Say Not to Say (Emerging Communication:Studies in New Technologies and Practices in Communication, 3)

SAY NOT TO SAY

Emerging Communication Studies in New Technologies and Practices in Communication Emerging Communication is publishing state-of-the-art papers that examine a broad range of issues in communication technology, theories. research, practices and applications. It presents the latest development in the field of traditional and computer-mediated canmunication with emphasis on novel technologies and theoretical work in this multidisciplinary area of pure and applied research. Since Emerging Communication seeks to be a general forum for advanced communication scholarship. it is especially interested in research whose significance crosses disciplinary and sub-field boundaries.

Editors-inchief Giuseppe Riva, Applied Technologv,for Neuro-P.yvcholog\*Lab.. Istituto L4ifi010giC~1 Italiano. Verbania,ha!\* Fabrizio Davide, Headquarters, Learning Senices, TELECOM ITALI.4 S.p.A., Rome. Italv Editorial Board Luigi Anolli, Universita Cattolica del Sacro Cuore. Milan, Itall. Cristina Botella, Universitat Jarrme I, Castellon, Spain Martin Holmberg, Linkoping Universih,, LinkGping, Sweden Ingemar Lundstrom,L inkoping Universiiy L inkoping, Sir lederI Salvatore Nicosia, Universih. oj-Tor Vergata. Rome, Itall. Brenda K . Wiederhold, CSPP Research and Senice Foundation. San Diego, C.4. CS.4

Volume 3 Earlier published in this series Vol. 1. G. Riva and F. Davide (Eds.), Communications Through Virtual Technologies: Identity. Community and

Technology in the Communication Age Vol. 2. G. Riva and C. Galimberti (Eds.). Towards CyberPsvchology: Mind. Cognition and Society in the Internet Age

ISSN: 1566-7677

Say Not to Say New Perspectives on Miscommunication Edited by

Luigi Anolli Centerfor Communication Psycho logy, Catholic University, Milan, Italy

Rita Ciceri Centerfor Communication Psychology, Catholic University, Milan, Italy

and

Giuseppe Riva Is tituto A uxologico Ita 1ian0, Milan, Italy

Amsterdam Berlin Oxford Tokyo Washington, DC

2002. The authors mentioned in the Table of Contents All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted, in any form or by any means. without the prior written permission from the publisher. ISBN I 58603 215 1 (10s Press) ISBN 4 274 90494 6 C3055 (Ohmsha) Library of Congress Control Number: 200 1099505

Publisher IOS Press Nieuwe Hemweg 6B 1013 BG Amsterdam The Netherlands fax: +3 1 20 620 34 19 e-mail: [email protected]

Di.vtributor in the UK and Ireland 1 0 s PresdLavis Marketing 73 Lime Walk Headington Oxford OX3 7.4D England fax: +44 1865 75 0079

Distributor in the USA and Canada 10s Press. Inc. 5795-G Burke Centre Parkway Burke. VA 2201 5 USA fax: + I 703 323 3668 e-mai I : iosbooks(ii iospress.com

Distributor in German?,..4ustria and Switzerland 1 0 s PresdLSL.de Gerichtsweg 28 D-04 103 Leipzig Germany fax: 149 341 995 4255

Distributor in Japan Ohmsha, Ltd. 3-1 Kanda Nishikicho C'hiyoda-ku. Tokyo 10 1 Japan fax: 181 3 3233 2426

LEGAL NOTICE The publisher is not responsible for the use which might be made of the following information. PRINTED IN THE NETHERLANDS

V

PREFACE

We live today in the Age of Information and of Communication because electric media instantly and constantly create a total field of interacting events in which all men participate. McLuhan, I964

Communication has come to be regarded as a symbol of the age in which we live. Talk is frequently of “communication explosion”, of “communication technology” revolution, even of “communication society”. But what is communication and how can we make sense of it? Many of us, when asked about communication, usually think of it as a mere information transmission from a source to a receiver by means of coding-encoding processes. However, different studies during the last fifty years clearly underlined that one thing is what is said, another is what is meant. In fact, many meanings are characterized by fuzzy boundaries that are defined by the context in which they are expressed. Within this perspective, meaning is not a fixed system of univocal correspondences between expression and content, but a set of possible inferences that have different probability degrees of realization. This vision has a simple consequence: it is possible to communicate only to the extent that speakers share beliefs, recognize reciprocal expectations, and accept rules for interaction which serve as necessary anchors in the development and understanding of utterances. To make this process even more complex is the observation that communicators have at their disposal a plurality of signaling systems, verbal and nonverbal. The acknowledgment of a plurality of communicative systems involves the necessity of examining how they may be related to each other and how they can make up a unitary totality. This is one of the main goals of Say not to Say: New perspectives on miscommunication. By defining a new research area - the “miscommunication psychology” - the Editors try to provide some guidelines to understand the processes by which we manage our experience - make meanings, construct identities and search for friends and partners through the analysis of miscommunication processes. According to the approach proposed by the Editors, miscommunication is considered not only as a defect or a mismatch, but also as a plus and as a powerful device in the hands of the communicators. In this sense, miscommunication can be considered as a chance, because it enhances the degrees of freedom available to communicators in their interaction.

