apply to grazing, mating, sucking, nest-building, dust-bathing, incubating eggs, playing and exploring? Would there not be a ‘need’ to perform these activities, just as there is a need to eat, drink and sleep? Lorenz’s thinking was probably more inﬂuential among German-speaking ethologists than among their English-speaking counterparts. This may help to explain why the idea of ‘ethological need’ seemed more readily adopted by German-speaking scientists, including Stolba. In contrast, many of the leading English-speaking animal behaviourists tended to view Lorenz’s ideas with immense scepticism. Thus it is perhaps no surprise that British-educated scientists like Dawkins, Duncan and Hughes looked upon ‘ethological needs’ as (in Dawkins’ words, quoted above) ill-deﬁned, misleading and obscure.33 Hughes and Duncan, in their critical analysis of ‘ethological needs’, asked how this concept could be interpreted if we reject Lorenz’s model and instead accept Hinde’s proposal that a given type of behaviour may be motivated by internal and external factors in varying degrees. They noted that some types of behaviour – escape and aggression were their examples – occur largely in response to external stimuli. Although internal factors (such as the animal’s state of arousal) are certainly relevant, the behaviour will not occur if the external stimulus is not present. In such cases, Hughes and Duncan proposed, there is no ‘ethological need’ to ﬂee 32A collection of critical commentaries is assembled in: Montagu, A. (editor) 1973. Man and Aggression. Oxford University Press, New York. 33Wood-Gush was South African, and to my surprise he appeared to hold Lorenz’s views in reasonably high regard. This may help to explain why he was able to co-author the paper with Stolba which accepted the concept of behavioural needs, at a time when many British ethologists would have distanced themselves from such thinking.

Dfraser_C010.indd 210

6/27/2008 12:44:23 PM

Preferences and Motivation

211

or ﬁght, and an animal’s welfare would not be compromised if it cannot carry out these types of behaviour as long as the animal is not exposed to the external stimuli that cause them. However (they noted) for certain other types of behaviour, internal factors play a determining role. Eating and sleeping are obvious cases, but there may be others, such as nest-building and other reproductive activities if these are strongly controlled by hormonal or other internal mechanisms that are largely independent of external cues. For Hughes and Duncan, then, the critical issue is the relative role of internal and external factors in causing the behaviour. If the motivation comes mainly from external cues, and if eliminating the external cues eliminates the motivation, then the animal has no ‘need’ to perform the behaviour as long as the cues are not present. However, if the motivation is largely internal, then one can reasonably speak of a ‘need’ to perform the behaviour. The argument was carried further by ethologist Per Jensen and psychologist Frederick Toates in a paper called, ‘Who needs behavioural needs?’ They argued that it is a mistake to use a ‘catalogue approach’ to classify certain types of behaviour as constituting needs and others as not. Instead, they argued that all behaviour is inevitably inﬂuenced by both internal and external factors, and that the relative importance of the different causal factors will vary from one situation to another. What is important for animal welfare, and what would justify speaking of a ‘need’ to perform a certain behaviour, is ‘the degree of suffering an animal will experience’ if the behaviour cannot be carried out.34 The debate over behavioural needs was nicely illustrated in a controversy about nest-building by sows. When a sow is about to farrow, she goes through an elaborate procedure of exploring the environment, ﬁnding a suitable nesting site, rooting out a depression, and then carrying straw and other material and fashioning it into a large nest.35 In many species, such nest-building is at least partly controlled by hormonal changes shortly before parturition. However, Mike Baxter, an engineer and designer who worked on improved animal housing, proposed that in the sow the behaviour is under external control – speciﬁcally, that if we provide the sow with a suitable pre-formed nest having all the correct features, then the sow will have no motivation to carry out nest-building herself. In other words, Baxter proposed, the sow needs a nest, but she does not have any ‘behavioural need’ to build one.36 This idea was put to an experimental test by a group of Baxter’s colleagues led by Dale Arey. They observed six sows making nests before farrowing, and they

34Jensen,

P. and Toates, F.M. 1993. Who needs behavioural needs? Motivational aspects of the needs of animals. Applied Animal Behaviour Science 37: 161–181. The quotations are from pages 176 and 177. 35Jensen, P. 1986. Observations on the maternal behaviour of free-ranging domestic pigs. Applied Animal Behaviour Science 16: 131–142. 36Baxter, M.R. 1983. Ethology in environmental design for animal production. Applied Animal Ethology 9: 207–220.

Dfraser_C010.indd 211

6/27/2008 12:44:23 PM

212

Studying Animal Welfare

quantiﬁed the time the animals spent in the major nest-building activities, namely rooting or pawing the substrate and carrying straw to the site. They also recorded the physical features of the nests the sows built. They found that the nests consisted of an elliptical depression averaging 1.6  1.1 metres in size and 0.2 metres deep, and containing an average of 23 kilograms of straw. Arey then provided another six sows with a pre-formed nest constructed to these speciﬁcations, and observed how the sows responded. Instead of just accepting the pre-formed nest and settling down to farrow, the sows still went through elaborate nest-building, although there were differences. The total duration of nest-building was actually somewhat longer, but with 23 kilograms of straw already present, the sows carried only 9.5 kilograms of additional straw. Arey and co-workers concluded that nest-building by sows is under both internal and external control, but clearly the motivation to build a nest is not eliminated by providing a suitable nest pre-formed. They proposed that ‘farrowing accommodation should therefore enable sows to perform nest building’, and that ‘nest material is essential to the behaviour and therefore to the environmental requirements of the sow at this time’.37 In this case, animals showed a sustained motivation to perform a given behaviour even when the ostensible goal of the behaviour was already met; but are there other criteria that could be used to justify speaking of ‘behavioural needs’? Dawkins outlined four possibilities.38 One is whether the animal performs the behaviour as a vacuum activity when the normal targets are not present. An example might be tongue-rolling (‘vacuum grazing’) by calves that are fed on a liquid diet and not provided with roughage. A second criterion is whether the animal shows a ‘rebound’ effect by performing high levels of the behaviour after a period of deprivation, as would occur with eating. A third criterion is whether the animal shows signs of frustration, including displacement and stereotyped behaviour, if there is no opportunity to perform the behaviour. An example is Duncan’s evidence of frustration among hens that cannot ﬁnd a nesting place in the hour before they lay an egg. And as a ﬁnal criterion, Dawkins suggested a high willingness to pay as evidenced, for example, by inelastic demand. Each of these criteria, Dawkins proposed, constitute evidence of strong motivation that cannot be prevented or assuaged in other ways and would justify applying the term ‘behavioural need’. Let us note one ﬁnal caution when speaking of ‘behavioural needs’. We have previously discussed how tongue-rolling by calves may be seen as vacuum grazing, and how sucking dry teats may constitute redirected sucking. A common assumption is that these abnormal activities reﬂect a strong internal motivation (in Hughes and Duncan’s terms, a ‘behavioural need’) to perform these actions. However, 37Arey, D.S., Petchey, A.M. and Fowler, V.R. 1991. The preparturient behaviour of sows in enriched pens and the effect of pre-formed nests. Applied Animal Behaviour Science 31: 61–68. The quotation is on page 67. 38Dawkins, M.S. 1988. Behavioural deprivation: A central problem in animal welfare. Applied Animal Behaviour Science 20: 209–225.

Dfraser_C010.indd 212

6/27/2008 12:44:23 PM

Preferences and Motivation

213

we have now seen that these actions may actually have signiﬁcant functional consequences: tongue-rolling appears to reduce ulceration of the stomach perhaps by stimulating the production of saliva, and sucking a dry teat results in the release of insulin and other digestive hormones. Thus, when we see an animal persisting in a seemingly functionless action, we should not assume that the behaviour is based simply on a ‘behavioural need’ to carry out the action; there may also be a need for yet-unknown consequences that follow from the behaviour.39 MUCH OF THE SCIENTIFIC literature on animal welfare has focused on unpleasant motivational states such as hunger, fear and frustration, but what if the motivation is based on pleasure? Can tests of motivation help us determine whether positive affect, as opposed to negative affect, is involved? Many experimental psychologists think of affect (if they consider it at all) as a single scale. For example, psychologist Michel Cabanac, who has championed the role of affect in motivation since the 1970s, portrays affect on a linear scale running from positive to negative,40 and Daniel Kahneman, a strong proponent of ‘hedonic psychology’, speaks simply of the ‘good-bad dimension’ of affect.41 However, in Chapter 8 we noted that in the case of food there are separate neural processes underlying ‘wanting’ (roughly, hunger) and ‘liking’ (the pleasure derived from eating); and we discussed the idea that negative affect may motivate animals to respond to threats to survival or reproduction, whereas positive affect may motivate potentially beneﬁcial behaviour such as play and exploration at times when there is little cost to engaging in these non-urgent activities. How might studies of motivation help separate these two types of situation? One candidate method comes from Marian Dawkins’ use of demand elasticity. If positive affect motivates animals to carry out beneﬁcial but low-priority behaviour at times when the cost of performing the behaviour is low, then we would expect such activities to be particularly sensitive to cost. Thus, the animal might eagerly perform the behaviour when the cost is low, but give up readily (showing ‘elastic demand’) when the price is increased. A classic review by psychologist Jerry Hogan and biologist Tim Roper provided several examples of behaviour that follows this pattern. For instance, Roper had observed that mice would learn to press a key to receive strips of paper which the animals would then manipulate into a nest. If the mice received a strip of paper every time they pressed the key, then they would often obtain 100 strips of paper in an experimental session. However, when the ‘price’ increased, the mice did not increase their rate of key pressing proportionally, and they received less and less 39For further discussion of behavioural needs see: Duncan, I.J.H. 1998. Behavior and behavioral needs. Poultry Science 77: 1766–1772. 40Cabanac, M. 1992. Pleasure: The common currency. Journal of Theoretical Biology 155: 173–200. 41Kahneman et al., 1999.

Dfraser_C010.indd 213

6/27/2008 12:44:23 PM

214

Studying Animal Welfare

paper. Moreover, if they had to walk more than a very short distance between the key and the paper dispenser, the level of effort they expended to obtain paper dropped noticeably. Summarizing an earlier interpretation of such behaviour by comparative psychologist Stephen Glickman, Hogan and Roper suggested that paper may differ from food as a reward because ‘the animal does not “feel” need for nest material in the way that it “feels” hungry or thirsty’.42 Instead (if the hypothesis proposed above is correct), manipulating fresh paper into a nest may be a pleasant activity for mice which will occupy them for long periods of time but only if the cost of doing so is low. A different way of using motivation to shed light on affect was developed by behavioural scientists Tina Widowski and Ian Duncan who asked whether hens that work for access to a dust-bath are ‘increasing pleasure’ or ‘reducing suffering’. They considered that if the motivation to dust-bathe is based on a negative affective state (analogous to hunger or thirst), then hens should show a rebound effect; that is, they should do more of the behaviour after a period of deprivation. In contrast, if dust-bathing is motivated by a form of enjoyment (analogous to playing or exploring), then the birds might be expected to perform the behaviour when the opportunity is present, whether or not they have been deprived. With this in mind, Widowski and Duncan trained ten hens to push against a door in order to gain access to peat moss in which they could dust-bathe, and they then added various weights to the door to determine the maximum weight that the hens would push to reach the peat moss. The hens were tested under two conditions: while housed in a cage where dust-bathing was impossible, and while housed in an enriched environment where they had constant access to a dust-bath. Widowski and Duncan found that although the hens did on average push more weight when deprived than when not deprived, several of the birds showed equal or greater motivation when they had been free to dust-bathe in the home environment. They noted that peat moss, which is a highly preferred substrate for dust-bathing, seems to ‘switch on’ the behaviour even in birds that have recently performed it in some other substrate. The results (in their words) ‘are very difﬁcult to explain using a “needs” model of motivation in which deprivation leads to state of suffering. They are much more consistent with an “opportunity” model of motivation in which performance of the behaviour, when the opportunity presents itself, leads to a state of pleasure’.43 INTUITIVELY IT MAY SEEM obvious that what an animal prefers, and what it will work to obtain, are relevant to its welfare. But if the term ‘animal welfare’ is used 42Hogan, J.A. and Roper, T.J. 1978. A comparison of the properties of different reinforcers. Pages 155–255 in Advances in the Study of Behavior, Volume 8 (J.S. Rosenblatt, R.A. Hinde, C. Beer and M.C. Busnel, editors). Academic Press, New York. The quotation is on page 192. 43Widowski, T.M. and Duncan, I.J.H. 2000. Working for a dustbath: Are hens increasing pleasure rather than reducing suffering? Applied Animal Behaviour Science 68: 39–53. The quotations are from pages 39 and 40.

Dfraser_C010.indd 214

6/27/2008 12:44:23 PM

Preferences and Motivation

215

to capture three broad concerns about the quality of life of animals – how well they feel, how well they function, and how free they are to live in a manner to which they are adapted – then could we state more precisely how preferences and motivation help us to understand animal welfare? A long-standing assumption, dating back to Aristotle’s History of Animals, is that much human and animal behaviour is motivated by states that are experienced as pleasant or unpleasant.44 As examples of unpleasant affect, we drink when dehydrated in order to reduce thirst, we rest a twisted ankle to relieve pain, and we step back from a snarling dog because of fear. As examples of positive affect, we drink port at the end of a meal, or go for a walk through the countryside, or play with a friendly dog, because of the pleasure that these activities provide. If we assume that such affective states have a strong inﬂuence on what animals prefer and what they are motivated to do, then surely preference and motivation are key avenues for understanding affect and hence animal welfare. Understanding motivation is also important for interpreting how natural behaviour relates to animal welfare. A hen’s natural behaviour, as we have discussed, includes eating when food is present and ﬂeeing when a fox is present. If we can show that the hen is highly motivated to ﬁnd food when no food is present, but is not motivated to ﬁnd a fox when no fox is present, then we can separate the elements of natural behaviour into those that contribute positively to animal welfare and those that do not. Understanding motivation is also relevant to the health and functioning of animals, but here the links need especially careful thought. Rats deprived of certain vitamins develop a preference for diets that contain the missing nutrients and (within limits) will work to obtain the kind of food that they need to restore their health. Chicks kept at temperatures that are too cold for their health will work for extra warmth. Pigs exposed to unhealthy levels of ammonia in the air will relocate to cleaner air. In these and many cases, much of motivation can be seen as helping to maintain the body in a healthy state. However, the link between an animal’s health and its preferences and motivations is far from perfect. Problems may arise when animals are kept in environments that include unnatural features such as rich foods that animals select in harmful amounts by following preferences that may have functioned well at an earlier time in the animal’s evolutionary ancestry. Problems may also arise if animals are exposed to dangers that are beyond the capacity of their sensory and affective systems. We have seen, for example, that pigs will detect and avoid lethal concentrations of carbon dioxide, but show little reaction to lethal concentrations of argon.45 Similarly, many ﬁsh species successfully avoid being harmed by certain aquatic pollutants such as copper simply by swimming away from contaminated water, but they fail to avoid certain other (especially human-made) pollutants 44See 45Raj

Dfraser_C010.indd 215

Fraser and Duncan, 1998, for a brief review. and Gregory, 1996.

6/27/2008 12:44:23 PM

216

Studying Animal Welfare

such as phenol,46 presumably because the ﬁsh never evolved the capacity to detect them. In these examples, evolution appears to have equipped animals to detect and avoid certain harms but not others, especially not harms that their ancestors would not have encountered during the evolutionary history of the species. As these examples show, what an animal prefers, and what it will work to obtain or avoid, will not always be what is best for its welfare as determined by other criteria. In these last examples we encounter a core problem in understanding animal welfare. We glimpsed the problem in the study by Hughes and Black which found that a type of ﬂooring that hens preferred also occasionally caused injuries to their feet. The problem, stated more generally, is how to integrate different types of information which appear relevant to animal welfare but which may lead us toward different conclusions. This will be a key theme for the ﬁnal part of the book.

46Giattina,

J.D. and Garton, R.R. 1983. A review of the preference-avoidance responses of ﬁshes to aquatic contaminants. Residue Reviews 87: 43–90; Hartwell, S.I., Jin, J.H., Cherry, D.S. and Cairns, J. Jr. 1989. Toxicity versus avoidance response of golden shiner, Notemigonus crysoleucas, to ﬁve metals. Journal of Fish Biology 35: 447–456.

Dfraser_C010.indd 216

6/27/2008 12:44:23 PM

Part III Drawing Conclusions about Animal Welfare

Dfraser_C011.indd 217

6/27/2008 4:58:41 PM

Introduction

Animal welfare science, in addition to the many practical ways that it can be used to improve the housing, handling and management of animals, has also been applied in matters of policy, for example in regulations, standards and other measures that have been used to ensure (and assure the public) that animal welfare is being safeguarded. These animal welfare assurance programmes have taken many forms. In the aftermath of Ruth Harrison’s Animal Machines, the government of the United Kingdom commissioned the writing of ‘codes of practice’ to outline how farm animals ought to be kept. The codes were voluntary, but an Act of Parliament passed in 1968 stipulated that failure to follow the codes could be used as evidence if a person was charged with the offence of causing animals unnecessary pain or distress. Then, in 1988 new regulations turned some of the provisions of the codes, such as the space allowance for hens in cages, into legal requirements.1 Similar codes, some purely voluntary and others with varying degrees of legal or other recognition, were subsequently created in a number of countries by governments, industry organizations and other bodies. The situation unfolded quite differently in Sweden. There, as we have seen, the impetus for change came from Astrid Lindgren, the famous author of such children’s classics as Pippi Longstocking and Emil and his Clever Pig, books that depicted children and animals living happy, independent lives and resisting the efforts of adults to limit their freedom. In her animal welfare advocacy, Lindgren was advised by no less a scientist than Ingvar Ekesbo who (as we saw in Chapter 5) had made a career of studying the health problems associated with different methods of housing and raising animals. Like Lindgren, Ekesbo was strongly in favour of more natural rearing conditions for animals. He had, in fact, convinced the Swedish government to ban the feeding of animal-derived products to ruminants, on grounds that the practice is unnatural, more than a decade before Mad Cow Disease caused other countries to do the same.

1Radford, M. 2001. Animal Welfare Law in Britain: Regulation and Responsibility. Oxford University Press, Oxford.

218

Dfraser_C011.indd 218

6/27/2008 4:58:41 PM

Introduction

219

With Ekesbo providing technical guidance, Lindgren began in the 1980s to write letters to the major Swedish newspaper in her chiding, satirical style, castigating the national government for allowing farmers to keep animals in systems that she saw as unnatural to the point of causing widespread misery. If the unknown Ruth Harrison could create huge political fallout in the United Kingdom, no Swedish government could afford to ignore persistent criticism from the country’s best known and best loved citizen. The government responded by passing a new Animal Protection Ordinance in 1988, and it followed Lindgren’s criticisms to a remarkable degree. Where Lindgren had written, ‘Laying hens are still jammed into inadequate cages, so that they can’t walk on their own two feet’, the Animal Protection Ordinance speciﬁed, ‘Hens for egg production shall not be housed in cages’. Where Lindgren had written, ‘legislation is necessary to guarantee dairy cows’ grazing rights in the summer months’, the Animal Protection Ordinance stipulated, ‘Cattle which are kept for milk production and which are older than six months shall be sent out to pasture in the summer’. Where Lindgren had written, ‘There are still windowless stalls where the animals are cooped up in perpetual darkness’, the Animal Protection Ordinance stated, ‘Livestock buildings shall be ﬁtted with windows that let in daylight’. In fact, so great was Lindgren’s inﬂuence that Prime Minister Ingvar Carlsson was reported to have visited Lindgren personally to gain her support in the hours after agreement on the new law had been reached.2 But national legislation could only do so much. When the United Kingdom banned the narrow veal calf crate in the 1980s, there was an existing export of calves from British farms to France and the Netherlands where crates were still in widespread use, and nothing in the new legislation could prevent the export from continuing. Moreover, trade agreements made it impossible for the United Kingdom to block the import of veal from France and the Netherlands, possibly veal from the very animals that had previously been shipped from the United Kingdom. The situation became particularly intolerable when the farmer-turnedpolitician who served as minister of agriculture could not provide assurances that his own calves were not involved in this ﬁasco. In response, an animal protection society took out a full-page advertisement in The Sunday Times with a photo of a pathetic calf and the caption: ‘Kept in the dark, abused and unable to move. And that’s just our agriculture minister’.3 The upshot has been a concerted effort to bring trading partners into line with a common set of standards. In the 1980s and 1990s, the European Union began passing directives – agreements that member countries are obliged to translate into 2Anonymous, 1989. How Astrid Lindgren Achieved Enactment of the 1988 Law Protecting Farm Animals in Sweden. Animal Welfare Institute, Washington. The quotations from Lindgren are on pages 17, 3 and 16–17, respectively. The quotations from the Swedish Animal Protection Ordinance of 1988 are from sections 9, 10 and 2, respectively. 3The Sunday Times, 15 January 1995.

Dfraser_C011.indd 219

6/27/2008 4:58:41 PM

220

Drawing Conclusions about Animal Welfare

national law – governing cages for hens and production methods for veal calves and pigs. For the most part, the directives speciﬁed quantitative standards such as minimum space allowances per animal and the maximum slope on the cage ﬂoor. But roughly ten years later, a new set of directives took a much more reformist stance by requiring countries to phase out the narrow crate for veal calves by 2007, the standard cage for hens by 2012, and the gestation stall for sows (during most of pregnancy) by 2013.4 But beyond the European Union, could different animal welfare standards become an impediment to international trade? An obvious solution would be to create global standards that trading partners would agree to follow. With this in mind the (then) 167 member countries of the World Organization for Animal Health voted in 2001 to begin developing global animal welfare guidelines, and the ﬁrst ones – a set of guidelines for slaughter, transport and killing of animals in disease eradication – were approved in 2005.5 Nonetheless, in many countries the national government lacked either the political will or the constitutional power to create national laws regarding animal welfare, and other sectors of society took up the lead. In the United States, where the national government has not exercised jurisdiction over how animals are kept on farms, McDonald’s Restaurants announced that the company would require animal welfare standards to be met by their suppliers of certain products. Burger King and other companies followed suit and began expanding their programmes to other countries. Soon it was realized that the situation would become confusing for both suppliers and consumers if each company had its own standards. Consequently, two national bodies in the United States – the Food Marketing Institute and the National Council of Chain Restaurants – began the controversial process of proposing a common set of standards on behalf of the supermarket and chain restaurant industries.6 Even in jurisdictions where legal standards were in place, certain producers, retail companies and animal welfare charities decided to take the further step of creating product-differentiation programmes that would allow conscientious consumers to purchase products produced according to speciﬁc standards. The best known of these is Freedom Food, begun in 1994 by the Royal Society for the Prevention of Cruelty to Animals in the United Kingdom. It used an active programme of 4Stevenson,

2004.

5Bayvel, A.C.D., Rahman, S.A. and Gavinelli, A. (editors). 2005. Animal welfare: Global issues,

trends and challenges. Revue Scientiﬁque et Technique de l’Ofﬁce International des Epizooties 24: 463–813. The guidelines can be found in the Terrestrial Animal Health Code, revised and published annually by the World Organization for Animal Health and available on its website at www.oie.int. 6Brown, K.H. and Hollingsworth, J. 2005. The Food Marketing Institute and the National Council of Chain Restaurants: Animal welfare and the retail food industry in the United States of America. Revue Scientiﬁque et Technique de l’Ofﬁce International des Epizooties 24: 655–663.

Dfraser_C011.indd 220

6/27/2008 4:58:41 PM

Introduction

221

inspection and labelling to guarantee that their products were from animals that had been raised, transported and slaughtered in accordance with speciﬁc standards.7 The result of all this activity is a wide range of animal welfare assurance programmes including voluntary codes, national or regional laws, international agreements, corporate programmes, and product-differentiation schemes.8 Almost all of these animal welfare programmes are claimed to be based on science. Yet as we have seen, animal welfare science is a multi-faceted activity, pulling information from different disciplines, building on different theories, informed by different views of what constitutes a good life for animals. It is one thing to put speciﬁc scientiﬁc ﬁndings to work in practical improvements, for example by identifying the cause of stomach torsion in sows, or allowing servals to use their natural behaviour to obtain food. But it is quite another thing to somehow combine the many and sometimes conﬂicting insights from different sources of information into comprehensive policy or standards designed to safeguard animal welfare. In the ﬁnal section of the book we turn to the difﬁcult issue of drawing conclusions about animal welfare from diverse types and sources of information. Chapter 11 proposes how we can conceptualize animal welfare in light of the different interpretations and scientiﬁc approaches, and some of the difﬁculties that arise when different scientiﬁc approaches lead to contradictory conclusions. Chapter 12 discusses several case studies where scientists have attempted to combine disparate criteria and types of evidence in order to draw conclusions about animal welfare. Chapter 13 develops some concepts and vocabulary for understanding the complex interaction of ‘facts’ and ‘values’ in the scientiﬁc study of animal welfare as well as in other ﬁelds of mandated science.

7RSPCA. 2007. Freedom Food – Welcome. Royal Society for the Prevention of Cruelty to Animals (RSPCA). Available at: http://www.rspca.org.uk/servlet/Satellite?pagenameRSPCA/RS PCARedirect&pgFreedomFoodHomepage, consulted April 2007. 8For a more detailed discussion see: Fraser, D. 2006b. Animal welfare assurance programs in food production: A framework for assessing the options. Animal Welfare 15: 93–104.

Dfraser_C011.indd 221

6/27/2008 4:58:41 PM

11

How Do the Different Measures Relate to Each Other?

In Part II we reviewed the various scientiﬁc approaches that have been used in the assessment of animal welfare. Some have tried to assess the basic health and functioning of animals, and have borrowed methods from pathology, epidemiology, animal science, stress biology and animal behaviour. Others have tried to understand the affective states of animals through methods drawn from physiology, neurobiology, behaviour, evolutionary theory and experimental psychology. Other studies, drawing on the idea that for animals to have good lives they must live in a manner for which they are adapted, have used methods of naturalistic observation, time budget analysis, motivation theory and environmental design. Now we need to turn to one of the most challenging issues in the ﬁeld: given the diversity of methods and approaches used in animal welfare research, how do the different types of information ﬁt together to create our understanding of animal welfare? PERHAPS THE SIMPLEST POSSIBILITY is that the different criteria of animal welfare – which I have grouped under the three broad categories of basic health and functioning, affective states and natural living – must inevitably agree because all are manifestations of the same single phenomenon that we call ‘animal welfare’. An analogy may help. When wildlife biologists began monitoring populations of moose, some of the ﬁrst statistics they collected were the ages of animals killed in the annual hunting season. For the males, a rough division into age classes was initially made from the antlers which ranged from the simple spikes of calves to the broad, complex antlers of older males. A second method was to classify animals according to the amount of wear on their teeth. Using this approach an experienced observer could divide animals into nine tooth-wear categories which were thought to reﬂect age, although imprecisely for older animals. Finally, biologists developed the method of extracting a tooth, preparing it in the laboratory, and counting the rings of dental cementum which were believed to be laid down at a rate of one per year. 222

Dfraser_C011.indd 222

6/27/2008 4:58:41 PM

How Do the Different Measures Relate to Each Other?

223

All three methods yielded broadly compatible results: there was no mistaking calves, one-year-olds or senior citizens by any of the methods, and there was broad agreement about other age groups. However, each method had its strengths and weaknesses. Antler classes were the least precise but could be determined just by looking at the head. Dental wear classes gave more categories, but could only be used if hunters allowed the jaw to be removed. Cementum ageing was assumed to be the most precise but it was by far the most laborious. Thus the different methods had their champions, although cementum ageing, once fully developed, eventually became the accepted standard. If the different approaches described in the earlier chapters constitute different ways of ‘measuring animal welfare’, in a way that is analogous to the different ways of determining the age of moose, then although each approach might have its practical advantages and disadvantages and be subject to certain errors, they should be in agreement.1 Thus, if we were to compare the welfare of animals in different housing systems or under different handling methods, we would expect the three approaches, although perhaps differing in precision or convenience, to provide the same answer as long as they are used correctly. This seems to be a widely held expectation. When biomedical researchers try to eliminate pain and distress in laboratory animals, they often claim (plausibly) that this will make the animals function more normally and that the experimental results will therefore be more valid. When zoo managers or organic farmers try to keep animals in natural surroundings, there is often an implicit assumption that this will result in the animals being more healthy and more happy. Using similar logic in reverse, animal producers operating high-health conﬁnement systems sometimes claim that because the animals are growing vigorously and are relatively free from disease, they must also be content, and their living conditions must be well suited to their needs. A tacit assumption in all these cases is that the three views of animal welfare, although involving different techniques, must ultimately agree in their conclusions. There is also a certain evolutionary logic to support such a view. Animals are adapted to live in environments with certain features: certain types of food, certain opportunities to ﬁnd and create shelter, certain types of social groupings and so on. We expect animals to possess adaptations that allow them to thrive in such environments, and conversely we expect that they may fail to thrive in environments 1I

ﬁnd the term ‘measure’ somewhat unsatisfactory when applied to animal welfare. If (to paraphrase the Concise Oxford Dictionary of Current English, 1982) ‘measure’ means ‘ascertain the extent or quantity of something by comparison with a ﬁxed unit or object of known size’, then we cannot ‘measure’ animal welfare in this commonly used sense of the word, because there is no ﬁxed unit or known quantitative standard. We can, of course, measure relevant variables such as mortality rates and disease incidence; hence, measurement is part of animal welfare science, but the phrase ‘measuring animal welfare’ seems to create a mistaken impression. My preference is to speak of ‘assessing’ animal welfare in the sense (again paraphrasing the Concise Oxford Dictionary) of evaluating or estimating its magnitude or quality.

Dfraser_C011.indd 223

6/27/2008 4:58:41 PM

224

Drawing Conclusions about Animal Welfare

for which they lack adaptations. Thus we may expect that putting animals in conditions that are more natural for the species should promote good health and functioning of the body. Moreover, if we view affective states such as fear, pain and pleasure as products of natural selection which evolved so that the animal should seek what is beneﬁcial and avoid what is harmful, then affective states should also be closely tied to the health and functioning of the animal.2 Nonetheless, we have encountered situations where the pursuit of animal health arguably does not improve – and indeed may reduce – welfare according to other criteria. In an effort to prevent serious respiratory disease in cats, managers of animal shelters often house cats individually in stainless-steel cages which are cleaned daily. This is done to protect the animals’ health, but the resulting environment is anything but natural, and the social isolation and frequent cleaning are likely to be distressing to the cats.3 Similarly, laboratory directors often try to prevent the spread of disease by keeping mice in individually ventilated cages where pathogens are excluded but where rapid air movement may cause discomfort.4 In the practice of ‘Segregated Early Weaning’ piglets are removed from the mother at two weeks of age or less, and are raised in a separate facility well isolated from the pathogens they would normally encounter from their mothers and other older animals. The practice is so successful at eliminating disease pathogens that the animals grow rapidly enough to justify the high cost of this demanding management system, but the early separation is clearly unnatural and appears to lead to prolonged separation distress among the piglets.5 Moreover, some of the surgical procedures done to farm animals – tail-docking of lambs, dehorning of calves – are done to prevent problems of disease and injury, but are both unnatural and presumably painful. The pros and cons of each of these cases could be debated, but clearly we cannot simply assume that measures to improve welfare as deﬁned by health will necessarily improve welfare as deﬁned by other criteria. Similarly, the pursuit of more natural living conditions does not necessarily promote welfare as judged in other ways. A study in Croatia compared sows in indoor, total-conﬁnement units and sows kept outdoors. The outdoor sows lived in a more natural environment but had more lameness, shorter lifespan, and a higher 2Duncan,

I.J.H. 1996. Animal welfare deﬁned in terms of feelings. Acta Agriculturae Scandinavica, Section A, Animal Science, Supplement 27: 29–35; Dawkins, M.S. 1998. Evolution and animal welfare. Quarterly Review of Biology 73: 305–328; Broom, D.M. 1998. Welfare, stress, and the evolution of feelings. Advances in the Study of Behavior 27: 371–403. 3Gourkow, N. and Fraser, D. 2006. The effect of housing and handling practices on the welfare, behaviour and selection of domestic cats (Felis sylvestris catus) by adopters in an animal shelter. Animal Welfare 15: 371–377. 4Baumans, V., Schlingmann, F., Vonck, M. and van Lith, H.A. 2002. Individually ventilated cages: Beneﬁcial for mice and men? Contemporary Topics in Laboratory Animal Science 41: 13–19. 5Robert, S., Weary, D.M. and Gonyou, H. 1999. Segregated early weaning and welfare of piglets. Journal of Applied Animal Welfare Science 2: 31–40.

Dfraser_C011.indd 224

6/27/2008 4:58:42 PM

How Do the Different Measures Relate to Each Other?

225

neonatal mortality rate.6 Poultry red mite (Dermanyssus gallinae) is a parasite that bites domestic chickens and can cause irritation, disease and even deaths in extreme cases. It is easily controlled in cages, but in more ‘natural’ systems such as aviaries it can become a signiﬁcant health problem.7 Organic systems provide many examples. They generally emphasize high animal welfare in the sense of natural living conditions, and they can be expected to eliminate health problems arising from crowding and poor air quality; but in eliminating unnatural elements, organic producers also minimize the use of pharmaceutical products that contribute to animal health in other ways. A review of studies comparing the health of organic and conventionally reared animals found more problems of parasitic disease on organic units, although other types of disease showed no consistent difference.8 Attempts to eliminate unpleasant affective states are no panacea either. Allowing Labrador Retrievers to eat as much as they want will presumably prevent hunger, but the result is an overweight animal whose longevity is reduced.9 A similar problem is seen in those farm animals (pigs, meat chickens) that have been bred speciﬁcally for high food intake and rapid weight gain; in these cases, many animals kept to breeding age would become obese and unhealthy unless their food is restricted to the point of causing chronic hunger.10 In many species of mammals, mothers wean their young partly by avoiding them and refusing to nurse. This almost certainly causes frustration and distress among the young, but if we tried to prevent these unpleasant states by giving the young free access to the mother, they would remain dependent on milk for longer than normal and might fail to develop a healthy level of solid food intake.11 If wild Canada geese in their breeding range are kept warm and well fed during the autumn and winter, many will fail to migrate south. The artiﬁcial feeding and shelter presumably prevent feelings of hunger and cold, but the result is a life that is far from natural and probably less healthy in certain respects. As the examples show, attempts to prevent negative affect may in some cases lead to a less natural life and in some respects a less healthy animal. 6Cox, B. and Bilkei, G. 2004. Lifetime reproductive performance of sows kept indoors and outdoors in Croatia. Veterinary Record 154: 569–570. 7Nordenfors, H. and Höglund, J. 2000. Long term dynamics of Dermanyssus gallinae in relation to mite control measures in aviary systems for layers. British Poultry Science 41: 533–540. 8Lund, V. and Algers, B. 2003. Research on animal health and welfare in organic farming – a literature review. Livestock Production Science 80: 55–68. 9Lawler, D.F., Evans, R.H., Larson, B.T., Spitznagel, E.L., Ellersieck, M.R. and Kealy, R.D. 2005. Inﬂuence of lifetime food restriction on causes, time, and predictors of death in dogs. Journal of the American Veterinary Medical Association 226: 225–231. 10Appleby, M.C. and Lawrence, A.B. 1987. Food restriction as a cause of stereotypic behaviour in tethered gilts. Animal Production 45: 103–110; Duncan, I.J.H. 2001. Animal welfare issues in the poultry industry: Is there a lesson to be learned? Journal of Applied Animal Welfare Science 4: 207–221. 11Pajor, E.A., Weary, D.M., Fraser, D. and Kramer, D.L. 1999. Alternative housing for sows and litters – 1. Effects of sow-controlled housing on responses to weaning. Applied Animal Behaviour Science 65: 105–121.

Dfraser_C011.indd 225

6/27/2008 4:58:42 PM

226

Drawing Conclusions about Animal Welfare

What are we to make of these seeming disagreements between the three types of evidence used to assess animal welfare? Is it simply that some of the measures are incorrect – like ﬁnding that one of the methods of estimating the age of moose is actually ﬂawed – or is there more to the issue? TO ANSWER THIS QUESTION, it helps to think of animals as facing certain challenges and possessing certain adaptations that help them deal with challenges.12 By challenges I am referring to conditions that would tend to harm the animal unless it makes some action or adjustment. In biological terms, harms would mean reducing the animal’s biological ﬁtness, for example by threatening its survival, its health or the viability of its offspring. Challenges may thus include cold or hot temperatures that threaten to disturb homeostasis, lack of food or water, exposure to pathogens or parasites, physical environments that may result in injury, and aggression from other animals. By adaptations I am referring to those features or strategies that the animal can use to avoid or mitigate harms. Adaptations reﬂect the animal’s evolutionary ancestry, although they may also be modiﬁed by genetic changes during domestication and they may involve the individual’s own learning and development. Some adaptations (as we have seen in Chapter 9) are anatomical and physiological, such as the ability to digest certain diets, to ward off certain pathogens and parasites, and to regulate body temperature under a certain range of environmental conditions. Other adaptations are behavioural; these may include a predisposition to build a nest before laying an egg, or to wallow in mud on a hot day. Adaptations may also involve pleasant or unpleasant affective states such as hunger, cold, pain and pleasure which motivate the animal to act in ways that reduce the unpleasant states and increase the pleasant ones, and cognitive abilities that allow animals to learn to respond to challenges in ways suited to the particular circumstances. As long as an animal is kept in an environment very similar to that in which the adaptations evolved and the individual developed, then we may expect a close correspondence between an animal’s adaptations and the challenges it faces. For example, when living in a wild state, a species of small mammal might face the challenges of predation by owls and of food shortages in the winter, and the animals might possess corresponding adaptations such as well developed burrowing, a strong motivation to store food, and the cognitive ability to remember the location of food it has cached. However, we often keep animals in environments very unlike those in which its adaptations evolved. Human-designed environments will almost inevitably use some materials, such as metal, concrete, or sawdust bedding, which are not natural for the animals. Simple practicality and economics may dictate that we keep 12This section reworks ideas developed by Fraser, D., Weary, D.M., Pajor, E.A. and Milligan, B.N. 1997. A scientiﬁc conception of animal welfare that reﬂects ethical concerns. Animal Welfare 6: 187–205. I am grateful to my co-authors for their valuable contributions to these ideas.

Dfraser_C011.indd 226

6/27/2008 4:58:42 PM

How Do the Different Measures Relate to Each Other?