This vision has a strong effect on the possibility of improving the communication process: communicators have the opportunity to manage their communicative strategs in the best possible way, given the contextual constraints and their respective encyclopaedia of knowledge. Through the choice of specific linguistic and extra-linguistic hints and clues - intonation, the word order in an utterance, an extra stress and so on - the speaker can convey different meanings and adapt them to a specific context. Starting f'rom these premises, the topics discussed directly involve critical issues for many application areas such as marketing, advertising, group management. media and new media development, and are presented with scientific competence and suggestions for actual use. In fact the Editors included throughout the book a wide variety of chapters, that discussed topics as seduction, irony or deception with a careful attention to the links between the cognitive and social aspects of communication. Real-world examples offer clear, relevant, and accessible demonstrations of the points discussed. Appropriate for psychologists, communication and media researchers, students and even the casual reader, this book provides a unique perspective on communication in today's society and enables us to see more clearly how analyzing miscommunication can help us in better understanding the different facets of the communication experience.

Giovanni Ferrero Fondazione "Piera,Pietro e Giovanni Ferrero " Aha, Itall.

vii

POSTFACE

Since teachers have the most direct, sustained contact with students and considerable control over what is taught and the climate for learning, improving teachers ’ knowledge, skill and disposition through professional development is a critical step in improving student achievement. King and Newman, 2000

Teaching is a very difficult and demanding profession. Good professors must have a deep knowledge of their subject areas, but this is not enough: they must be skilled, too, in the difficult process of communicating their knowledge and interacting with the students. However, a good researcher or a good professional is not always also an effective communicator. This is why, everyone who is concerned about the quality of education should consider carefully the adoption of policies and practices that support teachers in the acquisition of communicative skills. In particular, in everyday communication teachers face the challenge of effectively communicating and sharing their knowledge with students, in order to engage them actively in the process of making sense out of their educational experience. This activity can be thought of as a contingent and dynamic activity strictly related to the situation in which it happens. It is a situation where the meanings are by definition an instable entity: it is not taken and given in a fixed and automatic way once and for all, but is built and shared though social interaction. In this context a significant role is played by communication: understanding how to improve the communication skills is a significant step in enhancing the efficacy of the educational experience. This book is a first step toward this goal. By discussing the different aspects of communication, the Editors underline the complexity of this process. In order to be effective, the teacher and the students should be able to share the same core beliefs, recognize reciprocal expectations, and define common rules for interaction. Many different components are involved in this process - semiotic choices, discourse genre and conventions, the aims of the interlocutors and the like - and all of them mutually affect each other during all the teachindlearning processes. Within this framework, the novelty of Say not to Say: New perspectives on miscommunication is the different focus given to miscommunication: it is not only a possible problem but also an enhancement of the communicative possibilities available to

...

v111

the teachers. According to the Miscommunication as a CHance Theory - MaCHT - presented in this volume, a strategic use of miscommunication may enhance the degrees of freedom available to the teachers during their lessons. In fact, it includes a set of communicative acts like exaggerating, flattering, joking, kidding, speaking sarcastically, euphemistically or metaphorically, that when correctly used can improve the communicative eficacy. However, Say not to Sa-v: New perspectives on miscommunication does more than present and review current communication theories. It is also important to realize that it offers a series of recommendations, based on extensive evidence from research, about how communicators can improve the eficacy of their presentations. Following these, the teacher may target hisher communicative strategy by using simple linguistic and extralinguistic tools such as intonation or an extra stress. In conclusion, both the educators who want to improve their communication skills and the readers who are interested in exploring the miscommunication area. should consider this volume as an interesting starting point.

Sergio Zuninelli Rector Magn@cus, Universita Cattolica del Sacro Cuore Milan, Italt.

ix

INTRODUCTION Communication theorists generallyfocus more on the study of message-makingas a process, whereas semioticians center their attention more on what a message means and on how it creates meaning Danesi, 1994

What is communication? If we check Webster‘s New World Dictionary, 2nd College Edition, we can find the following definitions: “n. I . the act of transmitting 2. a) a giving or exchanging of information, signals or messages by talk, gestures, writing, etc. b) information, message, etc. 3. close, sympathetic relationship. 4. a means of communicating; specif., a) Epl.] a system for sending and receiving messages, as by telephone, telegraph, radio, etc. b) [PZ.] a system as of routes for moving troops and materiel c) a passage or way of getting from one place to another 5. [oftenpl. with sing. v.] a) the art of expressing ideas, esp. in speech and writing b) the science of transmitting information, esp. in symbols.” The number of definitions and the differences between them show clearly the difficulty of describing the communication structure. In fact, even after many attempts, a communication theory able to address all the features of this process is still lacking. The earliest communication frameworks were based on the process of signal transmission in telecommunication systems. Shannon and Weaver’s model is still the best known and most widely used. According to authors, communication is defined as the transfer of information - message - from a transmitter - source - to a receiver in the form of a signal, which is sometimes modified by disturbance - noise - in the transmission system itself. One undeniable advantage of technical communication models is that they provide a general description of communication. On the other hand, however, these models cannot take account either of the specifically linguistic features of verbal language, or of the process of sense-making typical of the communicative processes. One of the clearest theoretical rejections of the analogy between verbal communication and physical data transmission has come from Roman Jakobson, who links these six physical components of communication - sender, message, receiver, context, code, contact - to six linguistic functions - expressive, connotative, phatic, metalinguistic, denotative-referential,poetic - which establish the intentionality of the communicative act. However, the linguistic models of communication also proved unsatisfactory. In fact communication, though basically a linguistic phenomenon, is also affected by the psychosocial relationships between the subjects involved in it. Given these difficulties, some theorists used a different approach: they differentiated between standard communication and miscommunication. According to this vision, subjects are characterized by a “natural stance” towards language, as if communicating were a natural gift. However, this approach too has a strong limitation: the idea of a natural attitude towards language ends up by not recognizing the personal responsibility of the