Circle A Adaptations possessed by the animal

227

Circle B Challenges faced by the animal in its current circumstances

1

3

1 Adaptations that no longer serve an important function 3 Challenges for which the animal has corresponding adaptations

2

2 Challenges for which the animal lacks corresponding adaptations

Figure 11.1 A conceptual model illustrating three types of welfare problems that may arise when the adaptations possessed by the animal (Circle A) make an imperfect fit to the challenges it faces in the situation in which it is kept (Circle B). Adapted from Fraser et al. (1997). Reproduced with kind permission of the Universities Federation for Animal Welfare.

animals in social groupings and at densities that would not occur in nature, and in environments that are far smaller and less complex. Moreover (as we have seen) there can be good welfare arguments for departing from strictly natural environments by providing elements such as hygiene, shelter and warmth. For these various reasons, most human-designed environments are likely to differ signiﬁcantly from the environments in which the animals’ adaptations evolved. In such cases the adaptations that an animal possesses are unlikely to correspond exactly to the challenges that its actual environment creates. Figure 11.1 provides a way of visualizing the situation. Circle A in the ﬁgure represents the adaptations that the animal possesses, and Circle B represents the challenges that it faces in the circumstances in which it is kept. If the adaptations do not ﬁt closely with the challenges – if the circles do not overlap completely – then three different types of problems may arise. To the left of the ﬁgure are what we might call Type 1 problems where an adaptation possessed by an animal no longer serves a signiﬁcant role in meeting challenges in the environment. For example, as we have discussed, pregnant sows appear highly motivated to ﬁnd and prepare a nest in the days before they give birth, presumably because this behaviour was important for the survival of

Dfraser_C011.indd 227

6/27/2008 4:58:42 PM

228

Drawing Conclusions about Animal Welfare

their young in the environment in which the species evolved.13 Modern sows that farrow in a warm, protected barn retain this motivation even though the behaviour is no longer important for the survival of their piglets. Thus, a sow conﬁned in a stall where she cannot look for and prepare a nest site may become intensely restless, make persistent attempts to escape, and root vigorously on the solid ﬂoor, sometimes to the point of injuring herself.14 The inability to search for and build a nest may be very unpleasant for the animal (a negative affective state), even though the survival of the young (the behaviour’s original contribution to biological functioning) is not affected. In this case, the natural behaviour of nest-building, and the affective state that motivates it, appear to have become uncoupled from the beneﬁcial effect on biological functioning that they once served. To the right of the ﬁgure are what we might call Type 2 problems which involve challenges for which the animal lacks corresponding adaptations. For example (as we have seen) animals may be poorly equipped to avoid becoming obese if concentrated food is available, to avoid losing physical condition if daily life does not require exercise, to avoid pathogens when kept close to diseased animals, and to avoid certain environmental contaminants even at levels that may damage their health. In such cases we may see serious threats to health and functioning which are not accompanied by any corresponding adaptation such as a feeling of distress or discomfort that might motivate the animals to avoid the harm. Thus, once we depart from environments in which the animals’ adaptations evolved, and include artiﬁcial elements such as restrictive housing, heating, veterinary drugs and artiﬁcial feed, we may encounter a certain uncoupling of the adaptations that an animal possesses and the problems that the animal faces in its current circumstances.15 Under such conditions, different criteria of animal welfare may not align: there may be signiﬁcant effects on affective states with no commensurate effects on health and functioning (Type 1 problems), and signiﬁcant effects on health and functioning with no commensurate effects on affective states (Type 2 problems). What is best for the animal’s basic health and functioning may no longer be what the animal wants or what it would naturally do. In the above examples, the two circles in Figure 11.1 fail to correspond because of a change in the environment – a shift in circle B. However, an uncoupling of the circles can also occur if artiﬁcial selection changes the animal’s genetic makeup – a shift in circle A. For example, chickens bred for very high egg production have the capability of mobilizing large amounts of calcium from the bone to create egg shell, but this same capability leaves these birds prone to developing osteoporosis. Dairy cattle bred for very high milk production have various metabolic features that equip the animals to synthesize milk, but these features also leave the animals vulnerable 13Jensen,

1986. G.J. and de Lange, A. 1986. Pre- and post-farrowing behaviour in primiparous domestic pigs. Applied Animal Behaviour Science 15: 31–43; Arey et al., 1991. 15For further discussion see Fraser et al., 1997, and Barnard and Hurst, 1996. 14Lammers,

Dfraser_C011.indd 228

6/27/2008 4:58:42 PM

How Do the Different Measures Relate to Each Other?

229

to metabolic disorders. The result in these cases is a variant of Type 1 problems: the animals have a genetically determined capability that suits them for some commercial purpose (roughly, an ‘adaptation’ to commercial production) but creates harm, or risk of harm, to their welfare. The relative importance of Type 1 and Type 2 problems will vary from case to case. When people criticize restrictive ways of keeping animals, the focus is often on Type 1 problems: the caged monkey, the chained dog, and other animals that are predisposed to live active lives in complex environments but whose present circumstances, although perhaps meeting basic needs for food and disease prevention, provide little opportunity for the animals to exercise their behavioural adaptations. Type 2 problems are not greatly discussed in the animal welfare literature, but are common concerns in human welfare: miners who inhale damaging amounts of asbestos, ﬁshing communities who consume dangerous levels of heavy metals in ﬁsh, children who consume high-fat foods to the point of reducing their life expectancy, all without any evolved adaptation that would stop them from incurring such harm. However, the central part of Figure 11.1 is also important. It represents a third class of problems that can occur – let us call them Type 3 problems – where the challenges and adaptations correspond but welfare problems are nonetheless very real. Here we encounter challenges such as hot temperatures or aggressive groupmates for which the animal has corresponding adaptations including affective states (such as feeling hot or fearful) and natural mitigative actions (such as panting or hiding). In these cases the animal may be able to exercise its adaptations in a natural way, but we do not expect welfare problems to disappear. The expectation, rather, is that welfare problems should be more like the ones seen in nature rather than human-caused problems, and that affective states and natural behaviour should be better coupled with basic health and functioning. A simple Venn diagram (Figure 11.2), which has been proposed in somewhat different forms by Michael Appleby and Vonne Lund, allows us to picture the three criteria of animal welfare as three partially overlapping circles where an animal’s welfare might be judged by the three criteria in any combination.16 IF, THEN, THE THREE GENERAL criteria used to assess animal welfare do not always agree, perhaps one of the three can be said to represent ‘true’ animal welfare. The others might be well enough correlated with animal welfare to serve as useful indicators in some circumstances, but the one true criterion would trump the others whenever disagreement arises. There have been several attempts to argue that one or other of the three general criteria is correct and that the others are not.

16Appleby, M.C. 1999. What Should We Do About Animal Welfare? Blackwell Science, Oxford; Lund, V. 2006. Natural living – a precondition for animal welfare in organic farming. Livestock Science 100: 71–83.

Dfraser_C011.indd 229

6/27/2008 4:58:42 PM

Basic health and functioning

1

4

6 7

2

3 5

Affective states

Natural living

Figure 11.2 A Venn diagram showing three general criteria of animal welfare and the seven possible areas produced when they overlap imperfectly. Six dogs and a moose serve to illustrate how animals may be thought to enjoy good welfare by the different combinations of criteria. The figure is adapted from Appleby (1999) and Lund (2006) with kind permission of Elsevier. The examples are my own. (1)

(2)

(3)

(4)

(5)

(6)

(7)

Dfraser_C011.indd 230

Good welfare by basic health and functioning alone: a timid dog in a rescue shelter, receiving veterinary care and a healthy diet, but unable to play and explore, and fearful of other dogs and unfamiliar human visitors. Good welfare by affect alone: a well-cared-for dog with a terminal illness, receiving affectionate attention and analgesics to prevent pain, but debilitated by failing health and unable to continue its normal activities. Good welfare by natural living alone: a moose during a severe winter, living in a fully natural environment, becoming emaciated because of the cold temperatures and the high energetic cost of finding food in deep snow, probably suffering from the cold and from fear of predation, and surviving only if body reserves outlast the winter. Good welfare by basic health and affect: a Great Dane in an apartment, well cared for, walked on a leash enough to maintain physical fitness, showered with affection and fed sensibly, but kept in a highly restrictive environment. Good welfare by affect and natural living: a ‘spoiled’ Golden Retriever receiving lots of affection, veterinary care, and given ample time to run and fetch in a large yard by the sea, but over-fed on a rich diet which leads to obesity, hip dysplasia and congestive heart failure. Good welfare by natural living and basic health: a dog with a punitive but otherwise attentive owner, healthy, well fed, and living free in the countryside, but in chronic fear of the owner and occasionally in pain from being struck. Good welfare by all criteria: the beagle of my childhood home, showered with affection, receiving good but not excessive food, given veterinary care when needed, free to run and hunt throughout the farm and forest, and killed instantly by a car when running across a country road.

6/27/2008 4:58:42 PM

How Do the Different Measures Relate to Each Other?

231

A well known and persuasively argued position is Ian Duncan’s claim that animal welfare must be deﬁned strictly in terms of affective states. Duncan used plants – speciﬁcally a pine tree – to explore what he saw as the true meaning of welfare. He noted that we speak of welfare only in the case of beings that are capable of feeling. Functioning-based concepts such as health, stress, ﬁtness and longevity can be applied to a pine tree, but because we believe that a pine tree is incapable of feeling, we do not speak of these as affecting the tree’s welfare. Duncan concluded that welfare must be ‘all to do with what animals feel’. Thus, ‘it is not being ill that reduces welfare but feeling ill’.17 Stress, ﬁtness, longevity and natural behaviour inﬂuence welfare only inasmuch as they inﬂuence the animals’ affective experience. According to this view, if the assessment of animal welfare based on functioning variables such as health, or on the ability to live a natural life, does not agree with assessment based on affective states, it is the affective states that should be accepted as the true indicators. On this basis, Duncan boldly championed the study of affective states in animals as the fundamental element of animal welfare at a time when many scientists remained sceptical about studying affect. A competing proposal, which we encountered brieﬂy in Chapter 4, came from animal scientist Frank Hurnik.18 In an argument that went roughly the opposite of Duncan’s, Hurnik distinguished between an animal’s ‘needs’ and its ‘desires’. Noting that animals may desire things that are actually harmful to themselves or others, he concluded that the quality of life of animals is more closely related to the satisfaction of their needs than of their desires, speciﬁcally their needs for survival, health and comfort in that order of priority. The upshot of Hurnik’s argument (although Hurnik did not phrase it in these terms) is that functioning-based criteria, especially survival and health, take precedence over criteria based on affective states as evidenced by what animals actually desire and seek out. As a third and contrasting proposal, let us recall how Chris Barnard and Jane Hurst argued that animals have been ‘designed’ by natural selection to live according to certain decision rules which were adaptive in the environment where the species evolved, and that the welfare of animals should be deﬁned by whether they can exercise these evolved decision rules.19 Barnard and Hurst noted, for example, that although humans have evolved to preserve their lives in a way that allows slow reproduction, many other species (many ﬁsh and rodents for example) have evolved to expend themselves for reproduction over much shorter periods. For this and other reasons, Barnard and Hurst argued that when scientists use longevity, good health, lack of stress and lack of pain and suffering as criteria for animal welfare, they are simply projecting a human perspective onto other species.

17Duncan, I.J.H. 1993. Welfare is to do with what animals feel. Journal of Agricultural and Environmental Ethics 6(Suppl. 2): 8–14. The quotations are on pages 8 and 11. 18Hurnik, J.F. 1993. Ethics and animal agriculture. Journal of Agricultural and Environmental Ethics 6(Suppl. 1): 21–35. 19Barnard and Hurst, 1996.

Dfraser_C011.indd 231

6/27/2008 4:58:43 PM

232

Drawing Conclusions about Animal Welfare

What is important, they maintained, is that animals can live in accordance with the adaptations – speciﬁcally the evolved decision rules – that are characteristic of the species. From these examples we see that each of the three general criteria of animal welfare – welfare as deﬁned by affective states, welfare as deﬁned by basic health and functioning, and welfare as deﬁned by the animal’s ability to live in the manner for which it is adapted – has been championed by one or more scientists as representing the true approach to animal welfare which should trump the others in cases of disagreement. Each of the arguments has merit; each is based on different logic; and each contradicts the others. Much more could be said about these arguments and their underlying logic.20 For now, however, it is enough to note that no one proposal has emerged as a consensus position among scientists. BUT PERHAPS CONSENSUS COULD still be achieved through some form of consultation process. In environmental management, for example, it is sometimes possible for different ‘stakeholders’ or interested parties to reach a mutually acceptable vision of their goals through consultation. As we have seen, the different conceptions of animal welfare tend to be held by different groups, with many veterinarians and animal producers emphasizing basic health and functioning, many animal protectionists emphasizing prevention of suffering, and many consumers believing that animals should be able to lead natural lives. If we could organize a consultative process whereby these and other groups could be brought together, might the participants agree on a single, uniﬁed conception of animal welfare? Based on my own experience of dealing with different stakeholders from different parts of the world, all of whom bring valid viewpoints to the subject, here is how I imagine the process might unfold. First, a philosopher such as Lennart Nordenfelt, using his critical training to drill down to the philosophical essence of the welfare concept, might well conclude that welfare is fundamentally about happiness. Thus, good health or the ability to live in a natural manner might (in Nordenfelt’s words) be a ‘condition for’ welfare, but not a ‘part of’ welfare.21 On the other hand, a ﬁsh veterinarian concerned about a high rate of injury among salmon raised in net-cages, might feel that to equate welfare with ‘happiness’ would turn a useable, everyday concept into an exercise in the imponderable. For such a person, welfare is a term that is used to capture problems such as physical injuries, parasites, ‘stress’ caused by crowding, and related matters, and standards for the welfare of ﬁsh need to address these practical matters in a practical way, whether or not there is any discernible link to the ‘happiness’ of the ﬁsh. A European ethologist, engaged in assessing animal welfare standards on farms that use modern conﬁnement facilities, might well feel that the animal welfare 20For example, Fraser et al., 1997; Nordenfelt, L. 2006. Animal and Human Health and Welfare: A Comparative Philosophical Analysis. CAB International, Wallingford, UK. 21Nordenfelt, 2006, page 161.

Dfraser_C011.indd 232

6/27/2008 4:58:43 PM

How Do the Different Measures Relate to Each Other?

233

problems centre around the prevention of natural behaviour, and that any welfare standard worthy of the name must accommodate natural behaviour ahead of any other priority. In contrast, a veterinarian working in India, where unowned dogs number in the millions and where many of these animals die every day from starvation or disease, might be unimpressed by talk of natural behaviour and insist that any deﬁnition of welfare must be about health and nutrition above all else. Then again, a laboratory animal technician, raising primates in disease-free indoor colonies, might feel that the key issue is that animals in a severely limited environment fail to develop the cognitive and physical abilities that lie within their capacity. Such animals, like certain brain-damaged humans, may be in good physical health and may show little sign of suffering. The tragedy (as seen by the technician) lies in their failure to develop the capacities and live the kind of lives that they are capable of living. Thus different people, all using animal welfare to refer to the quality of life of animals, naturally focus on quite different elements because of their very different experiences. Each has valid logic and solid, experience-based reasons for believing as they do. Confronted by the different experiences of others, some people may well be persuaded to expand their conception of animal welfare to incorporate other points of view, but I have seen little evidence that people will be persuaded to abandon the emphasis that their own experience has produced, in favour of an alternative conception of animal welfare. I have described these possible outcomes as what might emerge if a consultative process were to take place in the future, but in a sense such consultations have already occurred. The animal welfare advisory bodies that have been formed in various jurisdictions typically bring together people with a broad range of interests and expertise, and some of these bodies have collectively articulated their conception of animal welfare. The most common example of such a collective vision is the well known ‘Five Freedoms’. These (as described by John Webster) are: freedom from thirst, hunger, and malnutrition; freedom from discomfort; freedom from pain, injury and disease; freedom from fear and distress; and freedom to express normal behaviour.22 The Five Freedoms were proposed in their current form by the Farm Animal Welfare Council of the United Kingdom, and have been adopted as guiding principles for animal welfare by a wide range of other bodies with diverse membership. However, instead of coming down on the side of any one of the three broad conceptions of animal welfare, the Five Freedoms incorporate all three, arranged in the form of a check-list that corresponds roughly to the way scientists might divide the relevant issues: veterinary issues (disease, injury), nutritional issues (thirst, hunger, malnutrition), environmental issues (discomfort), and behavioural issues (freedom to perform natural behaviour), with affective states captured in several places 22Webster,

Dfraser_C011.indd 233

J. 2005. Animal Welfare: Limping towards Eden. Blackwell Publishing, Oxford.

6/27/2008 4:58:43 PM

234

Drawing Conclusions about Animal Welfare

by terms such as pain, fear and distress. Thus, even the extensive discussion that has occurred over the development and acceptance of the Five Freedoms has tended more to conﬁrm than to simplify the three general conceptions of animal welfare. In short, whatever future potential there might be to resolve the conﬂicts through consultation and debate, discussion thus far has not led to one of the conceptions of animal welfare being generally regarded as correct and the others as incorrect or secondary. Before concluding this discussion, we need to draw a subtle but important distinction. When Lennart Nordenfelt proposed that welfare can best be conceptualized in terms of happiness, he was (I believe) trying to capture the philosophical essence of the concept of welfare. Much the same may be true of Ian Duncan’s argument that welfare must refer to affective states because we speak of welfare for animals but not for plants. However, when a given term is applied in practices, policies, regulations and international agreements, it often carries an everyday meaning (or meanings) that may not correspond exactly to the philosophical essence of the concept. Thus, for example, programmes and organizations pursuing ‘health care’ or ‘nature conservation’ may not ﬁt perfectly with the philosophical essence of the concepts ‘health’, ‘care’, ‘nature’ and ‘conservation’. Following similar logic, my goal (in this discussion and throughout this book) is not to determine the philosophical essence of the concept of animal welfare, but to identify what people try to capture by the term when they, for example, try to improve animal welfare conditions on their farms and zoos, create animal welfare standards, or include animal welfare in corporate policies. And when scientists do research in support of such actions, their work (I will argue below) should be guided by the everyday meaning(s) that the term carries in the practical world. ONE THING IS NOT A MATTER of debate: virtually all animal welfare scientists agree that animal welfare is a complex concept whose assessment requires attention to a number of variables. However, there are different types of complexity, and in the case of animal welfare we need to be clear on the nature of the complexity and its implications. Again, an analogy helps us explore the possibilities. Suppose that limnologists have been sent to study a remote lake and report on its characteristics. Among other things, the limnologists need to determine the lake’s biomass and its safety for recreational purposes. Both of these tasks involve studying a large number of variables, but there are important differences. To determine the biomass, the limnologists need to estimate the volume of water in the lake, collect water samples to estimate the density of suspended plankton, census the abundance and size classes of ﬁsh, and use quadrat sampling to estimate the abundance of plants. Then these measurements need to be combined according to some correct formula in order to estimate the total mass of living material. Determining the safety of the lake for recreational use also requires attention to a number of variables. These may include the presence of hidden rocks that might be

Dfraser_C011.indd 234

6/27/2008 4:58:43 PM

How Do the Different Measures Relate to Each Other?

235

a hazard to boaters, water contamination that might pose a risk to cottagers, and strong currents that could pull swimmers out of their depth. Determining safety, however, differs in important ways from determining biomass. The ﬁrst clue is that biomass is expressed in speciﬁc units such as kilograms of carbon. These units provide a ‘common currency’ whereby the mass of different types of living matter can meaningfully be combined into a single numerical value. In the case of safety, however, there are no units and hence no common currency that allows us, in a purely objective way, to establish equivalency of the different variables. There are no units, for example, that allow us to say whether ﬁve hidden rocks are equal to ten days per year when water exceeds acceptable levels of coliform bacteria. Second, in the case of biomass we assume that there is a single value which represents the true biomass of the lake and which we could in theory determine in a purely objective manner, for example by draining and sieving the lake and weighing all the living material. Thus, if different scientists propose different formulas for combining the variables, we think of these as more accurate and less accurate approximations to the true value. In the case of safety, even in theory there is no single correct numerical value. Safety will depend, for example, on whether the lake is mainly used for swimming or for cottaging or for boating. The different variables are thought to contribute to safety not because they allow us to calculate a number that expresses safety, and that could in theory be veriﬁed by purely objective means, but rather because they serve the same general function of making the lake a safer place. Nordenfelt refers to such concepts as ‘conglomerate concepts’. He deﬁnes these as concepts where ‘a number of more simple concepts have been put together to form a new, complex concept’.23 Thus, features that make the lake safe for boating, and features that make it safe for swimming, and features that make it safe for cottaging have been lumped together under the conglomerate concept of lake safety. A little reﬂection shows that animal welfare is, in important respects, more similar to safety than it is to biomass. The assessment of animal welfare involves many variables such as level of pain, freedom from illness, and the ability to perform behaviour that the animal is highly motivated to perform. However, animal welfare is not expressed in speciﬁc units that allow us to add and subtract the different variables in a mathematically correct manner. And there is no single value that could, in theory, be measured to give the ‘correct welfare’ of the animal, analogous to the correct biomass of the lake. Instead, animal welfare, like safety, appears to be a conglomerate concept wherein different elements have been grouped together because they serve the same general function of creating a better quality of life for animals. In fact, the broad criteria of animal welfare that we have discussed – basic health, affective states and natural living – may themselves best be seen as conglomerate 23Nordenfelt,

Dfraser_C011.indd 235

2006. The quotation is from page 129.

6/27/2008 4:58:43 PM

236

Drawing Conclusions about Animal Welfare

concepts. Health involves a combination of factors such as physical ﬁtness and freedom from various diseases, parasites and injuries, grouped together beneath the term ‘health’ because they are all relevant to the normal functioning of the body. The concept of natural living also involves many diverse elements including the ability of the animal to undergo normal ontogenic development, to go through normal postural adjustments, to groom, mate and raise young, and perhaps to have access to natural sunshine and unpolluted air. Affect is more controversial. Recall how Michel Cabanac portrays affect on a single linear scale running from positive to negative, and how Daniel Kahnemann writes of positive and negative affect as forming a simple ‘Good/Bad dimension’.24 According to such views of affect, reducing pain and inducing pleasure could be expressed on a common scale. However, other scientists have proposed that positive and negative affect are distinct phenomena that have separate neural processes and that serve different adaptive functions.25 Moreover, everyday human experience tends to bear out this interpretation: for example, a person can be simultaneously in pain from arthritis and joyful about the birth of a grandchild without the two hedonic experiences cancelling each other and resulting in some neutral midpoint on a single continuum. Thus, in the opinion of some scientists, different affective experiences should be seen as separate elements that cannot be reduced to a single dimension. Regardless of how this speciﬁc issue may be resolved, it is perhaps most appropriate to think of animal welfare as a conglomerate concept composed, at least in part, of other conglomerate concepts. When we try to make an ‘overall’ assessment of a conglomerate concept, we need to decide on the relative importance that we will attach to the different components. Thus the limnologists, if they were required to assess the ‘overall’ safety of the lake, would need to decide how much relative weight to attach to hidden rocks, strong currents and contaminated water. Different limnologists might agree on the factual matters – the strength of the currents, the number of days when the water is above a certain level of coliform contamination – but still disagree about the ‘overall’ safety of the lake. One limnologist, for example, might attach more importance to strong currents because they can kill swimmers suddenly and with little warning, and less importance to bacterial contamination because the risks can be reduced by careful monitoring. Another limnologist might attach less importance to strong currents because swimming deaths are relatively uncommon, whereas water contamination can affect many people at once. As the example illustrates, ‘overall’ assessment of a conglomerate concept involves value-based decisions in the sense of judging what is more important and less important.

24Cabanac,

1992; Kahneman, D. 1999. Objective happiness. Pages 3–25 in Well-Being: The Foundations of Hedonic Psychology (D. Kahneman, E. Diener and N. Schwarz, editors). Russell Sage Foundation, New York. 25Berridge, 1996 and 1999; Fraser and Duncan, 1998.

Dfraser_C011.indd 236

6/27/2008 4:58:43 PM

How Do the Different Measures Relate to Each Other?

237

But there is an important difference between the safety of a lake and the welfare of an animal. When we speak of the safety of a lake we are referring speciﬁcally to its safety for people. Hidden rocks are dangerous for boaters; undertows are dangerous for swimmers. Arriving at an overall assessment of safety is complicated by the fact that the various factors differ in importance to different people, and we cannot weight the various interests in a purely objective way because of the involvement of human interests. In the study of animal welfare, however, the point is not to determine what is best from the viewpoint of different people; the point is to determine what is best for the animals themselves. If we could ﬁnd a purely objective way to determine what is best for the animals themselves – or from the animals’ own ‘point of view’, as Marian Dawkins has put it26 – would that not provide the ‘correct’ assessment of their welfare, much as the actual biomass of a lake is the correct value which any estimation procedure tries to approximate? Thus our ﬁnal question is whether we can ﬁnd a purely objective way to determine the correct weighting of the many factors that enter into animal welfare by establishing what is best for animals from their own viewpoint and independent of human values. HOW MIGHT THIS BE DONE? Of the various methods we have reviewed, preference and motivation research provide the one approach that might conceivably be used to establish how animals themselves would rate the relative importance of the very diverse factors that may inﬂuence their welfare. If animals are given the opportunity to choose among different options, or to demonstrate the strength of their motivation to obtain or avoid certain options, would this not tell us how they themselves weight the different alternatives? Let us now brieﬂy recall some of the features and limitations of these methods (discussed in Chapter 10) and what they imply for establishing a view of animal welfare that is purely objective and value-free. We think of animals as having evolved the capacity to make adaptive choices between options if such choices commonly occurred during the evolution of the species. For example, when searching for food in unprotected areas, many birds choose between foraging and ﬂeeing in a way that appears to be sensitive to the risk of predation. Presumably these species evolved in an environment where such a choice was reasonably common, and birds that chose ‘appropriately’ left more descendants than those that did not. However, to achieve a fully objective view of the animal’s own priorities on all aspects of animal welfare, we would need to ask many questions that do not correspond to choices that occurred during the evolutionary history of the animals. For example, is it better to be indoors where there may be a build-up of pathogens that cause pneumonia, or outdoors where there is a risk of exposure to an exotic but deadly disease introduced by wildlife? In other cases, the cues that animals used 26Dawkins, M.S. 1990. From an animal’s point of view: Motivation, ﬁtness, and animal welfare. Behavioural and Brain Sciences 13: 1–9 and 54–61.

Dfraser_C011.indd 237

6/27/2008 4:58:43 PM

238

Drawing Conclusions about Animal Welfare

when choosing options in a natural state may no longer correlate with beneﬁcial or harmful outcomes. Thus an animal may prefer sweet foods because sweetness was correlated with high energy value in the situation in which the species evolved, even though sweetness is now achieved by non-nutritional sweeteners. We also need to compare options over long time periods such as the animal’s entire life, whereas the animal may have evolved the capacity to select only over shorter time scales. For example an animal might choose an environment based on a comfortable temperature and resting surface, even though that environment lacks the resources that the animal will later need for raising its young; or the animal may select a diet rich in energy and protein even though such a diet will ultimately prove harmful because it lacks certain nutrients or contributes to obesity. Within certain limits, as we have seen, preference research can help us identify what an animal considers a comfortable pen, a palatable food, or a congenial social group, and motivation research can provide insights into how keenly an animal wants to obtain particular beneﬁts at particular times. However, this type of research is not capable of answering many of the questions we would need to answer in order to establish, in a purely objective way, how animals themselves would balance the many different elements that may contribute to their welfare. Moreover, for many practical questions of animal welfare, we need not only to judge what balance of factors is best for a given animal, but also to balance the competing interests of different animals. Keeping lactating sows constrained in a heated pen may be good for the welfare of their piglets but not for the welfare of the sow who might prefer more freedom and cooler temperatures. Or we may need to decide whether to dehorn all the calves in a herd because of the likelihood that some of them might use their horns to injure other animals in the future. Obviously knowledge about an animal’s individual preferences do not allow us to balance, in a purely objective manner, the competing interests of different animals. The discussion thus far may leave the impression that we need to invoke human values in assessing animal welfare because we have only an imperfect understanding of animals, but in fact, the role of values in assessing quality of life runs far deeper. When people have tried to articulate what constitutes a good life for humans, they have proposed competing ideas that have much in common with the various conceptions of animal welfare.27 One line of thought, which we see in the Greek philosopher Epicurus, the English reformer Jeremy Bentham, and the modern ethicist Peter Singer, holds that a good life is a hedonically pleasant life wherein pleasures predominate and pains are at a minimum.28 Another line of thought, 27This point is nicely explored by Appleby, M.C. and Sandøe, P. 2002. Philosophical debate on the nature of well-being: Implications for animal welfare. Animal Welfare 11: 283–294. 28The Epicurean emphasis on wholesome pleasures and avoiding pain is described by Gaskin, J.C.A. 1995. Epicureanism. Pages 239–240 in The Oxford Companion to Philosophy (T. Honderich, editor). Oxford University Press, Oxford; the ideas of Bentham and Singer are cited in Chapters 1 and 4 respectively.

Dfraser_C011.indd 238

6/27/2008 4:58:44 PM

How Do the Different Measures Relate to Each Other?

239

extending from Aristotle to Amartya Sen, holds that each person has certain valuable capabilities, and that a good life involves being able to exercise those capabilities.29 A third line of thought, which we see in the Latin poet Virgil, the Romantic poets we encountered in Chapter 3, and in the children’s stories of Astrid Lindgren, sees a good life as one that is lived in harmony with nature and is not constrained or corrupted by the artiﬁciality that pervades much of human society.30 Much the same debate has arisen among social scientists who try to create scientiﬁc measures of human well-being. Some rely on basic measures related to functioning of the body such a longevity, disease incidence, and indicators of adequate nutrition. Others use a hedonic approach, and try to quantify people’s reported level of happiness and other subjective states. Others use a preferencebased approach and try to assess the degree to which people are able to live according to their own preferences.31 Here again we see parallels with the diverse methods used by animal welfare scientists to assess quality of life of animals. Thus, even with our intimate knowledge of what it is like to be a human being, different people have very different conceptions of what constitutes a good life; and corresponding to these different conceptions are different scientiﬁc approaches that have been used to assess human well-being. These are not disagreements about facts; they are value-based disagreements about what is good and important in life, and corresponding value-based disagreements about what are the most germane measures of well-being. TO SUMMARIZE, THE DIFFERENT criteria that people invoke when assessing animal welfare – which I have grouped as the basic health and functioning of animals, their affective states, and their ability to lead the kind of life for which they are adapted – although often correlated, do not always agree with each other, especially in cases where animals are kept in circumstances very unlike those in which the species evolved, or where the genetic makeup of animals has been modiﬁed in certain ways through artiﬁcial selection. Despite the various arguments that have been advanced in favour of one or other set of criteria being the true and deﬁnitive indicators of animal welfare, no one criterion clearly trumps the others. Instead, animal welfare (at least as the term is deﬁned in dictionaries and used in everyday speech) appears to be what Nordenfelt has called a ‘conglomerate concept’ made up of distinct and incommensurable elements which we group together 29Rollin,

B.E. 1993. Animal welfare, science, and value. Journal of Agricultural and Environmental Ethics 6(Suppl. 2): 44–50; Sen, A. 1993. Capability and well-being. Pages 30–53 in The Quality of Life (M. Nussbaum and A. Sen, editors). Oxford University Press, New York. 30Virgil’s Georgics (J. Lembke, translator, 2005, Yale University Press, New Haven) idealized rural life; Astrid Lindgren’s Pippi Longstocking (F. Lamborn, translator, 1950, Viking Press, New York) described how a young girl and her animal friends foiled the attempts of adults to force them to live more conventional lives. 31Brock, D. 1993. Quality of life measures in health care and medical ethics. Pages 95–132 in The Quality of Life (M. Nussbaum and A. Sen, editors). Clarendon Press, Oxford.

Dfraser_C011.indd 239

6/27/2008 4:58:44 PM

240

Drawing Conclusions about Animal Welfare

because they serve the same general function of contributing to the quality of life of animals. And although preference and motivation research allow us to ask many questions about what the animals themselves consider most desirable, these methods do not provide a purely objective way of answering all the questions we would need to answer to understand what is best from the animals’ own ‘point of view’. Thus, when we try to balance or combine different elements of animal welfare, for example to arrive at an ‘overall’ evaluation, we inevitably draw on value-based ideas about the relative importance of the different elements. In reality, of course, people do devise ways of combining different measures. These are seen, for example, in evaluations of alternative housing systems, in programmes of animal welfare standards, and in various multi-variable systems for scoring ‘overall’ animal welfare. How, then, are we to understand the interplay of factual and value-based elements when we use such multi-variable approaches? We will explore this question in the next chapter.

Dfraser_C011.indd 240

6/27/2008 4:58:44 PM

Selecting and

12

Combining Criteria of Animal Welfare

When people use animal welfare science as a basis for standards, policies and auditing systems, they frequently draw on some combination of criteria that may include basic health and functioning, affective states and natural living. In this chapter let us examine several examples that combine different criteria of animal welfare, and in so doing analyse the interplay of factual and value-based elements that arise.1 The ﬁrst example involves comparing animal welfare in housing systems that differ in a large number of features. Speciﬁcally, compared to sows kept indoors in individual stalls, sows kept in group housing systems have more freedom of movement but they may also be exposed to aggressive group-mates, competition for food, and (if kept outdoors) to variable weather. In cases such as these, can we conclude that one type of housing is better than another in providing for the animals’ welfare? The second example involves animal welfare assurance programmes – programmes used by governments, industries, animal protection organizations and other agencies to assure the public that animal welfare is being safeguarded. Scientiﬁc evidence is invoked as the basis of many of these programmes, yet the programmes actually create very different requirements. Some, for example, require that hens have larger cages, while others require that cages not be used at all. If the programmes really are science-based, how could they possibly differ so much? Is it simply that they provide different ‘levels’ of animal welfare, or is there more to the story? A third example involves scoring systems for animal welfare that are based on a variety of variables. Scoring systems play a role in product certiﬁcation and marketing, and they are used in some countries to assess compliance with animal

1Portions of this chapter are reworked from two papers: Fraser, D. 2003. Assessing animal welfare at the farm and group level: The interplay of science and values. Animal Welfare 12: 433–443; and Fraser, D. 2006b. Animal welfare assurance programs in food production: A framework for assessing the options. Animal Welfare 15: 93–104. I am grateful to the publishers for allowing me to use sections of the text.

241

Dfraser_C012.indd 241

6/27/2008 12:44:48 PM

242

Drawing Conclusions about Animal Welfare

Figure 12.1 A bank of ‘gestation stalls’ – a widely used method of housing sows during most (roughly three months) of pregnancy. The stalls allow efficient individual feeding of animals, and they eliminate fighting and competition over food, but the animals cannot walk or turn around during the entire period that they spend in the stalls. Photo reproduced by kind permission of Global Action Network, Montreal.

protection laws; but can comprehensive, multi-variable scoring systems be made fully objective, or do these scoring systems inevitably involve some combination of empirical measurement and values? TO KEEP THINGS INTERESTING, let’s start with a ﬁght. In 1997 a scientiﬁc committee created by the European Union published a review of the scientiﬁc literature on the welfare of intensively kept pigs. The committee asked (among other questions) whether welfare problems are caused by gestation stalls – the stalls where sows are often kept, unable to walk or turn around, during most of pregnancy (Figure 12.1). The review concluded that, ‘Some serious welfare problems for sows persist even in the best stall-housing system’.2 With this review in hand, the European Union decided in 2001 to greatly restrict the use of gestation stalls from 2013 onward. 2Scientiﬁc

Veterinary Committee, 1997. The Welfare of Intensively Kept Pigs. Report of the Scientiﬁc Veterinary Committee, European Union, Brussels, Belgium. The quote is from section 5.2.11. I have identiﬁed passages according to section numbers because an online version of the report does not have obvious page numbers.

Dfraser_C012.indd 242

6/27/2008 12:44:48 PM

Selecting and Combining Criteria of Animal Welfare

243

Shortly thereafter, a group of Australian scientists made a second review of the scientiﬁc literature on the welfare of sows, and asked much the same question, but they came up with essentially the opposite conclusion. They concluded that, ‘Both individual [including stalls] and group housing can meet the welfare requirements of pigs’.3 Various swine industry spokespersons used that review to argue that there is no scientiﬁc basis for eliminating the gestation stall.4 The two reviews were done with great thoroughness by very accomplished scientists, and each group likely felt that they were doing the most thorough and objective job possible. What, then, should we conclude when different groups of scientists, conscientiously reviewing the scientiﬁc literature on the same subject, reach such different conclusions? A careful reading of the articles shows that the two groups differed, ﬁrst and foremost, in what they regarded as important for animal welfare. The Australian reviewers adopted what they called ‘the homeostasis approach’ for assessing animal welfare. This involved relying on measures of basic health and functioning or, as the report put it, ‘relative changes in biological … responses and corresponding decreases in ﬁtness’ including ‘widely accepted criteria of poor welfare such as health, immunology, injuries, growth rate, and nitrogen balance’. The reviewers did not deny that affective states are involved in animal welfare. In fact, they noted that ‘animal emotions’ play a role in animal welfare ‘as they would have evolved on the basis of their survival values and contribution to biological ﬁtness’. However, they took the view that all signiﬁcant risks to welfare would have ‘consequent effects on ﬁtness variables such as growth, reproduction, injury, and health’.5 Thus, the Australian reviewers did not regard negative affective states as indicating a welfare problem unless these were accompanied by changes to health or functioning, and this decision contributed to their conclusion that gestation stalls ‘can meet the welfare requirements’ of the animals. In contrast, the European reviewers used a conception of welfare that involved basic health and functioning, but they also saw affective states as directly relevant to animal welfare. ‘Suffering’ they wrote, ‘is one of the most important aspects of poor welfare and we should investigate the existence of good or bad feelings wherever possible when trying to assess welfare’. Thus they included in their assessment ‘the effects of fear and the behavioural and physiological consequences of lack of control, especially frustration’, without assuming that unpleasant affective states should be taken into account only if accompanied by differences in 3Barnett, J.L., Hemsworth, P.H., Cronin, G.M., Jongman, E.C. and Hutson, G.D. 2001. A review of the welfare issues for sows and piglets in relation to housing. Australian Journal of Agricultural Research 52: 1–28. The quote is from page 13. 4Further developments include a review by McGlone et al., 2004, and one by a team assembled by the American Veterinary Medical Association (Rhodes et al., 2005). For simplicity I will deal with just the original two reviews in order to bring out some relevant points without a welter of detail. 5Barnett et al., 2001. The quotations are on page 3.

Dfraser_C012.indd 243

6/27/2008 12:44:50 PM

244

Drawing Conclusions about Animal Welfare

functioning-based variables such as reproduction and health. For example, their evidence of welfare problems included signs that sows are ‘frustrated’ in stalls and that sows ﬁnd stalls ‘aversive’.6 A similar divergence in approach can be seen in how the two groups handled freedom to carry out natural behaviour. The European group regarded the opportunity to carry out natural behaviour as directly relevant to welfare. They stated that ‘sow welfare will be worse in conditions where exploration of a complex environment, rooting in a soft substratum and manipulation of materials such as straw are not possible’7; and they saw high levels of abnormal behaviour as indicative of poor welfare. In part it was evidence of abnormal behaviour, and the inability of sows in stalls to carry out natural behaviour, that led the European reviewers to conclude that serious welfare problems occur in even the best stall systems. The Australian reviewers, in relying on functioning-based criteria, evidently viewed the freedom to perform natural behaviour as being only instrumentally important for animal welfare, not inherently important; that is, the ability to walk, root and engage in normal social behaviour would be important for welfare if (but only if) it led to beneﬁts in terms of health, reproduction, or other ‘homeostatic’ variables. Given that stalls do not generally result in poorer performance on such measures, it followed that sow welfare is not signiﬁcantly impaired by an inability to walk and perform other natural behaviour. In addition to these basic differences in what they considered important for animal welfare, the Australian and European reviewers also differed on an even more fundamental issue. The culture of science typically attaches great value to ‘objectivity’. Variables such as growth rate, survival and incidence of infectious diseases can be measured in very ‘objective’ ways in the sense that there is relatively little scope for scientists to use personal ‘subjective’ interpretation when they make these measurements. Operationally speaking, we expect to see a high level of agreement between observers when they quantify these variables. Moreover, these are long-established, traditional measurements within the animal sciences, and their use as indicators of poor welfare is relatively uncontroversial. Hence, what we might call a ‘scientiﬁcally conservative’ approach which favours traditional, well-established measures that can be scored in very objective ways, would rely on measures such as these. The assessment of affective states such as pain, frustration and distress is somewhat different. Pain, for example, is often scored by subjective scaling methods which are open to disagreement among observers,8 or by variables such as plasma 6Scientiﬁc

Veterinary Committee, 1997. The quotations are from sections 1.2, 5.2.2 and 5.2.1. Veterinary Committee, 1997, section 5.2.1. 8Beynen, A.C., Baumans, V., Bertens, A.P.M.G., Havenaar, R., Hesp, A.P.M. and van Zutphen, L.F.M. 1987. Assessment of discomfort in gallstone-bearing mice: A practical example of the problems encountered in an attempt to recognize discomfort in laboratory animals. Laboratory Animals 21: 35–42. 7Scientiﬁc

Dfraser_C012.indd 244

6/27/2008 12:44:50 PM

Selecting and Combining Criteria of Animal Welfare

245

cortisol levels whose relationship to pain is open to interpretation. Moreover (as we have seen) the mid-twentieth century saw considerable debate over whether the affective states of animals are amenable to scientiﬁc investigation at all. Hence, within the culture of science, measures of affective states such as pain and distress are not as traditional, well established and uncontroversial as are measures of growth rate, survival and other aspects of basic health and functioning. In our case study, then, the Australian reviewers took a scientiﬁcally more conservative approach than the Europeans. They used measures that are well established, that can be measured with little subjectivity, and that are uncontroversially related to animal welfare. Indeed, the Australian reviewers remarked that their reliance on such criteria ‘affords this approach credibility within scientiﬁc circles’.9 In so doing, however, the Australian reviewers sacriﬁced kinds of evidence (signs of frustration, inability to carry out basic bodily movements) which many others see as relevant to animal welfare. In contrast, the European reviewers, in embracing measures reﬂecting affective states and natural behaviour, appeared to use a conception of animal welfare that comes closer to the everyday meaning of the term, but they may have incurred the scepticism of those scientists who view such measures as less objective or not scientiﬁcally respectable. In summary, both the European and Australian review teams reached plausible conclusions within the context of the value assumptions that underlay their approaches. The Australian reviewers made a good case that gestation stalls can meet the welfare requirements of sows if we assume that the important indicators of animal welfare are basic health and reproductive success, that any welfare problems involving affective states and restrictions on natural behaviour will be reﬂected in measures of health and functioning, and that it is appropriate to restrict the assessment of animal welfare to measures that are uncontroversial within the culture of science. The European reviewers made a good case that the welfare of sows is jeopardized by gestation stalls, given that affective states, abnormal behaviour and an ability to behave in a relatively natural way are inherently relevant to animal welfare, and that it is better to include such considerations than to limit the assessment of animal welfare to measures that are scientiﬁcally uncontroversial. There was another important difference between the two studies. The Australian review concluded with an endorsement: that gestation stalls (and other systems) ‘can meet the welfare requirements’ of the animals. With the conclusion couched in these positive terms, readers could conclude that there is no need to move away from gestation stalls, even if other systems might have welfare advantages. The European review concluded with a criticism: that there are ‘serious welfare problems’ with stalls. Given this conclusion, it would seem rational to call for an end to stalls even if existing alternatives have problems of their own. As a personal note, it has been interesting to watch this debate play out in the worlds of science and policy, and to hear how the participants have invoked the 9Barnett

Dfraser_C012.indd 245

et al., 2001, page 3.

6/27/2008 12:44:50 PM

246

Drawing Conclusions about Animal Welfare

concepts of objectivity and values to explain the disagreements. Supporters of the European review have tended to look upon it as reﬂecting ‘true’ animal welfare because of its more comprehensive nature, and to view the Australian review as having been biased by a concern to protect the ﬁnancial interests of an industry that wants to continue using the gestation stall. Supporters of the Australian review have tended to see it as reﬂecting ‘true’ animal welfare because it uses ‘objective’ measures, and to see the European review as having been biased by public misconceptions about animal welfare. Thus some participants on both sides have viewed their preferred interpretation as objective and correct, and opposing views as having been tainted by values. Given this confusion, let us brieﬂy summarize the roles that values actually played in this debate. The ﬁrst role was in deciding what to regard as important for animal welfare: if sows found their environment ‘aversive’ and ‘frustrating’, was that important enough to count as an animal welfare problem, or should we consider that a welfare problem exists only if measures of basic health and functioning are affected? The second role had to do with judging scientiﬁc evidence: is it better to use only traditional measures such as growth and reproduction that can be measured easily and have ‘credibility in scientiﬁc circles’, or to include signs of frustration which may be controversial among some scientists and whose scoring requires some subjective judgement on the part of the observer? The third role of values came in bridging from the scientiﬁc evidence to morally signiﬁcant conclusions. If each system has advantages and disadvantages, to conclude that one involves ‘problems’ is, in effect, a call for corrective action, whereas to conclude that all systems ‘can meet welfare requirements’ implies that no such action is needed. In summary, the debate nicely illustrated how value-based considerations played an important and complex role in what appeared to be straightforward reviews of scientiﬁc evidence. Before leaving this topic, let us consider a further reservation about comparing the relative merits of different housing systems. If (as we have seen in several examples) the quality of human care is an important factor for animal welfare, and if there are important genetic differences that are relevant to animal welfare, then there are limits to our ability to conclude that one housing system is, on net, better than another. For example, in cases where farm staff have the time and skill to assemble sows into compatible groups, where the diet is high enough in ﬁbre that the animals can eat a substantial volume each day and have little need to compete for food, and where the environment is sufﬁciently spacious and well designed so that the sows can rest comfortably, then the animals may well thrive if housed and fed in groups. However, if staff time and skill are minimal, if groups are formed without regard for their compatibility, if the animals are crowded and uncomfortable, and if they are left to compete for limited quantities of a concentrated diet, then it seems very plausible that at least some of the animals would fare badly in group housing. Moreover, the situation may also vary depending on whether the sows are of a placid nature or are excitable and prone to aggression. If we think

Dfraser_C012.indd 246

6/27/2008 12:44:50 PM

Selecting and Combining Criteria of Animal Welfare

247

of the welfare of animals as resulting from a complex combination of the housing system, the level of management, the nutritional strategy, and the type of animals, then any general conclusion about the superiority of one type of housing over another would need to be heavily qualiﬁed. IN THE ABOVE EXAMPLE, we see that scientiﬁc evidence led to different conclusions about animal welfare at least partly because of different underlying assumptions, especially different assumptions about what is important for animal welfare. A similar analysis helps us make sense of the differences that have arisen between various programmes designed to ensure the welfare of food animals. Toward the end of the twentieth century (as we have seen) many organizations and governments developed programmes to assure the public that they were taking appropriate actions for the welfare of food animals. These took the form of voluntary codes of practice, regulations, international agreements, corporate standards, and product-differentiation programmes. All of these programmes have been described as protecting animal welfare, and most are described as ‘science-based’. However, the programmes actually set out very different requirements for how animals should be treated. To give just one example, Table 12.1 summarizes some of the requirements of three programmes for the welfare of laying hens. One programme is the set of voluntary guidelines ﬁrst adopted in the late 1990s by the United Egg Producers in the United States, mentioned in Chapter 5. It requires that hens in cages have 432–555 cm2 of ﬂoor space per hen; it also requires sufﬁcient space at the feed trough ‘to allow all birds to eat at the same time’, and one water cup or nipple for every 12 birds. The guidelines also make recommendations about ‘forced moulting’ by food withdrawal. This is the practice of forcing hens to stop laying by withholding all food, usually for several days, to the point that the birds lose their feathers and roughly 30% of their body weight. Restoring the food supply after such a moult causes a renewed period of laying. The practice is highly controversial partly because it is thought to cause an extreme level of hunger.10 The guidelines recommend that food withdrawal no longer be used for forced moulting, and that when moulting is practiced by other means, the number of birds that die during the moult ‘should not substantially exceed normal variations in ﬂock mortality’.11 The second programme listed in Table 12.1 appeared as a European Union directive created in 1999. The directive required that conventional cages for hens be phased out over a period of years, and that hens be kept instead either in non-cage systems or 10Bell,

D., Chase, B., Douglass, A., Hester, P., Mench, J., Newberry, R., Shea-Moore, M., Stanker, L., Swanson, J. and Armstrong, J. 2004. UEP uses scientiﬁc approach in its establishment of welfare guidelines. Feedstuffs 76: 1–9. 11United Egg Producers. 2008. United Egg Producers Animal Husbandry Guidelines for U.S. Egg Laying Flocks, 2008 Edition. United Egg Producers, Atlanta. Available at: http://www.uepcertiﬁed. com/docs/UEP-Animal-Welfare-Guidelines-2007–2008.pdf, accessed January 2008. The quotation is on page 6.

Dfraser_C012.indd 247

6/27/2008 12:44:50 PM

248

Drawing Conclusions about Animal Welfare

Table 12.1 Some of the main quantitative requirements for housing systems used for laying hens as specified in the Animal Husbandry Guidelines of the United Egg Producers in the United States, the European Union requirements for furnished cages, and the RSPCA Freedom Food Program in the United Kingdom.

Requirement Space (cm2 per bird) Feed trough space (cm per bird) Water cups or nipples Perch (cm per bird) Nest boxes Littered area (cm2 per bird) Daily light (h per day) Dark (h per day)

United Egg Producers

European Union furnished cage

Freedom Food

432555 —b

600750a 12

1100 10c

1/12 birds — — —

2 within reach 15 Yes Yes

1/10 birds 15 1/5 birds 250

— —

— —

8 6, gradual onsetd

Notes: a 600 of the 750 cm2 must be ‘usable’ meaning that the area is at least 30 cm wide, has a floor slope not exceeding 14%, and head room of at least 45 cm. b Described as ‘sufficient to allow all birds to eat at the same time’. c Or equivalent if circular feeders are used. d Artificial light to be turned off ‘in a stepped or gradual manner to allow the hens to prepare for darkness’. Sources: United Egg Producers. 2008. United Egg Producers Animal Husbandry Guidelines for U.S. Egg Laying Flocks, 2008 Edition. United Egg Producers, Atlanta. Available at: http://www.uepcertified.com/docs/UEP-Animal-WelfareGuidelines-2007-2008.pdf, accessed January 2008. Council of the European Union. 1999. Council Directive 1999/74/EC of 19 July 1999 laying down minimum standards for the protection of hens. Official Journal of the European Communities L203: 53–57. RSPCA. 2006. Welfare standards for laying hens and pullets, February 2006. Available at: http://www.rspca.org.uk/ servlet/Satellite?blobcol=urlblob&blobheader=application%2Fpdf&blobkey=id&blobtable=RSPCABlob&blobwhere= 998045492811&ssbinary=true, accessed February 2007.

in ‘furnished’ cages where each hen would have 600–750 cm2 of ﬂoor space, 12 cm of space at the feed trough, and access to a perch where they could roost, to a nest box where they could lay, and to an area with litter to allow dust-bathing. A third programme was created by Freedom Food, the company originated by the Royal Society for the Prevention of Cruelty to Animals in the United Kingdom. Its standard for hens prohibits all use of cages. It requires that hens have about 1100 cm2 of space, 10 cm of space at a feed trough and a water cup or nipple for every 10 birds. It also requires that birds have perches, nest boxes, an area with litter for dust-bathing as described in Table 12.1, plus eight hours of continuous light and six hours of continuous darkness each day. Producers certiﬁed as meeting the requirements are allowed to identify their eggs with the Freedom Food logo, and the eggs tend to sell for a modest premium.

Dfraser_C012.indd 248

6/27/2008 12:44:50 PM

Selecting and Combining Criteria of Animal Welfare

249

Here, then, we see three very different programmes, each of which is claimed to protect the welfare of laying hens, and each of which is claimed to be based on science. But how could this possibly be? How could science support both 432 cm2 per bird and 1100? How could science justify both the use and the abolition of cages? Is it simply that the proponents of some standards were mistaken in their reading of the science, or (worse) that they manipulated the science to justify production methods that suited their industry or their ideology? I believe it helps if we see the various requirements created by animal welfare assurance programmes as serving several distinct goals corresponding approximately to the three broad animal welfare concerns discussed previously and shown in Figure 11.2 on page 230. Certain requirements (which I will call ‘Type 1 requirements’ for ease of communication) are designed primarily to achieve basic health and functioning of the body as reﬂected by a low incidence of disease and high rates of survival, reproduction and growth. For example, Type 1 requirements require that animals be raised with enough space to prevent crowding-related reductions in survival and productivity, and that ammonia concentration in the air not reach levels that impair respiratory health and growth rate. Such requirements ﬁgured prominently in some of the earliest animal welfare assurance programmes such as the early codes of practice, early European Union directives, and the initial programmes developed by chain restaurant and retail companies in the United States. A second group of requirements (Type 2) are focussed on the affective states of animals, and are intended especially to eliminate or reduce unpleasant states such as pain, distress and hunger. Examples include requirements that local anaesthetic be used to mitigate pain during procedures such as dehorning or castration, that electric prods (instruments that deliver electric shocks to animals to cause them to move) not be applied to sensitive parts of the body where they are likely to cause the most pain, and that a high percentage of animals at slaughter plants be stunned successfully on the ﬁrst attempt. Type 2 requirements have a long history in welfare standards at slaughter plants where reduction of pain and distress during the killing process is widely seen as a major animal welfare concern. They are also used to some degree in on-farm standards, although in many jurisdictions pain management is not used for some presumably painful procedures such as castration of young animals. Both the third and fourth groups of requirements are related to the goal of allowing animals to lead reasonably natural lives, but I have separated them into two types because of some important differences. One group of requirements (Type 3) attempt to provide animals with the opportunity to carry out certain elements of their natural behaviour. Examples include requirements for hens to be able to perch and enter a nest box for laying, for sows on restricted diets to have access to roughage for foraging, and for calves and sows to be able to walk and turn normally and hence not to be kept in restrictive crates or stalls. Type 3 requirements are basic to many alternative production systems such as organic and free-range, and also occur in certain national regulations and European Union directives.

Dfraser_C012.indd 249

6/27/2008 12:44:51 PM

250

Drawing Conclusions about Animal Welfare

The ﬁnal group of requirements (Type 4) stipulate that animals should have some level of access to natural components in their environment such as natural light, fresh air and the outdoors. Examples include requirements for hens to have daily access to the outdoors in free-range systems, for barns to have windows that admit natural daylight, and for dairy cows to be kept on pasture in the summer months. Type 4 requirements are widely used in alternative production systems such as organic and free-range, and occasionally occur in national regulations as in Sweden. Although the four broad goals lead to very different requirements, all four types of requirements can have at least a certain level of scientiﬁc justiﬁcation within the goals that they serve. Drawing on evidence reviewed in earlier chapters, here is a very brief summary. Type 1 requirements are generally based on studies demonstrating that basic aspects of animal health and functioning are impaired if the requirements are not met. For example, as we saw in Chapter 5, the requirement for laying hens to have roughly 450 cm2 of ﬂoor space is based on extensive research showing increased mortality rate and decreased rate of lay if less space is provided. Similarly, the common requirement that ammonia concentration in the air not exceed 25 ppm for broiler chickens is based on research showing reduced growth and survival at higher levels. Type 2 requirements are generally based on studies that use behavioural and physiological indicators of pain, distress, hunger and other negative states. For example (as noted previously) research has shown that dehorning of calves leads to many behavioural and physiological signs believed to denote pain, and that these can be reduced or eliminated by certain pain-management techniques. Research also suggests that forced moulting by withdrawing food from laying hens causes extreme hunger and frustration in the birds. Such research is expanding rapidly, but in cases where the science is poorly developed, some requirements are based on seemingly common-sense assumptions about what is likely to cause pain and distress in animals, for example that incorrect stunning will cause pain if it fails to render the animal unconscious. Type 3 requirements are generally based on studies showing that animals are motivated to carry out certain types of natural behaviour, and in some cases that behavioural or physiological indicators of distress are present if such behaviour is prevented. For example (as we have seen) hens show a very strong motivation to enter a nest box in the hour before they lay an egg, and they show signs of frustration if they are prevented from doing so; and sows on restricted diets often develop repetitive, stereotyped behaviour that can be reduced by providing access to roughage (such as straw) that permits more normal foraging. Although the science is developing rapidly, some issues remain poorly researched. Hence, some requirements have been based on seemingly common-sense assumptions, for example that calves kept in narrow crates will be motivated to move more freely than the stall allows.

Dfraser_C012.indd 250

6/27/2008 12:44:51 PM

Selecting and Combining Criteria of Animal Welfare

251

For Type 4 requirements, which focus on natural components in the environment, the scientiﬁc rationale is somewhat different and (compared to the types of evidence just described) tends to be less direct. The main support comes from the many studies showing health and welfare problems (respiratory illness, lameness, stereotyped behaviour, etc.) among animals conﬁned to indoor and other artiﬁcial environments, and the fact that keeping animals in more ‘natural’ and less restrictive environments is plausibly seen as a way to avoid such problems. However, there has been relatively little critical evaluation of the actual effects of Type 4 requirements. For example, keeping animals outdoors should prevent respiratory disease caused by dust and manure gases in the air of conﬁnement buildings, but does it make them more vulnerable to other diseases transmitted from wild animals or the soil? There has also been little research to determine whether animals are actually motivated to be outdoors and under what conditions, and to what extent outdoor environments really accommodate the animals’ natural behaviour. Moreover, less research has generally been done to determine appropriate quantitative values for Type 4 requirements: in contrast to the dozens of studies comparing different amounts of space for hens in cages, there has been much less scientiﬁc investigation of the space requirements of birds on free range. In short, despite the very real animal welfare problems that can arise in artiﬁcial environments, the scientiﬁc justiﬁcation for speciﬁc Type 4 requirements often appears less well developed than for requirements of Types 1–3. To summarize, many animal welfare requirements, different as they are, do have a scientiﬁc basis, especially for Types 1, 2 and 3. There is a scientiﬁc basis for requiring 450 cm2 in cages for laying hens because less space jeopardizes the birds’ basic health. There is also a scientiﬁc basis for requiring that such cages be replaced by environments that allow perching and nesting because hens are strongly motivated to carry out these types of behaviour. These requirements differ not because one is based on science and the other is not, but because they are designed to achieve different objectives. Speciﬁcally, in this case the more comprehensive requirements are designed to meet more comprehensive welfare objectives. With this in mind, let us return to the three examples outlined in Table 12.1. We see that the programme of the United Egg Producers is based mainly on Type 1 requirements (enough space to avoid crowding-related reductions in survival, health and production) plus one Type 2 requirement: to eliminate forced moulting by feed withdrawal in order to prevent extreme hunger. The European Union standard for furnished cages contains, in addition to these, signiﬁcant Type 3 requirements, speciﬁcally to accommodate perching, dust-bathing and nesting as key types of natural behaviour. And the Freedom Food standard includes (in addition) more extensive Type 3 and 4 requirements that require more amenities for natural behaviour including freedom of movement, and reasonably natural daily periods of light and dark. Which type of programme is the best for animal welfare, all things considered? If we accept the arguments above, then well researched requirements of

Dfraser_C012.indd 251

6/27/2008 12:44:51 PM

252

Drawing Conclusions about Animal Welfare

Types 1, 2 and 3 should all contribute to animal welfare. Hence, we would expect that a more comprehensive programme – one that protects basic health and functioning (Type 1), prevents pain and distress (Type 2), and allows animals to behave in accordance with their motivation (Type 3) – should be better for the animals than a programme that accomplishes only one or two of these goals, as long as there are no important trade-offs whereby pursuing one of these goals jeopardizes another. Programmes that involve only Type 1 requirements are likely to accomplish only limited animal welfare objectives, and are likely to be perceived as only minimal standards. I feel we are generally on less ﬁrm ground with certain Type 4 requirements, partly because the implications for animal welfare are less clear, partly because the potential for unresearched trade-offs is particularly large, and partly because the science needed to set a clear standard is often lacking. A challenge created by the wide variety of animal welfare assurance programmes is to prevent widespread confusion and misunderstanding. When producers see space requirements increased from 450 to 750 cm2 per hen and are told that both ﬁgures are based on science, they could be forgiven for concluding that the decisions are more political than scientiﬁc. When consumers ﬁnd that there are several very different standards all claiming to protect animal welfare, they could well assume that the less stringent standards are little more than window-dressing. In order to mitigate this kind of confusion, we need to be clear that animal welfare assurance programmes serve a number of different goals. If one programme is designed to provide enough space and necessities so that survival and health are not jeopardized, while another programme is designed to provide unrestricted movement and exposure to the outdoors, they will inevitably create confusion if they both claim to be protecting ‘animal welfare’. But if the speciﬁc goals are made clear, it may be possible for both programmes to be seen for what they are. IN THE EXAMPLES DESCRIBED ABOVE, we have seen that scientiﬁc information plays a key role but that its application is underlain by certain value-related assumptions and decisions, especially about what is important for animal welfare. Let us now look at how values and scientiﬁc information interact in two multi-variable scoring systems designed to assess animal welfare on different farms or slaughter plants. The ﬁrst example is an animal welfare auditing system designed for slaughter plants, initially in the United States. When cattle, pigs or sheep arrive at slaughter plants, they are typically unloaded from the vehicle into holding pens where they may remain for several hours. They are then required to walk along a passageway into a restraining chute where they are stunned. For cattle, stunning commonly involves a ‘captive-bolt’ pistol or similar device which is placed against the forehead of the animal and ﬁres a steel bolt that strikes the skull and penetrates into the brain, resulting in concussion and brain damage that should render the animal instantly unconscious. For pigs, a more common technology involves two electric ‘paddles’ that are typically placed on either side of the head. These deliver a brief, powerful electric current across the head which disrupts the electrical activity of the brain and is

Dfraser_C012.indd 252

6/27/2008 12:44:51 PM

Selecting and Combining Criteria of Animal Welfare

253

believed to render the animal unconscious. (An electro-encephalogram taken after electrical stunning resembles that of a human patient during a ‘grand-mal’ epileptic seizure, which is known to involve a loss of consciousness.) Once unconscious, the animal is typically shackled by a chain around one hind leg and hoisted so that it hangs upside-down on the bleeding rail, whereupon a cut is made to the neck (for cattle) or the heart area (for pigs). It is intended that the animal will die from rapid loss of blood without regaining consciousness. From her extensive experience at slaughter plants, animal behaviourist Temple Grandin was aware of several problems that can occur during the moving and killing of animals. In some cases animals slip or fall as they are walking to the stunning area, and they may sustain painful injuries as a result. Staff of the slaughter plant often use electric prods when moving animals. These are known (and intended) to be aversive to the animals. In a well-designed facility, prods should be needed only in exceptional cases, but with poorly designed passageways or poorly trained staff, prods may be used on many animals and result in unnecessary distress. If the stunning equipment is not correctly positioned or fails to function normally, the animal may receive a painful blow or electric shock without losing consciousness. Perhaps worst of all is the possibility that the animal has some level of consciousness when hung on the bleed rail or even when bleeding begins. Grandin also observed that vocalizations of cattle and pigs provide a useful indicator of animal welfare problems. In one study she monitored the vocalizations of cattle at six commercial slaughter plants. Of 1125 animals studied, 112 vocalized during handling and stunning, and all except two of the vocalizations occurred ‘immediately after a stressful event’, notably after the animals were prodded with an electric prod, after slipping on the ﬂoor, after experiencing excessive pressure on the body from restraining devices, or after an unsuccessful attempt at stunning. Grandin concluded that the proportion of animals vocalizing while in the passageway or restraining chute provides a useful indirect indicator of various presumably aversive events.12 Based on these observations, Grandin developed a multi-variable system for auditing the welfare of animals in slaughter plants. In plants killing beef cattle, for example, her system involves scoring the percentage of animals (typically from a sample of 100 or more) that are stunned correctly on the ﬁrst attempt, the percentage (if any) that show signs of sensibility when hung on the bleeding rail, the percentage that are prodded with an electric prod while being moved, and the percentage that slip, fall, or vocalize during handling or stunning (Table 12.2). The variables were selected because they denote obvious welfare problems (or, in the case of vocalizations, are closely related to welfare problems), and because they can be scored objectively for a large number of animals in a single visit. 12Grandin, T. 1998a. The feasibility of using vocalization scoring as an indicator of poor welfare during cattle slaughter. Applied Animal Behaviour Science 56: 121–128. The quotation is on page 121.

Dfraser_C012.indd 253

6/27/2008 12:44:51 PM

254

Drawing Conclusions about Animal Welfare

Table 12.2 Numerical scoring elements used to assess the welfare of cattle at slaughter plants, based on the guidelines of the American Meat Institute. Measure Stunned correctly on first attempt Insensible when on the bleeding rail Not prodded with an electric prod Move without slipping Move without falling Not vocalizing

Minimum acceptable (%)

Excellent (%)

95 99.8 75 97 99 97

99 100 95 100 100 99

Source: Grandin, T. 2001. Cattle Slaughter Audit Form (Updated October 2001) Based (with some reorganization) on American Meat Institute Guidelines. Available at: http://www.grandin.com/cattle.audit.form.html, consulted July 2007.

To set target values for most of the variables, Grandin tried to identify achievable values that would represent how the better slaughter plants could be expected to perform. For example, in a study of 11 plants killing beef cattle she found that the percentage of cattle prodded with an electric prod varied from 5% at one plant to 90% at another, and that four plants were achieving 95% success in stunning cattle on the ﬁrst attempt. These data then served as the basis for setting targets for the audit. However, for what Grandin considered the worst animal welfare problems, the target values were very stringent. In particular, plants were allowed only two animals per thousand that showed any sign of sensibility while hanging from the bleeding rail, and the plant would fail automatically if any animal showed signs of sensibility when bleeding began or if the staff moved any conscious animal by dragging it.13 Grandin’s audit system was adopted by the American Meat Institute – an association of slaughter plants in the United States – as voluntary guidelines for their members, and beginning around the year 2000 has also been required by several chain restaurant companies as a condition for purchasing products. Grandin’s audit clearly relied on scientiﬁc research for two purposes: to identify achievable target values, and to establish the relation between vocalizations and welfare problems. However, pragmatic considerations were also involved; in particular, she chose variables that could be measured easily by an auditor, and avoided measures (heart rate, activation of stress response systems) that would require specialized facilities. Her views of the relative importance of the variables were reﬂected in part by the target values; for example, plants were allowed three animals per hundred that slipped but none that showed signs of sensibility when bled. Beyond these differential targets, however, Grandin made no attempt to weight the variables according to their relative importance. She did not try, for example, to give more points for one variable than for another, but required rather that all the targets be met. 13Grandin,

T. 1998b. Objective scoring of animal handling and stunning practices at slaughter plants. Journal of the American Veterinary Medical Association 212: 36–39.

Dfraser_C012.indd 254

6/27/2008 12:44:51 PM

Selecting and Combining Criteria of Animal Welfare

255

A quite different approach is used in the ‘Animal Needs Index’ or ‘ANI’ (in German, ‘Tiergerechtheitsindex’) developed by Austrian animal scientist Helmut Bartussek. The ANI is a method for scoring animal housing for its effect on animal welfare.14 The ANI has proven very inﬂuential in Austria. In particular it is referenced in Austrian animal protection legislation and it is used to assess animal welfare in the Austrian system of organic production standards. The ANI is based on ﬁve ‘components’ listed in Table 12.3: the animal’s ability to move; social contact with other animals; the quality of the ﬂooring for lying, standing and walking; conditions within the building such as light and air quality; and the quality of human care. For each of these ﬁve components, several ‘criteria’ are scored. For example, when the component ‘quality of human care’ is scored for cattle facilities, the criteria (Table 12.3) include the cleanliness of the housing, the state of the equipment and other features. Each of the criteria is assigned points up to a maximum that varies from one criterion to another. Under component 4, for example, a cattle facility can receive a score ranging from 0.5 to 2.0 for light, from 0.5 to 1.5 for air quality and so on. The points are then added to give the numerical value of the Index. Total scores of 28 and higher (more than 75% of the maximum number of points) indicate that an environment is ‘very suitable with respect to welfare’, whereas scores less than 11 indicate that the environment is ‘not suitable’. Bartussek was admirably clear in describing the mixture of science, values and practical considerations that went into the design of the ANI. He noted that the assessment of animal welfare often involves ‘physiological, behavioural, pathological and production-related parameters’, but that ‘we do not know how the parameters equate with each other’. Therefore, he argued, the design of on-farm welfare assessment tools ‘must primarily be the result of negotiation’ but the underlying principles should be ‘justiﬁed by as much scientiﬁc background as possible’. In deciding what elements to score and their relative importance, Bartussek relied on several considerations. First, extensive review of the scientiﬁc literature identiﬁed various features that are closely associated with animal welfare. These became the ‘criteria’, and Bartussek then assigned a maximum score to each one based on how important he judged it to be. A second consideration was the ease with which a parameter could be scored; this was important because thousands of farms needed to be assessed efﬁciently under the programme. Third, Bartussek described ‘historic and pragmatic’ reasons for certain features of the system. Speciﬁcally, because the earliest version of the ANI had been adopted into the animal protection law of a region of Austria, later versions needed to maintain a degree of consistency with the numbers generated by the early version. A ﬁnal consideration was more philosophical: Bartussek noted that producers have an economic incentive to safeguard certain basic aspects of animal 14Bartussek,

H. 1999. A review of the animal needs index (ANI) for the assessment of animals’ well-being in the housing systems for Austrian proprietary products and legislation. Livestock Production Science 61: 179–192. My discussion deals with what Bartussek calls the ANI-35. The quotations are on pages 180, 182 and 186.

Dfraser_C012.indd 255

6/27/2008 12:44:51 PM

256

Drawing Conclusions about Animal Welfare

Table 12.3

Elements of the Animal Needs Index for cattle.

Component

Criterion scored

1. Ability to move

Area per animal Rising and lying down Exercise outdoors Access to pasture Area per animal Social structure of the herd Integration of followers Exercise outdoors Access to pasture Resilience of the lying area Cleanliness of the lying area Slip resistance of the lying area Floor condition, moving area Floor condition, exercise area Access to pasture Light Air quality Draughts in the lying area Equipment noise Days outside/year Hours outside/day Cleanliness of the housing State of the equipment State of the animals’ coat Cleanliness of the animals State of the animals’ hooves Injuries from housing/equipment Animal health

2. Social contact

3. Quality of flooring

4. Conditions within the building

5. Quality of human care

a For

Lowest and highest scores assigneda 0 – 3.0 0 – 3.0 0 – 3.0 0 – 1.5 0 – 3.0 0.5 – 2.0 0.5 – 1.0 0 – 2.5 0 – 1.5 0.5 – 2.5 0.5 – 1.0 0.5 – 1.0 0.5 – 1.0 0.5 – 1.5 0 – 1.0 0.5 – 2.0 0.5 – 1.5 0.5 – 1.0 0.5 – 1.0 0 – 2.0 0 – 2.0 0.5 – 1.0 0.5 – 1.0 0.5 – 1.0 0.5 – 0.5 0.5 – 1.5 0.5 – 1.5 0.5 – 1.5

certain of the criteria, a farm could receive a negative score of 0.5 if conditions were poor.

Source: Based (with some rewording) on Bartussek (1999).

welfare such as nutrition, health and hygiene. With this in mind Bartussek decided to emphasize ‘housing parameters clearly contrary to the short-term economic interests of producers’. For this reason he attached relatively low importance to variables such as basic hygiene, access to food and water, and freedom from disease; these were either not included or were given low weightings in the scoring system. To assess the success of the ANI, Bartussek did a survey of ﬁrms that used the ANI system to evaluate farms, and asked them (among other things) to indicate

Dfraser_C012.indd 256

6/27/2008 12:44:51 PM

Selecting and Combining Criteria of Animal Welfare

257

how well the ANI scores represent the actual ‘animal-welfare situation’ on the farms they assessed. Using a 5-point scale running from 1 (very good) to 5 (not sufﬁcient), the different ﬁrms gave very different evaluations, but the two ﬁrms that had assessed the largest number of facilities gave a rating of only 4. One could imagine, for example, that with the ANI awarding very few points for basic nutrition and disease status, assessors who saw these as key elements of animal welfare might well perceive only a modest relationship between animal welfare (as they saw it) and the ANI scores. Bartussek’s description nicely clariﬁes the interplay that occurs between scientiﬁc information and other considerations in animal welfare assessment by the ANI system. The selection and weighting of criteria was clearly guided by the available science, combined with pragmatic considerations such as ease of scoring. But the science did not determine, for example, that ‘light’ (with a maximum score of 2.0) is exactly twice as important to cows as the ‘state of the animals’ coat’ (with a maximum score of 1.0). Clearly these decisions were inﬂuenced by Bartussek’s judgement of what is more important and less important for cows to have a good life. In addition, the system was underlain by the (in essence) ethical decision to reward farmers much more for those elements of animal welfare that are contrary to their short-term economic beneﬁt (things like area per animal and exercise outdoors) than for those elements (good nutrition, basic health) that are likely to yield ﬁnancial returns through greater productivity. But if the decisions about the selection and weighting of criteria could be made not by an individual but by a panel of ‘experts’, could this eliminate the involvement of values in animal welfare assessment? A Dutch team led by Marc Bracke has made use of expert opinion in various ways in programmes designed to assess the ‘overall’ welfare of animals. In one example, the team identiﬁed a group of 29 individuals who met their criteria for a ‘pig-welfare expert’. The team also selected a list of 20 ‘attributes’ that could be scored for various housing systems for sows. The experts were then asked to assign a weighting factor to each attribute on a scale from 0 (‘the least important attribute’) to 10 (‘the most important attribute’). The experts showed reasonable agreement on some of the attributes. For example, the comfort of the resting surface was consistently scored as being of medium importance: the weighting factors assigned by the experts had a median value of 5, and the inter-quartile range (the range around the median that included half the experts) was between 4 and 6. For other attributes, however, the experts showed much less agreement. For example, the amount of space per pen received a median score of 8, but half of the experts were spread over the range of 5–10, and one expert assigned a value of only 2. Several attributes (air quality, light and the opportunity to wallow) received scores covering the full range from 0 to 10.15 15Bracke, M.B.M., Metz, J.H.M., Spruijt, B.M. and Schouten, W.G.P. 2002. Decision support system for overall welfare assessment in pregnant sows B: Validation by expert opinion. Journal of Animal Science 80: 1835–1845. The quotations are from page 1836.

Dfraser_C012.indd 257

6/27/2008 12:44:52 PM

258

Drawing Conclusions about Animal Welfare

Why would ‘experts’ disagree so greatly about the relative importance of different factors? Part of the answer may lie in the role played by values in people’s conception of animal welfare. If experts in bicycle repairs were asked the fastest way to adjust the tension of a brake cable, we expect that they will have a shared understanding of the goal, and that their opinions will reﬂect their technical knowledge of how this goal can be accomplished. However, if experts in animal welfare are asked how to achieve (in effect) a good life for animals, their responses will reﬂect their value-based beliefs about what kind of life is most desirable, together with their technical knowledge about how to achieve such a life. Thus, although the use of pooled expertise can prevent one person’s views from unduly inﬂuencing a system of animal welfare assessment, and although it may provide a transparent process for assigning weighting scores to diverse elements, this should not be mistaken for eliminating the role of values in welfare assessment. The ANI and other attempts to combine variables in ‘overall’ welfare assessment have many parallels in other ﬁelds of mandated science. For example, the Environmental Sustainability Index is used to compare the progress of different nations toward achieving ‘environmental sustainability’. Like the ANI, it assembles a large number of variables into ﬁve ‘components’, and these are then weighted and combined mathematically to produce the index.16 The Human Development Index, which is used to identify countries with greater or lesser degrees of ‘human development‘, weights and combines four variables (life expectancy, adult literacy, proportion of the population enrolled in education, and gross domestic product per capita) to produce the value of the index for a given country.17 As social scientists Stephen Morse and Evan Fraser point out, in devising such multi-variable concepts and measurement devices, knowledge and scientiﬁc research play important roles, but there are also important political questions over who deﬁnes the concept and whose interests or viewpoints are privileged by a given deﬁnition.18 IN THE THREE TASKS we have reviewed in this chapter – comparing animal welfare in different housing systems, devising welfare assurance programmes, and creating scoring systems – scientiﬁc evidence about animal welfare has played a key and essential role. However, when we probe into the use of scientiﬁc information, we ﬁnd that the application of the science is underlain, in part at least, by valuebased decisions. Some of these decisions are simply pragmatic; for example, we may decide to favour methods of welfare assessment that are easily scored. Others are epistemological in that they draw on what we regard as good or reliable sources

16Morse, S. and Fraser, E.D.G. 2005. Making ‘dirty’ nations look clean? The nation state and the problem of selecting and weighting indices as tools for measuring progress towards sustainability. Geoforum 36: 625–640. 17Sagar, A.D. and Najam, A. 1998. The human development index: A critical review. Ecological Economics 25: 249–264. 18Morse and Fraser, 2005, page 626.

Dfraser_C012.indd 258

6/27/2008 12:44:52 PM

Selecting and Combining Criteria of Animal Welfare

259

of knowledge; for example, we may feel that certain kinds of evidence are better than others because they can be scored with less subjectivity. But other decisions invoke our assumptions and beliefs about what is important for an animal to have a good life. We invoke these beliefs whenever we decide to treat measures (longevity, limping, stereotyped behaviour) as ‘indicators’ of animal welfare. And when we combine different measures into some overall evaluation, we are (either tacitly or explicitly) using these same values to make quantitative claims about the relative importance of different elements. But let us be clear on what is not implied by the above discussion. To say that values are involved in assessing animal welfare is not to say that animal welfare is a purely subjective concept that is not amenable to scientiﬁc study, nor that the welfare of animals depends on the human culture in which the animals are kept, nor that animal welfare is just a matter of opinion and that welfare requirements should therefore be decided by public opinion poll. These ideas often arise from the mistaken view that concepts are either objective and value-free, or else subjective and based merely on opinion. To be clear on how our understanding of animal welfare is both values-based and science-based, let us attempt in the next chapter to develop a more nuanced understanding of the role of values in science.