communicator, given that the message would be automatically guaranteed by objective conditions. This book is intended to be a first attempt to overcome these frameworks. In particular, the book, accepting a difficult challenge, tries to define the background of a unitary theory of communication where communication and miscommunication are simply the two faces of the same coin. In particular, the main message of the book is the following: the borderline between what is communicated and what is miscommunicated cannot be split up and partitioned in two separate and discrete fields. In fact, miscommunication is everywhere, because it belongs to the communicative exchanges of everyday life. Miscommunication is central to interpersonal communication. Every day we are constantly confionted with the handling of miscommunication at every stage of the communication process, handling that will influence the effectiveness of the message and the relationships on which those messages impact. This is why this book tries to provide the reader with worthwhile options for a vast array of interpersonal situations - deceprion, irony. humor, seduction, computer mediated communication - and discusses the theory. research, and evidence bearing on the miscommunication process. After completing this text, the reader should thus be better equipped to make more reasoned. more reasonabfe and more effective communication decisions. We have put a great deal of thought and effort into the definition of the structure of the book and the sequence of the contributions, so that those in search of a specific reading path will be rewarded. To this end we have divided the book into four main Sections comprising I 1 chapters overall: 1 Towards a definition of miscommunication:A theoretical approach 2 Pretending to communicate:Deception, seduction and equivocation 3 Communicate to pretend: Iron?,and humor 4 Communicate in pretending: Computer mediated communication.

Each chapter begins with a brief abstract and a table of contents that help the reader to identify the relationships between the section's chapters. The starting point of the book is the assumption that a viable theory of communication has also to explain miscommunication in its different forms. Following this perspective, in Section 1 - Towards a definition of miscommunication:A theoretical approach - Anolli and Reboul try to outline some general principles that can connect communication and miscommunication processes in a coherent theoretical perspective. In particular Anolli, in the Chapter I , sketches out the miscommunication as a chance theon. (MaCHT). His purpose is to suggest a new unitary model of communication that embraces both communicative and miscommunicative phenomena. Within this perspective, he proposes a definition of miscommunication as "say not to say": the chapter tries to overcome the standard concept of miscommunication as a lack, fault and violation of rules, considering not only its negative aspects, but also the positive ones in the interaction. Reboul, expanding this view in Chapter 2, analyzes two well-known views of linguistic communication: the code model and the hypothesis of semantic transparency. After a quick analysis of their contents, the author describes how the most recent pragmatic theories have discarded them. This enables them to account for the whole range of communication mishaps. from mere misunderstandings to sophisticated deception strategies. In Section 2 - Pretending to communicate: Deception, seduction and eqirr\>ocation- the attention of the authors moves to the analysis of deception. seduction and equivocation. In

xi

Chapter 3, Anolli, Balconi and Ciceri discuss a general model of deception - the Deceptive Miscommunication Theory (DeMiT) - that can explicate both the main characteristics of deceptive communication and the local management of the deceptive message in its different expressions. Seductive interaction is the focus of Chapter 4, proposed by Ciceri. This process is here analyzed as a flexible plurality of behavioral patterns, corresponding to the variety of communicative intentions: exhibition, approaching the partner, deepening reciprocal knowledge, and reaching a level of intimacy. Particularly, this approach makes it possible to analyze the connections between different systems of expression (verbal and nonverbal) and to describe several seductive strategies of obliquity and to disguise tactical communication, which are defined as miscommunication forms. In the final chapter of this Section, Chapter 5, Fernandez-Dols, Carrera and Casado try to give an explanation of a relevant inconsistency: the difference between the patterns of representation of emotional facial expression in most of the available art, and the patterns of emotional facial expression outlined by contemporary studies on recognition of emotions. Starting from the theoretical background presented in the first Section, Section 3 Communicate to pretend: Irony and humor - presents a framework for the analysis of Irony and Humor. Anolli, Infantino and Ciceri open this Section with a new theoretical perspective by proposing the fencing game (or irony situation) model. In Chapter 6 irony is not considered only as a comment or remark at a linguistic level but also as a complex communicative interaction between interlocutors, depending on contextual constraints and opportunities. Humor and Irony are the focus of Chapter 7. Attardo investigates the “performance” of humor by examining the motivations for a subject to use irony, and the responses that the interactant produces to it, which range from mode adoption to ignoring it. The last chapter of Section 3, Chapter 8, deals with the risks and rewards of ironic communication. In this chapter, Gibbs and Colston argue that irony cannot be characterized simply as having positive or negative social impact, but can serve multiple communicative purposes, depending on the social context and aims of the conversational participants. The final Section - Communicate in pretending: Computer mediated communication is devoted to the analysis of Computer Mediated Communication, CMC. In fact, this emerging form of communication may be described as a particular form of miscommunication: a necessarily “pared-down” form of conversation, which lacks the many features and some rules on which traditional forms of interaction depend. What are its characteristics? In Chapter 9, Riva outlines how CMC users are able to make order and create relationships out of the miscommunication processes typical of this medium. Moreover, it presented the emerging forms of CMC - instant messaging, shared hypermedia, weblogs and graphical chats - and their possible social and communicative effects. In Chapter 10, Mantovani addresses the effects of CMC on the development of interpersonal attraction, aiming at identifying the specific features that this process has in cyberspace. In fact, as clearly underlined by the author, CMC does not only support emotional and intensely involving communication between people, but it is also characterized by a specific allure, which makes it intriguing for many users. The last chapter, Chapter 11, focuses on the concealment in self-presentation, typical of CMC. To understand the characteristics of this process, Mabry analyzed the message contents of men and women participating in asynchronous online discussion. Research findings indicated that gender identity and communicator status were strongly related to message openness.