Dfraser_C012.indd 259

6/27/2008 12:44:52 PM

Animal Welfare, Values and Mandated Science

13

During the nineteenth century, when the boundaries of science were still being drawn, scientists like Henri Poincaré, Max Weber and Auguste Comte insisted that there is a clear demarcation between science and ethics. In Poincaré’s terms (as we saw in the Introduction), they ‘touch’ but do not ‘interpenetrate’. And in the natural sciences of the time (geology, chemistry, biological taxonomy) this ﬁrst-order approximation worked well enough that generations of scientists were taught to view their work as clearly separable from values and ethics. Of course, the application of science might raise ethical issues. Alfred Nobel, who founded the prizes that bear his name, was an inventor of explosives, and lived to see his scientiﬁc discoveries result in vast numbers of human deaths through warfare. But even if science could be applied in good or evil ways, many people contended that the research itself remained clearly separate from ethical matters. In the twentieth century, however, science came to be applied to topics such as food safety, environmental sustainability and animal welfare – topics where notions of good and bad, better and worse, are embedded in the concepts themselves. Here the simple demarcation between science and ethics proved so inadequate that it stimulated a vigorous debate about how exactly ethics and other value-based considerations inﬂuence science, and vice versa. To understand animal welfare, and other concepts that are studied through science while rooted in values, we need to go beyond Poincaré’s model of two separate domains that touch but do not interpenetrate. TO BRING THIS RATHER CONCEPTUAL issue alive, let’s start with another ﬁght.1 In the one corner, wearing impeccable scientiﬁc credentials as the world’s most senior 1For simplicity I have cast the debate as one between Tannenbaum and Broom, but many scientists and philosophers have made important contributions. Other relevant works listed in the bibliography include: Sandøe and Simonsen, 1992; Rollin, 1993 and 1995; Fraser, 1995 and 1999; Staﬂeu et al., 1996; Appleby, 1999; Nordenfelt, 2006; and Lassen et al., 2006. Portions of this chapter are reworked from Fraser, 2003.

260

Dfraser_C013.indd 260

6/27/2008 12:45:01 PM

Animal Welfare, Values and Mandated Science

261

holder of an academic chair in animal welfare science is Donald Broom. Broom deﬁned an animal’s welfare as ‘its state as regards its attempts to cope with its environment’.2 Broom, as we have seen, proposed some semantic conventions: that ‘animal welfare’ be used to refer to the animal’s state, not to external beneﬁts given to the animal, and that welfare be considered as a scale running from good to bad, not just the positive end of the scale. He also made the more substantive claim that ‘welfare can be measured in a scientiﬁc way that is independent of moral considerations’. In another work, co-authored by Andrew Fraser, Broom elaborated on this idea: The assessment of animal welfare can be carried out in an objective way which is quite independent of any moral considerations. Mortality rate, reproductive success, extent of adrenal activity, amount of abnormal behaviour, severity of injury, degree of immunosuppression, or level of disease can all be measured … . In addition to measures of poor welfare, it is possible to investigate the preferences of animals and the value which they place on various resources or other aspects of their environment … . When scientiﬁc evaluation of welfare has been carried out, there remains the moral question of how poor welfare should be before it is regarded as unacceptable. This is an issue where the farmer, the veterinary surgeon, the welfare research worker, or the member of the public are equally entitled to have an opinion.3

Thus, Broom appears to be saying, the assessment of animal welfare is a scientiﬁc task that is not (or should not be) inﬂuenced by moral considerations, whereas ethical decisions about how to treat animals are matters of opinion, and the opinions of animal welfare scientists should count for no more and no less than those of other people who want to be heard on the issue. In the opposite corner, wearing the trunks of a philosopher highly experienced in issues of animal ethics, is Jerrold Tannenbaum. Tannenbaum referred to the view taken by Broom as ‘the pure science model’ of animal welfare, and he undertook to show that it is incorrect. In particular, Tannenbaum argued that the study of animal welfare ‘is inextricably connected with ethical concerns’ and hence that ‘the pure science model fundamentally misconstrues the nature of animal welfare’.4 Tannenbaum pointed out several cases, progressing from the more obvious to the more subtle, where he saw ethical considerations underlying the scientiﬁc study of animal welfare. 2Broom,

D.M. 1991. Animal welfare: Concepts and measurement. Journal of Animal Science 69: 4167–4175. The quotations are on page 4168. 3This quotation is from Fraser and Broom, 1990, a collaboration between Broom and Andrew F. Fraser. The passages I am quoting are nearly identical to ideas published elsewhere by Broom, so for simplicity I refer to them as his. 4Tannenbaum, J. 1991. Ethics and animal welfare: The inextricable connection. Journal of the American Veterinary Medical Association 198: 1360–1376. The quotation is on page 1360.

Dfraser_C013.indd 261

6/27/2008 12:45:02 PM

262

Drawing Conclusions about Animal Welfare

First, and most obviously, he noted that in animal welfare science we use ethical considerations when choosing which animals to study. Thus, we study the welfare of laboratory rats because we feel we have some ethical responsibilities toward them, but we do not study the welfare of sewer rats because we feel we are not responsible for their welfare. Indeed, studying the welfare of sewer rats, although it might be just as valid scientiﬁcally, would strike many scientists as a waste of time. Next, the levels of welfare we study are determined by what levels we feel we should consider providing. Thus we study how the welfare of pigs is inﬂuenced by straw bedding but not by feather mattresses because we would entertain an obligation to provide them with straw but not with mattresses. Third, and most germane to our present discussion, the variables we include in the study of animal welfare depend on the kind of life we think we ought to provide for animals. Scientists who feel that we owe animals a life free from unnecessary suffering are likely to study states of suffering; those who feel that we owe animals nothing more than survival and basic health will study survival and basic health. Thus, Tannenbaum argued, when scientists study animal welfare, even the decisions they make over which data to collect are determined by ethical considerations. The case looks convincing, yet Broom also made valid points about the separation of ethics and science: that collecting data about animal welfare is one thing, and making ethical decisions about how to treat animals is another; that scientists should not try to reduce ethical questions to purely technical ones; and that the debate about what constitutes appropriate animal care is a matter for society broadly, not just for scientists. If both sides of the debate have merit, how are we to resolve the apparent conﬂict? PART OF THE CONFUSION results from using an overly simple dichotomy between ‘facts’ (as the province of science) and ‘values’ (as the province of ethics, philosophy and personal opinion). We can do more justice to both sides of the Broom– Tannenbaum debate if we think of statements as falling into three types instead of two. For simplicity, let us continue to call the ﬁrst type of statements ‘facts’, not to imply that such statements are ‘true’ in some absolute way, but rather that they refer to factual matters or the sort of matters that can be judged as true or false. Factual statements (in this sense) are sometimes called ‘descriptive’ (as opposed to prescriptive) in everyday speech, and ‘constative’ in philosophical analysis. Thus an example of a factual (or descriptive, or constative) statement would be: ‘Pigs are bigger than chickens’. A second type of statement we might call ‘preference values’. These are ‘evaluative’ statements in the sense that they involve some notion of better or worse, more important or less important, more desirable or less desirable. For example, ‘Pigs are more important than chickens’.

Dfraser_C013.indd 262

6/27/2008 12:45:02 PM

Animal Welfare, Values and Mandated Science

263

The third level we might call ‘moral values’. These consist of ‘normative’ or ‘prescriptive’ statements by which we indicate ideas of right and wrong, and how people ought to behave. For example, ‘We ought to eat chickens rather than pigs’.5 Of course there are signiﬁcant connections between our preference values and our moral values. What we think is good or important will inﬂuence what we think is right, and what we think is right will inﬂuence what we think is good and important. Indeed, if there were no other competing considerations, we ought to act so as to achieve what is good and avoid what is bad. In real life, however, competing considerations are almost always involved. A decision about animal care may encounter competing moral considerations such as how an action will affect the safety and availability of food, or whether the action is prohibited by law.6 Hence, our moral values – what we think ought to be done – do not follow in any simple manner from our preference values. Separating values into preference values (captured in evaluative statements) and moral values (captured in normative statements) helps us to be clearer on the role that values play in the scientiﬁc assessment of animal welfare. In assessing the welfare of a group of laboratory rats, we may study the incidence of disease because we believe that a low incidence of disease is good for the animals (an evaluative claim), but the data-gathering exercise does not tell us whether we ought to vaccinate the animals against certain diseases or house them in sterile facilities (a normative claim). Likewise, we assess the level of pain associated with lameness in dairy cattle because we feel that pain is an important consideration for their quality of life (an evaluative claim); but the scientiﬁc study does not by itself determine when we ought to treat cattle for lameness or cull them from the herd (a normative claim). In these cases we are basing our selection of variables on preference values – on what we view as good, or desirable, or important for animals. Of course, the preference values that underlie the science are related to ‘ethical concerns’ (Tannenbaum’s point), but the scientiﬁc study remains distinct from making moral decisions about correct action (Broom’s point). FOOD SAFETY, ENVIRONMENTAL INTEGRITY, agricultural sustainability and animal welfare – these phrases (along with many others) identify subjects of modern applied research where the science is underlain to an important degree by preference values. 5I am using the terms constative, evaluative and normative in a way that follows the Oxford Dictionary of Philosophy, 1994. Thus (paraphrasing this source) a constative statement declares something to be the case, an evaluative statement attributes positive or negative value to something, and a normative statement refers to rules whose violation would make a person liable to censure. The term ‘evaluative’ has been used in different ways by different philosophers. For example, philosopher John Searle uses ‘evaluative statements’ to include normative statements. See Searle, J.R. 1967. How to derive ‘ought’ from ‘is’. Pages 101–114 in Theories of Ethics (P. Foot, editor). Oxford University Press, Oxford. 6For discussion see Hurnik, 1993, and Sandøe, P. and Simonsen, H.B. 1992. Assessing animal welfare: Where does science end and philosophy begin? Animal Welfare 1: 257–267.

Dfraser_C013.indd 263

6/27/2008 12:45:02 PM

264

Drawing Conclusions about Animal Welfare

Notice ﬁrst that each of the phrases includes two words. The ﬁrst (food, environmental, agricultural, animal) identiﬁes a general area of enquiry. The second term (safety, integrity, sustainability, welfare) is at least partly an evaluative term in that it conveys some notion of merit or worth. To say that safety or integrity or welfare has increased implies not simply a change but a change for the better. The terms do not speciﬁcally invoke moral values: to say that safety has increased does not imply that the improvement is morally required. Thus these concepts – which I will call ‘evaluative concepts’ – incorporate preference values but not speciﬁcally moral values. Let us be clear on how evaluative concepts like food safety and animal welfare differ from other concepts used in science such as atomic weight and metabolic rate. These latter terms are not evaluative concepts in that a greater or lesser atomic weight or metabolic rate does not by itself carry any sense of better or worse. Description, in those cases, does not by itself imply relative value or merit. The study of animal welfare obviously includes concepts such as plasma cortisol levels and duration of tonic immobility whose measurement is purely factual or descriptive in much the same way that measuring atomic weight or metabolic rate is factual or descriptive. However, when we use descriptive measures to draw conclusions about animal welfare, then we are going beyond the level of simple description and entering a level that involves notions of better and worse. To state how long a hen remained in tonic immobility is descriptive; to conclude from this that the bird’s welfare is good or bad involves evaluation. There is another difference between such scientiﬁc concepts as atomic weight and many evaluative concepts such as animal welfare. The concept of atomic weight arose in science and takes its meaning from science. It had no everyday meaning before it was invented by scientists. In contrast, many evaluative concepts arose in everyday language and had everyday meanings before scientists began paying attention to them. And when these concepts are used in public debate, policy and legislation, it is presumably their everyday meanings that people understand. When a term with an everyday meaning comes to be adopted into science, there is a risk of confusion if scientists try to give the term a new, technical meaning that does not correspond to its everyday meaning. Here is a cautionary tale: In a (ﬁctional) nutrition laboratory, scientists decided to conduct a scientiﬁc assessment of bread quality in order to help consumers to buy good bread. They were equipped to measure standard nutrients such as protein and minerals, but they did not have an assay for mould-derived toxins, and they were skeptical of the less objective methods commonly used to assess freshness, texture, and ﬂavour. Noting that nutrients are important components of bread quality and that nutrient analysis has an established history in science, they combined their various nutrient measurements into a ‘bread quality index’ and showed (scientiﬁcally, and using the most objective measures available) that stale, mouldy bread is equal in quality to freshly baked bread.

Dfraser_C013.indd 264

6/27/2008 12:45:02 PM

Animal Welfare, Values and Mandated Science

265

In this anecdote, the scientists adopted an everyday term (bread quality) but gave it a scientiﬁc meaning that did not correspond to its everyday meaning. Science can, of course, be applied to bread quality and other topics that have everyday meanings, but if scientists try to deﬁne everyday concepts in terms of scientiﬁc variables, they must be careful not to miss or distort the everyday meanings of the terms, or their research may prove irrelevant to its intended social purpose, if not actually misleading. Clearly animal welfare has an everyday meaning (or meanings). It is widely used in everyday speech to refer to the quality of life of animals, especially when ethical concerns are being discussed. When scientists attempt to assess animal welfare, if the meaning of animal welfare that they adopt does not correspond to the everyday meaning(s) of the term, then their conclusions may be as irrelevant to public policy and debate (or as misleading) as the ﬁctional scientists’ conclusion about mouldy bread. If, however, we treat animal welfare exactly like bread quality, we might make a different error: In a (also ﬁctional) classroom, a young boy took the class’s pet frog home to care for it for the holidays. He shared the class’s concern that the frog should be kept healthy and comfortable. Therefore, he carefully cleaned the terrarium and rubbed the frog every day with sanitizing hand wash. He had been learning about the food groups, so decided to feed the frog on fresh vegetables and whole grains as well as animal products. And he installed a heat lamp over the terrarium so that the frog would stay warm and dry.

This anecdote reminds us that animal welfare is unlike bread quality in that animals, unlike loaves of bread, have interests of their own, and these interests provide the ultimate criteria for animal welfare. If the frog does not want to be dry, if it is harmed by being dry, and if it does not derive any possible beneﬁt from being dry, then the boy is simply mistaken in thinking that being dry is good for the frog. Through scientiﬁc knowledge, we can make better judgements about what is good or bad for animals, and thus improve on uninformed opinion or simplistic extrapolation from humans to other species. The mouldy bread error and the dry frog error convey two messages about the scientiﬁc study of animal welfare that may at ﬁrst appear to be in conﬂict: that the use of the term ‘animal welfare’ in science needs to respect the everyday meaning(s) of the term so that scientiﬁc conclusions will be relevant to the related public debate; and that our conception of animal welfare needs to be based on a scientiﬁc understanding of what is good for the animals themselves. If these ideas appear to be in conﬂict, the resolution lies partly in thinking in terms of a two-by-two matrix whereby conclusions about animal welfare (or other evaluative concepts) may involve (1) either a sound or faulty understanding of the relevant factual issues, and (2) a meaning of the evaluative concept that does or does not conform to its widely held meaning(s) in everyday language. The mouldy

Dfraser_C013.indd 265

6/27/2008 12:45:02 PM

266

Drawing Conclusions about Animal Welfare

bread error falls into one cell of the matrix: it uses correct facts about bread, but it redeﬁnes the evaluative concept ‘bread quality’ in a way that does not conform to the widely understood, everyday meaning of the term. The dry frog error falls into the diagonally opposite cell: the young boy shared the class’s understanding of the value issue – that the frog should be kept healthy and comfortable – but he was mistaken on the factual issue of how to achieve this goal. Thus scientists working in a mandated ﬁeld need to test their recommendations against two criteria: recommendations need to be based on a sound scientiﬁc understanding of the phenomena, and they need to reﬂect the value issues captured in the everyday meanings of the terms. To summarize, a good deal of contemporary scientiﬁc research focuses on topics such as food safety, environmental conservation and animal welfare. These cannot be viewed as merely descriptive concepts (like atomic weight) because they incorporate notions of what is better or worse. Nor should they be seen speciﬁcally as ethical concepts: a ﬁnding about food safety or animal welfare does not by itself tell us how we ought to act. Rather, they can be regarded as evaluative concepts which incorporate preference values. Although evaluative concepts are valuesbased in this sense, they are not simply matters of opinion or philosophy. Rather, science is of fundamental importance for understanding such concepts. Indeed, the preference values embedded in evaluative concepts help form the rationale for collecting and interpreting scientiﬁc information. Finally, many evaluative concepts have everyday meanings that reﬂect preference values commonly held in society. When scientists study an evaluative concept, they need to recognize the everyday meaning(s) of the concept, while at the same time providing information that is relevant to understanding it. THE IDEA OF SCIENTISTS STUDYING concepts that are both values-based and sciencebased may seem strange to readers who think of science as value-free investigation. To help put this view of science ﬁrmly to rest, let us look brieﬂy at some of the many ways that values inﬂuence the conduct of science, including the scientiﬁc study of animal welfare. Perhaps the most obvious case involves values based on vested interests. As philosopher Geert Munnichs has pointed out, in many debates that involve science, the scientists themselves have vested interests in the subject because of their particular mandate, employment or sources of funding.7 Those working for pharmaceutical companies, for example, will have a primary interest in bringing proﬁtable products to the market; those working for government bodies that regulate new drugs will have a primary interest in protecting the public from harm. Ethicist Conrad Brunk and co-workers used an analysis of a pesticide registration hearing to point out how company scientists and government regulators approached the seemingly 7Munnichs, G. 2004. Whom to trust? Public concerns, late modern risks, and expert trustworthiness. Journal of Agricultural and Environmental Ethics 17: 113–130.

Dfraser_C013.indd 266

6/27/2008 12:45:02 PM

Animal Welfare, Values and Mandated Science

267

technical issue of pesticide safety in fundamentally different ways that had much to do with their respective roles.8 Paul Thompson provides the unappealing metaphor of scientists associated with different interest groups, lining up on either side of contentious issues and giving allegedly expert but directly contradictory testimony, in a manner reminiscent of lawyers for the prosecution and the defence.9 Such thinking certainly applies in the case of animal welfare. Thus, scientists who receive research funding from the animal industries may use (and insist on the validity of) measures of animal welfare that correspond to industry goals such as rate of growth and production, whereas scientists funded by organizations devoted to reducing animal suffering may use measures of animal welfare based on affective states. As Munnichs notes, the only solution in such cases may be to ensure that scientists aligned with all the relevant interest groups participate in resolving issues so that no one approach or set of assumptions comes to be regarded as ‘scientiﬁc’ to the exclusion of others. For example, when the American Veterinary Medical Association appointed a committee to advise them on the appropriateness of gestation stalls for sows, it assembled a group of scientists and other technical specialists with very diverse afﬁliations. Some were funded by the swine industry, some by the animal protection movement, some by government and so on. In such cases, as Munnichs noted, science achieves legitimacy not because the individual scientists are disinterested, value-free providers of information, but because the process engages scientists who reﬂect a wide range of vested interests and associated values. In addition to vested interests of this sort, scientists also have ‘epistemological values’ – beliefs that certain sources of knowledge are better than others – and many of these arise from the culture of science itself. Most obviously, scientists tend to value knowledge that has arisen within science, and may ignore or discount matters that have not been studied scientiﬁcally. For example, there is now a large body of scientiﬁc research on pain and fear in animals, and when scientists provide input into animal welfare policy, they are likely to draw on this knowledge. In contrast there has been relatively little scientiﬁc study of boredom and pleasure, and scientists may tend to ignore or de-emphasize these states even though they may be very relevant to animal welfare. Scientists may also value types of evidence that conform to the traditions of their particular disciplines. In assessing fear, for example, physiologists may trust physiological measures such as increased heart rate which can be measured using the standard tools of physiology, but they may doubt the validity of behavioural measures such as increased vigilance. In contrast, animal behaviourists may trust measures based on vigilance because behavioural theory provides a strong rationale for linking vigilance to fear, but they may doubt the validity of measures such as heart rate. 8Brunk, C.G., Haworth, L. and Lee, B. 1991. Value Assumptions in Risk Assessment: A Case Study of the Alachlor Controversy. Wilfrid Laurier University Press, Waterloo, Canada. 9Thompson, 1998.

Dfraser_C013.indd 267

6/27/2008 12:45:02 PM

268

Drawing Conclusions about Animal Welfare

Many scientists also attach a high value to certainty. Some measures of basic health and functioning – variables such as rates of growth, survival and reproduction – can be measured with a high level of certainty. In contrast, affective states such as frustration need to be inferred indirectly from other variables. Perhaps as a result, some scientists have assigned more weight to measures of growth and reproduction than to states like frustration, even though frustration is, by some accounts, more directly relevant to animal welfare. Scientists in many ﬁelds also attach particular value to evidence derived from experimentation. When scientiﬁc investigation of an issue has reached the point of creating different explanatory hypotheses, an experimental test of the hypotheses is often the most deﬁnitive step. When Louis Pasteur proposed that the disease anthrax is caused by a speciﬁc pathogen, and that inoculation with an attenuated form of the pathogen could protect animals from developing the disease, the matter was settled in a dramatic and convincing fashion by an experiment in which sheep were exposed to the pathogen after some had been inoculated and others had not.10 The power of experimentation has understandably led many scientists to value experimental evidence ahead of other types of evidence such as correlations or other means of recognizing patterns. However, experimentation also carries certain risks. Experiments often involve some degree of simpliﬁcation or abstraction. An experiment may test the effect of a single factor or a small number of factors under controlled conditions, rather than addressing the full range of factors and variation seen in real life. Hence, an experiment may give conclusive results concerning the variables tested, but actually provide a very incomplete understanding of the phenomenon under study. Moreover, if there are multiple causal factors, scientists may focus their attention on those factors that are amenable to experimentation and ignore those that are not. Experiments on stereotyped behaviour, for example, have identiﬁed hunger and restriction of movement as contributing factors, but are there other factors (social tension, genetic selection) whose importance has been overlooked because they are not so amenable to experimentation? Scientists in many ﬁelds also value quantitative data ahead of qualitative or narrative data. Recall how George Romanes used narrative accounts of animal behaviour to build up an understanding of the mental capacities of animals. Such use of narrative data went out of favour in twentieth-century science, and was replaced by more quantitative approaches, such as studies of how quickly animals would learn to perform a simple response for a food reward. Later, when Jane Goodall and others began describing speciﬁc events in the lives of chimpanzees – events that demonstrated the capacity to plan deceptions and to respond to the death of a mother with prolonged depression – these narrative accounts created a much richer and more complex picture of the mental lives of chimpanzees than had 10Vallery-Radot, R. 1919. The Life of Pasteur (R.L. Devonshire, translator). Constable & Company, London. The passage is on pages 311–323.

Dfraser_C013.indd 268

6/27/2008 12:45:02 PM

Animal Welfare, Values and Mandated Science

269

ever been revealed by controlled laboratory studies, but many scientists discounted these ﬁndings initially because the data were qualitative and narrative rather than quantitative and experimental. There is also a tension in science between what have been loosely called ‘reductionistic’ versus ‘holistic’ approaches. Recall that one of the key elements of Positivism was the belief that the sciences are built on each other in a hierarchical manner – that sociological phenomena could ultimately be ‘reduced’ to biology, that biology could be ‘reduced’ to chemistry and so on. Scientists who hold this mental model of science tend to seek explanations at a more ‘fundamental’ level. They might believe, for example, that complex types of behaviour are best explained by identifying the simpler processes (nerve impulses, hormonal changes, genetic differences) that underlie them. For scientists in this mind-set, it might seem only logical and scientiﬁc to ‘reduce’ animal welfare to simple measures such as levels of cortisol in the blood. In contrast, scientists favouring more ‘holistic’ approaches – that is, approaches that do not see complex phenomena as reducible to constituent parts – are likely to favour measures taken at the level of the entire animal, such as the animal’s preferences and motivations, or even (like Alex Stolba) at the level of the animal interacting with its natural environment. Many of the preferences that scientists show for particular types of evidence – for measurements that can be made with a high level of certainty, for results derived from planned experiments, for quantitative data ahead of narrative data, and so on – can be justiﬁed on some rational grounds. We can arguably achieve more reliable understanding by doing controlled experiments than by calculating correlations, and by drawing inferences from repeatable events rather than unique events. But epistemological values that favour certain types of data ahead of others also exert a complex inﬂuence on the phenomena that scientists study and the conclusions that they draw. Finally, in addition to epistemological values and values based on vested interests, scientists are also inﬂuenced by other kinds of personal and cultural values. In particular, in addressing such culturally rooted concepts as animal welfare, scientists will naturally draw on their understanding of ‘animals’ and ‘welfare’, and these (as we have seen) are concepts that mean rather different things in different societies and to different people in the same society. For example, scientists whose beliefs are strongly inﬂuenced by the world-view of Industrialism are likely to attach greater value to measures of animal welfare that are based on productivity and efﬁciency, and they may be unimpressed by research that tries to assess the emotions of animals or how well animals perform their natural behaviour. In contrast, scientists whose beliefs are more strongly inﬂuenced by the world-view of Romanticism are likely to value measures based on affective states and natural behaviour, and to discount measures based on growth and productivity. Here we need to amplify Munnichs’ argument about scientists and vested interests. When we see scientists with close ties to (say) the animal industries or the animal protection movement, we need not assume that these scientists have a

Dfraser_C013.indd 269

6/27/2008 12:45:02 PM

270

Drawing Conclusions about Animal Welfare

vested ﬁnancial interest in the success of those sectors. Rather, the scientists may simply share the values and world-views of the sectors with which they work. When, for example, scientists associated with the pig industry argue that growth and reproduction are the most trustworthy measures of animal welfare (see, for example, quotations in Box 5.1 on page 101), they may not be trying to protect the economic interests of the industry; rather they may simply share beliefs about animals and animal welfare that are common among pig producers. There is one further pitfall for those who consider that concepts based on science are value-free. When dealing with animal welfare, food safety and other evaluative concepts, scientists may believe that if they relate the concept to particular scientiﬁc theories or measures, this will make their interpretation of the concepts ‘science-based’ and hence free from values. In the debate over sow gestation stalls, for example, the Australian reviewers, in deﬁning animal welfare in terms of homeostasis and related scientiﬁc measures, felt that their approach was a ‘scientiﬁc’ approach, in contrast to ‘public perception’ and ‘emotional views’ of animal welfare.11 Barnard and Hurst linked their very different conception of animal welfare to the theory of evolution, and they considered that this distinguished their conception of animal welfare from other conceptions that they regarded as ‘anthropomorphic’ in the sense of mistakenly invoking human values.12 These and other conceptions of animal welfare are science-based in that they invoke scientiﬁc concepts and measures. However (as we have seen) they are also values-based in that they rest on what their proponents consider desirable or important or ethically signiﬁcant for animals to have a satisfactory quality of life. IF VALUES AND FACTUAL ISSUES interact in the conduct of animal welfare research, the interplay is particularly complex when the science comes to be applied. Henri Poincaré (as we saw in the Introduction) proposed a two-part process whereby an ethical decision identiﬁes ‘to what goal we should aspire’ and science then ‘teaches us how to attain it’. This view has merit as a rough, ﬁrst-order approximation in understanding the application of science. For example, when a government or industry decides on a set of animal welfare standards, there is, very roughly, an ethical decision about what goals to pursue – to protect the basic health of animals, to prevent avoidable suffering and so on – and then there are technical decisions about how to achieve these goals. Thus, when the United Egg Producers decided to require 432–555 cm2 of cage space per laying hen, there was (in effect) a tacit ethical decision to provide enough space to prevent crowding-related reductions in survival and production, but not to provide for such natural behaviour as nesting and perching; and then there was a technical decision on how to achieve the goal they had set. When the European Union decided to require that standard cages be replaced with ‘furnished’ cages, there was (in effect) an ethical decision that hens 11Barnett

et al., 2001. The words quoted appear on pages 2, 13 and 21 respectively. and Hurst, 1996, page 405.

12Barnard

Dfraser_C013.indd 270

6/27/2008 12:45:03 PM

Animal Welfare, Values and Mandated Science

271

should be able to carry out key elements of their natural behaviour, followed by a technical decision about how this could be achieved. On deeper reﬂection, however, we can recognize a much closer interplay between the ethical and technical elements. Even after deciding that a particular animal welfare standard will try to minimize pain, there are further, more detailed ethical decisions that cannot be made without technical knowledge about the possible options and trade-offs. Tail-docking of piglets likely causes some degree of pain, but may make the animals less likely to experience painful tail-biting later in life. Removing the horn buds from young calves is clearly a painful procedure, but may reduce the chance that other animals will be injured by mature horns in the future. In such cases, making the ethical decisions about which options to pursue is bound up with technical matters such as the level of pain that they cause, options for mitigating the pain, and likely future consequences. Much the same interplay of ethical and technical elements occurs in decisions about promoting health. Conﬁning animals indoors may help to prevent the entry of infectious diseases but lead to poorer lung health because of bad air quality. Keeping cats in hygienic steel cages may reduce exposure to respiratory pathogens but cause a stress response that could be sufﬁcient to lower their immunity. Here again, after the ethical decision is made to pursue a high level of health, there are more detailed ethical decisions about which options to pursue, and these decisions cannot be made without a technical understanding of the available options, likely outcomes and potential trade-offs. This intertwining of ethical and technical elements occurs partly because animal welfare is not a single phenomenon but a ‘conglomerate concept’ consisting of different elements, some of which are themselves conglomerate concepts. Thus, in creating programmes or policies to protect animal welfare, there are high-level ethical decisions about which broad animal welfare goals to pursue (health, suffering, natural living) and more detailed ethical decisions about which speciﬁc objectives to pursue within each of the broad goals. Especially at the more detailed level, the decisions require an understanding of the technical aspects of the subject. Thus, the two-step process implied by Poincaré serves only as a crude ﬁrst approximation. The actual application of animal welfare science is a much more iterative process in which ethical and technical elements are more intimately intertwined. CREATING A DIVISION BETWEEN FACTS (as the province of science) and values (as the province of philosophy and ethics) also fails to capture the role that scientists can and do play in helping society arrive at ethical decisions. In matters of ethics and policy, scientists frequently function not simply as providers of information, but as engaged participants who make unique contributions. Of the normative roles played by scientists, perhaps the least controversial has been to develop and promote the use of methods designed to improve animal welfare. When Ragnar Tauson did his classic comparisons of injuries to hens kept in different types of cages, he also worked with cage designers and Swedish

Dfraser_C013.indd 271

6/27/2008 12:45:03 PM

272

Drawing Conclusions about Animal Welfare

regulators to ensure that cages were improved in the egg industry. Later, when research had shown that hens are highly motivated to retreat to a secluded area to lay, Tauson and others developed new designs for ‘furnished’ cages that contain a nesting area to accommodate this behaviour, together with a perch and an area for dust-bathing. In these examples we see scientists stepping beyond the role of information provider, and engaging instead as promoters and facilitators of change. More controversially, scientists have also tried to assess the merits and even the ethical acceptability of alternative practices. When Helmut Bartussek developed his rating system for different farms, his objective was not simply to describe the differences but to rate them according to their merit in promoting animal welfare. When the Australian team of scientists reviewed evidence on sow housing, they did not merely judge the different systems as equivalent, but concluded that the housing systems can provide ‘acceptable’ animal welfare. Here we see scientists acting as active participants in ethically charged issues. Scientists have also played an active role in identifying animal welfare problems that had not previously been recognized. When Matthew Leach and colleagues discovered that carbon dioxide is aversive to rats at concentrations that were much too low to induce unconsciousness, they sounded an alarm by claiming that carbon dioxide killing of rodents is likely to cause unacceptable distress to the animals, and this led to major reconsideration of methods for killing rodents.13 After Paul Hemsworth recognized that poor handling of farm animals can produce a learned fear of humans severe enough to affect basic health and reproduction, he and his co-workers drew attention to this overlooked topic and developed a body of theory and practice for improving animal welfare through better handling.14 Similarly, scientists have helped focus attention on the welfare implications of animal breeding by showing, for example, that cattle bred for very high milk production have increased problems of lameness and metabolic disorders, and that breeding chickens for rapid growth has led to problems of chronic hunger in birds that are kept for breeding.15 In these various cases, scientists have played an important role in expanding animal welfare concerns into areas that had been barely visible in the public debate. Even more fundamentally, scientists have helped to clarify and articulate the nature of public concerns over animal welfare. For example, John Webster played a key role in articulating the ‘Five Freedoms’ which now serve as guiding principles for the housing and handling of animals, and scientists William Russell and Rex Burch proposed the ‘Three R’s’ of Replacement, Reduction and Reﬁnement,

13Leach

et al., 2002.

14Hemsworth, P.H. and Coleman, G.J. 1998. Human–Livestock Interactions: The Stockperson

and the Productivity and Welfare of Intensively-Farmed Animals. CAB International, Oxford. 15Van Zutphen, L.F.M. and Bedford, P.G.C. (editors). 1999. Genetics and Animal Welfare. Animal Welfare (Special Issue) 8: 309–438.

Dfraser_C013.indd 272

6/27/2008 12:45:03 PM

Animal Welfare, Values and Mandated Science

273

which became fundamental ethical principles in laboratory animal use for the past half century.16 Thus, in the case of animal welfare, scientists have taken a strong lead in articulating public concerns and translating these concerns into workable principles that can be pursued in practical ways. In the above examples, we see scientists moving well beyond the role of information providers, and acting instead as informed and engaged participants in ethical and policy issues. Although this may jar the expectations of those who see science as a value-free activity, it seems only natural that scientists who devote their careers to understanding the welfare of animals should also take a leading role in matters of animal welfare policy and ethics. LET US ATTEMPT TO SUMMARIZE how ethics and other value-based considerations inﬂuence science in general and animal welfare science in particular. We began with a debate between those who consider that the scientiﬁc assessment of animal welfare is ‘independent of moral considerations’17 and those who hold that the study of animal welfare is ‘inextricably connected with ethical concerns’.18 We saw that both sides of this debate have valid points: the scientiﬁc study of animal welfare is indeed a different process from making ethical decisions about how animals should be treated, yet ethical considerations do profoundly inﬂuence the scientiﬁc study of animal welfare. The apparent conﬂict can be resolved in part if we divide statements not simply into ‘facts’ and ‘values’, but into three types: ‘facts’, ‘preference values’ and ‘moral values’. In this way we can see that animal welfare science is underlain by preference values – speciﬁcally, by what is considered good or bad, better or worse, for animals. Animal welfare, along with food safety, environmental sustainability and many other topics of scientiﬁc study, can be seen as ‘evaluative concepts’ – concepts that are not merely descriptive but, rather, incorporate notions of good and bad, better and worse. And when scientists investigate evaluative concepts, preference values – what they consider ‘good’ for animals, what they consider ‘safe’ to eat – will guide what variables they choose to study and how they weight different types of evidence. This does not mean that evaluative concepts are merely matters of personal philosophy or opinion; rather, science is highly relevant to understanding these concepts. In this sense, animal welfare and many other evaluative concepts are both science-based and values-based. The idea of a concept being both values-based and science-based may seem foreign to those who view science as value-free investigation. On reﬂection, however, we see that values underlie much of the scientiﬁc enterprise. Values arise 16Russell, W.M.S. and Burch, R.L. 1959. The Principles of Humane Experimental Technique. Methuen, London. 17Broom, 1991, page 4168. 18Tannenbaum, 1991, page 1360.

Dfraser_C013.indd 273

6/27/2008 12:45:03 PM

274

Drawing Conclusions about Animal Welfare

from the vested interests of scientists; scientists apply epistemological values in selecting and interpreting the variables they study; and especially when studying evaluative concepts such as animal welfare, the personal and cultural values of scientists inﬂuence their understanding of the subject and how they study it. When such science is applied, there is often a particularly intricate interplay of ethical and technical elements, and the scientists, rather than being mere providers of information, play active and diverse roles in the associated ethical issues.

Dfraser_C013.indd 274

6/27/2008 12:45:03 PM

Coda

At the end of the 1700s, when the anatomical similarity of humans and other vertebrate animals was well established, the German poet Goethe incorporated this understanding into his cosmological view of animals and nature. Goethe (as historian Rod Preece points out) took issue with the view that humans are the most ‘perfect’ of God’s creations, and that animals should be viewed as lesser beings created for human ends. In his ‘Metamorphosis of Animals’ he wrote: Each animal is an end in itself, it emerges fully formed From Nature’s loins, and produces perfect offspring. All its limbs are formed in accord with eternal laws, And the most unusual form preserves the secret of the primal pattern. Thus is every mouth skilled at seizing food which is Appropriate for the body, even if it is weakly and toothless Or is as powerfully dentured as the jaw; in each case An appropriate organ dispatches the food to the other body parts. As well, each foot, whether long or short, moves In harmony with the purpose of the animal and its needs.1

Goethe’s poem was written roughly sixty years before Darwin’s Origin of the Species, but it clearly captured two ideas that we today associate with the theory of evolution: that different vertebrate animals represent variations on the same ‘primal pattern’, and that each species is uniquely adapted to its speciﬁc manner of living. However, some of Goethe’s contemporaries staunchly defended the view that humans alone represent the pinnacle of the created order, and having seen their ideas undermined by centuries of anatomical research, they produced an 1This translation is from Preece, R. 2002. Awe for the Tiger, Love for the Lamb: A Chronicle of Sensibility to Animals. University of British Columbia Press, Vancouver. The translation is by Preece, and appears on page 181.