xii

The eleven contributions selected are among the first scientific attempts to take a serious look at the different aspects of miscommunication. However, the authors did not start from scratch. Social sciences have a broad knowledge of the different factors that characterize communication processes. So, the chapters proposed for this book are descriptive and theoretical-oriented in nature. The wide array of research areas represented here strengthens the idea of miscommunication as a research field strictly integrated with the main flow of communication research. The Editors gratefully acknowledge the assistance of a number of people and institutions without whose help this project could not have been carried out. We have benefited from the friendship with Paul Ekman, Magnus Magnusson and Klaus Scherer and from their long research in this field. We also acknowledge the Fondazione Piera. Pietro e Giovanni Ferrero, Alba, Italy, for the continuous support given to the Centre-for Studies and Research in Communication Psycholop research (http://www.psicocomunicazione.net). We are grateful to Valentino Zurloni, too, who helped the Editors in the difficult editorial process required in the preparation of this book. A final thank you goes to Prof. Eugenia Scabini, Dean of the Faculty of Psychology at the Universitti Cattolica del Sacro Cuore, Milan, Italy, and to Prof, Fernando Dogana, Head of the Department of Psychology at the same university, who supported the Editors in their research work. We want to thank the colleagues of the Centre .for Studies and Research in Communication Psycholoar for their help in carrying out the research lines discussed in this book. In particular, we are grateful to Alessia Agliati, Paola Colombo. Claudio Lucchiari, and Olivia Realdon for their support. Moreover, Riva wants to thank his directors and colleagues at the Istituto Auxologico Italiano, one of the leading research and health care center in Italy. In particular, his thanks go to the President of the Institute, Prof. Giovanni Ancarani, to the General Manager. Dr. Mario Colombo. and to the Scientific Director, Prof. Albert0 Zanchetti. Some of the chapters included in this book are the results of the following Italian Ministry of University and Research grants: Processi comunicativi in ambienti ad aha lirtualita (D 1 -2000: Communicative processes in high virtualih environments), and Cornurticazione linguistica ed extra-linpistica ( D I - 200 I Linguistic and extra-linguistic communication). In the end, we hope that the contents of this book will stimulate further integrated research in the communication area. In particular, we hope that in the near future miscommunication will be considered not only as a defect or a mismatch, but also as a plus and as a powerful device in the hands of communicators. Luigi Anolli, Ph.D. Universita Cattolica Milan, Itall.

Rita Ciceri, Ph.D. Universita CattoI ica Milan, Itall.

Giuseppe Riva. Ph.D. Istituto Aurologico Italiano. Milan, itall. Wniversita Cattolica Milan, Itail,

CONTRIBUTORS Luigi ANOLLI Centerf o r Communication Psychology, Catholic University Milan, Italy Salvatore ATTARDO Department of English, Youngstown State University Youngstown, OH, USA Michela BALCONI Centerf o r Communication Psychology, Catholic University Milan, Italy Pilar CARRERA Departamento de Psicologia Social y Metodologia, Universidad Autbnoma de Madrid Madrid, Spain Cristina CASADO Departamento de Psicologia Social y Metodologia, Universidad Autbnoma de Madrid Madrid, Spain Rita CICERI Centerf o r Communication Psychology, Catholic University Milan, Italy Herbert L. COLSTON Psychology Department, University of Wisconsin Kenosha, WI, USA Jose-Miguel FERNANDEZ-DOLS Departamento de Psicologia Social y Metodologia, Universidad Autonbma de Madrid Madrid, Spain Raymond W. GIBBS Psychology Department, University of California Santa Cruz, CA, USA Maria Giaele INFANTINO Centerf o r Communication Psychology, Catholic University Milan, Italy Edward A. MABRY Department of Communication, University of Wisconsin Milwaukee, WI, USA

Fabrizia MANTOVANI Center.forCommunication Psycholo
xv

Contents Preface, G. Ferrero Postface, S. Zaninelli Introduction, L. Anolli, R. Ciceri and G. Riva Contributors

V

vii ix

...

Xlll

Section I - Towards a Definition of Miscommunication: A Theoretical Approach 1. MaCHT - Miscommunication as CHance Theory: Toward a Unitary Theory of 3 Communication and Miscommunication, L. Anolli 2. Semantic Transparency, Semantic Opacity, States of Affairs, Mental States and Speech Acts, A. Reboul 43 Section I1 - Pretending to Communicate: Deception, Seduction and Equivocation 3. Deceptive Miscommunication Theory (DeMiT): A New Model for the Analysis 73 of Deceptive Communication, L. Anolli, M. Balconi and R. Ciceri 4. Seductive Communication: Paradoxical Exhibition, Obliquity and Non Verbal Synchronization, R. Ciceri 101 5. The Meaning of Expression: Views fi-om Art and Other Sources, 117 J.-M. Fernandez-Doh, P. Carrera and C. Casado Section I11 - Communicate to Pretend: Irony and Humor 6. “You’re a Real Genius!”: Irony as a Miscommunication Design, L. Anolli, 135 M.G. Infantino and R. Ciceri 7. Humor and Irony in Interaction: From Mode Adoption to Failure of Detection, S. Attardo 159 8. The Risks and Rewards of Ironic Communication, R. W. Gibbs and H.L. Colston 181 Section IV - Communicate in Pretending: Computer Mediated Communication 9. Communicating in CMC: Making Order Out of Miscommunication, G. Riva 197 10. Cyber-Attraction: The Emergence of Computer-Mediated Communication in the Development of Interpersonal Relationships, F. Mantovani 229 1 1. Ambiguous Self-Identification and Sincere Communication in CMC, E.A. Mabry 247 Author Index

265

This page intentionally left blank

SECTION I TOWARDS A DEFINITION OF MISCOMMUNICATION: A THEORETICAL APPROACH The view that word meanings are given by sets of attributes, or features, has dominated discussions of language and mind for centuries. Psychological research on the mental representation of meaning has focused primarily on qualities of the attributes. Many studies have examined, for example, whether semantic representations only contain necessary attributes. Answers to this question have ranged from a qualified yes to a resounding no. Although there have been disagreements on the nature of attributes, most investigators have assumed that word meanings are composed of attributes, semantic markers, or the like. But even this idea has been challenged. One assault on decompositional theories of meaning proposes that word meanings are not given by semantic attributes; another proposes that some words, such as proper names, do not have meanings at all. From a psychological perspective, two issues are at stake, and they need not stand or fall together: One is the status of componential semantic representations in the language system's permanent repository of word meanings; the other is the status of these representations in comprehension. McNamara and Miller, I988

This page intentionally left blank

3

Say Not to Say L. Anolli et al. (Eds.) IOS Press, 2002

1

MaCHT - Miscommunication as CHance Theory: Toward a unitary theory of communication and miscommunication Luigi ANOLLI

Abstract: Traditional psychology considers communication in terms of semantic and intentional transparency. This idea should be overcome, since the borderline between what is communicated and what is miscommunicated cannot be split up and partitioned in two separate and discrete domains. The starting point of this chapter is based on the assumption that a viable theory of communication has also to explain miscommunication in its different forms. In this perspective, communication sets up a unique and global category, including miscommunication phenomena, coming to the eventuality of not communicating. In particular, our aim is to outline some general principles that might connect communication and miscommunication processes to each other in a global, parsimonious and coherent theoretical perspective. This chapter intends to sketch out a miscommunication as a chance theoty (MaCHT). The purpose is to suggest a new definition of miscommunication as "say not to say". In this perspective, the problem at issue is to overcome the traditional concept of miscommunication as a lack, fault and violation of rules, considering not only its negative aspects, but also the positive ones.

Contents

1.6 1.7 1.8

The reasons for studyng miscommunication. .................... Beyond the communicative transparency ideal. ................. Semantic instability as communicative flexibility and a source of miscommunication.. ..................................... Meaning stability as context regularity and a source of mutual understanding. ......................................................... The plurality of signaling systems and the semantic synchrony. ............................................................. The intention design and communicative synchrony.. .......... Conclusions: toward a "re-definition" of miscommunication.. . Acknowledgements. ..................................................

20 29 37 38

1.9

References. ............................................................

38

1.1 1.2 1.3

1.4 1.5

4

7 9 17

4

1.1

L . Anoili ,’MaCHT

The reasons for studying miscommunication

Nowadays miscommunication has become a significant topic in linguistic and communication psychology, because it is a subject of interesting debate in the border area between semantics and pragmatics with strong relational implications. Among other scholars, we can mention Cupach and Spitzberg [ 11, Fraser [2], Mortensen [3]. Parret [4]. Young I S ] , Markova, Graumann, and Foppa [ 6 ] . Certainly, it is a truism to assume that communication can be successfbl without being explicit, perfect and evident. People can satisfactorily communicate with each other without being aware of the full transfer of information and without being capable of fulfilling completely their communicative intentions. Moreover, to understand the meaning of an utterance, it is not only sufficient to understand the transparent semantic content on the basis of its truth-conditional features. The common-sense idea of communication in terms of semantic and intentional transparency should be overcome, since communication is always a risky activity. According to folk psychology, persons are characterized by a “natural stance” towards language, as if communicating were a natural gift, grounded on the transparency of signs, as if communication were a datum, not a process in which the assumptions are neither totally foreseeable nor deducible from the preceding interaction. The idea of a natural attitude towards language ends up by not recognizing the personal responsibility of the communicator, given that the message would be automatically guaranteed by objective conditions. The borderline between what is communicated and what is miscommunicated cannot be split up and partitioned in two separate and discrete fields. This territory is rich in mystery. and is a challenge for any and every communication theory. In fact, miscommunication is everywhere, because it belongs to the communicative exchanges of everyday life. So, we need to specify the theoretical perspective within which we take the meaning of “miscommunication”. We intend “miscommunication” in a broad sense as a very large set of communicative phenomena and processes. According to this perspective. it not only includes its standard meaning such as a lacking, defect and violation of communicative rules [3], and “the dark side of interpersonal communication” [4], but also mismatching interpretation, as well as misrepresentation of information. Miscommunication also covers misunderstanding between speaker and addressee when they do not share the same culture and have different interpretative patterns [ 5 ] . Fraser [2, pp. 144-1451 widens the meaning of miscommunication, to the extent that it comprehends a set of communicative acts like exaggerating, flattering, insulting, joking, kidding, lying, speaking sarcastically, euphemistically or metaphorically, defaming with innuendo, fabricating, dissembling, misleading, understating and so on. Concerning this. he introduces the notion of infelicitous communication for cases of misrepresentation of the message. In turn. Parret [4], although claiming to hold a standard view of miscommunication, gives more significance to the pretending to communicate, that tackles a range of communicative phenomena like seduction, simulation, deception and manipulation which could belong to the miscommunication field. However, according to the point of view herein adopted, it is necessary to enlarge the meaning of miscommunication and take into consideration not only its negative aspects. but also the positive ones. Miscommunication is not plainly a defect or a mismatch: actually, it is also a chance, because it enhances the degrees of freedom available to the communicators in their interaction. Communication is almost always a mixture of explicit and implicit aspects. because the meaning of any communicative act is not a mathematical formula but a clwter of pemantic pieces more or less connected to each other and more or