275

Dfraser_Coda.indd 275

6/27/2008 12:45:12 PM

276

Coda

anatomical argument of their own. They pointed out that all other mammalian species possess a distinct intermaxillary bone – the bone in the front, central part of the upper jaw which holds the incisors. In the human skeleton, the upper jaw appears to be a solid structure with no separate intermaxillary. The defenders of human uniqueness claimed that this anatomical feature sets humans apart from the rest of creation, and indeed that it makes human speech possible. The challenge posed by this idea was enough to cause Goethe to set down his pen and take up a scalpel in order to prove it wrong. The intermaxillary bone, Goethe determined, does indeed exist in the young human, but it simply becomes fused with the rest of the jaw as the body grows.2 Historians sometimes dispute the scientiﬁc importance of Goethe’s anatomical work, but when we see participants on both sides of the debate about human and animal nature looking to a rather obscure anatomical fact to support their arguments, we appreciate how closely the philosophical debate was linked to science. Today there are few intellectuals like Goethe who contribute to both the sciences and the arts. One recent candidate, however, was the late C.P. Snow – a scientist who wrote novels about scientists. Snow entitled his most controversial essay ‘The two cultures’, and in it he bemoaned the fact (as he saw it) that people engaged in science have little understanding of the arts, and vice versa.3 While this isolation may well exist at the level of the individual scientist or artist, at the broader societal level science is an important part of culture, and it both shapes, and is shaped by, popular beliefs, interests and values. In the relationship between science and questions about the proper treatment of animals, the links are particularly rich. We ﬁrst encountered how, during the rebirth of learning in Europe, the study of anatomy both stimulated and responded to the great public interest in the vertebrate body, and eventually helped to reshape cultural beliefs away from emphasizing the differences between humans and other species, and toward emphasizing the similarities. Thus, when Charles Darwin tried to make sense of the biological diversity he encountered during his years of ﬁeld research, the ground had been well prepared for evolutionary thinking by literary ﬁgures such as Pope and Goethe, and by scientists such as Lamarck and Darwin’s own grandfather, Erasmus Darwin.4 Charles Darwin turned the philosophical idea of continuity among species into a scientiﬁc theory about how this continuity came about. But it would be a mistake to think that the theory of evolution arrived de novo from Darwin’s genius and then transformed popular understanding of animals. More probably, an already

2Goethe,

Johann Wolfgang von. 1786. An intermaxillary bone is present in the upper jaw of man as well as in animals (Jena, 1786). Republished 1995 as pages 111–116 in Goethe – the Collected Works, Volume 12: Scientiﬁc Studies (D. Miller, editor and translator). Princeton University Press, Princeton. 3Snow, C.P. 1965. The Two Cultures: and a Second Look. Cambridge University Press, Cambridge, UK. The passages are on pages 12–15. 4Preece, 1999.

Dfraser_Coda.indd 276

6/27/2008 12:45:12 PM

Coda

277

established change in popular understanding of animals laid the groundwork, both in society and in Darwin’s own mind, which made the theory of evolution by natural selection seem plausible. The tendency, in both science and philosophy, to see a continuity between humans and animals almost certainly contributed to, and reinforced, the interest in animals that blossomed in the arts during the 1700s and 1800s. Many writers and visual artists of these centuries clearly saw animals as beings worthy of artistic and popular attention, and they portrayed sympathetic relations between humans and animals as something worthy of celebration. And this focus on animals and human–animal relations joined with other trends in society – notably the English push to achieve moral progress by stamping out cruelty, and the focus on suffering and pain as the basis for moral action – to turn the ethical treatment of animals into a signiﬁcant social concern. As ethical concern over animals increased, debate arose over the appropriate framework in which such concerns could be expressed. Did concern for animals require an entirely new approach such as a recognition of animal rights, as John Lawrence proposed in 1791, together with a meatless diet as George Nicolson and Joseph Ritson proposed soon after? Or, as the Reverend Thomas Young believed, was it enough simply to cultivate a traditional Biblical ethic of care and stewardship, and to treat animals with traditional Christian virtues of duty and compassion? When concern about animals resurfaced in the second half of the twentieth century, after a hiatus corresponding roughly to the two World Wars and the Great Depression, the key preoccupations of the culture had changed somewhat. The earlier goal of improving the moral tone of society by stamping out cruelty had been replaced by other popular moral ideas. On the one hand were concepts like rights, liberation, and a desire to end discrimination on the basis of race and gender; on the other hand were concerns about suffering, well-being, quality of life, and protection of the vulnerable. The ﬁrst way of formulating the ideas was seized on by philosophers and (usually radical) reformers who framed the issues in terms of animal rights, animal liberation and ‘speciesism’. The second way of formulating the issues opened the door for science to contribute by helping society to understand, assess and improve animal welfare. When scientists undertook these tasks, the questions they asked and the types of data they collected were underlain by their views on what constituted a good or ethically satisfactory life for animals. These views were themselves underlain by attitudes and beliefs that had deep cultural roots. Some scientists approached the study of animal welfare with a set of values typical of a rational, industrial world-view which sees the application of knowledge and technology as means to achieve a better life by overcoming the vicissitudes of nature. For such scientists, a good life for animals must be marked by good health and functioning of the body, and it will lead to high productivity and efﬁcient reproduction. Other scientists reﬂected the values of the Romantic Movement, perhaps in two different ways. Some, taking up the Romantic emphasis on emotion, combined with the Utilitarian

Dfraser_Coda.indd 277

6/27/2008 12:45:12 PM

278

Coda

ethic of Jeremy Bentham, saw animals as emotional beings, and saw a good life for animals as a hedonically pleasant life with a minimum of suffering and at least some opportunity for pleasure and contentment. Others took up the reverence for nature inherent in much Romantic thought, and saw a good life for animals as a life lived close to nature and in accordance with the natures of the animals themselves. In their research on animal welfare, some of the scientists appeared to believe that what they were doing was value-free investigation – simply establishing the facts about animal welfare. If there were disagreements about animal welfare, the resolution lay in more rigorous experiments and more objective evaluation of the data. In reality, some of the disagreements could better be explained by recognizing that the science was itself underlain by different value-based assumptions about what constitutes a good life for animals. And when the science came to be applied in practical ways through various animal welfare practices, policies and standards, these too reﬂected the different value-based assumptions that we see both in animal welfare science and in society at large. If we accept that our scientiﬁc understanding of animal welfare is underlain by the cultural context in which the issues arose, and by value-related beliefs that vary from scientist to scientist and from culture to culture, what are the practical implications? These warrant a great deal of reﬂection, not only in animal welfare science but in mandated science generally. As a small contribution to this process, let me begin the discussion with three points. First, we need to view the scientiﬁc approach to animal welfare as one framework, but not the only framework, that society uses to resolve questions about the proper treatment of animals. In some senses, the scientiﬁc approach competes with other modern approaches that focus on animal rights, liberation and discrimination on the basis of species. For those who view the caging of chimpanzees as a violation of their rights, and for those who see the raising and eating of animals as acts of oppression, scientiﬁc studies of animal welfare may seem (at best) a side issue. The scientiﬁc study of animal welfare also competes with much older approaches to the ancient problems of animal ethics. Rural people who retain a traditional pastoralist ethic of animal care may view all talk of animal ‘welfare’ with suspicion, perhaps as a form of urban cultural imperialism that tries to foist foreign terms and perspectives on an agricultural issue that rural people had long since resolved in a perfectly satisfactory way. For people with an agrarian mindset, a focus on animal welfare may seem a distraction that competes for attention with the real issues of family values, rural lifestyle, and respect for the land, which are themselves sufﬁcient to ensure that animals have satisfactory lives. Certainly the scientiﬁc study of animal welfare makes important and unique contributions to issues of animal ethics. It can be used to indicate and clarify problems, identify trade-offs, evaluate alternatives, develop solutions, and build up an understanding of how life is experienced by animals themselves. But these contributions need to be thoughtfully located within the broad range of ethical approaches to the issues they are meant to address.

Dfraser_Coda.indd 278

6/27/2008 12:45:12 PM

Coda

279

Second, especially as companies and international agencies develop standards for animal welfare that are intended to apply in different parts of the world, we need to be clear about how science and other elements, including value-based, culturally rooted elements, interact in policy and practical decisions. The concept of welfare – of what constitutes a good life – will almost certainly vary from one society to another. In countries where many humans are focussed on obtaining adequate nutrition, shelter and health care, these are likely to be seen as key concerns for animal welfare as well; in countries where such necessities can be taken for granted, concern over animal welfare is likely to extend to other issues such as comfort and mitigation of pain. In cultures that value individual freedom, the idea of keeping animals in stalls and cages where they can barely move may seem abhorent; in cultures that value order and discipline, such practices may evoke a very different response. These various differences in values will inevitably lead to disagreements over what should be included in animal welfare standards. This is not because animal welfare is simply a matter of opinion which varies from culture to culture and has no basis in science, but rather because animal welfare involves many components, some of which will have more resonance in certain cultures than in others. In the debates that follow, it will be important that these value-based disagreements not be mistaken for scientiﬁc disagreements, with each side claiming that their particular proposals uniquely bear the stamp of science and objectivity. Third, in the study of animal welfare as in other ﬁelds of science, we need a much more nuanced understanding of the interplay between ‘facts’ and ‘values’. Some scientists seem to think in terms of a simple dichotomy: that a statement is either factual, objective and scientiﬁc, or else it is a subjective value-judgement arising from personal opinion. Moreover, if an approach (e.g., in studying animal welfare) can be linked to a scientiﬁc theory, or makes use of common scientiﬁc measurements, then they think that approach must be ‘scientiﬁc’ and therefore objective and valuefree. For a more nuanced understanding to develop will require some adjustment to typical science education. If scientists are to make solid contributions to ethical concerns over animal welfare, food safety, climate change, environmental sustainability and other topics that they address in their research, they will need to be educated not only in the technical aspects of the subjects, but also in the cultural, ethical and value-based elements that underlie their science and shape its application. THERE IS, OF COURSE, ENORMOUS potential to apply animal welfare science in practical ways to improve the lives of animals. The research (as we have seen) has identiﬁed ways to prevent injury and death, to reduce the incidence of disease and abnormal behaviour, to reduce ‘stress’ responses, to manage pain, to remove impediments to normal rates of growth and reproduction, and to allow animals to carry out behaviour that they are highly motivated to perform. As with any science, there is often a gap between research and application. We now know, for example, that high humidity and ammonia levels are important

Dfraser_Coda.indd 279

6/27/2008 12:45:12 PM

280

Coda

factors for the survival and growth of chickens; that the welfare of clouded leopards is improved by tall cages located well away from lions and tigers; that individual housing of macaque monkeys in laboratories increases the risk of self-biting; and that successful pain management after surgical dehorning of calves requires analgesics as well as local freezing. But how well this information is put into practice on farms, in laboratories and in zoos around the world is far from clear. Fortunately, there are often practical incentives to improve animal welfare. Ragnar Tauson’s cage designs for hens not only improved the welfare of the birds by preventing injuries and feather loss; they also increased production efﬁciency because well feathered birds do not need to eat extra food to keep warm. As Paul Hemsworth’s research has shown, positive handling of animals not only improves their welfare by eliminating chronic fear, it also leads to greater commercial productivity. Keeping laboratory animals in enriched environments which prevent basal ganglia dysfunction is good not only for the animals, but likely for the validity of experimental results. Controlling rabies in stray dogs in India makes a major contribution to human welfare as well as animal welfare. One of the challenges for scientists in the ﬁeld is to inform people about such practical beneﬁts; but with the small number of scientists and the vast numbers of people who keep animals worldwide, communication is a mammoth challenge. Perhaps (as Andrew Fraser saw in the 1960s) incorporating animal welfare science in the education of veterinary, biological and agricultural students, who will then implement the research and carry the messages to other potential users, is the one promising strategy. But the practical beneﬁts of animal welfare science raise a perplexing question: in a world where science has played such a large role in modern methods of keeping animals in zoos, laboratories and farms, how has it happened that such basic problems as cage-related injuries, stereotypical behaviour, and fear-related reductions in productivity remained under-researched and unsolved for so long? Part of the answer, at least in the case of agricultural animals, may lie in the transition from ‘animal husbandry’ to ‘animal science’. In the early 1900s, veterinary and agricultural students studied a subject known as ‘animal husbandry’ which included the feeding, breeding, handling, management and housing of animals. With the development of the biological sciences during the 1900s, animal husbandry was replaced by what was called ‘animal science’. The feeding of animals became the science of nutrition; the breeding of animals became the sciences of genetics and reproductive physiology. The other elements of animal husbandry – handling, management and housing – were not completely ignored as topics for scientiﬁc research: recall, for example, that David Wood-Gush had already created a research team in poultry ethology in the 1960s. However, the level of research was miniscule compared to the amount devoted to nutrition, genetics and physiology. Indeed, only when the treatment and living conditions of animals became a focus for protest and debate did research into animal welfare begin in earnest. In an important sense, then, animal welfare science is helping to ﬁll the substantial gaps that were left when animal husbandry was replaced

Dfraser_Coda.indd 280

6/27/2008 12:45:12 PM

Coda

281

by animal science. And given the advances that have already been made in nutrition and other ﬁelds, animal welfare science may well provide the next generation of improvements in animal management and production. However, the approach will differ from that used in traditional animal science. Traditionally, animal science has tried to improve the productivity, efﬁciency and proﬁtability of animal production, whether or not the animals themselves beneﬁt from the changes. Thus, for example, animal science research showed that crowding animals to a degree that clearly reduces their welfare may still provide the best return on the producer’s investment in buildings and equipment,5 and that feeding pigs on ﬁnely ground grain can improve the animals’ commercial performance even though it causes them stomach ulcers.6 In genetics, the use of science to achieve rapid gains in commercially important traits like growth rate and egg production has often been at the expense of the animals’ basic health and welfare.7 In these and other cases, traditional animal science has sometimes proven detrimental to the animals themselves. Because animal welfare research focuses ﬁrst and foremost on beneﬁts for the animals – their health, comfort, emotions and motivations – it should provide practical improvements for the raising and handling of animals, but in ways that make a better ﬁt to public concern over animal welfare. IN CHAPTER 3, I PROPOSED that there are four world-views that continue to have a major inﬂuence on how people in the West view animals and what constitutes a good life for animals. Three of these – Pastoralism, Agrarianism and Industrialism – correspond to different socio-economic systems that have played major roles in Western culture. The fourth – Romanticism – arose at least partly as a reaction against the values and type of society associated with Industrialism. There are other important world-views, of course. One is the Ojibway view of animals as highly capable beings whose sympathy for people makes human survival possible. Another is the Jain view that animals and humans are fundamentally similar players in the endless cycles of life and death. Nonetheless, my hypothesis (pending further study) is that the four world-views outlined in Chapter 3 are major drivers of contemporary Western thought about animals and how we should treat them. Does any of these four provide a basis for developing a fully adequate approach to animals and animal ethics in today’s world, and is any of them a satisfactory spring-board for implementing the insights of animal welfare science? Certainly, each world-view has a contribution to make. The Romantic world-view, with its 5Kornegay, E.T. and Notter, D.R. 1984. Effects of ﬂoor space and number of pigs per pen on performance. Pig News and Information 5: 23–33. 6Wondra, K.J., Hancock, J.D., Kennedy, G.A., Hines, R.H. and Behnke, K.C. 1995. Reducing particle size of corn in lactation diets from 1,200 to 400 micrometers improves sow and litter performance. Journal of Animal Science 73: 421–426. 7Rauw, W.M., Kanis, E., Noordhuizen-Stassen, E.N. and Grommers, F.J. 1998. Undesirable side effects of selection for high production efﬁciency in farm animals: A review. Livestock Production Science 56: 15–33.

Dfraser_Coda.indd 281

6/27/2008 12:45:12 PM

282

Coda

emphasis on emotion, nature and personal connection to animals, has much to contribute to the conservation of wild animals and to fostering empathy with animals in homes, but it tends to be viewed with scepticism by many people who raise domestic animals for a living in a competitive market economy. The Pastoralist ethic of care continues to inﬂuence the thinking of many animal caregivers including many farmers, ranchers, veterinarians, and zoo personnel, but it makes an uncomfortable ﬁt to ﬁelds such as commercial slaughter, high-throughput laboratories, and some forms of industrialized animal production. Agrarianism provides an admirable set of values for treating farm animals humanely as part of sustainable food production systems, but its application is mainly limited to agriculture, indeed to that fraction of agriculture where people retain a strong connection to the land. The values of Industrialism, although often reviled as fostering the heartless exploitation of animals, actually provide a basis for addressing such grave animal welfare problems as disease, injury and malnutrition, but they provide little guidance on, for example, wildlife management or the care of animal companions. In short, although all four world-views have a distinct contribution to make, none seems to provide a broad enough basis for addressing the variety of issues to which animal welfare science tries to contribute. However, the last century saw the ﬂourishing of another mode of socioeconomic life. This was the rise of the professions and a sense of what it means to be a professional. Professionalism involves, among other things, acquiring the requisite knowledge and skills through training and supervision, recognizing (and sometimes helping to shape) accepted standards of professional behaviour, conforming to those standards and helping the profession to ensure that all its members do the same. Professionalism also entails keeping abreast of new developments in the ﬁeld, and implementing those developments in practice. Perhaps the values associated with Professionalism, if applied to animal care activities in agriculture, biomedical research, wildlife management and other ﬁelds, would provide the most promising basis for a practical ethic of animal use and for translating animal welfare science into action. I PROPOSED IN PART I that three major concerns arise in discussions of animal welfare: that animals should be healthy and thriving, that they should be able to lead reasonably natural lives, and that they should be spared signiﬁcant suffering and be able to enjoy life. Part II then showed how science has ampliﬁed and clariﬁed these ideas, and showed the care that must be used when these ideas are translated into practical ways of housing and managing animals. Let us end by brieﬂy reviewing the different scientiﬁc approaches and the understanding of animal welfare that they provide. Measures based on disease or injury are perhaps the most straightforward type of evidence. By deﬁnition, disease or injury denote poor health and functioning. We also expect that many injuries will involve pain, and that many illnesses and infections will lead to feelings of sickness perhaps mediated by release of cytokines.

Dfraser_Coda.indd 282

6/27/2008 12:45:13 PM

Coda

283

Therefore, disease and injury are likely to involve unpleasant affective states. Moreover, disease and injury can prevent animals from living their lives in the way they would desire. Few scientiﬁc passages are more heart-rending than Jane Goodall’s description of the chimpanzee McGregor who, after being afﬂicted with polio, could no longer participate in the normal social behaviour of his group.8 In summary, then, increased incidence of disease or injury is easy to classify as an important indicator of animal welfare because it relates to at least one type of animal welfare concern (basic health and functioning) and potentially to all three. In many countries of the world, the most obvious animal welfare problems involve basic issues of disease, injury, parasitism, malnutrition and dehydration. Animal welfare science cannot ignore such issues. Measures based on the productivity of animals (growth rate, rate of laying eggs, milk yield) may seem, on the surface, to also rank as measures of how well the body is functioning, but their interpretation is much less straightforward. In some cases, declines in productivity measures may indeed indicate problems in basic health such as illness or injury, or that the animal is experiencing fear or distress. In other cases reduced production may indicate that the animal is receiving an unsuitable diet or is unable to eat in its normal manner. In such cases, declines in productivity may well serve as indicators of animal welfare. On the other hand, high rates of growth, egg laying, or milk production may simply be the result of strong genetic selection for speciﬁc types of performance that are achieved at the expense of other aspects of bodily functioning. Very high milk yield in cattle is often correlated with a high incidence of diseases like mastitis; very high rates of weight gain by broiler chickens are often associated with lameness because the skeletal development cannot keep pace with the growth of muscle. In these cases, high levels of production may bear little relation to the health and welfare of the animal. The idea that animals should not be caused undue ‘stress’ – deﬁned in everyday speech as ‘physical, mental or emotional strain or tension’ – has played a key role in attempts to understand and improve animal welfare. Thanks to a century of research on the ‘stress response systems’ of the body, we now have some remarkable tools for understanding how the body responds to physical challenges and emotional states. In particular, increased activation of the HPA system, as evidenced by high production of glucocorticoids such as cortisol, is widely used as an indicator of animal welfare, but the interpretation of such measures is actually not straightforward. In the clouded leopard example in Chapter 6, the basic welfare problem appeared to be a kind of chronic agitation among leopards that were conﬁned in fear-producing situations (close to predators, exposed to unfamiliar people) but could not use their natural behaviour (climbing and perching above ground level) to escape. In this case, activation of the HPA system appeared to correlate with on-going distress and thus could be seen as a welfare indicator. Likewise, if an event such as dehorning is suspected to cause pain, a sharp increase in 8Goodall,

Dfraser_Coda.indd 283

1971. The passage is on pages 213–220.

6/27/2008 12:45:13 PM

284

Coda

cortisol production may conﬁrm and help to quantify the level of pain. Moreover, a sustained period of high cortisol secretion may lead to reduced immune competence or interfere with growth and reproduction and thus lead to problems of basic health and functioning. In these various cases, high levels of cortisol secretion (or other indicators of increased activity of the stress response systems) could be viewed as indicators of impaired welfare. In the absence of such logic, however, we cannot assume that different levels of activation of the stress response systems have any bearing on animal welfare. Higher cortisol production may simply mean that the animals are in cooler temperatures and are compensating by eating and metabolizing more food, or that they are more active, for example because they are exercising or playing more. Thus, if we want to treat activation of the stress response systems as a welfare indicator, we need to be clear on how we believe it signals some difference in basic health and functioning, in affective states, or in the animal’s ability to live in the manner to which it is adapted. Without such links, we cannot assume that changes in the activation of these systems are related to animal welfare. Abnormal behaviour, especially dysfunctional or repetitive behaviour, has played a large role in people’s thinking about the well-being of both people and animals, and there are various ways in which abnormal behaviour may relate to animal welfare. As we have seen, for example, persistent wire-chewing by voles may reﬂect an abnormality in the central nervous system analogous to abnormalities seen in autism and Parkinson’s disease, and repetitive chain-chewing by sows can indicate chronic hunger. In these cases, the link between the abnormal behaviour and basic health or affective states allows us to argue that the abnormal behaviour is a welfare indicator. On the other hand, persistent performance of an odd behaviour may be a habit that the animal formed in a previous environment; or the behaviour may be a way of reducing arousal, and the absence of the behaviour may simply indicate that the animal has not yet learned to use the behaviour in arousing situations. As with activation of stress response systems, if we want to claim that the presence or absence of abnormal behaviour has some bearing on animal welfare, we need to be clear on the links. The affective states of animals play a central role in concerns about animal welfare. For example, animal protectionists often see their goal as preventing ‘suffering’, and laws commonly make it an offence to cause animals unnecessary ‘pain and distress’. Science has helped to apply these ideas. Some affective states (separation distress, hunger in dependent young) have corresponding signals which we can identify and then use to monitor these states. In other cases (fear, pain, frustration) there are reliable behavioural or physiological signs that commonly accompany the affective state, at least for a particular species in a particular type of situation. With appropriate research we can devise ways to quantify these states and compare the effectiveness of mitigative actions. However, other states that may be relevant to animal welfare (boredom, pleasure) have attracted much less scientiﬁc attention and may prove much more difﬁcult to monitor and quantify.

Dfraser_Coda.indd 284

6/27/2008 12:45:13 PM

Coda

285

And several decades when the behavioural sciences were strongly inﬂuenced by Positivism have left the scientiﬁc study of affect somewhat under-developed, without the body of theoretical and methodological advances that animal welfare scientists could then simply apply to practical problems. The result is a somewhat patchy body of science, where some states (notably pain) can often be monitored through well researched methods, while others cannot. Especially in the industrialized countries, where animals are commonly kept in highly restrictive environments on farms or in laboratories, the unnaturalness of animals’ living conditions has become a prominent animal welfare concern. Indeed, there are many animal welfare problems that arise from keeping animals under unnatural conditions: health problems resulting from polluted air or unnatural diets, lameness and other injuries resulting from unnatural walking and lying surfaces, frustration and behavioural abnormalities when natural behaviour is prevented, and even a failure of normal social and neurological development if young animals are kept in impoverished environments. Many such problems can be solved by restoring animals to more natural conditions: fresh air, diets suited to their digestive systems, environments that permit young animals to play, and so on. And often science and human ingenuity can identify practical ways to modify artiﬁcial environments to incorporate important elements of the animals’ natural lives. Some critics appear to idealize ‘natural’ conditions by assuming that animals kept in such situations will inevitably be healthier and happier than those kept under artiﬁcial conditions. In fact, with harsh weather, predators, parasites and other challenges, ‘natural’ living conditions can be a two-edged sword. Here science has an important role to play in assembling the data to disentangle the factors and understand how different rearing conditions actually affect animals. There is also a need to disentangle what genuinely ﬁts with the adaptations of animals versus conditions that merely appear natural to the human on-looker. Probably no amount of science will completely resolve the disagreements: there will likely always be people who resolutely hold that a good life must be a natural life even if there is some cost in terms of health and comfort. However, we can certainly do better at raising animals in ways that ﬁt their adaptations while still providing them with health and comfort beneﬁts that they would not enjoy in nature. On the surface, allowing animals to choose for themselves between different options would seem a kind of royal road to understanding and improving animal welfare, and certainly the preferences, motivations and aversions of animals are important sources of information. They can be used, for example, to determine what animals regard as a comfortable temperature or ﬂooring product, how strongly animals are motivated to carry out certain types of behaviour such as nesting before laying an egg, and whether they ﬁnd a management practice aversive. There are, however, many important cautions that need to be observed when using such methods. These include controlling for the animals’ previous experience and providing choices that ﬁt within the species’ evolved perceptual and cognitive ability so that the choices made will in fact correlate with beneﬁcial outcomes.

Dfraser_Coda.indd 285

6/27/2008 12:45:13 PM

286

Coda

Determining the strength of the animal’s motivation can greatly improve the value of the results, but great care is needed in selecting and interpreting measures of motivation strength. What, then, do we understand by animal welfare, after viewing the question through the lens of scientiﬁc knowledge and research? An animal has a good life if it is healthy and thriving, without distortions (genetic, hormonal, dietary or other) that enhance one aspect of its bodily functioning to the detriment of others; if its behavioural and physiological systems are not pushed to such extremes that there is a breakdown, or signiﬁcant risk of breakdown, in health or development; if it can enjoy life, and if negative states – pain, fear, frustration and others – are not so severe or prolonged as to constitute suffering; if it is free to live under circumstances that it might itself choose; and if it is not prevented from doing those things that it is strongly motivated to do. In this prescription I believe we see the major values that underlie our understanding of animal welfare interpreted in light of the knowledge and analysis that animal welfare science provides.

Dfraser_Coda.indd 286

6/27/2008 12:45:13 PM

References

Adams, A.W. and Craig, J.V. 1985. Effect of crowding and cage shape on productivity and proﬁtability of caged layers: A survey. Poultry Science 64: 238–242. Adams, G.J. and Johnson, K.G. 1994. Sleep, work, and the effects of shift work in drug detector dogs Canis familiaris. Applied Animal Behaviour Science 41: 115–126. Albin, R.L., Young, A.B. and Penney, J.B. 1989. The functional anatomy of basal ganglia disorders. Trends in Neurosciences 12: 366–375. Alroe, H.F., Vaarst, M. and Kristensen, E.S. 2001. Does organic farming face distinctive livestock welfare issues? A conceptual analysis. Journal of Agricultural and Environmental Ethics 14: 275–299. Altmann, J. 1980. Baboon Mothers and Infants. Republished 2001, University of Chicago Press, Chicago. Anonymous, 1989. How Astrid Lindgren Achieved Enactment of the 1988 Law Protecting Farm Animals in Sweden. Animal Welfare Institute, Washington. Arey, D.S., Petchey, A.M. and Fowler, V.R. 1991. The preparturient behaviour of sows in enriched pens and the effect of pre-formed nests. Applied Animal Behaviour Science 31: 61–68. Appleby, M.C. 1999. What Should We Do about Animal Welfare? Blackwell Science, Oxford. Appleby, M.C. and Hughes B.O. (editors). 1997. Animal Welfare. CAB International, Wallingford, UK. Appleby, M.C. and Lawrence, A.B. 1987. Food restriction as a cause of stereotypic behaviour in tethered gilts. Animal Production 45: 103–110. Appleby, M.C. and Sandøe, P. 2002. Philosophical debate on the nature of well-being: Implications for animal welfare. Animal Welfare 11: 283–294. Appleby, M.C., Weary, D.M. and Chua, B. 2001. Performance and feeding behaviour of calves on ad libitum milk from artiﬁcial teats. Applied Animal Behaviour Science 74: 191–201. Atkinson, E. 1912. Greyfriar’s Bobby. Harper and Brothers, New York. Baenninger, R. 1990. Consciousness and comparative psychology. Pages 249–269 in Reﬂections on the Principles of Psychology: William James after a Century (M.G. Johnson and T.B. Henley, editors). Erlbaum, Hillsdale, USA.

287

Dfraser_References.indd 287

6/27/2008 12:45:37 PM

288

References

Barnard, C.J. and Hurst, J.L. 1996. Welfare by design: The natural selection of welfare criteria. Animal Welfare 5: 405–433. Barnett, J.L. and Hemsworth, P.H. 1990. The validity of physiological and behavioural measures of animal welfare. Applied Animal Behaviour Science 25: 177–187. Barnett, J.L., Hemsworth, P.H., Cronin, G.M., Jongman, E.C. and Hutson, G.D. 2001. A review of the welfare issues for sows and piglets in relation to housing. Australian Journal of Agricultural Research 52: 1–28. Barnett, J.L., Hemsworth, P.H. and Winﬁeld, C.G. 1987. The effects of design of individual stalls on the social behaviour and physiological responses related to the welfare of pregnant pigs. Applied Animal Behaviour Science 18: 133–142. Barnett, J.L., Winﬁeld, C.G., Cronin, G.M., Hemsworth, P.H. and Dewar, A.M. 1985. The effect of individual and group housing on behavioural and physiological responses related to the welfare of pregnant pigs. Applied Animal Behaviour Science 14: 149–161. Bartussek, H. 1999. A review of the animal needs index (ANI) for the assessment of animals’ well-being in the housing systems for Austrian proprietary products and legislation. Livestock Production Science 61: 179–192. Bateson, P. 1997. The Behavioural and Physiological Effects of Culling Red Deer. Report to the Council of the National Trust, unpublished. Bateson, P. and Bradshaw, E.L. 2000. Physiological effects of hunting red deer (Cervus elaphas). Proceedings of the Royal Society, London, B 264: 1707–1714. Bauman, D.E. 1992. Bovine somatotropin: Review of an emerging animal technology. Journal of Dairy Science 75: 3432–3451. Baumans, V., Schlingmann, F., Vonck, M. and van Lith, H.A. 2002. Individually ventilated cages: Beneﬁcial for mice and men? Contemporary Topics in Laboratory Animal Science 41: 13–19. Baxter, M.R. 1983. Ethology in environmental design for animal production. Applied Animal Ethology 9: 207–220. Baxter, S. 1984. Intensive Pig Production: Environmental Management and Design. Granada, London. Bayvel, A.C.D., Rahman, S.A. and Gavinelli, A. (editors). 2005. Animal Welfare: Global Issues, Trends and Challenges. Special issue, Revue Scientiﬁque et Technique de l’Ofﬁce International des Epizooties 24: 463–813. Bell, D., Chase, B., Douglass, A., Hester, P., Mench, J., Newberry, R., Shea-Moore, M., Stanker, L., Swanson, J. and Armstrong, J. 2004. UEP uses scientiﬁc approach in its establishment of welfare guidelines. Feedstuffs 76(11): 1–9. Bentham, J. 1789. Introduction to the Principles of Morals and Legislation. Republished 1961 as pages 5–398 in The Utilitarians. Dolphin Books, Garden City, USA. Berridge, K.C. 1996. Food reward: Brain substrates of wanting and liking. Neuroscience and Biobehavioral Reviews 20: 1–25. Berridge, K.C. 1999. Pleasure, pain, desire, and dread: Hidden core processes of emotion. Pages 525–557 in Well-Being: The Foundations of Hedonic Psychology (D. Kahneman, E. Diener and N. Schwarz, editors). Russell Sage Foundation, New York.

Dfraser_References.indd 288

6/27/2008 12:45:38 PM

References

289

Beynen, A.C., Baumans, V., Bertens, A.P.M.G., Havenaar, R., Hesp, A.P.M. and van Zutphen, L.F.M. 1987. Assessment of discomfort in gallstone-bearing mice: A practical example of the problems encountered in an attempt to recognize discomfort in laboratory animals. Laboratory Animals 21: 35–42. Bierens de Haan, J.A. 1947. Animal psychology and the science of animal behaviour. Behaviour 1: 71–80. Bizup, J. 2003. Manufacturing Culture: Vindications of Early Victorian Industry. University of Virginia Press, Charlottesville. Blake, W. 1794. The Fly. Republished 1989 as pages 220–221 in Blake: The Complete Poems, Second Edition (W.H. Stevenson, editor). Longman, London. Blake, W. about 1805. Auguries of Innocence. Republished 1989 as pages 589–592 in Blake: The Complete Poems, Second Edition (W.H. Stevenson, editor). Longman, London. Blum, D. 2002. Love at Goon Park: Harry Harlow and the Science of Affection. Perseus Publishing, Cambridge, USA. Bodson, L. 2000. Motivations for pet-keeping in ancient Greece and Rome: A preliminary survey. Pages 27–41 in Companion Animals and Us (A.L. Podberscek, E.S. Paul and J.A. Serpell, editors). Cambridge University Press, Cambridge. Bracke, M.B.M., Metz, J.H.M., Spruijt, B.M. and Schouten, W.G.P. 2002. Decision support system for overall welfare assessment in pregnant sows B: Validation by expert opinion. Journal of Animal Science 80: 1835–1845. Bradshaw, E.L. and Bateson, P. 2000. Welfare implications of culling Red Deer (Cervus elaphus). Animal Welfare 9: 3–24. Brambell, F.W.R. (chairman). 1965. Report of the Technical Committee to Enquire into the Welfare of Animals kept under Intensive Livestock Husbandry Systems. Her Majesty’s Stationery Ofﬁce, London. Bratton, S.P. 1984. Christian ecotheology and the Old Testament. Environmental Ethics 6: 195–209. Brock, D. 1993. Quality of life measures in health care and medical ethics. Pages 95–132 in The Quality of Life (M. Nussbaum and A. Sen, editors). Oxford University Press, Oxford. Broom, D.M. 1991. Animal welfare: Concepts and measurement. Journal of Animal Science 69: 4167–4175. Broom, D.M. 1998. Welfare, stress, and the evolution of feelings. Advances in the Study of Behavior 27: 371–403. Brown, K.H. and Hollingsworth, J. 2005. The Food Marketing Institute and the National Council of Chain Restaurants: Animal welfare and the retail food industry in the United States of America. Revue Scientiﬁque et Technique de l’Ofﬁce International des Epizooties 24: 655–663. Brunk, C.G., Haworth, L. and Lee, B. 1991. Value Assumptions in Risk Assessment: A Case Study of the Alachlor Controversy. Wilfrid Laurier University Press, Waterloo, Canada. Burgdorf, J. and Panksepp, J. 2001. Tickling induces reward in adolescent rats. Physiology & Behavior 72: 25–38. Burkhardt, R.W. Jr. 1997. The founders of ethology and the problem of animal subjective experience. Pages 1–13 in Animal Consciousness and Animal Ethics (M. Dol,

Dfraser_References.indd 289

6/27/2008 12:45:38 PM

290

References

S. Kasanmoentalib, S. Lijmbach, E. Rivas and R. van den Bos, editors). Van Gorcum, Assen, The Netherlands. Burnet, J. 1930. Early Greek Philosophy, 4th Edition. Adam and Charles Black, London. Burns, R. 1786. Poems, Chieﬂy in the Scottish Dialect. Republished 1955 in Poems and Songs of Robert Burns (J. Barke, editor). Collins, London. Cabanac, M. 1992. Pleasure: The common currency. Journal of Theoretical Biology 155: 173–200. Campbell, N. 2004. What was Huxley’s epiphenomenalism? Biology and Philosophy 16: 357–375. Canadian Veterinary Medical Association. 1998. Report of the Canadian Veterinary Medical Association Expert Panel on rbST. Health Canada, Ottawa. Cannon, W.B. 1929. Bodily Changes in Pain, Hunger, Fear and Rage, 2nd Edition. AppletonCentury Co, New York. Carpenter, E. 1980. Animals and Ethics. A Report of the Working Party Convened by Edward Carpenter. Watkins and Dulverton, London. Carson, R. 1962. Silent Spring. Houghton Mifﬂin, Boston. Cavalieri, P. and Singer, P. (editors). 1993. The Great Ape Project: Equality beyond Humanity. St. Martin’s Press, New York. Celli, M.L., Tomonaga, M., Udono, T., Teramoto, M. and Nagano, K. 2003. Tool use task as environmental enrichment for captive chimpanzees. Applied Animal Behaviour Science 81: 171–182. Chapple, C.K. 2006. Inherent value without nostalgia: Animals and the Jaina tradition. Pages 241–249 in A Communion of Subjects: Animals in Religion, Science and Ethics (P. Waldau and K. Patton, editors). Columbia University Press, New York. Cheney, D.L. and Seyfarth, R.M. 1990. How Monkeys See the World: Inside the Mind of Another Species. University of Chicago Press, Chicago. Chrousos, G.P. 1998. Stressors, stress and neuroendocrine integration of the adaptive response. The 1997 Hans Selye Memorial Lecture. Annals of the New York Academy of Sciences 851: 311–335. Colpaert, F.C., Tarayre, J.P., Alliaga, M., Slot, L.A.B., Attal, N. and Koek, W. 2001. Opiate self-administration as a measure of chronic nociceptive pain in arthritic rats. Pain 91: 33–45. Companion Animal Working Group 2006. Breeding and Welfare in Companion Animals. Companion Animal Welfare Council, Sidmouth, UK. Concise Oxford Dictionary of Current English, 7th Edition, 1982. Oxford University Press, Oxford. Connor, M.L. (chairperson). 1993. Recommended Code of Practice for the Care and Handling of Farm Animals: Pigs. Agriculture and Agri-Food Canada, Ottawa. Cooper, J.J. and Appleby, M.C. 1996. Demand for nest boxes in laying hens. Behavioural Processes 36: 171–182. Cottingham, J. 1993. A Descartes Dictionary. Blackwell, Oxford. Council for Agricultural Science and Technology (CAST). 1981. Scientiﬁc Aspects of the Welfare of Food Animals. Report 91. CAST, Ames, USA.