L. Anolli /MaCHT

5

less consistent with each other. In this sense, it is not an exaggeration to say that miscommunication is a question of luck, since it makes the communicative flow among the participants interesting, and sometimes intriguing. Roughly put, miscommunication has to be conceived not only as a simple minus but also as a plus and as a powerful device in the hands of the communicators. The starting point of this chapter is based on the assumption that a viable theory of communication must also explain miscommunication in its different forms, starting from figurative language to ironic and deceptive communication, arriving at pathological communication such as the schizophrenic one. In fact, processes and mechanisms which explain miscommunication cannot be different from those that explain communication. In this perspective, communication sets up a unique and global category, including miscommunication phenomena, reaching as far as the eventuality of not communicating. But, to not communicate, one should communicate that one does not want to communicate. Crudely put, we do not believe that miscommunication is simply a communication failure, since, in our opinion, a failure involves a sort of arrest of the communicative exchange. But, by default, communication can neither fail nor be silent. It consists in a continuous and articulated system of processes and phenomena which vary in a changeable and flexible manner in terms of participants’ conditions as far as the context of use allows. In the past, scholars have focused their attention on two main theoretical domains, unrelated to each other, as if they had tackled different communicative phenomena and processes. On the one hand, some scientists have been engaged in the philosophical or experimental field; on the other, many others have referred to clinical psychology or psychiatry. Within the first group, some scholars were extremely interested in the analysis and study of the so-called “default (or standard) communication”. They intended to examine and understand the patterns and structures involved in communicative exchanges between “normal” speakers in everyday life according to the cultural standards and models of a specific community, although using and following different points of view. They devised different and interesting frameworks, among which it is worth mentioning herein the mathematical theory of communication proposed by Shannon and Weaver [7], the ethnomethodological perspective (based on everyday-conversation) advanced by Cicourel [S], Garfinkel [S], Sacks [lo], and Schegloff [ l l , 121, Goffinank theory of social interaction [ 13, 14, 151, Grice’s theory of conversational implicatures [16], the speech act theory developed by Austin [17] and Searle [ls, 191, as well as the relevance theory of Sperber and Wilson [20]. The assumptions of this basic, broad line of thought claim that default communication is the standard, and miscommunication processes take place because of a deviation, or a violation - more or less significant - of this standard paradigm in any part of the communicative process and sequence. For instance, in Shannon and Weaver’s model [7] it is expected that noise as an interference could hinder or impede - in part or as a whole signal transmission from the sender to the receiver. According to McCornack [21], Jacobs, Brashers and Dawson [22], and Jacobs, Dawson, and Brashers [23], in Grice’s theory [ 161 the violation of one or more maxims can lead to the appearance of communicative distorted forms, ,like deceptive communication. Similarly, in Sperber and Wilson’s paradigm [20] the lack of consideration of the relevance principle produces reiterative, trite and sterile communicative acts, with no additional information, and under given circumstances, it generates problematic communication. On the other side, many clinical psychologists were extremely interested in the analysis and study of the so-called “pathological communication”, which reveals disturbed patterns

6

L. ..inolli /MaCHT

in communication and interaction. These patterns are forms which have more or less deviated from the cultural models of a determinate society, and are involved in different kinds of psychological disease. Among the different fixmeworks proposed to explicate the link between communicative patterns and psychopathological disturbance it is worth mentioning herein the double bind theory, advanced by Bateson [24, 251, Bateson, Jackson, Haley, and Weakland [26], the structural family therapy treatment approach. developed by Minuchin [27], Minuchin and Nichols 1281, the systemic model of family therapy, proposed by Selvini [29]. Selvini, Cirillo, Selvini, and Sorrentino [30], Satir's model of family therapy [31]. On the whole, these researchers had determined to identi6 the pathological phenomena of communication involved in psychological disorders, and to treat the former as the cause of the latter. These scholars were not really interested in the study of communication as a standard process. With reference to these two basic lines of thought, divergent from each other, our aim is to examine the communicative act as a whole, trying to explain the continuity of communicative processes, from the understanding of literal meaning to pathological forms of communication. As a principle, we think that the mechanisms and processes involved in default communication design are not different from the ones involved in miscommunication design. In particular, the aim of this chapter is to outline some general principles that might connect together communication and miscommunication processes in a global. parsimonious and coherent theoretical perspective. The purpose herein is to suggest a new definition of miscommunication as "sav not to sa-v". In this perspective, as already said, the problem at issue is to overcome the traditional concept of miscommunication as a lack. fault, and violation of rules [3], and as the "dark side of interpersonal communication" [ I ] . This standard view can be integrated with the concept of miscommunication as a communicative chance in everyday interaction. It could also be an opportunih. in the interaction between the participants both in the Machiavellian sense and in a more positive perspective as an enhancement of the communicative possibilities available to the participants. Within this framework we are in line with the paradigm of the "staged communication act" proposed by Clark [32] and pursued by Gibbs [33]. The key ingredient in staged communication and miscommunication acts is pretence as the basic condition in ironic, deceptive or evasive communication. This chapter intends to sketch out a miscommunication as a chance theon, (MaCHT). In order to delineate this theoretical model, the following subjects will be tackled and analyzed: a) the overcoming of the explicit and transparent communication ideal. suggested by the standard communication conception; in this topic we will examine the flexibility and variability of meaning as a source of communication and miscommunication; b) the analysis of context regularity as an assumption of meaning regularity; such regularity has been intended as a source of mutual understanding also in miscommunication conditions; c) the multiplicity of communication systems as a condition for stressing or weakening the communicative meaning according to the speaker's intention, the interaction with the addressee, and the context of use; in fact, in every communicative act we can determine and find out different verbal, prosodic. mimic. and gesture signaling systems; however. this multiplicity is co-ordinated by a central system. like a "central communicative processor", in order to guarantee the necessary semantic synchrony among the different signaling systems as a condition to get an unitary and consistent message; d) the intentional "game" between the participants; in this topic we will examine the intentional convergence between the speaker and the addressee. because intentional stance expresses itself in a graduated manner in the former, and in a multiplicit? of interpretations in the latter On thi\ ?round n.e will analye the communicati\e

L. Anolli /MaCHT

7

synchrony process as a general condition to participate effectively in communicative interaction.

1.2

Beyond the communicativetransparency ideal

Shannon and Weaver [7] have elaborated a mathematical model of communication as mere information transmission from a source to a receiver by means of coding-encoding processes. This framework presupposes a substantial semantic transparency, because the message, defined by the source is, in theory, the same as the one that reaches the receiver. Ideally and by default, the message received is a reliable and faithful photocopy of the one sent. Noise (i.e., the interference with another message that is going along the same channel) and a filtering activity (i.e., selection of signals as they reach the receiver) are the main risks for this information transmission process. The assumptions of the communication engineering of Shannon and Weaver [7] are given by the truth-conditional conception of meaning proposed by the Vienna Circle, in particular by Frege [34, 351 and Carnap [36, 371. According to truth-conditional (objectivistic) semantics, thought and language reflect the real world (or a possible world), and concepts are mental representations of objects and events in the world. As a consequence, linguistic expressions get their meaning only via their capacity to correspond to the real (or possible) world in a fixed and objective manner, independent of individual cognition. Meaning is grounded on truth conditions as objective features which a certain state of affairs in the real world (or in a possible world) must have, so that a certain sentence is true in that world. In this way meaning is independent of individual minds and it is built up by a complex of necessary and sufficient conditions, constant and formally codified (as definitional knowledge) by dictionaries which describe lemmas in a natural language. This is a perspective based on semantic clearness as the expression of an objectivistic point of view of reality (a kind of "God's eye" according to Lakoff's metaphor) [38, p. 1641. Austin [ 171 pointed out how this theoretical point of view seems quite unsustainable, by introducing the concept of descriptive fallacy for what pertains to the representation of a state of affairs: utterances are not a tool for reality description and cannot be analyzed and valued in terms of truth-conditions. They set up a locutionary act (of saying something), an illocutionary act (in saying something) and a perlocutionary act (by saying something). However, in his speech acts theory, Austin [ 171 maintained semantic transparency for what concerns mental states, since his framework has become an important device to study and value such states. In turn, Grice [ 161 has criticized semantic transparency in its objectivistic version, and has underlined the semantic indeterminacy of language. One thing is what is said, another is what is meant (that is, what it is intended to say), since the second one is more extended than the first one. The distance and, sometimes, the discrepancy between what is said and what is meant are overwhelmed by the so-called " implicatures", i.e. "authorized" inferential processes based on linguistic indexes [ 161. To determine what a speaker means, a listener must go beyond hisher understanding of what that speaker simply says. Furthermore, he has oriented scholars' attention towards the comprehension of meaning as an assumption of any communicative interaction in a deeply different perspective in comparison with the Vienna Circle one. In fact, Grice [ 161 has considered meaning in a subjective manner as what the speaker means, that is, hidher communicative intention. According to him, communicative exchange is a process in which a speaker has the intention to cause the receiver to think or do something, in such a way that the latter may,

8

L . Anolli / MaCHT

at the same time, recognize the former intention to cause that thought or action. This process of mutual and shared knowledge is based on the Co-operative principle ("Make your conversation a contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged"). This principle carries with it four conversational maxims: Maxim of Quantip: Make your contribution as informative as is required. but not more so, for the current purpose of the exchange. Maxim of Quality: Do not say anything you believe to be false or for which you lack adequate evidence. Maxim of Relation: Say only what is relevant for the current purposes of the conversation. Maxim qf Manner: Be brief, but avoid ambiguity and obscurity of expression.

But in this way, even if the semantic transparency is conceptually overcome, the risk emerges of an intentional transparenn,. In fact, the Co-operative principle involves a conscious sharing of the speaker's communicative intention (so called rn-intention as the speaker's intention to produce an effect in the listener by means of the hearer's recognition of that intention). P knows that A knows that P knows that A knows (and so forth endlessly) that P has a definite intention. In such a way, according to Clark and Clark [39], the listener successfully recognizes the speaker's intention ("authorized inference"). The Grice [ 161 account of communication involves mutual awareness between the speaker and the listener. This is not an awareness of the physical presence of each other, but an awareness of the mental state of each partner involved in the communicative act, that is. an awareness of the other's awareness. In the opinion of Gomez [40], this mental situation has the same mutually embedding structure as two mirrors confionted: each mirror reflects the other reflecting the other's reflection, ad infinitum. Within this perspective the Gricean maxims lead to an open, transparent, clear and explicit communication. Moreover, according to Grice [41], "communication is a complex kind of intention that is achieved or satisfied just by being recognized". That is. communication consists of a sender intending to cause a receiver to think or do something just by getting the receiver to recognize that the sender is trying to cause that thought or action. Within this theoretical perspective, communication is like a dance where one partner leads and the other follows. The relevance theory, outlined by Sperber and Wilson [20], also assumes that a single perspective, i.e. the speaker's thought and intention, is privileged in the interpretation of utterances. Speaker and addressee assume the speaker's thought as a common baseline. so that the addressee's inferences will converge with the speaker's thought. But this theoretical perspective does not seem to explain the dynamic and contingent nature of meaning-making. In particular, this communicative transparency ideal, both at semantic and intentional level. has been questioned by many researchers who have underlined communication complexity and difficulty. They have emphasized a set of communicative phenomena and processes, such as semantic opacity and condensation. meaning ambiguity and indeterminacy, vagueness and fuzziness. intentional gradation. the contingent nature of meaning-making. intertextual thematic systems and the like.