Dfraser_References.indd 290

6/27/2008 12:45:38 PM

References

291

Council for Agricultural Science and Technology (CAST). 1997. The Well-Being of Agricultural Animals. Report 130. CAST, Ames, USA. Cowper, W. 1785. The Task, Book VI. Republished 1994 as The Task, and Selected Other Poems (J. Sambrook, editor). Longman, New York. Cox, B. and Bilkei, G. 2004. Lifetime reproductive performance of sows kept indoors and outdoors in Croatia. Veterinary Record 154: 569–570. Cronin, G.M. 1985. The Development and Signiﬁcance of Abnormal Stereotyped Behaviours in Tethered Sows. Doctoral thesis, Agricultural University of Wageningen, Wageningen, The Netherlands. Curtis, S.E. 1987. Animal well-being and animal care. Veterinary Clinics of North America: Food Animal Practice 3: 369–382. Dahrendorf, R. 1987. Max Weber and modern social science. Pages 574–580 in Max Weber and His Contemporaries (W.J. Mommsen and J. Osterhammel, editors). Allen & Unwin, London. Dantzer, R. 2001. Cytokine-induced sickness behavior: Where do we stand? Brain, Behavior, and Immunity 15: 7–24. Dantzer, R. and Mormède, P. 1983a. Stress in farm animals: A need for reevaluation. Journal of Animal Science 57: 6–18. Dantzer, R. and Mormède, P. 1983b. De-arousal properties of stereotyped behaviour: Evidence from pituitary–adrenal correlates in pigs. Applied Animal Ethology 10: 233–244. Darwin, C. 1872. The Expression of the Emotions in Man and Animals. Republished 1965, University of Chicago Press, Chicago. Darwin, E. 1796. Zoonomia; Or, the Laws of Organic Life, Volume 1. Republished 1801, J. Johnson, London. Davidson, A.B. 1995. It isn’t all black and white. Pages 25–30 in Decision Making and Agriculture: The Role of Ethics (K.B. Beesley, S. Burns, M. Campbell and P. Sanger, editors). Rural Research Centre, Truro, Canada. Dawkins, M.S. 1977. Do hens suffer in battery cages? Environmental preferences and welfare. Animal Behaviour 25: 1034–1046. Dawkins, M.S. 1980. Animal Suffering: The Science of Animal Welfare. Chapman & Hall, London. Dawkins, M.S. 1983. Battery hens name their price: Consumer demand theory and the measurement of ethological ‘needs.’ Animal Behaviour 31: 1195–1205. Dawkins, M.S. 1988. Behavioural deprivation: A central problem in animal welfare. Applied Animal Behaviour Science 20: 209–225. Dawkins, M.S. 1990. From an animal’s point of view: Motivation, ﬁtness, and animal welfare. Behavioural and Brain Sciences 13: 1–9 and 54–61. Dawkins, M.S. 1998. Evolution and animal welfare. Quarterly Review of Biology 73: 305–328. Dawkins, M.S. 2004. Using behaviour to assess animal welfare. Animal Welfare 13: S3–S7. Dawkins, M.S. and Beardsley, T. 1986. Reinforcing properties of access to litter in hens. Applied Animal Behaviour Science 15: 351–364.

Dfraser_References.indd 291

6/27/2008 12:45:38 PM

292

References

Dawkins M.S., Donnelly, C.A. and Jones, T.A. 2004. Chicken welfare is inﬂuenced more by housing conditions than by stocking density. Nature 427: 342–344. de Passillé, A.M.B., Christopherson, R. and Rushen, J. 1993. Nonnutritive sucking by the calf and postprandial secretion of insulin, CCK, and gastrin. Physiology & Behavior 54: 1069–1073. Descartes, R. 1637. Discourse on the method for properly conducting reason and searching for truth in the sciences. Republished 1984–1991 in The Philosophical Writings of Descartes, Vol. 1 (J. Cottingham, R. Stoothoff, D. Murdoch and A. Kenny, editors and translators). Cambridge University Press, Cambridge. Dobel, P. 1977. The Judeo-Christian stewardship attitude toward nature. Reprinted 1998 as pages 26–30 in Environmental Ethics: Readings in Theory and Application, Second Edition (L. Pojman, editor). Wadsworth Publishing Company, Belmont, USA. Dohoo, I.R., DesCôteaux, L., Leslie, K., Fredeen, A., Shewfelt, W., Preston, A. and Dowling, P. 2003. A meta-analysis review of the effects of recombinant bovine somatotrophin. 2. Effects on animal health, reproductive performance, and culling. Canadian Journal of Veterinary Research 67: 252–264. Duncan, I.J.H. 1970. Frustration in the fowl. Pages 15–31 in Aspects of Poultry Behaviour (B.M. Freeman and R.F. Gordon, editors). British Poultry Science Ltd., Edinburgh. Duncan, I.J.H. 1978. The interpretation of preference tests in animal behaviour. Applied Animal Ethology 4: 197–200. Duncan, I.J.H. 1987. The welfare of farm animals: An ethological approach. Science Progress, Oxford 71: 317–326. Duncan, I.J.H. 1992. Behavioral assessment of welfare. Pages 62–68 in The Well-being of Agricultural Animals in Biomedical and Agricultural Research (J.A. Mench, S.J. Mayer and L. Krulisch, editors). Scientists’ Center for Animal Welfare, Washington. Duncan, I.J.H. 1993. Welfare is to do with what animals feel. Journal of Agricultural and Environmental Ethics 6 (Suppl 2): 8–14. Duncan, I.J.H. 1996. Animal welfare deﬁned in terms of feelings. Acta Agriculturae Scandinavica, Section A, Animal Science, Suppl 27: 29–35. Duncan, I.J.H. 1998. Behavior and behavioral needs. Poultry Science 77: 1766–1772. Duncan, I.J.H. 2001. Animal welfare issues in the poultry industry: Is there a lesson to be learned? Journal of Applied Animal Welfare Science 4: 207–221. Duncan, I.J.H., Slee, G., Kettlewell, P., Berry, P. and Carlisle, A.J. 1986. Comparison of the stressfulness of harvesting broiler chickens by machine and by hand. British Poultry Science 27: 109–114. Dunlap, T.R. 1988. Saving America’s Wildlife: Ecology and the American Mind, 1850–1990. Princeton University Press, Princeton. Edinboro, C.H., Scott-Moncrieff, J.C., Janovitz, E., Thacker, H.L. and Glickman, L.T. 2004. Epidemiologic study of relationships between consumption of commercial canned food and risk of hyperthyroidism in cats. Journal of the American Veterinary Medical Association 224: 879–886. Edsall, J.T. 1975. Scientiﬁc Freedom and Responsibility: A Report of the AAAS Committee on Scientiﬁc Freedom and Responsibility. American Association for the Advancement of Science, Washington.

Dfraser_References.indd 292

6/27/2008 12:45:38 PM

References

293

Edwards, S.A. 1995. Designing systems to meet behavioral needs: The Family Pen system for pigs. Pages 115–125 in Animal Behavior and the Design of Livestock and Poultry Systems. Northeast Regional Agricultural Engineering Service, Ithaca, USA. Ekesbo, I. 1966. Disease incidence in tied and loose housed dairy cattle and causes. Acta Agriculturae Scandinavica 15 (Suppl): 1–74. Elgar, M.A. 1989. Predator vigilance and group size in mammals and birds: A critical review of the empirical evidence. Biological Reviews 64: 13–33. Erwin, J. and Deni, R. 1979. Strangers in a strange land: Abnormal behaviors or abnormal environments? Pages 1–28 in Captivity and Behavior: Primates in Breeding Colonies, Laboratories, and Zoos (J. Erwin, T.L. Maple and G. Mitchell, editors). Van Nostrand Reinhold, New York. Ewer, R.F. 1973. The Carnivores. Weidenfeld and Nicolson, London. Falk, J.L. 1961. Production of polydipsia in normal rats by an intermittent food schedule. Science 133: 195–196. Falk, J.L. 1971. The nature and determinants of adjunctive behavior. Physiology & Behavior 6: 577–588. Feddes, J.J.R. and Fraser, D. 1994. Non-nutritive chewing by pigs: Implications for tail-biting and behavioral enrichment. Transactions of the American Society of Agricultural Engineers 37: 947–950. Feigl, H. 1995. Positivism and logical empiricism. Pages 630–636 in The New Encyclopedia Britannica, 15th Edition, Volume 25. Encyclopedia Britannica Inc., Chicago. Fox, M.W. (editor). 1968. Abnormal Behavior in Animals. W.B. Saunders Company, Philadelphia. Fraser, A.F. 1959. Displacement activities in domestic animals. British Veterinary Journal 115: 195–200. Fraser, A.F. and Broom, D.M. 1997. Farm Animal Behaviour and Welfare, 3rd Edition. CAB International, Wallingford, UK. Fraser, D. 1975. Vocalizations of isolated piglets. I. Sources of variation and relationships among measures. Applied Animal Ethology 1: 387–394. Fraser, D. 1985. Selection of bedded and unbedded areas by pigs in relation to environmental temperature and behaviour. Applied Animal Behaviour Science 14: 117–126. Fraser, D. 1995. Science, values and animal welfare: Exploring the “inextricable connection”. Animal Welfare 4: 103–117. Fraser, D. 1999. Animal ethics and animal welfare science: Bridging the two cultures (The D.G.M. Wood-Gush Memorial Lecture). Applied Animal Behaviour Science 65: 171–189. Fraser, D. 2001. The ‘new perception’ of animal agriculture: Legless cows, featherless chickens, and a need for genuine analysis. Journal of Animal Science 79: 634–641. Fraser, D. 2003. Assessing animal welfare at the farm and group level: The interplay of science and values. Animal Welfare 12: 433–443. Fraser, D. 2005. Animal Welfare and the Intensiﬁcation of Animal Production: An Alternative Interpretation. FAO Readings in Ethics No. 2, Food and Agriculture Organization of the United Nations, Rome.

Dfraser_References.indd 293

6/27/2008 12:45:38 PM

294

References

Fraser, D. 2006a. Caring for farm animals: Pastoralist ideals in an industrialized world. Pages 547–555 in A Communion of Subjects: Animals in Religion, Science and Ethics (P. Waldau and K. Patton, editors). Columbia University Press, New York. Fraser, D. 2006b. Animal welfare assurance programs in food production: A framework for assessing the options. Animal Welfare 15: 93–104. Fraser, D., Bernon, D.E. and Ball, R.O. 1991. Enhanced attraction to blood by pigs with inadequate dietary protein supplementation. Canadian Journal of Animal Science 71: 611–619. Fraser, D. and Duncan, I.J.H. 1998. ‘Pleasures’, ‘pains’ and animal welfare: Toward a natural history of affect. Animal Welfare 7: 383–396. Fraser, D., Friendship, R.M. and Martineau, G.-P. 1994. Aristotle on pigs: Husbandry, health and natural history of pigs in ancient Greece. Pig News and Information 15: 77N–80N. Fraser, D. and Matthews, L.R. 1997. Preference and motivation testing. Pages 159–173 in Animal Welfare (M.C. Appleby and B.O. Hughes, editors). CAB International, Wallingford, UK. Fraser, D. and Preece, R. 2004. Animal ethics and the scientiﬁc study of animals: Bridging the ‘is’ and the ‘ought’. Essays in Philosophy 5, available at: http://sorrel.humboldt. edu/~essays/fraser.html). Fraser, D. and Waddell, M.S. 1974. The importance of social and self-grooming for the control of ectoparasitic mites on normal and dystrophic laboratory mice. Laboratory Practice 23: 58–59. Fraser, D. and Weary, D.M. 2005. Applied animal behavior and animal welfare. Pages 345–366 in The Behavior of Animals: Mechanisms, Function and Evolution (J.J. Bolhuis and L.-A. Giraldeau, editors). Blackwell Scientiﬁc, Oxford. Fraser, D., Weary, D.M., Pajor, E.A. and Milligan, B.N. 1997. A scientiﬁc conception of animal welfare that reﬂects ethical concerns. Animal Welfare 6: 187–205. Gallup, G. 1981. Hypnosis. Pages 294–297 in The Oxford Companion to Animal Behaviour (D. McFarland, editor). Oxford University Press, Oxford. Garner, J.P. and Mason, G.J. 2002. Evidence for a relationship between cage stereotypies and behavioural disinhibition in laboratory rodents. Behavioural Brain Research 136: 83–92. Gaskin, J.C.A. 1995. Epicureanism. Pages 239–240 in The Oxford Companion to Philosophy (T. Honderich, editor). Oxford University Press, Oxford. Gentle, M.J. 2001. Attentional shifts alter pain perception in the chicken. Animal Welfare 10: S187–S194. Giattina, J.D. and Garton, R.R. 1983. A review of the preference–avoidance responses of ﬁshes to aquatic contaminants. Residue Reviews 87: 43–90. Goethe, J.W. von. 1806. Metamorphosis of animals. Republished 1994 as pages 161–163 in Goethe – the Collected Works, Volume 1: Selected Poems (C. Middleton, editor). Princeton University Press, Princeton. Goethe, J.W. von. 1786. An intermaxillary bone is present in the upper jaw of man as well as in animals (Jena, 1786). Republished 1995 as pages 111–116 in Goethe – the Collected

Dfraser_References.indd 294

6/27/2008 12:45:38 PM

References

295

Works, Volume 12: Scientiﬁc Studies (D. Miller, editor and translator). Princeton University Press, Princeton. Goldsmith, O. 1770. The Deserted Village, A Poem. N. Douglas, London. Goodall, J. 1971. In the Shadow of Man. William Collins, London. Gourkow, N. and Fraser, D. 2006. The effect of housing and handling practices on the welfare, behaviour and selection of domestic cats (Felis sylvestris catus) by adopters in an animal shelter. Animal Welfare 15: 371–377. Grandin, T. 1998a. The feasibility of using vocalization scoring as an indicator of poor welfare during cattle slaughter. Applied Animal Behaviour Science 56: 121–128. Grandin, T. 1998b. Objective scoring of animal handling and stunning practices at slaughter plants. Journal of the American Veterinary Medical Association 212: 36–39. Grandin, T. 2001. Cattle Slaughter Audit Form (Updated October 2001). Based on American Meat Institute Guidelines. Available at: http://www.grandin.com/cattle.audit.form.html, consulted July 2007. Gregory, N.G. 1998. Animal Welfare and Meat Science. CAB International, Wallingford, UK. Grifﬁn, D.R. 1976. The Question of Animal Awareness. Rockefeller University Press, New York. Grifﬁn, D.R. 1984. Animal Thinking. Harvard University Press, Cambridge, USA. Grifﬁn, D.R. 1992. Animal Minds. University of Chicago Press, Chicago. Haldane, J.B.S. 1923. Daedalus, or Science and the Future. Republished 1930, Kegan Paul, Trench, Trubner & Co, London. Harding, E.J., Paul, E.S. and Mendl, M. 2004. Animal behaviour: Cognitive bias and affective state. Nature 427: 312. Harlow, H.F. and Harlow, M.K. 1962. Social deprivation in monkeys. Scientiﬁc American 207: 136–146. Harrison, R. 1964. Animal Machines. Vincent Stuart Ltd., London. Hartwell, S.I., Jin, J.H., Cherry, D.S. and Cairns, J. Jr. 1989. Toxicity versus avoidance response of golden shiner, Notemigonus crysoleucas, to ﬁve metals. Journal of Fish Biology 35: 447–456. Harwood, D. 1928. Love for Animals and How it Developed in Great Britain. Republished 2002 as Dix Harwood’s Love for Animals and How it Developed in Great Britain (1928) (R. Preece and D. Fraser, editors). Edwin Mellen Press, Lewiston, USA. Hediger, H. 1950. Wild Animals in Captivity (G. Sircom, translator). Butterworths Scientiﬁc Publications, London. Hediger, H. 1955. Studies of the Psychology and Behaviour of Captive Animals in Zoos and Circuses (G. Sircom, translator). Butterworths Scientiﬁc Publications, London. Held, R. and Hein, A. 1963. Movement-produced stimulation in the development of visually guided behavior. Journal of Comparative and Physiological Psychology 56: 872–876. Hemsworth, P.H. and Coleman, G.J. 1998. Human–Livestock Interactions: The Stockperson and the Productivity and Welfare of Intensively-Farmed Animals. CAB International, Oxford. Hemsworth, P.H., Coleman, G.J., Barnett, J.L. and Borg, S. 2000. Relationships between human–animal interactions and productivity of commercial dairy cows. Journal of Animal Science 78: 2821–2831.

Dfraser_References.indd 295

6/27/2008 12:45:39 PM

296

References

Hennessy, M.B., Davis, H.N., Williams, M.T., Mellott, C. and Douglas, C.W. 1997. Plasma cortisol levels of dogs at a county animal shelter. Physiology & Behavior 62: 485–490. Hennessy, M.B., Williams, M.T., Miller, D.D., Douglas, C.W. and Voith, V.L. 1998. Inﬂuence of male and female petters on plasma cortisol and behaviour: Can human interaction reduce the stress of dogs in a public shelter? Applied Animal Behaviour Science 61: 63–77. Herscovici, A. 1996. Food for Thought: Facts about Food and Farming in Canada. Ontario Farm Animal Council, Mississauga, Canada. Hiebert, T. 1957. The Yahwist’s Landscape: Nature and Religion in Early Israel. Oxford University Press, New York. Hinde, R.A. and Stevenson-Hinde, J. (editors). 1973. Constraints on Learning: Limitations and Predispositions. Academic Press, London. Hirstein, W., Iversen, P. and Ramachandran, V.S. 2001. Autonomic responses of autistic children to people and objects. Proceedings of the Royal Society, London, B 268: 1883–1888. Hogan, J.A. and Roper, T.J. 1978. A comparison of the properties of different reinforcers. Pages 155–255 in Advances in the Study of Behavior, Volume 8 (J.S. Rosenblatt, R.A. Hinde, C. Beer and M.C. Busnel, editors). Academic Press, New York. Hughes, B.O. 1976. Behaviour as an index of welfare. Proceedings of the 5th European Poultry Conference, Malta, pages 1005–1012. Hughes, B.O. and Black, A.J. 1973. The preference of domestic hens for different types of battery cage ﬂoor. British Poultry Science 14: 615–619. Hughes, B.O. and Duncan, I.J.H. 1988. The notion of ethological ‘need’, models of motivation and animal welfare. Animal Behaviour 36: 1696–1707. Hurnik, J.F. 1993. Ethics and animal agriculture. Journal of Agricultural and Environmental Ethics, 6 (Suppl 1): 21–35. Hurst, S.R., Raslear, T.G., Shurtleff, D., Bauman, R. and Simmons, L. 1988. A cost–beneﬁt analysis of demand for food. Journal of the Experimental Analysis of Behavior 50: 419–440. Inge, M.T. (editor). 1969. Agrarianism in American Literature. The Odyssey Press, New York. Jefferson, T. 1787. Notes on the State of Virginia, Second American Edition, Query XIX: The present state of manufactures, commerce, interior and exterior trade? Republished 1969 as pages 9–10 in Agrarianism in American Literature (M.T. Inge, editor). The Odyssey Press, New York. Jensen, P. 1986. Observations on the maternal behaviour of free-ranging domestic pigs. Applied Animal Behaviour Science 16: 131–142. Jensen, P. and Toates, F.M. 1993. Who needs behavioural needs? Motivational aspects of the needs of animals. Applied Animal Behaviour Science 37: 161–181. Johnson, S. 1755. A Dictionary of the English Language. J. & P. Knapton, T. & T. Longman, C. Hitch and L. Hawes, A. Millar, and R. & J. Dodsley, London. Johnston, B. 1976. Ojibway Heritage. University of Nebraska Press, Lincoln. Jones, R.B. 1986. The tonic immobility reaction of the domestic fowl: A review. World’s Poultry Science Journal 42: 82–96.

Dfraser_References.indd 296

6/27/2008 12:45:39 PM

References

297

Kahneman, D. 1999. Objective happiness. Pages 3–25 in Well-Being: The Foundations of Hedonic Psychology (D. Kahneman, E. Diener and N. Schwarz, editors). Russell Sage Foundation, New York. Kahneman, D., Diener, E. and Schwarz, N. (editors). 1999. Well-Being: The Foundations of Hedonic Psychology. Russell Sage Foundation, New York. Kerr, S.G.C., Wood-Gush, D.G.M., Moser, H. and Whittemore, C.T. 1988. Enrichment of the production environment and the enhancement of welfare through the use of the Edinburgh Family Pen System of Pig Production. Research and Development in Agriculture 5: 171–186. Kessler, M.R. and Turner, D.C. 1997. Stress and adaptation of cats (Felis silvestris catus) housed singly, in pairs and in groups in boarding catteries. Animal Welfare 6: 243–254. Keverne, E.B., Levy, F., Poindron, P. and Lindsay, D.R. 1983. Vaginal stimulation: An important determinant of maternal bonding in sheep. Science 219: 81–83. Kiley, M. 1972. The vocalizations of ungulates, their causation and function. Zeitschrift für Tierpsychologie 31: 171–222. Kiley-Worthington, M. 1989. Ecological, ethological, and ethically sound environments for animals: Toward symbiosis. Journal of Agricultural Ethics 2: 323–347. Kirkden, R.D., Edwards, J.S.S. and Broom, D.M. 2003. A theoretical comparison of the consumer surplus and the elasticities of demand as measures of motivational strength. Animal Behaviour 65: 157–178. Kiyokawa, Y., Shimozuru, M., Kikusui, T., Takeuchi, Y. and More, Y. 2006. Alarm pheromone increases defensive and risk assessment behaviors in male rats. Physiology & Behavior 87: 383–387. Knierim, U., Carter, C.S., Fraser, D., Gartner, K., Lutgendorf, S.K., Mineka, S., Panksepp, J. and Sachser, N. 2001. Good welfare: Improving quality of life. Pages 79–100 in Coping with Challenge: Welfare in Animals including Humans (D.M. Broom, editor). Dahlem Workshop Report 87. Dahlem University Press, Berlin. Knight, E.M. 1940. Lassie Come-Home. Holt, Rinehart and Winston, New York. Knowles, T.G. and Wilkins, L.J. 1998. The problem of broken bones during the handling of laying hens – a review. Poultry Science 77: 1798–1802. Kolakowski, L. 1968. The Alienation of Reason: A History of Positivist Thought (N. Guterman, translator). Doubleday, New York. Kornegay, E.T. and Notter, D.R. 1984. Effects of ﬂoor space and number of pigs per pen on performance. Pig News and Information 5: 23–33. Kyriakis, S.C., Bourtzi-Hatzopoulou, E., Alexopoulos, C., Kritas, S.K., Polyzopoulou, Z., Lekkas, S. and Gardey, L. 2002. Field evaluation of the effect of in-feed doxycycline for the control of ileitis in weaned piglets. Journal of Veterinary Medicine, Series B 49: 1439–1450. Lacey, H. 1999. Is Science Value Free? Values and Scientiﬁc Understanding. Routledge, London. Ladewig, J. and Smidt, D. 1989. Behavior, episodic secretion of cortisol, and adrenocortical reactivity in bulls subjected to tethering. Hormones and Behavior 23: 344–360.

Dfraser_References.indd 297

6/27/2008 12:45:39 PM

298

References

LaFond, E., Breur, G.J. and Austin, C.C. 2002. Breed susceptibility for developmental orthopedic diseases in dogs. Journal of the American Animal Hospital Association 38: 467–477. Lamarck, J.B. 1809. Philosophie Zoologique. Republished 1984 as Zoological Philosophy: An Exposition with Regard to the Natural History of Animals (H. Elliot, translator). University of Chicago Press, Chicago. Lammers, G.J. and de Lange, A. 1986. Pre- and post-farrowing behaviour in primiparous domestic pigs. Applied Animal Behaviour Science 15: 31–43. Lassen, J., Sandøe, P. and Forkman, B. 2006. Happy pigs are dirty! – conﬂicting perspectives on animal welfare. Livestock Science 103: 221–230. Lawler, D.F., Evans, R.H., Larson, B.T., Spitznagel, E.L., Ellersieck, M.R. and Kealy, R.D. 2005. Inﬂuence of lifetime food restriction on causes, time, and predictors of death in dogs. Journal of the American Veterinary Medical Association 226: 225–231. Leach, M.C., Bowell, V.A., Allan, T.F. and Morton, D.B. 2002. Aversion to gaseous euthanasia agents in rats and mice. Comparative Medicine 52: 249–257 with erratum on 52: 449 and 572. Lembke, J. (translator). 2005. Virgil’s Georgics. Yale University Press, New Haven, USA. Lenzer, G. (editor). 1975. Auguste Comte and Positivism: The Essential Writings. University of Chicago Press, Chicago. Leopold, A. 1948. A Sand County Almanac. Republished 1987, Oxford University Press, New York. Lester, S.J., Mellor, D.J., Holmes, R.J., Ward, R.N. and Stafford, K.J. 1996. Behavioural and cortisol responses of lambs to castration and tailing using different methods. New Zealand Veterinary Journal 44: 45–54. Levy, D.M. 1944. On the problem of movement restraint: Tics, stereotyped movements, hyperactivity. American Journal of Orthopsychiatry 14: 644–671. Lindgren, A. 1985. Letter to Expressen, Sept 22, 1985. Republished as pages 1–4 in Anonymous, 1989. How Astrid Lindgren Achieved Enactment of the 1988 Law Protecting Farm Animals in Sweden. Animal Welfare Institute, Washington. Linzey, A. 1991. Christianity and the Rights of Animals. Crossroad, New York. Lorenz, K. 1937. Über die Bildung des Instinktbegriffes. Die Naturwissenschaften 25: 289–300, 307–318, 324–331. Republished 1957 as: The nature of instinct: The conception of instinctive behavior. Pages 129–175 in Instinctive Behavior: The Development of a Modern Concept (C.H. Schiller, editor and translator). International Universities Press, New York. Lorenz, K. 1939. Comparative study of behavior. Report given at the Zoological Convention, Rostock, 1939. Republished 1957 as pages 239–263 in Instinctive Behavior: The Development of a Modern Concept (C.H. Schiller, editor and translator). International Universities Press, New York. Lorenz, K. 1966. On Aggression (M.K. Wilson, translator). Harcourt, Brace & World, New York. Lund, V. 2006. Natural living – a precondition for animal welfare in organic farming. Livestock Science 100: 71–83.

Dfraser_References.indd 298

6/27/2008 12:45:39 PM

References

299

Lund, V. and Algers, B. 2003. Research on animal health and welfare in organic farming – a literature review. Livestock Production Science 80: 55–68. Malcolmson, R.W. and Mastoris, S. 1998. The English Pig: A History. Hambledon Press, London. Manning, A. 1972. An Introduction to Animal Behaviour, 2nd Edition. Edward Arnold, London. Markowitz, H. 1982. Behavioral Enrichment in the Zoo. Van Nostrand Reinhold, New York. Markowitz, H., Aday, C. and Gavazzi, A. 1995. Effectiveness of acoustic “prey”: Environmental enrichment for a captive African leopard (Panthera pardus). Zoo Biology 14: 371–379. Mason, G.J. 1991. Stereotypies: A critical review. Animal Behaviour 41: 1015–1037. Mason, G.J. 1993. Forms of stereotypic behaviour. Pages 7–40 in Stereotypic Animal Behaviour: Fundamentals and Applications to Welfare (A.B. Lawrence and J. Rushen, editors). CAB International, Wallingford, UK. Mason, G.[J.] 2007. Stereotypic behaviour in captive animals: Fundamentals, and implications for welfare and beyond. Pages 325–356 in Stereotypic Animal Behaviour: Fundamentals and Applications to Welfare, 2nd Edition (G. Mason and J. Rushen, editors). CAB International, Wallingford, UK. Mason, G.J., Cooper, J. and Clarebrough, C. 2001. Frustrations of fur-farmed mink. Nature 410: 35–36. Mason, G.J. and Latham, N. 2004. Can’t stop, won’t stop: Is stereotypy a reliable animal welfare indicator? Animal Welfare 13: S57–S69. Mason, G.[J.], McFarland, D. and Garner, J. 1998. A demanding task: Using economic techniques to assess animal priorities. Animal Behaviour 55: 1071–1075. Mason, J.W. 1968. The scope of psychoendocrine research. Psychosomatic Medicine 30: 565–575. Mason, J.W. 1974. Speciﬁcity in the organization of neuroendocrine response proﬁles. Pages 68–80 in Frontiers in Neurology and Neuroscience Research 1974 (P. Seemand and G.M. Brown, editors). University of Toronto Press, Toronto. Mason, J.W. 1975a. A historical view of the stress ﬁeld. Part 1. Journal of Human Stress 1: 6–12. Mason, J.W. 1975b. A historical view of the stress ﬁeld. Part 2. Journal of Human Stress, 1: 22–36. McGlone, J.J. 1993. What is animal welfare? Journal of Agricultural and Environmental Ethics 6 (Suppl 2): 26–36. McGlone, J.J., Salak-Johnson, J.L., Nicholson, R.I. and Hicks, T. 1994. Evaluation of crates and girth tethers for sows: Reproductive performance, immunity, behavior and ergonomic measures. Applied Animal Behaviour Science 39: 297–311. McGlone, J.J., von Borell, E.H., Deen, J., Johnson, A.K., Levis, G., Meunier-Salaün, M., Morrow, J., Reeves, D., Salak-Johnson, J.L. and Sundberg, P.L. 2004. Compilation of the scientiﬁc literature comparing housing systems for gestating sows and gilts using measures of physiology, behavior, performance, and health. The Professional Animal Scientist 20: 105–117.

Dfraser_References.indd 299

6/27/2008 12:45:39 PM

300

References

McGreevy, P. and Nicol, C. 1998. Physiological and behavioral consequences associated with short-term prevention of crib-biting in horses. Physiology & Behavior 65: 15–23. McGreevy, P.D. and Nicholas, F.W. 1999. Some practical solutions to welfare problems in dog breeding. Animal Welfare 8: 329–341. Mella, C.M., Perez-Oliva, O. and Loew, F.M. 1976. Induction of bovine polioencephalomalacia with a feeding system based on molasses and urea. Canadian Journal of Comparative Medicine 40: 104–110. Mellen, J.D. 1991. Factors inﬂuencing reproductive success in small captive exotic felids (Felis spp.): A multiple regression analysis. Zoo Biology 10: 95–110. Mench, J. A. and Stricklin, W. R. 1990. Consumer demand theory and social behavior: All chickens are not equal. Behavioral and Brain Sciences 13: 28. Mendl, M. 1991. Some problems with the concept of a cut-off point for determining when an animal’s welfare is at risk. Applied Animal Behaviour Science 31: 139–146. Meyer-Holzapfel, M. 1968. Abnormal behavior in zoo animals. Pages 476–503 in Abnormal Behavior in Animals (M.W. Fox, editor). W.B. Saunders Company, Philadelphia. Midgley, M. 1983. Animals and Why they Matter. University of Georgia Press, Athens. Miles, D.M., Branton, S.L. and Lott, B.D. 2004. Atmospheric ammonia is detrimental to the performance of modern commercial broilers. Poultry Science 83: 1650–1654. Mill, J.S. 1863. Utilitarianism. Republished 1961 as pages 399–472 in The Utilitarians. Dolphin Books, Garden City, USA. Minero, M., Canali, E., Ferrante, V., Verga, M. and Odberg, F.O. 1999. Heart rate and behavioural responses of crib-biting horses to two acute stressors. Veterinary Record 145: 430–433. Moberg, G.P. 1985. Biological response to stress: Key to assessment of animal well-being? Pages 27–49 in Animal Stress (G.P. Moberg, editor). American Physiological Society, Bethesda, USA. Molony, V., Kent, J.E. and Robertson, I.S. 1993. Behavioural responses of lambs of three ages in the ﬁrst three hours after three methods of castration and tail docking. Research in Veterinary Science 55: 236–245. Moncrief, L.W. 1970. The cultural basis of our environmental crisis. Science 170: 508–512. Montagu, A. (editor). 1973. Man and Aggression. Oxford University Press, New York. Mormède, P., Andanson, S., Aupérin, B., Beerda, B., Guémemé, D., Malmkvist, J., Manteca, X., Manteuffel, G., Prunet, P., van Reenen, C.G., Richard, S. and Veissier, I. 2007. Exploration of the hypothalamic–pituitary–adrenal function as a tool to evaluate animal welfare. Physiology & Behavior 92: 317–339. Morse, S. and Fraser, E.D.G. 2005. Making ‘dirty’ nations look clean? The nation state and the problem of selecting and weighting indices as tools for measuring progress towards sustainability. Geoforum 36: 625–640. Munnichs, G. 2004. Whom to trust? Public concerns, late modern risks, and expert trustworthiness. Journal of Agricultural and Environmental Ethics 17: 113–130. Nelson, S.A. 1998. God and the Land: The Metaphysics of Farming in Hesiod and Vergil. Oxford University Press, Oxford.

Dfraser_References.indd 300

6/27/2008 12:45:39 PM

References

301

Newman, M.A. 1993. The History of the Vancouver Aquarium. Vancouver Public Aquarium Association, Vancouver. Nicol, C.J., Davidson, H.P.D., Harris, P.A., Waters, A.J. and Wilson, A.D. 2002. Study of crib-biting and gastric inﬂammation and ulceration in young horses. Veterinary Record 151: 658–662. Niel, L. and Weary, D.M. 2006. Rats avoid exposure to carbon dioxide and argon. Applied Animal Behaviour Science 107: 100–109. Nordenfelt, L. 2006. Animal and Human Health and Welfare: A Comparative Philosophical Analysis. CAB International, Wallingford, UK. Nordenfors, H. and Höglund, J. 2000. Long term dynamics of Dermanyssus gallinae in relation to mite control measures in aviary systems for layers. British Poultry Science 41: 533–540. Nussbaum, M.C. 2004. Beyond “compassion and humanity”: Justice for nonhuman animals. Pages 299–320 in Animal Rights: Current Debates and New Directions (C.R. Sunstein and M.C. Nussbaum, editors). Oxford University Press, Oxford. Olsen, J. 1971. Slaughter the Animals, Poison the Earth. Simon & Schuster, New York. Ottosson, U., Bäckman, J. and Smith, H.G. 1997. Begging affects parental effort in the pied ﬂycatcher, Ficedula hypoleuca. Behavioral Ecology and Sociobiology 41: 381–384. Ovid. The Teachings of Pythagoras. Republished 1955 as pages 367–379 in Ovid’s Metamorphoses (R. Humphries, translator). Indiana University Press, Bloomington. Owen, R. 1813–1816. Address preﬁxed to Third Essay, in A New View on Society. Republished 1991 as pages 4–7 in Robert Owen: A New View of Society and Other Writings (G. Claeys, editor). Penguin Classics, London. Oxford Dictionary of Philosophy. 1994. Oxford University Press, Oxford. Pajor, E.A., Weary, D.M., Fraser, D. and Kramer, D.L. 1999. Alternative housing for sows and litters: 1. Effects of sow-controlled housing on responses to weaning. Applied Animal Behaviour Science 65: 105–121. Panksepp, J. 1998. Affective Neuroscience: The Foundations of Human and Animal Emotions. Oxford University Press, New York. Parker, H.G. et al. (10 authors). 2007. Breed relationships facilitate ﬁne-mapping studies: A 7.8-kb deletion cosegregates with Collie eye anomaly across multiple dog breeds. Genome Research 17: 1562–1571. Passmore, J. 1975. The treatment of animals. Journal of the History of Ideas 36: 195–218. Pauling, L. 1976. Vitamin C, the Common Cold, and the Flu. W.H. Freeman and Company, San Francisco. Penman, D. 1996. The Price of Meat. Victor Gollancz, London. Peterson, R.L. 1955. North American Moose. University of Toronto Press, Toronto. Petrie, N.J., Mellor, D.J., Stafford, K.J., Bruce, R.A. and Ward, R.N. 1996. Cortisol response of calves to two methods of disbudding used with or without local anaesthetic. New Zealand Veterinary Journal 44: 9–14. Phillips, P.A., Thompson, B.K. and Fraser, D. 1988. Preference tests of ramp designs for young pigs. Canadian Journal of Animal Science 68: 41–48.

Dfraser_References.indd 301

6/27/2008 12:45:40 PM

302

References

Plutarch. The cleverness of animals, both of the sea and of the land. Republished 1971 as pages 97–158 in Plutarch Moral Essays (R. Warner, translator). Penguin Books, Harmondsworth, UK. Plutarch. On the use of reason by ‘irrational’ animals. Republished 1992 as pages 383–399 in Plutarch Essays (R. Waterﬁeld, translator). Penguin Books, London. Pollard, S. 1968. The Idea of Progress: History and Society. Penguin Books, Harmondsworth, UK. Pope, A. 1734. An Essay on Man. Republished 1948 as pages 97–137 in Alexander Pope: Selected Works (L. Kronenberger, editor). The Modern Library, New York. Popper, K.R. 1959. The Logic of Scientiﬁc Discovery. Basic Books, New York. Porphyry. On Abstinence from Killing Animals (G. Clark, translator). Republished 2000, Duckworth, London. Post, A. 1974. Popular Freethought in America, 1825–1850. Octagon Books, New York. Preece, R. 1999. Animals and Nature: Cultural Myths, Cultural Realities. University of British Columbia Press, Vancouver. Preece, R. 2001. Introduction. Pages 1–37 in An Essay on Humanity to Animals (1798) by Thomas Young (R. Preece, editor). Edwin Mellen Press, Lewiston, USA. Preece, R. 2002. Awe for the Tiger, Love for the Lamb: A Chronicle of Sensibility to Animals. University of British Columbia Press, Vancouver. Preece, R. 2003. Darwinism, Christianity, and the great vivisection debate. Journal of the History of Ideas 64: 399–420. Preece, R. and Fraser, D. 2000. The status of animals in Biblical and Christian thought: A study in colliding values. Society and Animals 8: 245–263. Radford, M. 2001. Animal Welfare Law in Britain: Regulation and Responsibility. Oxford University Press, Oxford. Raj, A.B.M. and Gregory, N.G. 1996. Welfare implications of the gas stunning of pigs 2. Stress of induction of anaesthesia. Animal Welfare 5: 71–78. Random House Dictionary of the English Language, College Edition. 1968. Random House, New York. Randrup, A. and Munkvad, I. 1967. Stereotyped activities produced by amphetamine in several animal species and man. Psychopharmacologia (Berlin) 11: 300–310. Rauw, W.M., Kanis, E., Noordhuizen-Stassen, E.N. and Grommers, F.J. 1998. Undesirable side effects of selection for high production efﬁciency in farm animals: A review. Livestock Production Science 56: 15–33. Regan, T. 1985. The case for animal rights. Pages 13–26 in In Defense of Animals (P. Singer, editor). Harper & Row, New York. Reinhardt, V. and Rossell, M. 2001. Self-biting in caged macaques: Cause, effect, and treatment. Journal of Applied Animal Welfare Science 4: 285–294. Reynolds, J. 1983. The Great Paternalist: Titus Salt and the Growth of Nineteenth Century Bradford. Maurice Temple Smith, London. Rhodes, R.T. et al. (14 authors). 2005. Task Force report: A comprehensive review of housing for pregnant sows. Journal of the American Veterinary Medical Association 227: 1580–1590.

Dfraser_References.indd 302

6/27/2008 12:45:40 PM

References

303

Ritvo, H. 1987. The Animal Estate: The English and other Creatures in the Victorian Age. Harvard University Press, Cambridge, USA. Robbins, J. 1987. Diet for a New America. Stillpoint Publishing, Walpole, USA. Robert, S., Matte, J.J., Farmer, C., Girard, C.L. and Martineau, G.P. 1993. High-ﬁbre diets for sows: Effects on stereotypies and adjunctive drinking. Applied Animal Behaviour Science 37: 297–309. Robert, S., Weary, D.M. and Gonyou, H. 1999. Segregated early weaning and welfare of piglets. Journal of Applied Animal Welfare Science 2: 31–40. Rollin, B.E. 1990. The Unheeded Cry. Oxford University Press, Oxford. Rollin, B.E. 1993. Animal welfare, science, and value. Journal of Agricultural and Environmental Ethics 6 (Suppl 2): 44–50. Rollin, B.E. 1995. Farm Animal Welfare: Social, Bioethical, and Research Issues. Iowa State University Press, Ames, USA. Romanes, G.J. 1891. Animal Intelligence. D. Appleton and Company, New York. Rosenwald, A.S. 1981. First European Symposium on Poultry Welfare. Poultry Digest 40: 576–582. Roughan, J.V. and Flecknell, P.A. 2003. Evaluation of a short duration behaviour-based post-operative pain scoring system in rats. European Journal of Pain 7: 397–406. Rowan, A.N. and Rosen, B. 2005. Progress in animal legislation: Measurement and assessment. Pages 79–94 in State of the Animals III (D.J. Salem and A.N. Rowan, editors). Humane Society of the United States, Washington. Rozin, P. and Kalat, J.W. 1971. Speciﬁc hungers and poison avoidance as adaptive specializations of learning. Psychological Review 78: 459–486. RSPCA. 2007. Freedom Food – Welcome. Royal Society for the Prevention of Cruelty to Animals (RSPCA). Available at: http://www.rspca.org.uk/servlet/Satellite?pagename RSPCA/RSPCARedirect&pgFreedomFoodHomepage, consulted April 2007. Ruesch, H. 1978. Slaughter of the Innocent. Bantam Books, New York. Rushen, J. 1985. The scientiﬁc status of animal consciousness. Applied Animal Behaviour Science 13: 387–390. Rushen, J. 1991. Problems associated with the interpretation of physiological data in the assessment of animal welfare. Applied Animal Behaviour Science 28: 381–386. Rushen, J. 1996. Using aversion learning techniques to assess the mental state, suffering, and welfare of farm animals. Journal of Animal Science 74: 1990–1995. Rushen, J. and de Passillé, A.M.B. 1992. The scientiﬁc assessment of the impact of housing on animal welfare: A critical review. Canadian Journal of Animal Science 72: 721–743. Russell, W.M.S. and Burch, R.L. 1959. The Principles of Humane Experimental Technique. Methuen, London. Russon, A.E., Bard, K.A. and Parker, S.T. (editors). 1996. Reaching into Thought: The Minds of the Great Apes. Cambridge University Press, New York. Rutter, S.M. and Duncan, I.J.H. 1991. Shuttle and one-way avoidance as measures of aversion in the domestic fowl. Applied Animal Behaviour Science 30: 117–124. Rutter, S.M. and Duncan, I.J.H. 1992. Measuring aversion in domestic fowl using passive avoidance. Applied Animal Behaviour Science 33: 53–61.

Dfraser_References.indd 303

6/27/2008 12:45:40 PM

304

References

Sachser, N., Dürschlag, M. and Hirzel, D. 1998. Social relationships and the management of stress. Psychoneuroendocrinology 23: 891–904. Sagan, C. and Druyen, A. 1992. Shadows of Forgotten Ancestors: A Search for Who We Are. Random House, New York. Sagar, A.D. and Najam, A. 1998. The human development index: A critical review. Ecological Economics 25: 249–264. Sainsbury, D. 1986. Farm Animal Welfare. Cattle, Pigs and Poultry. Collins, London. Salak-Johnson, J.L., Anderson, D.L. and McGlone, J.J. 2004. Differential dose effects of central CRF and effects of CRF astressin on pig behavior. Physiology & Behavior 83: 143–150. Sambraus, H.H. 1985. Mouth-based anomalous syndromes. Pages 391–422 in Ethology of Farm Animals (A.F. Fraser, editor). Elsevier Science Publishers, Amsterdam. Samuel, W.M. and Welch, D.A. 1991. Winter ticks on moose and other ungulates: Factors inﬂuencing their population size. Alces 27: 169–182. Sandøe, P., Nielsen, B.L., Christensen, L.G. and Sørensen, P. 1999. Staying good while playing god – the ethics of breeding farm animals. Animal Welfare 8: 313–328. Sandøe, P. and Simonsen, H.B. 1992. Assessing animal welfare: Where does science end and philosophy begin? Animal Welfare 1: 257–267. Sanford, S.E., Waters, E.H. and Josephson, G.K.A. 1984. Gastrosplenic torsions in sows. Canadian Veterinary Journal 25: 364. Sapolsky, R.M. 1992. Neuroendocrinology of the stress-response. Pages 287–324 in Behavioral Endocrinology (J.B. Becker, S.M. Breedlove, and D. Crews, editors). MIT Press, Cambridge, USA. Saunders, M. 1893. Beautiful Joe: An Autobiography. C.H. Banes, Philadelphia. Savoury, C.J., Seawright, E. and Watson, A. 1992. Stereotyped behaviour in broiler breeders in relation to husbandry and opioid receptor blockade. Applied Animal Behaviour Science 32: 349–360. Savoury, C.J. and Mann, J.S. 1999. Stereotyped pecking after feeding by restricted-fed fowls is inﬂuenced by meal size. Applied Animal Behaviour Science 62: 209–217. Schochet, E.J. 1984. Animal Life in Jewish Tradition: Attitudes and Relationships. KTAV Publishing House, New York. Schuppli, C.A. and Fraser, D. 2000. A framework for assessing the suitability of different species as companion animals. Animal Welfare 9: 359–372. Schwartzkopf-Genswein, K.S., Stookey, J.M. and Welford, R. 1997. Behavior of cattle during hot-iron and freeze branding and the effects on subsequent handling ease. Journal of Animal Science 75: 2064–2072. Schweitzer, A. Reverence for Life. Republished 1971 as pages 25–32 in Albert Schweitzer: Reverence for Life (P. Seymour, editor). Hallmark Editions, Kansas City. Scientiﬁc Veterinary Committee. 1997. The Welfare of Intensively Kept Pigs. Report of the Scientiﬁc Veterinary Committee. European Union, Brussels. Searle, J.R. 1967. How to derive “ought” from “is”. Pages 101–114 in Theories of Ethics (P. Foot, editor). Oxford University Press, Oxford. Selye, H. 1956. The Stress of Life. McGraw-Hill Book Co, New York.

Dfraser_References.indd 304

6/27/2008 12:45:40 PM

References

305

Selye, H. 1974. Stress without Distress. McClelland and Stewart, Toronto. Sen, A. 1993. Capability and well-being. Pages 30–53 in The Quality of Life (M. Nussbaum and A. Sen, editors). Oxford University Press, New York. Seo, T., Sato, S., Kosaka, K., Sakamoto, N. and Tokumoto, K. 1998. Tongue-playing and heart rate in calves. Applied Animal Behaviour Science 58: 179–182. Sequoia, A. 1990. 67 Ways to Save the Animals. HarperCollins, New York. Sewell, A. 1877. Black Beauty: The Autobiography of a Horse. Republished 1945, Grosset & Dunlap, New York. Shaftesbury, Third Earl of (Anthony Ashley Cooper). 1711. Characteristics of Men, Manners, Opinions, Times. Republished 1964 (J.M. Robertson, editor). Bobbs-Merrill, Indianapolis. Shah, H. 2005. Jainism. Available at: http://www.jcnc.org/jainism/reference_detail.asp? category=Non%2Dviolence+%2D+Questions+%26+Answers, accessed July 2006. Sherwin, C. 1998. The use and perceived importance of three resources which provide caged laboratory mice the opportunity for extended locomotion. Applied Animal Behaviour Science 55: 353–367. Singer, C. 1957. A Short History of Anatomy from the Greeks to Harvey. Dover Publications, New York. Singer, P. 1990. Animal Liberation, 2nd Edition (1st Edition 1975). Avon Books, New York. Smith, A. 1776. The Wealth of Nations. Republished 1904, Dent and Sons, London. Snow, C.P. 1965. The Two Cultures: And a Second Look. Cambridge University Press, Cambridge. Sonenshine, D.E. 1993. Biology of Ticks, Volume 2. Oxford University Press, New York. Sorabji, R. 1993. Animal Minds and Human Morals: the Origins of the Western Debate. Cornell University Press, Ithaca, USA. Speck, F.G. 1935. Naskapi, the Savage Hunters of the Labrador Peninsula. University of Oklahoma Press, Norman. Špinka, M. 2006. How important is natural behaviour in animal farming systems? Applied Animal Behaviour Science 100: 117–128. Špinka, M., Newberry, R.C. and Bekoff, M. 2001. Mammalian play: Training for the unexpected. Quarterly Review of Biology 76: 141–168. Stafford, K.J. and Mellor, D.J. 2005. Dehorning and disbudding distress and its alleviation in calves. The Veterinary Journal 169: 337–349. Staﬂeu, F.R., Grommers, F.J. and Vorstenbosch, J. 1996. Animal welfare: Evolution and erosion of a moral concept. Animal Welfare 5: 225–234. Steiner, G. 1998. Descartes on the moral status of animals. Archiv für Geschichte der Philosophie 80: 268–291. Stephens, D.B. 1982. A review of some behavioral and physiological studies which are relevant to the welfare of young calves. Pages 47–67 in Welfare and Husbandry of Calves (J.P. Signoret, editor). Martinus Nijhoff, The Hague. Stevenson, P. 2004. European Union Law on the Welfare of Farm Animals. Compassion in World Farming Trust, Petersﬁeld, UK. Available at: http://www.ciwf.org.uk/publications/ reports/EU_Law_2004.pdf, accessed November 2004.

Dfraser_References.indd 305

6/27/2008 12:45:40 PM

306

References

Stolba, A. and Wood-Gush, D.G.M. 1984. The identiﬁcation of behavioural key features and their incorporation into a housing design for pigs. Annales de Recherches Vétérinaires 15: 287–298. Sudarshan, M.K. et al. (10 authors). 2007. Assessing the burden of human rabies in India: Results of a national multi-center epidemiological survey. International Journal of Infectious Diseases 11: 29–35. Taneja, V., Behrens, M., Mangalam, A., Grifﬁths, M.M., Luthra, H.S. and David, C.S. 2007. New humanized HLA-DR4-transgenic mice that mimic the sex bias of rheumatoid arthritis. Arthritis & Rheumatism 56: 69–78. Tannenbaum, J. 1991. Ethics and animal welfare: The inextricable connection. Journal of the American Veterinary Medical Association 198: 1360–1376. Tauson, R. 1998. Health and production in improved cage designs. Poultry Science 77: 1820–1827. Taylor, G.B. 1972. One man’s philosophy of welfare. Veterinary Record 91: 426–428. Terlouw, E.M.C., Lawrence, A.B. and Illius, A.W. 1991. Inﬂuences of feeding level and physical restriction on development of stereotypies in sows. Animal Behaviour 42: 981–991. te Velde, H., Aarts, N. and van Woerkum, C. 2002. Dealing with ambivalence: Farmers’ and consumers’ perceptions of animal welfare in livestock breeding. Journal of Agricultural and Environmental Ethics 15: 203–219. The New English Bible. 1970. Oxford University Press, Oxford. Thomas, K.V. 1983. Man and the Natural World: Changing Attitudes in England, 1500–1800. Allen Lane, London. Thomson, J. 1728. Spring. From The Seasons. Republished 1951 as pages 3–46 in The Complete Poetical Works of James Thomson (J.L. Robertson, editor). Oxford University Press, London. Thompson, P.B. 1998. Agricultural Ethics: Research, Teaching, and Public Policy. Iowa State University Press, Ames, USA. Thorpe, W.H. 1965. The assessment of pain and distress in animals. Appendix III in Report of the Technical Committee to Enquire into the Welfare of Animals kept under Intensive Livestock Husbandry Systems, F. W. R. Brambell (chairman). Her Majesty’s Stationery Ofﬁce (HMSO), London. Thorpe, W.H. 1979. The Origins and Rise of Ethology: The Science of the Natural Behaviour of Animals. Heinemann Educational Books, London. Tinbergen, N. 1951. The Study of Instinct. Clarendon Press, Oxford. Tinbergen, N. 1952. “Derived” activities; their causation, biological signiﬁcance, origin, and emancipation during evolution. Quarterly Review of Biology 27: 1–32. Toates, F.[M.] 1986. Motivational Systems. Cambridge University Press, Cambridge. Toates, F.[M.] 1995. Stress: Conceptual and Biological Aspects. John Wiley & Sons, Chichester. Toates, F.M. 2004. Cognition, motivation, emotion and action: A dynamic and vulnerable interdependence. Applied Animal Behaviour Science 86: 173–204. Troisi, A. 2002. Displacement activities as a behavioral measure of stress in nonhuman primates and human subjects. Stress 5: 47–54.

Dfraser_References.indd 306

6/27/2008 12:45:40 PM

References

307

Tuber, D.S., Miller, D.D., Caris, K.A., Halter, R., Linden, F. and Hennessy, M.B. 1999. Dogs in animal shelters: Problems, suggestions, and needed expertise. Psychological Science 10: 379–386. Tucker, C.B., Weary, D.M. and Fraser, D. 2003. Effects of three types of freestall surfaces on preferences and stall usage by dairy cows. Journal of Dairy Science 86: 521–529. Turner, E.S. 1964. All Heaven in a Rage. Michael Joseph, London. United Egg Producers. 2008. United Egg Producers Animal Husbandry Guidelines for U.S. Egg Laying Flocks, 2008 Edition. United Egg Producers, Atlanta. Available at: http:// www.uepcertiﬁed.com/docs/UEP-Animal-Welfare-Guidelines-2007–2008.pdf, accessed January 2008. Uvnäs-Moberg, K. 1998. Oxytocin may mediate the beneﬁts of positive social interaction and emotions. Psychoneuroendocrinology 23: 819–835. Vallery-Radot, R. 1919. The Life of Pasteur (R.L. Devonshire, translator). Constable & Company, London. Van Zutphen, L.F.M. and Bedford, P.G.C. (editors). 1999. Genetics and Animal Welfare. Animal Welfare 8 (Special Issue): 309–438. Veasey, J.S., Waran, N.K. and Young, R.J. 1996. On comparing the behaviour of zoo housed animals with wild conspeciﬁcs as a welfare indicator, using the giraffe (Giraffa camelopardalis) as a model. Animal Welfare 5: 139–153. Virgil. Georgics (J. Lembke, translator) Republished 2005, Yale University Press, New Haven, USA. Voltaire, F.-M.A. 1764. A Philosophical Dictionary. Republished 1924, A.A. Knopf, New York. Waiblinger, S., Baumgartner, J., Kiley-Worthington, M. and Niebuhr, K. 2004. Applied ethology: the basis for improved animal welfare in organic farming. Pages 117–161 in Animal Health and Welfare in Organic Agriculture (M. Vaarst , S. Roderick , V. Lund and W. Lockeretz, editors). CAB International, Wallingford, UK. Watson, J.B. 1924. Behaviorism. Reprinted 1970, W.W. Norton & Company, New York. Watson, J.D. 1968. The Double Helix: A Personal Account of the Discovery of the Structure of DNA. Atheneum, New York. Weary, D.M. and Fraser, D. 1995. Signalling need: Costly signals and animal welfare assessment. Applied Animal Behaviour Science 44: 159–169. Weary, D.M., Ross, S. and Fraser, D. 1997. Vocalizations of isolated piglets: A reliable indicator of piglet need directed towards the sow. Applied Animal Behaviour Science 53: 249–257. Webster, J. 1994. Animal Welfare: A Cool Eye towards Eden. Blackwell Science, Oxford. Webster, J. 2005. Animal Welfare: Limping towards Eden. Blackwell Publishing, Oxford. Weiss, J.M. 1972. Psychological factors in stress and disease. Scientiﬁc American 226: 104–113. Welp, T., Rushen, J., Kramer, D.L., Festa-Bianchet, M. and de Passillé, A.M.B. 2004. Vigilance as a measure of fear in dairy cattle. Applied Animal Behaviour Science 87: 1–13. Wemelsfelder, F. 1993. The concept of animal boredom and its relationship to stereotyped behaviour. Pages 65–95 in Stereotypic Animal Behaviour: Fundamentals and Applications to Welfare (A.B. Lawrence and J. Rushen, editors). CAB International, Wallingford, UK.

Dfraser_References.indd 307

6/27/2008 12:45:40 PM

308

References

Wemelsfelder, F., Hunter, T.E.A., Mendl, M.T. and Lawrence, A.B. 2001. Assessing the ‘whole animal’: A free choice proﬁling approach. Animal Behaviour 62: 209–220. White, L. 1967. The historical roots of our ecological crisis. Science 155: 1203–1207. Widowski, T.M. and Duncan, I.J.H. 2000. Working for a dustbath: Are hens increasing pleasure rather than reducing suffering? Applied Animal Behaviour Science 68: 39–53. Wiedenmayer, C. 1997. Causation of the ontogenetic development of stereotypic digging in gerbils. Animal Behaviour 53: 461–470. Wielebnowski, N.C., Fletchall, N., Carlstead, K., Busso, J. M. and Brown, J.L. 2002. Noninvasive assessment of adrenal activity associated with husbandry and behavioral factors in the North American Clouded Leopard population. Zoo Biology 21: 77–98. Wiepkema, P., van Hellemond, K., Roessingh, P. and Romberg, H. 1987. Behaviour and abomasal damage in individual veal calves. Applied Animal Behaviour Science 18: 257–268. Williams, R. 1958. Culture and Society 1780–1950. Republished 1966 (Harper Torchbook Edition), Harper & Row, New York. Wondra, K.J., Hancock, J.D., Kennedy, G.A., Hines, R.H. and Behnke, K.C. 1995. Reducing particle size of corn in lactation diets from 1,200 to 400 micrometers improves sow and litter performance. Journal of Animal Science 73: 421–426. Wood-Gush, D.G.M. 1971. The Behaviour of the Domestic Fowl. Heinemann Educational Books, London. Wood-Gush, D.G.M., Duncan, I.J.H. and Savoury, C.J. 1978. Observations on the social behaviour of domestic fowl in the wild. Biology of Behaviour 3: 193–205. World Health Organization (WHO). 2004. WHO Expert Consultation on Rabies. First Report. Technical Report Series, No. 931. WHO, Geneva. World Society for the Protection of Animals (WSPA). 1994. Stray Dog Control. WSPA, London. Wright, R. 1992. Stolen Continents: Conquest and Resistance in the Americas. Houghton Mifﬂin, Boston. Würbel, H. 2001. Ideal homes? Housing effects on rodent brain and behaviour. Trends in Neurosciences 24: 207–211. Würbel, H. 2007. The motivational basis of caged rodents’ stereotypies. Pages 86–120 in Stereotypic Animal Behaviour: Fundamentals and Applications to Welfare, 2nd Edition (G. Mason and J. Rushen, editors). CABI, Wallingford, UK. Yanacopoulo, A. 1992. Hans Selye, ou, La Cathédrale du Stress. Le Jour, Montreal. Yerkes, R.M. 1927. The mind of a gorilla. Genetic Psychology Monographs 2: 1–193. Young, P.T. 1959. The role of affective processes in learning and motivation. Psychological Review 66: 104–125. Young, R.J. 2003. Environmental Enrichment for Captive Animals. Blackwell Publishing, Oxford. Young, T. 1798. An Essay on Humanity to Animals. Republished 2001 as An Essay on Humanity to Animals (1798) by Thomas Young (R. Preece, editor). Edwin Mellen Press, Lewiston, USA. Zuckerman, S. 1932. The Social Life of Monkeys and Apes. Kegan Paul, Trench, Trubner & Co, London.

Dfraser_References.indd 308

6/27/2008 12:45:40 PM

Index

abnormal behaviour, 114, 125–45, 184, 279. See also: self-injury; stereotyped behaviour and adaptation to a new environment, 156–7 adjunctive behaviour, 130–31, 139, 144 and animal welfare, 125, 172, 244–5, 284 beneﬁcial consequences, 140–44 displacement activities, 129–30 and HPA system, 135 and mental health, 125 and pain, 154, 157 redirected behaviour, 127–8, 131, 139, 144 vacuum activity, 128–9, 131, 209, 212 Adams, A.W., 97, 102, 187, 287 adaptations and animal welfare, 176–90, 226–9 adjunctive behaviour. See: abnormal behaviour adrenal gland cortex, 106–7, 119 medulla, 105–6 adrenalin. See: epinephrin adrenocorticotrophic hormone (ACTH), 106–7, 119 Affective Neuroscience: The Foundations of Human and Animal Emotions (book), 168 affective states, 104, 146–68, 179, 214, 284 and abnormal behaviour, 130, 135–7, 143, 145 as adaptations, 108, 177–9, 224, 226, 228, 229 and animal welfare, 69–78, 81, 146–68, 174, 180, 183, 188–90, 222, 225–36, 239, 243, 245, 267, 283, 284–5 in animal welfare standards, 249–52 and conscious experience, 162–5 and ﬁtness, 148, 150, 161 and moral concern, 69–72, 77 positive, 159–62, 213–14

and Positivism, 166–7, 269 scientiﬁc study of, 146–68, 244, 268 scientiﬁc understanding of, 78, 81–3, 146–68, 244–5 signals of, 146–50. See also: signals and stress responses, 108–10, 112–14, 119, 123 aggression, 125, 209–10, 226 baboons, 131 chickens, 89, 155 sows, 115–16, 119, 120, 122, 241, 246 Agrarianism, 41, 44–8, 52, 58, 59–60, 63, 278, 281–2 ahimsa, 27 air quality, 225, 255–7, 271. See also: ammonia alarm calls. See: vocalizations Algers, B., 225, 299 Altmann, J., 131, 287 ammonia, 72, 96–7, 215, 249–50, 279 amphetamines, 134, 136–7, 145 anaesthesia, 112–14, 119, 120, 156, 249 analgesia, 76, 114, 119, 156, 162–3, 206, 230, 280 anatomy, 31–3, 276 ANI. See: Animal Needs Index animal behaviour, 35–8, 64–5, 73, 75, 82, 268. See also: abnormal behaviour; ethology and animal welfare science, 64–5, 80, 222 and anxiety, 163–4 and fear, 99, 152, 156–7 and frustration, 155–6 and pain, 153–5, 165 and Positivism, 36, 166–7 animal care, 27–8, 42–4, 47, 60, 70, 278, 282 animal husbandry, 42, 64, 280. See also: animal care; animal production

309

Dfraser_Index.indd 309

6/27/2008 12:45:25 PM

310

animal liberation, 65, 277 Animal Liberation (book), 61 Animal Machines (book), 61–5, 66, 218 animal mythology, 28–9 Animal Needs Index, 252–8 animal producers, 43–4, 59–60, 77, 100–101, 223, 232 animal production, 5, 43, 47, 58–9, 61–5, 75–6, 279–82 and animal welfare, 75–6, 96–103 animal protectionists, 4, 60, 71, 232 animal rights, 14, 19–20, 22, 51, 65, 67, 219, 277–8 animal sacriﬁce, 9–10, 15 animal science, 64, 75–6, 80, 222, 280–81 animal shelters cats, 156–7, 224, 271 dogs, 117–20, 122 Animal Suffering (book), 73, 168 animal welfare. See also: animal welfare assessment and abnormal behaviour, 125–45, 244–5, 284 and adaptations, 176–90, 226–9 and affective states, 69–78, 81, 146–68, 174, 180, 183, 188–90, 222, 225–36, 239, 243, 245, 267, 283–5 and agrarianism, 48 as conglomerate concept, 234–9 and culture, 41–60 deﬁnitions, 66, 74, 72–8, 81, 231, 232–4, 261, 286 different criteria, 66–78, 80, 83, 222–40, 282–6 as evaluative concept, 263–6 everyday meanings, 232, 234, 245, 263–6 and health/functioning, 70–78, 84–96, 282–4. See also: health and legislation, 5, 20–21, 218–20 and natural behaviour, 169–72, 175–6, 179–90. See also: natural behaviour and natural environments, 172–4 and natural living, 69, 169–90, 285 and preferences/motivations of animals, 191–216, 285–6. See also: preference research and productivity, 68–70, 96–103, 283. See also: productivity and “stress”, 73–6, 111–24, 283–4. See also: stress and values, 234–9, 258–9, 260–63, 266–74, 278–9. See also: values animal welfare assessment, 222 comparing housing systems, 247

Dfraser_Index.indd 310

Index

versus “measurement”, 223 “overall” assessment programs, 252–8 and science, 250–59 and values, 258–9, 261–2 animal welfare science, 5–6, 280, 281, 282 and ethics, 271–3 as mandated science, 8, 260–74, 278–9 and values, 270–71, 273–4, 277–9 animal welfare standards, 85, 241. See also: legislation different types, 218–21, 247–59, 270–71 for laying hens, 97–8, 248, 247–9 animals compared to humans. See: human beings compared to animals antibiotics, 63, 76, 102 anxiety, 119, 130, 163 appetite, 87, 105, 108, 161 Appleby, M.C., 135, 188, 201, 225, 229–30, 238, 260, 287, 290 approach test, 99, 152 Arey, D.S., 211–12, 287 Aristotle, 12, 44, 68, 215, 239, 294 on pigs, 9–10 Armstrong, J., 98, 247, 288 avoidance, 162, 205, 216, 228 learning, 178, 197, 204–7 passive, 207 shuttle, 207 baboon, 35, 131 Baenninger, R., 167–8, 287 bank vole, 136–7 Barnard, C.J., 73, 75–7, 190, 228, 231, 270, 288 Barnett, J.L., 99, 115–17, 119, 121–2, 243, 245, 270, 288, 295 review of sow housing, 242–7 Bartussek, H., 252–8, 272, 288 basal ganglia, 136–7, 183, 280 Bateson, P., 4, 111–12, 288, 289 Baumans, V., 224, 244, 288, 289 Baxter, M.R., 211, 288 Bayvel, A.C.D., 220, 288 bear baiting, 17 in Innu culture, 26 Beardsley, T., 197, 291 Beautiful Joe (dog), 51 bedding, 68, 70, 102, 199, 202, 203, 226 bulls, 117 cats, 156 dairy cattle, 87–8, 195 gerbils, 138 pigs, 193–5, 262

6/27/2008 12:45:26 PM

Index

begging calls. See: vocalizations Behavioral Enrichment in the Zoo (book), 184 Behaviorism (book), 81 behavioural needs. See: needs behaviourism, 81. See also: Watson, J.B. Bekoff, M., 160, 305 Bell, D., 98, 247, 288 Bentham, J., 18–19, 68–9, 238, 278, 288 Berridge, K.C., 161, 236, 288 Bessei, W., 207 Beynen, A.C., 244, 289 Bible, 25–8, 32, 41–4, 60, 277 Bierens de Haan, J.A., 82, 289 biomedical research, 4, 15, 66, 282 bird, 27, 30, 35, 64, 125, 150, 237, 248. See also: chicken; ﬂycatcher; hen; skylark; starling in art, 48 imprinting, 149, 178 Black, A.J., 191–2, 195, 216, 296 Black Beauty (horse), 51 Blake, W., 51–2, 289 Blum, D., 82, 289 body condition, 100 body weight, 90, 96–7, 247 Bolingbroke, Lord, 19 boredom, 67, 122, 141, 181, 183, 188, 267 bovine growth hormone, 90–92, 102 Bracke, M.B.M., 257, 289 Bradshaw, E.L., 4, 111–12, 288, 289 Brambell, F.W.R., 64, 101, 289 Brambell Committee, 69, 80, 100–101, 191 Broom, D.M., 72–4, 76, 162, 201, 224, 260–63, 273, 289, 293, 297 Brown, K.H., 220, 289 Brunk, C.G., 267, 289 BST. See: bovine growth hormone buffalo (African), 191–2 bull, 117–19, 120, 122 baiting, 17, 20–21 Burch, R.L., 273, 303 Burgdorf, J., 159, 289 Burich, S., 2, 61 Burkhardt, R.W. Jr., 82, 289 Burns, R., 25, 44, 45, 51, 290 Cabanac, M., 213, 236, 290 cages, 4, 64, 119, 126, 137, 143, 232 bank voles, 136 cats, 156, 224, 271 chimpanzees, 181 gerbils, 138–9

Dfraser_Index.indd 311

311

hens, 5, 8, 63, 71, 85–6, 90, 97, 100, 155, 191–3, 218–20, 225, 241, 247–52, 270–72, 280 mice, 182–3, 199–201, 224 mink, 203–4 monkeys, 169, 181–2 rats, 164 calf, 10, 47, 102, 119, 149, 219 animal welfare standards, 220, 249 branding, 155 crates, 5, 219, 250 dehorning, 112–14, 119, 120, 122, 123, 224, 238, 249–50, 271, 280 sucking, 127, 141, 180, 187–8, 209 tongue-rolling, 129, 141–3, 212 ulceration, 140, 169 calving, 87, 96, 100, 150 Canadian Veterinary Medical Association, 90–92, 102, 290 Canali, E., 142, 300 Cannon, W.B., 69, 104–5, 107, 123, 162, 290 caracal, 98 carbon dioxide, 177, 204–6, 215, 272 caretakers. See: human behaviour toward animals caribou, 176 Carlstead, K., 114, 308 Carpenter, E.R., 67, 68, 70, 77, 290 Carson, R., 62, 290 Carter, C.S., 162, 297 cat (domestic), 21, 105, 160 Cat Stress Score, 156–7 hyperthyroidism, 87, 89 in shelters, 156–7, 224, 271 visual perception, 182–3 cats (wild), 98. See also: leopard; lion; serval; tiger cattle, 43, 44, 47, 58, 63, 65, 67, 87, 90, 99, 110, 174, 283. See also: bull; calf animal welfare standards, 255–6 in art, 48, 49 bedding preferences, 195–6 bovine growth hormone. See: bovine growth hormone branding, 155 diet, 179 fear, 99, 152 genetic selection, 100, 228, 272 handling, 99 housing, 87–8 lameness, 263, 272 legal protection, 21, 67, 219, 250 nursing, 187 polioencephalomalacia, 180 temperature adaptation, 179

6/27/2008 12:45:26 PM

312

Cavalieri, P., 5, 290 Celli, M.L., 184, 290 central nervous system, 95, 105–6, 139, 182 chain-pulling pigs, 138 sows, 135–6 Chapple, C., 26, 27, 290 Chaucer, G., 44 Cheney, D.L., 37, 149, 290 chicken, 17, 61, 63, 169, 173, 190. See also: hen ammonia, 96–7, 250 animal welfare standards, 250 fear, 150–52, 162, 207 ﬂooring, 191 genetic selection, 96, 138, 225, 228, 272 harvester, 151 housing, 90, 280 imprinting, 178 lameness, 283 pain, 165 space allowance, 89 in the wild, 170, 176 chimpanzee, 4–5, 36–7, 40, 158, 160, 177–8, 180, 184, 268, 278, 283 food-ﬁnding, 177 termite-ﬁshing, 177, 184 Christianity, 20, 30, 32, 41–4, 277 Christopherson, R., 141, 292 Chrousos, G.P., 107–8, 290 Chua, B., 188, 287 cognition, 36–7, 64, 81, 142, 178–9, 183, 189, 226, 233 Coleman, G.J., 99, 272, 295 Colpaert, F.C., 162–3, 290 comfort, 6, 76, 81, 117, 190, 231, 257, 279, 281 competition, 59, 97, 115, 241, 242 Comte, A., 166 conﬁnement housing systems, 58, 61–5, 69–70, 71–2, 169–70, 223–4, 232 conglomerate concepts, 235–6, 239, 271 Congo (gorilla), 36, 168 Connor, M.L., 85, 290 conservation, 114, 234, 266, 282 Constable, J., 47 constative statements, 262–3 consumer’s surplus, 199, 203–4 control (by animals), 109 Cooper, J.J., 18, 201, 204, 290 corticosteroids, 115–16, 121, 135, 142, 163. See also: cortisol; glucocorticoids; HPA system corticotrophin releasing hormone, 106–7

Dfraser_Index.indd 312

Index

cortisol, 110, 115, 117, 118–20, 156, 264, 269, 284, 296. See also: HPA system amount of change, 120 diurnal variation, 117, 121 and human handling, 99 in hunted deer, 111 and oxytocin, 161 and pain, 112–13, 120, 154, 245, 284 in unfamiliar environments, 117–18, 119, 186, 187 Cottingham, J., 30, 290 Council for Agricultural Science and Technology (CAST), 100–101, 290–91 cow (domestic). See: cattle Cowper, W., 52, 291 Craig, J.V., 97, 102, 287 creatine kinase, 111 creation stories, 25, 31 crib-biting (horses), 141, 144 Croatia, 224 Cronin, G.M., 116, 132–3, 139, 243, 288, 291 cruelty, 6, 15, 18, 20, 22, 43, 63, 277 Cuba, 180 culture, 8 and animal welfare science, 8, 23, 277–8 and attitudes toward animals, 5, 21, 24–40, 41–60, 65, 276, 279, 281 culture of science, 244–6, 267–70 and science, 6, 38–40, 276 Curtis, S.E., 101, 172, 291 Cuyp, A., 48 cytokines, 108, 282 Dahrendorf, R., 7, 291 Dantzer, R., 108–10, 121, 138, 142, 201, 291 Darwin, Charles, 34–5, 81–2, 167, 276, 291 Darwin, Erasmus, 34, 276, 291 David (biblical character), 28 Dawkins, M.S., 73–4, 76–7, 89–90, 96, 123, 167, 168, 193, 197–202, 207–10, 212–13, 224, 237, 291, 292 de Passillé, A.M.B., 120–21, 141, 292, 303, 307 death. See also: survival/longevity chickens, 96–7, 102, 225 dogs, 95 pigs, 128, 172 sows, 84–5, 224 Deen, J., 299 deer, 174 hunting, 111–12, 119, 120, 122, 123 in literature, 51 defaecation, 148

6/27/2008 12:45:26 PM

Index

dehorning, 112–14, 119, 120, 123, 224, 249–50, 271, 280 demand measures, 197–204, 213 democracy, 44–5, 65 Descartes, R., 30, 292 development/maturation of animals and animal welfare, 71, 189, 236 chickens, 90, 96 neural, 183 and play, 159 social, 181–2 visual perception, 182–3 diarrhoea, 89, 102 digestive system, 141, 187 digging (gerbils), 138–9 dingo, 132–4 discomfort bulls, 117, 122 in Five Freedoms, 233 mice, 224 sheep, 206 stereotyped behaviour, 139 disease, 229, 233, 263 and animal welfare, 68, 70–72, 84–96, 100, 103, 172, 188–9, 223–5, 233, 236–7, 244, 261, 271, 282 and animal welfare standards, 249–51, 256–7 and cytokines, 108 eradication programs, 220 in organic farming, 225 prevention, 95, 279 and productivity, 100, 283 and stress responses, 75, 108, 121–3 displacement behaviour, 129–31, 144 hens, 129, 155 primates, 129 dissecting theatre, 32 distress, 29, 123, 225, 228, 244 and animal welfare, 6, 81, 83, 101, 189, 233, 234, 283 in animal welfare standards, 249–52 and carbon dioxide (rats), 205–6, 272 deer, 111 and HPA system, 283 and legislation, 146, 218 and moral concern, 19, 67 and pain, 154 separation, 146–8, 150, 159, 162, 224 in Thorpe’s essay, 62, 64, 81 dog (domestic), 29, 229–30 in ancient Greece, 9–10 in art, 48–50 and baiting, 17

Dfraser_Index.indd 313

313

expectation in, 167–8 as experimental animals, 61, 105 and fear, 152, 173 and human contact, 118, 186–7 and hunting, 4, 8, 111, 112, 123 legal protection, 61 in literature, 28–9, 51 over-eating, 225 petting, 118 play signal, 160 and rabies, 95 in shelters, 117–20, 122 stray, 233, 280 Dohoo, I.R., 91, 92, 292 Dolendo, B., 33 dolphin (in Plutarch), 13 domestication, 169, 173, 176, 179 dominion, 42, 51 Donnelly, C.A., 89, 292 Douglass, A., 98, 247, 288 drinking, 209 adjunctive (rats), 130, 138 adjunctive (sows), 130–31 analgesics (rats), 162 calves, 141, 209 giraffes, 175 in hyperthyroidism (cats), 89 location for (pigs), 194 motivation for, 210, 215 Duncan, I.J.H., 73–4, 76, 77, 129, 151–2, 155, 160, 170, 193, 207, 209–11, 212, 213, 214, 215, 224, 231, 234, 236, 292, 294, 296, 303, 308 Dunlap, T.R., 3, 292 dust-bathing (hens), 75, 272 in animal welfare standards, 248, 251 motivation, 198, 202, 214 Edinboro, C.H., 89, 292 Edinburgh Family Pen. See: Family Pen (for pigs) Edwards, S.A., 172, 201, 293 egg production, 85, 97, 100, 101, 219, 228 Ekesbo, I., 87, 88, 90, 96, 100, 101, 218–19, 293 elasticity of demand. See: demand measures electric shock, 109, 207, 249 electro-immobilization (sheep), 206 elephant (in Romanes), 35–6, 168 Elgar, M.A., 152, 293 emotion, 69. See also: affective states; feelings and abnormal behaviour, 125, 126, 130, 142 and affect, 69 and animal welfare, 146, 243, 281

6/27/2008 12:45:26 PM

314

emotion (Contd.) in animals, 12, 30–31, 35, 38, 51–2, 58–9, 64, 80–82, 278 in culture, 8, 49, 60, 277, 282 and motivation, 147 neural processes, 162 and Positivism, 36, 166–7 and stress responses, 104–5, 109–10, 112, 119, 122, 123, 142, 148 England, 15, 17, 21, 45, 60, 66, 68 Enlightenment, 18, 22, 30, 60 environmental enrichment, 144, 184–9, 201 Epicurus and followers, 12, 18, 30, 238 epidemiology, 80, 87–90, 95, 222 epinephrin, 105–6 ethics, 12, 18, 60, 64, 261–6, 270–74, 277–9, 290 and animal welfare science, 262–6 and concern for animals, 6–8, 21–3, 26–9, 38, 51–2, 60, 261, 265 and facts, 24, 28, 270 and science, 6, 7, 39, 270–74 ethological needs. See: needs ethology, 75, 166, 169, 208, 280. See also: animal behaviour deﬁned, 82 Europe, 5, 15, 32, 45, 46, 276 European Union, 5, 86, 219, 242, 247–8, 249, 251 evaluative concepts, 263–6 evaluative statements, 262–3 everyday meanings, 264 of animal welfare, 232, 234, 245, 263–6 of need, 208 of stress, 104, 123 Ewer, R.F., 160, 293 Ewer, T., 64 exercise, 228 mice, 199 pigs, 171 and stress responses, 108, 112, 120 exotic pets, 179 exploration, 189, 230 and affect, 213–14 and animal welfare, 244 mice, 183 mink, 203 motivation for, 135, 160–62, 210 rats, 163 sows, 211, 244 and stress responses, 120 facts (versus values), 7, 262–3 Falk, J.L., 130, 138, 293

Dfraser_Index.indd 314

Index

falsiﬁcation, 38 Family Pen (for pigs), 169–72 farm animals, 76, 169. See also: entries for various species; conﬁnement housing systems in agrarian agriculture, 47, 282 and animal welfare, 62, 68–9, 96, 173, 218, 224–5 in culture, 29, 47 human contact, 99, 124, 272 productivity, 96–103 Farmer, C., 130, 303 fear, 35, 159, 162, 280 and abnormal behaviour, 139 as adaptation, 153, 224 in animal shelters, 156 and animal welfare, 70, 77, 230, 233–4, 243, 283 approach test, 99, 152 and avoidance learning, 207 and corticotrophin releasing hormone, 135 and heart rate, 151, 267 of humans, 64, 80, 98–9, 124, 127, 152, 230, 272, 280 and moral concern, 19, 67, 72 and motivation, 160 as negative affect, 215 and stress responses, 105, 112, 119–20, 123, 153, 156 and tonic immobility, 150–51 and vigilance, 152, 267 Feddes, J.J.R., 128, 293 feelings, 17, 73–4, 77, 282. See also: affective states; emotion and affect, 69 and animal welfare, 73–4, 77, 146, 243 and motivation, 208, 225 and Positivism, 166–7 Feigl, H., 166, 293 Feminist thought, 168 Ferrante, V., 142, 300 Festa-Bianchet, M., 153, 307 ﬁght or ﬂight response, 105. See also: SAM system ﬁsh, 34, 35, 232 avoidance, 215 self-expenditure, 231 ﬁtness, 73, 153 and affect, 148, 150, 161 and animal welfare, 73–4, 77, 231, 243 and signals, 148, 150 Five Freedoms, 233, 272 ﬁxed action pattern, 177

6/27/2008 12:45:26 PM

Index

ﬂamingo, 185 Flecknell, P.A., 156, 158, 303 Flint (chimpanzee), 37, 168 Flo (chimpanzee), 37 ﬂycatcher, 149 Fölsch, D., 207 food safety, 63 as evaluative concept, 263–4, 266, 270, 273 as mandated science, 7, 260 Forkman, B., 298 fox hunting, 4 in Plutarch, 12 Fox, M.W., 125, 293 France, 30, 65, 219 Francis of Assisi (Saint), 30, 53 Fraser, A.F., 125, 261, 280, 293 Fraser, D., 10, 31, 42, 44, 47, 59, 81, 92, 128, 147–8, 160, 162, 166, 179, 191, 193–5, 196, 225, 226–7, 241, 293, 294, 295, 297, 301, 302, 304, 307 Fraser, E.D.G., 258, 300 free choice proﬁling, 158 Freedom Food, 220, 248, 251 freezing (fear response) by hens, 207 by sows, 135 frustration, 145, 183, 244, 268 and abnormal behaviour, 135, 139, 145, 155–6 and animal welfare, 180, 183, 189, 212–13, 225, 243, 245–6, 285–6 in animal welfare standards, 250 in hens, 129, 155–6, 212 indications of, 129, 155, 212, 284 functioning, 179, 183, 189, 215, 268. See also: health and abnormal behaviour, 126, 136–7, 141, 145, 183 and animal welfare, 71–8, 83, 84–103, 122–3, 179, 189, 222, 228–32, 239, 243–6, 268, 277, 282–4, 286 in animal welfare standards, 249–52 and genetic selection, 91–4, 283 and natural conditions, 224 and stress responses, 121, 284 fur-plucking (clouded leopards), 114 Gainsborough, T., 47 gait chickens, 90 rats, 158 Galileo, 6, 38 Gallup, G., 151, 294

Dfraser_Index.indd 315

315

Garner, J.P., 136–7, 294, 299 Gaskin, J.C.A., 238, 294 Gavinelli, A., 220, 288 genetic selection, 268, 283 and animal welfare, 91–4, 283 cattle, 94 chickens, 96, 100, 138 dogs, 93–4 mice, 91–3 Gentle, M.J., 165, 294 gerbil, 138, 139 Géricault, T., 50 gestation stall, 5, 58, 241–7, 267, 270 compared to other housing, 98, 115–17, 241–7 legislation, 5, 220 scientiﬁc disagreement, 242–7 and stereotyped behaviour, 132–3, 135–6 giraffe, 175–6 Girard, C.L., 130, 303 Glickman, S., 214 glucocorticoids, 106–7, 114–15, 118–19. See also: corticosteroids; cortisol glucose, 107, 111, 116, 121 glycogen, 107, 111 God, 25, 27, 31–3, 35, 42–3, 45–6, 275 Goethe, J.W. von, 31, 275–6, 294 Goldsmith, O., 45, 295 Gonyou, H., 224, 303 Goodall, J., 4–5, 36–7, 168, 177, 268, 283, 295 goose, 178, 225 gorilla, 36, 40 Gourkow, N., 224, 295 Grandin, T., 253–4, 295 grazing, 67, 127, 128, 129, 184, 210, 212, 219 Great Chain of Being, 33, 35 Greece, 9–15, 21–2, 30, 60 Gregory, N.G., 100, 206, 295, 302 Greyfriar’s Bobby (dog), 28 Grifﬁn, D.R., 37, 167, 295 Grommers, F.J., 102, 281, 302, 305 grooming, 163 mice, 91–2 growth, 89, 188, 279–80 and ammonia, 96–7 and animal welfare, 58, 71, 75, 77, 84, 89–90, 100–101, 103, 121–2, 172, 243–6, 249, 267–70, 281, 283 and animal welfare standards, 249–50 and environmental conditions, 58, 90 and genetic selection, 96, 138, 272 and natural behaviour, 188 and stress responses, 107, 124, 284

6/27/2008 12:45:26 PM

316

growth hormone, 107. See also: bovine growth hormone guinea-pig, 134, 176, 185–7 Haartman, L. von, 149 Haldane, J.B.S., 6, 295 Ham (chimpanzee), 4 Harding, E.J., 165, 295 hare, 51 Harlow, H.F., 82, 181–2, 295 Harrison, R., 61–3, 66–7, 70, 77, 146, 218–19, 295 Hartley, D., 19 Harwood, D., 15, 17, 18, 19, 30, 33, 50, 295 health, 71, 84–6, 141, 160, 172, 179, 181, 215, 281, 289 and abnormal behaviour, 126–7, 284 and affective states, 108, 225, 282–3 and animal adaptations, 179, 183 and animal welfare, 22, 68, 70–78, 81, 83, 84–96, 101–3, 122, 146, 169, 171, 208, 222–4, 228–30, 231–3, 235, 239, 243–5, 256, 262, 271, 277, 282–4 and animal welfare standards, 249–51, 256, 270 in artiﬁcial environments, 190 in conﬁnement systems, 58, 218 and human handling, 272 under natural conditions, 174, 189, 223–4, 233 in organic farming, 225 and preferences, 215 and productivity, 56, 60, 68, 101–3 and science, 268 and social contact, 162 and stress responses, 123, 143, 284 heart rate, 105, 142, 148, 152, 153, 161, 267 Hediger, H., 125, 180–81, 184, 186, 295 Held, R., 182, 295 Hemsworth, P.H., 99, 116–17, 121–2, 124, 152, 243–6, 272, 280, 288, 295 hen, 5, 70 animal welfare standards, 219–20, 241, 247–52 broken bones, 100 cages, 5, 63, 71, 85–6, 193, 219–20, 247–52, 270–72, 280 dust-bathing, 75, 214 fear, 207 feather damage, 86 ﬂooring, 191, 196–7 forced moulting, 247, 250 frustration, 129, 155–6, 212 genetic selection, 100

Dfraser_Index.indd 316

Index

instrumental conditioning, 197, 202 lesions (foot, neck), 85–6 nesting, 155, 201, 212, 272 overgrown claws, 86 pain, 165 preference research, 191, 195, 198 red poultry mite, 225 space allowance, 97, 102, 247–52 vocalizations, 160 Hennessy, M.B., 117–20, 186, 296, 307 Henry VIII (King), 15 herding, 25–8, 43–4 Hermarchus, 14 Herscovici, A., 68, 70, 296 hiding, 115, 163, 188, 229 Hinde, R.A., 197, 210, 214, 296 Hinduism, 11 Hirstein, W., 142, 296 Hogan, J.A., 213–14, 296 Hogarth, W., 15, 16, 50 Hollingsworth, J., 220, 289 Holy Isle, 170, 176 hormones, 141, 144, 213. See also: bovine growth hormone; growth hormone in stress responses, 75, 104–10, 122–4, 148 horse, 47 in ancient Greece, 9 in art, 15–16, 48–50 and baiting, 17 crib-biting, 141–2, 144 and hunting, 4, 111 legal protection, 19, 21, 61 ulceration, 141 weaving, 125 HPA system, 106–24, 135, 137, 142, 153–4, 161, 185–7, 283–4 Hughes, B.O., 72–3, 191–2, 195, 209–12, 216, 287, 296 human behaviour toward animals, 124 chickens, 90 clouded leopards, 114–15, 134, 283 cows, 99 dogs, 118, 186 rats, 159 wild cats, 98 human beings compared to animals, 10–14, 19, 24, 29–38, 40, 52, 60, 275–7, 281 fear of, 64, 80, 99, 152, 272 mental health, 125, 134, 135, 136, 144, 164–5, 233 and rabies, 95 relation to animals, 25–8, 41–60 stress in, 130

6/27/2008 12:45:26 PM

Index

Human Development Index, 258 humidity (ambient), 90, 279 hunger, 213 as affective state, 69–70, 72, 160, 177–9, 180, 214, 226 and animal welfare, 233 in animal welfare standards, 249–51 chickens, 138, 247, 272 dogs, 167–8, 225 and forced moulting, 247, 250 hens, 198, 249–51 signals of, 149–50 and stereotyped behaviour, 85, 135, 137, 268, 284 sows, 85, 135, 169 hunting, 3–4, 14, 24–7, 111–12, 119, 123, 222 by chimpanzees, 177 and Innu culture, 26–7 by servals, 184 Hurnik, J.F., 76, 231, 263, 296 Hurst, J.L., 73, 75–7, 190, 228, 231, 270, 288 Hurst, S.R., 202, 296 Hutson, G.D., 243, 288 Huxley, J., 82 Huxley, T., 167 hyperactivity guinea-pigs, 134 sows, 135 hyperthyroidism (cats), 87, 89, 95 hypothalamus, 106–7 Illius, A.W., 135, 306 illness. See: disease immunity, 271 and animal welfare, 101, 121–3, 243, 261 and cytokines, 108 and stress responses, 75, 107, 108, 118, 121–3, 284 imprinting, 178 In the Shadow of Man (book), 4 India, 95, 233, 280 Industrial Revolution, 45, 54 Industrialism, 41, 53–9, 60, 277, 281–2 inelastic demand. See: demand measures inﬂammation, 87, 108, 113, 141 Inge, T.M., 44, 296 Innu, 26–8 insect, 22–7, 34, 35, 194 catching 128, 209–10 parasitic, 153, 173 instrumental learning, 197–204, 213 insulin, 107, 213 intelligence, 19, 28–9, 35–7, 64

Dfraser_Index.indd 317

317

Jainism, 26–8, 60, 281 Jefferson, Thomas, 45–6, 296 Jensen, P., 211, 296 Johnson, A.K., 299 Johnson, Samuel, 65, 296 Jones, R.B., 151, 296 Jones, T.A., 89, 292 Jongman, E.C., 243, 288 Judaism, 41 junglefowl, 173 Kahneman, D., 161, 168, 213, 236, 297 Kant, I., 30 keeper. See: human behaviour toward animals Kessler, M.R., 156–7, 297 Keverne, E.B., 161, 297 Kiley-Worthington, M., 75–7, 297, 307 kinship, 10–11, 43 Kirkden, R.D., 201, 297 Kiyokawa, Y., 163, 164, 297 Knierim, U., 162, 297 Knowles, T.G., 100, 297 koala, 179 Kornegay, E.T., 281, 297 Kramer, D.L., 153, 225, 301, 307 Kyriakis, S.C., 102, 297 laboratory animals, 8, 126, 169, 184, 280, 282. See also: gerbil; guinea-pig; mouse; rat Lacey, H., 6–7, 297 lactation, 90 Ladewig, J., 117, 119, 121, 297 Lamarck, J.-B., 34, 276, 298 lamb, 70, 127, 153–4, 161, 224 lameness, 285 in animal welfare standards, 251 in cattle, 91–2, 94, 102, 263, 272 in chickens, 89–90, 96, 283 in sows, 224 Lammers, G.J., 228, 298 Landseer, E.H., 47–9 Lassen, J., 298 Lassie (dog), 28 Latham, N., 143–4, 299 Lawler, D.F., 225, 298 Lawrence, A.B., 135, 158, 287, 306, 308 Lawrence, J., 19, 277 Leach, M.C., 205, 272, 298 legislation, 7, 19, 67, 146, 208, 264 Austria, 255 European Union, 5, 86, 219, 242, 247–9, 251, 270 Netherlands, 5

6/27/2008 12:45:26 PM

318

legislation (Contd.) New Zealand, 5 Sweden, 86, 219, 250 United Kingdom, 4, 17, 20–21, 55–6, 61, 218–19 United States of America, 5, 220 leopard African, 188 clouded, 114–15, 119, 120, 122, 134, 280, 283 Leopold, A., 3, 298 Lester, S.J., 154, 298 Levis, G., 299 Levy, D.M., 125, 161, 298 Lindgren, Astrid, 66–70, 77, 218–19, 239, 298 Lindsay, D.R., 161, 297 Linnaeus, C., 34 lion, 13, 115, 119, 280 in art, 49 and baiting, 17 in literature, 52 Loew, F.M., 180, 300 longevity. See: survival/longevity Lorenz, K.Z., 128, 169, 177, 209–10, 298 Lund, V., 225, 229–30, 298, 299 Lutgendorf, S.K., 162, 297 MacGregor (chimpanzee), 4 malaise, 108 Malcolmson, R.W., 58, 299 malnutrition, 70, 233, 282 mandated science animal welfare science as, 8, 260–74, 278–9 deﬁned, 7 everyday concepts in, 264–6 and “overall” measurement, 258 values in, 260–74, 278–9 Manning, A., 178, 299 Markowitz, H., 184, 185, 188, 299 Martin, R., 21 Martineau, G.P., 10, 130, 294, 303 Mason, G., 128, 132, 136–7, 139, 143–4, 203–4, 294, 299 Mason, J.W., 108–9, 299 mastitis, 87–8, 91, 102, 283 mating, 129, 161 motivation for, 129, 202–3, 210 and stress responses, 120 Matte, J.J., 303 Matthews, L.R., 191, 294 McCune, S., 156 McFarland, D., 299 McGlone, J.J., 74–6, 98, 135, 243, 299, 304

Dfraser_Index.indd 318

Index

McGreevy, P., 144, 300 Mella, C.M., 180, 300 Mellen, J.D., 98, 300 Mellor, D.J., 112, 114, 154, 298, 301, 305 Mench, J., 98, 247, 288, 300 Mendl, M., 121, 158, 165, 295, 300, 308 meta-analysis, 90–92, 143–4 Metz, J.H.M., 257, 289 Meunier-Salaün, M., 299 Meyer-Holzapfel, M., 125, 132–4, 137, 139, 300 Midgley, M., 167, 300 Mike (chimpanzee), 4 Miles, D.M., 96–7, 300 milk fever, 87–8, 96 milk production/yield, 47 in ancient Greece, 9–11 and animal welfare, 283 and bovine growth hormone, 90–91, 102 and genetic selection, 94, 228, 272, 283 and handling/stress, 99 Mill, J.S., 19, 300 Millet, J.-F., 47 Milligan, B.N., 226, 294 Mineka, S., 162, 297 Minero, M., 142, 300 mink, 203–4 Moberg, G.P., 73–6, 121–3, 300 Moby Doll (orca), 2, 61 Molony, V., 153, 300 Mongolian gerbil. See: gerbil monkey, 35, 130, 229, 280 and baiting, 17 self-injury (macaques), 126–7, 169, 280 signals (vervet monkeys), 149 social isolation (macaques), 82, 181–2 and stress (macaques), 108 moose, 176, 179, 230 adaptations, 174, 176, 179 age determination analogy, 222–3, 226 natural living conditions, 172–4 and ticks, 174, 177 Morland, G., 47 Mormède, P., 109–10, 115, 120–21, 138, 142, 291, 300 Morrow, J., 299 Morse, S., 258, 300 mortality rate. See: survival/longevity Morton, D.B., 205, 298 motivation, 178 as adaptation, 178, 226 and affect, 82, 160–61, 208–9, 213–14 and animal welfare, 64, 80, 197–216, 281 and animal welfare standards, 250–52

6/27/2008 12:45:27 PM

Index

and displacement activities, 129 and needs, 207–12 for dust-bathing (hens), 197, 198 for nest-building (sows), 211, 212, 228 for nesting (hens), 155, 201, 250 rebound effect, 212, 214 and stereotyped behaviour, 135, 140 strength, 197–207 for sucking (calves), 141 theories, 80, 209–10 willingness to pay, 197–204 mouse, 25 exercise, 199–201 grooming, 91–2 and mites (Myobia musculi), 91–2 as model of arthritis, 92–3 as model of muscular dystrophy, 91–2 nest-building, 213–14 stereotyped behaviour, 183 ventilated cages, 224 whisker-trimming, 126 mule (in Plutarch), 12 multivariate analysis, 87–90, 114–15 Munnichs, G., 266–7, 300 narrative data, 35, 168, 268–9 National Trust (United Kingdom), 4, 111–12 natural behaviour, 176 and animal welfare, 64, 67–71, 75–8, 81, 169–72, 175–90, 207, 215, 228–33, 244–5 and animal welfare standards, 249–52 of chickens, 170 of clouded leopards, 115, 283 and environmental enrichment, 184–8 of giraffes, 175–6 and motivation, 210–16 of pigs, 169–72, 211–12 and stereotyped behaviour, 143–4 natural living, 68–71, 76, 81, 222, 224–5, 230–32. See also: natural behaviour instrumentally or inherently important, 169, 189–90 natural selection, 34, 77, 80, 119, 127–8, 147–8, 224, 231, 277 nature of animals (beliefs about), 24–40, 60 needs behavioural, 207–13 versus desires, 208, 231 proximate, 209 nest-building, 211, 226 mice, 213–14 sows, 170, 209, 211–12, 227 nesting, 208 hens, 155, 201, 212, 248–51, 272

Dfraser_Index.indd 319

319

Netherlands, the, 5, 82, 219 New Zealand, 5, 112 Newberry, R., 98, 160, 170, 247, 288, 305 Nicol, C.J., 141–2, 144, 300, 301 Nicolson, G., 19, 277 Niel, L., 204–6, 301 Noah (biblical character), 27 noradrenalin. See: norepinephrin Nordenfelt, L., 232, 234–5, 239, 260, 301 Nordenfors, H., 225, 301 norepinephrin, 105–6 normative statements, 263 nursing, 120, 160, 161 and HPA system, 120 Nussbaum, M.C., 68–70, 239, 289, 301 objectivity, 235, 244, 245, 246, 279 in animal welfare assessment, 70, 72, 101, 123, 261–2, 278 in “overall” assessment, 235–7 in science, 6–7, 244 and the study of affective states, 82, 163 use of preference research to achieve, 192–3, 201, 237–40 Odberg, F.O., 142, 300 OIE. See: World Organization for Animal Health Ojibway, 25, 60, 281, 296 Olsen, J., 61 operant conditioning. See: instrumental learning orca, 2–3, 5, 8 organic farming animal welfare standards, 249–50, 255 and health, 225 and naturalness, 71–2, 75–6, 190, 223 otter, 126 overgrown claws (hens), 86 Ovid, 10–11, 301 Owen, R., 56–7, 64, 301 pacing, 144 by African leopard, 188 by clouded leopards, 114–15, 135 by dingo, 132–4 by hens, 129, 155 by servals, 184 pain, 150, 223, 271 and animal welfare, 6, 69, 70, 73–4, 81, 230–36, 263, 279, 282 and animal welfare standards, 249–52 and carbon dioxide, 204–5 conscious experience of, 165 and legislation, 146, 218 management of, 112–14, 119, 153–5, 156–8, 279–80. See also: anaesthesia; analgesia

6/27/2008 12:45:27 PM

320

pain (Contd.) and moral concern, 12, 18–20, 22, 67–72, 146, 277 scientiﬁc study, 35–6, 64, 112–14, 153–5, 156–8, 162–3, 165, 244, 267 and self-administration of analgesics, 162–3 and stress responses, 112–14, 120, 123, 153–4 Pajor, E.A., 225–6, 294, 301 Panksepp, J., 159, 162, 289, 297, 301 paralysis cattle, 100 pigs, 128 parasites, 71, 153, 179, 225, 226, 232, 236. See also: ticks Passmore, J., 41, 301 Pasteur, Louis, 6, 95, 268 Pastoralism, 25–8, 41–4, 47, 52, 59, 70, 278, 281 pasture cattle, 65, 87–8, 219, 250, 256 horses, 141 sows, 84, 224–5 pathology and abnormal behaviour, 126–7, 136–7, 139, 142–3, 145 and animal welfare, 75–6, 84–6, 95, 121–2, 123, 222, 255 and stress responses, 75–6, 121–2 and unnatural diet, 180 Paul, E.S., 165, 295 Pauling, L., 39, 176, 301 Pavlov, I.P., 104 pecking, 86, 139, 170, 197, 207 perching, 115, 193, 201, 202, 203, 248–51, 270, 272, 283 Petrie, N.J., 112, 301 pheromone (rats), 163–4 Phillips, P.A., 196, 301 physiology, 280. See also: stress and adaptations, 176, 179, 226 and affect, 73, 105, 148, 153–4, 156, 222, 250, 267 and animal welfare, 64, 67, 73–6, 80, 104–24, 243, 255 and anxiety, 163 and pain, 112–14, 153–4 and stereotyped behaviour, 134, 140–43 pig (domestic), 29, 58, 61, 63, 71, 170, 179, 181, 196, 242. See also: sow adjunctive behaviour, 138–9, 142 in ancient Greece, 9–10 animal welfare standards, 220 bedding, 193–5, 262 and evil spirits (Biblical story), 43 Family Pen, 170–72

Dfraser_Index.indd 320

Index

and free-choice proﬁling, 158 gas euthanasia, 215 genetic selection, 225 natural behaviour, 170–72, 176 preference research, 193–6 ramps, 196 Segregated Early Weaning, 224 separation distress, 146–8, 150, 162 tail-biting, 128 ulceration, 281 pigeon, 59, 130 Pippi Longstocking (books), 66, 70, 218, 239 pituitary gland, 106, 107, 108 plants, 13, 25, 27, 34, 231, 234, 249 play, 43, 159–61, 189, 210, 213–15, 230, 285 dogs, 186 monkeys, 181–2 pigs, 196 rats, 159 pleasure, 104 and moral concern, 19, 60, 66–7, 69–70, 146, 278 in motivation, 159–62, 178, 213, 226, 236 scientiﬁc study of, 35, 159–62, 213–14, 267 Plotinus, 15 Plutarch, 12–14, 302 Poincaré, H., 7, 260, 270–71 Poindron, P., 161, 297 polioencephalomalacia (cattle), 180 Pollard, S., 54, 302 Pope, A., 33–4, 276, 302 Popper, K., 38, 302 Porphyry, 14, 43 Positivism, 36, 166, 269 Potter, P., 48 poultry. See: chicken; hen predation and animal welfare, 58, 71, 230 and behaviour of prey species, 150, 163, 226, 237 in natural environments, 170, 176, 189 and stress responses, 114–15 and tonic immobility, 153 and vigilance, 152 prediction (by animals), 109 Preece, R., 15, 19–20, 30–31, 34, 41–2, 275–6, 294, 302 preening, 129, 155 preference research, 64, 191–7, 237–9, 261 and animal welfare, 191–216, 237–9, 285–6 for bedding (cows), 195 for bedding (pigs), 193–4 for enclosure type, 193 for ﬂooring (hens), 191

6/27/2008 12:45:27 PM

Index

for food versus litter (hens), 198 limitations, 193, 214–16 for ramp design features (pigs), 196 prepathological state, 74, 122–3 prescriptive statements. See: normative statements primate, 5, 8, 35, 129, 176, 233 productivity and ammonia (chickens), 96–7 and animal welfare, 68–70, 74–6, 96–103, 121, 277, 283 and animal welfare standards, 249–52 in conﬁnement systems, 58–9, 63, 70 and fear, 99, 280 and human handling (cattle), 99 in industrial activities, 54–6 and space allowance (hens), 97–8 and stress responses, 124 proﬁtability, 96, 101–2, 281 progress, 54, 59 Pythagoras and followers, 10–12 quality of life and animal welfare, 6, 22, 66, 74–6, 96, 122, 189, 215, 228–9, 231, 235, 258 rabies, 95, 280 Radford, M., 218, 302 Rahman, S.A., 220, 288 Raj, A.B.M., 206, 302 rat, 262, 263 adjunctive drinking, 130 affect, 162, 164 alarm pheromone, 163–4 avoidance learning, 178, 207 and carbon dioxide, 204–6, 272 demand for food, 202 pain, 156, 162 self-administration of drugs, 162–3 speciﬁc appetite, 215 stress responses, 109, 163 ulceration, 109–10 vocalizations, 159 rationality (of animals), 12–14, 19, 30, 35 Rauw, W.M., 102, 281, 302 Rebecca (biblical character), 28 red deer. See: deer Reeves, D., 299 Regan, T., 65, 66, 302 reincarnation, 10–11 Reinhardt, V., 126–7, 302 reproduction, 84 and animal welfare, 74–8, 98–101, 103, 122, 243–4, 249, 261, 268, 277

Dfraser_Index.indd 321

321

and fear, 272 and ﬁtness, 73 and natural selection, 80, 159, 231 and needs, 208 and stress responses, 75, 107–8, 121–4, 284 reservation price, 199, 202–4 Reynolds, J., 50, 56, 302 Rhodes, R.T., 243, 302 Ritson, J., 19, 277 Ritvo, H., 41, 303 Robert, S., 130, 224, 303 robin (in literature), 52 rodent, 27, 231, 272. See also: bank vole; gerbil; guinea-pig; mouse; rat Rollin, B.E., 36, 68–70, 81, 166, 167, 239, 260, 303 Romanes, G., 35, 36, 81–2, 168, 268, 303 Romanticism, 41, 48–54, 58, 59–60, 69–70, 239, 277–8, 281–2 rooting, 84, 135, 160, 170–71, 211–12, 244 Roper, T.J., 213–14, 296 Rosenwald, A.S., 101, 303 Ross, S., 147, 307 Rossell, M., 126–7, 302 roughage, 84–5, 129, 140, 144, 212, 249 Roughan, J.V., 156–8, 303 Rowan, A.N., 5, 303 Rowell, T., 131 Royal Society for the Prevention of Cruelty to Animals, 21, 220–21, 248 Rozin, P., 178, 303 Ruesch, H., 61, 303 running wheel, 201 Rushen, J., 120, 121, 141, 153, 167, 206, 292, 303, 307 Russell, W.M.S., 273, 303 Russon, A.E., 37, 303 Rutter, S.M., 207, 303 Sachser, N., 162, 186–7, 297, 304 Sagar, A.D., 258, 304 Sainsbury, D., 68, 70–71, 304 Salak-Johnson, J.L., 98, 135, 299, 304 salmon, 77, 232 Salt, Titus, 56 SAM system, 104–9, 118, 120–21, 123–4, 153, 161 Sambraus, H.H., 127, 304 Samuel, W.M., 174, 304 Sandøe, P., 238, 263, 287, 298, 304 Sanford, E., 84–5, 90, 96, 304 Sapolsky, R.M., 105, 304 Savoury, C.J., 138, 170, 304, 308 Schouten, W.G.P., 257, 289

6/27/2008 12:45:27 PM

322

Schuppli, C.A., 179, 304 Schwartzkopf-Genswein, K.S., 155, 304 Schweitzer, A., 53, 304 Scientiﬁc Veterinary Committee (European Union) review of sow housing, 242–7, 304 self-administration of drugs (rats), 162–3 self-expenditure, 75–7, 231 self-injury, 126–7, 131, 145 clouded leopards, 114–15 macaque monkeys, 126–7, 169, 280 and stress responses, 114, 135 Selye, H., 105–9, 122–3, 304, 305 Sen, A., 239, 305 Seo, T., 142, 305 serval feeding behaviour, 184, 221 reproductive success, 98 Seyfarth, R.M., 37, 149, 290 Shaftesbury, Third Earl of, 18–19, 305 Shea-Moore, M., 98, 247, 288 sheep adaptations, 179 in ancient Greece, 9–11 and anthrax, 268 in art, 15–16, 48 attachment to lambs, 161 in the Bible, 28, 43 electro-immobilization, 206 in literature, 51 slaughter, 252 on winter pasture, 173 wool-pulling, 127 shelters. See: animal shelters Sherwin, C., 199–201, 305 signals, 146–50 alarm calls, 149 begging calls, 149–50 honest signalling, 150 of positive affect, 159–60 separation calls, 146–9 Simonsen, H.B., 263, 304 Singer, C., 31, 305 Singer, P., 5, 61, 66–9, 77, 146, 238, 290, 305 skylark, 129 in literature, 51 slaughter, 9, 15–16, 29, 61 animal welfare standards, 220–21, 249, 252–4 in the Bible, 42 Smidt, D., 117, 121, 297 Smith, A., 54–5, 305 Smith, W., 6 Smuts, B., 131

Dfraser_Index.indd 322

Index

Snow, C.P., 276, 305 Snyders, F., 48–9 Sorabji, R., 10, 12–15, 43, 305 soul, 30, 34, 44 sow, 75, 225, 238. See also: gestation stall adjunctive drinking, 130–31 aggression, 116, 119–20, 122 alternate-day feeding, 84–5 animal welfare standards, 220, 249–50 Family Pen, 170–72 gastrosplenic torsion, 84–5, 90, 169, 221 hunger, 135–7, 284 natural behaviour, 170–72 nest-building, 211–12, 227–8 outdoor systems, 224 “overall” welfare assessment, 257 restricted feeding, 84, 130, 135–7, 249, 250 separation calls, 147 stereotyped behaviour, 132–3, 135–6, 284 tethering, 98, 116, 119–20, 122 space allowance, 100, 102, 220 chickens, 89–90 clouded leopards, 115 hens, 97, 218, 247–52 Speck, F., 26, 305 Špinka, M., 160, 190, 305 Spruijt, B.M., 257, 289 Stafford, K.J., 112, 114, 154, 298, 301, 305 Staﬂeu, F.R., 305 starling, 128, 209–10 Steiner, G., 30, 305 Stephens, D.B., 127, 305 stereotyped behaviour, 131–45, 212, 250, 251, 268. See also: abnormal behaviour and amphetamines, 134 of bank voles, 136–7 of calves, 140–42 of clouded leopards, 135 deﬁnition, 132, 137, 139 of a dingo, 134 diversity, 144 functions, 140–44 of gerbils, 138–9 of horses, 141–2 and hunger, 135, 138 and neural damage, 136–7 of pigs, 138, 142 of sows, 132–3, 135–6 and stress responses, 134–5 stereotypies. See: stereotyped behaviour Stevenson, P., 5, 305 Stoics, 12, 19, 30 Stolba, A., 170–72, 175, 207, 210, 269, 306 stomach ulcers. See: ulceration

6/27/2008 12:45:27 PM

Index

Stookey, J.M., 155, 304 stress, 104–24, 162. See also: HPA system; SAM system and abnormal behaviour, 145 and aggression, 116 and alarm pheromone, 163–4 in animal shelters, 117–18, 156–7, 271 and animal welfare, 64, 73–7, 80, 101, 111–24, 190, 222, 231, 279, 283–4 Cannon, 104–5, 123 and crowding, 232 and cytokines, 108 deﬁnitions, 104, 107 and displacement activities, 129–30 everyday meaning, 104, 123 and fear, 119, 279 heat, 100 and housing, 114–18 HPA system, 106–9, 112–24, 135, 137, 142, 153, 161, 186, 283 and human handling, 99, 115, 124, 189 and hunting, 111–12, 119 and pain, 112–13, 119 prediction and control, 109 SAM system, 105–7, 109, 118, 120–21, 123–4, 153, 161 Selye, 105–9, 122–3 and social isolation, 127, 185–6 and stereotyped behaviour, 134, 137, 143 stress-induced hyperthermia, 163 in unfamiliar enclosures, 186 Stricklin, W.R., 300 Stubbs, G., 48–9 subjective experience, 82, 166, 167. See also: affective states sucking, 128, 132, 209–10 calves, 121, 188 non-nutritive, 141 redirected, 121, 132, 141, 184, 212 suffering, 60, 68, 71, 233, 267, 271 and animal production, 58, 63 and animal welfare, 67, 68–9, 73–6, 101, 173, 230–32, 262, 278 and moral concern, 18, 20, 68, 277 Sundberg, P.L., 299 survival/longevity 280. See also: death and animal welfare, 71, 74–7, 89–90, 97–8, 100, 121, 172, 225, 231, 243–5, 262, 268 in animal welfare standards, 249–52, 270 and natural selection, 159, 160, 213, 227 Swanson, J., 98, 247, 288 Sweden, 34, 66, 85–7, 88, 218–19, 271 sympathetic nervous system, 105, 106, 108, 142, 152, 163

Dfraser_Index.indd 323

323

tail-biting (pigs), 128 tail-chewing (clouded leopards), 114 tail-docking (lambs), 153, 154, 224 Tannenbaum, J., 260, 261, 262–3, 273, 306 Tauson, R., 85, 90, 271–2, 280, 306 Taylor, G.B., 68, 70, 306 temperature (ambient), 130, 196, 226, 238 for chickens, 90 for pigs, 195 shivering, 176–7, 179 in stress responses, 108–10 Terlouw, E.M.C., 135, 306 termite, 177, 184, 186 tethering and bull-baiting, 17 bulls, 117, 119–20, 122 dairy cows, 87 sows, 98, 115–16, 135 Theophrastus, 11, 13 Thomas Aquinas (Saint), 30, 31 Thomas, K.V., 41, 306 Thompson, B.K., 196, 301 Thompson, P.B., 45, 47, 267, 306 Thomson, J., 51–2, 306 Thorpe, W.H., 62, 64–5, 73, 81, 191, 192, 306 threshold value (in stress responses), 121–2 thymus, 106 ticks, 174, 177 tie-stalls (dairy cows), 87–8 tiger, 115, 119, 280 Tinbergen, N., 82–3, 129, 166, 306 Toates, F.M., 82, 123, 146, 211, 296, 306 tongue-rolling, 129, 140–42, 144, 212 tonic immobility, 150–53, 156, 264 torsion of the stomach and spleen (sows), 84–5, 95, 169, 221 trampled teats (dairy cows), 87–8 trapping (of animals), 65, 95 Troisi, A., 130, 306 Tuber, D.S., 186, 307 Tucker, C.B., 195–6, 307 Turner, D.C., 156–7, 297 Turner, E.S., 4, 20–21, 61, 63, 307 ulceration calves, 140–41, 169, 213 horses, 141–2 pigs, 281 rats, 109–10 and stress responses, 106 The Unheeded Cry (book), 36, 81, 166 United Egg Producers, 247–51, 270, 307 United Kingdom, 46, 61, 89, 111, 218–20, 233, 248

6/27/2008 12:45:27 PM

324

United States (of America), 5, 56, 65, 100, 102, 127, 220, 247–9 urination, 89, 148 Utilitarianism, 18–19, 68, 277 Uvnäs-Moberg, K., 161, 307 vaccines, 59, 70, 95 vacuum activity. See: abnormal behaviour values and animal welfare assessment, 242, 243–6, 249–59, 260–71 and animal welfare science, 23, 72–8, 221, 234–40, 277–9 versus facts, 7, 262–3, 271, 279 moral values, 263–4 preference values, 262–3, 266 in science, 6–8, 39–40, 266–74 van Zutphen, L.F.M., 102, 244, 272, 289, 307 Vancouver Aquarium, 2–3 veal calf. See: calf Veasey, J.S., 175–6, 307 vegetarianism, 10–11, 13–15, 19 Verga, M., 142, 300 veterinarians, 59, 60, 70, 77, 84, 89, 125, 232, 282 veterinary medicine, 80, 84–96 Victoria (Queen), 21 vigilance, 152, 153, 267 Virgil, 9, 44, 239, 298, 307 vitamin C, 176, 179, 208 vocalizations, 76 alarm calls, 149–50 begging calls, 149–50 cats, 160 ﬂycatchers, 150 hens, 160 pigs, 76, 146–8 rats, 159 separation calls, 146–8 vervet monkeys, 149 Voith, V.L., 118, 296 vole, 136–7 Voltaire, F.-M.A., 30, 307 von Borell, E.H., 299 Waddell, M.S., 91–2, 294 Waiblinger, S., 75–7, 307 Wallace, A.R., 34 Waran, N.K., 175, 307 Watson, J.B., 39, 81–3, 138, 166, 307 Weary, D.M., 146–8, 188, 195, 204–6, 224, 225, 226, 287, 294, 301, 303, 307

Dfraser_Index.indd 324

Index

weaving circus animals, 137 horses, 125 Weber, M., 7 Webster, J.[A.J.F.], 173, 233, 272, 307 Wegner, R.-M., 207 Weiss, J.M., 109–10, 307 Welch, D.A., 174, 304 well-being and animal welfare, 6, 22, 74, 101, 122, 267 Well-Being: The Foundations of Hedonic Psychology (book), 168 Welp, T., 153, 307 Wemelsfelder, F., 158, 181, 307, 308 Western thought regarding animals, 9–23, 28–40, 41–60, 61, 281 White, L., 41, 308 Whittemore, C.T., 172, 297 Widowski, T.M., 214, 308 Wiedenmayer, C., 138–9, 308 Wielebnowski, N.C., 114–15, 135, 308 Wiepkema, P., 140–44, 308 wildebeest, 150, 163 wolf in literature, 13, 29, 52 as predator, 111, 173 in Western culture, 3–5, 29 Wollaston, W., 19 Wondra, K.J., 281, 308 Wood-Gush, D.G.M., 155, 169–73, 175, 191, 207, 280, 297, 306, 308 Wordsworth, W., 53 World Health Organization, 95, 308 World Organization for Animal Health, 220 world-views, 8, 41–60, 81, 281–2. See also: Agrarianism; Industrialism; Pastoralism; Romanticism Wright, R., 28, 308 Würbel, H., 139, 183, 308 Yanacopoulo, A., 123, 308 Yerkes, R.M., 36–7, 81–2, 168, 308 Young, P.T., 82, 308 Young, R.J., 175, 185, 307, 308 Young, T., 20, 277, 308 zebra (grooming), 184, 186 zoo, 84, 98, 114, 119, 125, 126, 132–4, 175–6, 179, 184–6, 190, 223, 234, 280, 282 Zuckerman, S., 131, 308

6/27/2008 12:45:27 PM