L. Anolli / MaCHT

9

1.3 Semantic instability as communicative flexibility and a source of miscommunication

In principle, any communication theory must face and solve the dilemma of opposition between meaning stability and meaning instability of a word, an utterance, or a gesture and the like. In the present paragraph we intend to analyze some significant aspects concerning semantic instability, deferring the analysis of meaning stability to the next paragraph. De Saussure had already considered meaning evolution in the course of time, and differentiated between the synchronic and diachronic levels of a natural language. But it is not necessary to resort to a long temporal period to see the meaning variability - and it is a large variability - of a word or an idiom. Very many communicative phenomena are implicated in this process. The meaning of a word or an utterance is not an easy, simple, not further decomposable (meaning as a semantic atom) reality, but a complex reality, articulated in features and decomposable in a limited number of semantic traits. Katz and Fodor [42], by following the method of Componential Analysis applied by Trubeczoij to the phoneme analysis, considered this limited set of semantic traits as a whole of necessary and suficient conditions (NSC) in order to define the meaning extension to a real world (or to a possible world). Such NSCs are governed by some basic principles: a) no trait can be cancelled, b) no trait can be added, c) every trait has the same semantic weight (absence of hierarchical relations within the semantic space), d) meaning is never graduated, but has clear-cut borders (according to binary logic). As a consequence, meaning either exists in its totality or it does not exist. This framework, which also follows the Hjelmslev’s [43] structuralist analysis of meaning, involves a neat separation between dictionary knowledge (as meaning constitutive properties) and encyclopaedia knowledge (as secondary and additional properties). Such a separation leads to the further sharp distinction between necessary (analytical) features and accidental ones: the former are essential and not cancelable, while the latter are contingent and factual. Within this perspective of trait semantics, meaning is conceived as a univocal, absolute, and fixed structure in its constitutive components. It seems to be a mathematical, discrete, and defined unit which can be combined to compose an utterance and build a discourse according to an additive logic. Moreover, meaning is understood as an objective entity, completely context-free and independent from speakers’ use in contingent situations. In fact, truth-conditions, with which meaning is identified, are objective features of a world (real or possible one). In actual fact, subsequent studies have pointed out that the NSC model cannot be sustained any longer, and have underlined semantic and communicative processes that lead to meaning variability and instability. Meaning is not taken in a fixed and automatic way, once and for all, but meaning-making can be thought of as a heterogeneous, contingent and dynamic activity in the world between two or more participants. Many different components which are involved in this process, such as semiotic choices, discourse genre and conventions, the aims of the interlocutors and the like, mutually affect each other in real-time to produce the emergent communicative act. 1.3.1 Some determinants of semantic instability

In order to uphold the MaCHT herein proposed, we need to point out some significant processes and phenomena of semantic instability and flexibility. In fact, miscommunication involves, by default, the possibility of continuous variations and shades of meaning. Without this semantic plasticity participants cannot “play” with the meaning,

L. .4nolli / MaCHT

10

and fit their communicative expressions to their intentions and desires. as well as to contextual constraints. 1.3.I . I

The possibiliv of canceling semantic traits

Following Cruse [44] and Violi [45], we firstly take into consideration the possibilit?, qf canceling semantic traits or defeasibility. Putnam [46] already suggested that it is not possible to determine semantic definitional features for natural categories or for artifacts, and that the NSC model cannot explain individual exceptions and variations. For instance, if we say that cats have four legs and a tail, how can we manage with a cat without a tail and with only three legs? Is it still a cat or not? Recently, it has been underlined by Geeraerts [47], Kleber [48], and Violi [45] that the vpical properties of a meaning are practically all subject to the possibility of cancellation. without changing the meaning itself. We briefly reaffirm that typical features are those that can be distinguished by diagnostic test of the adversative hut. For instance: ( 1)

(2)

* It is a bird but it.flies It is a bird hut it does not,f!\.

There is a semantic anomaly in (1) but not in (2). Within this perspective, for instance, in the meaning of bird there are typical features - and therefore cancelable or defeasible like flying ability (there are birds that do not fly, like ostriches), feathers (there are birds without feathers, like penguins), wings (there are birds without wings. like kiwis) [47]. But even the essential properties, that is those properties which are shared by every member of a category, can be cancelled in extraordinary situations. For example, in the past the whale was considered a fish, because it lives in the sea, while now is considered a mammal: in this case the essential property "fish" has been cancelled and substituted by "mammal". The possibility of canceling semantic traits is based on the conventional nature of meaning as a historically and culturally defined entity, and involves the overcoming of every ontological and natural conception of meaning. As for being conventional. it can also be negotiated. modified and culturally transformed. 1.3.I . 2 The f i c q * boirndaries of'meaning

Even the NSC model assumption that meaning has clear-cut borders as a discrete semantic unit has been criticized. After several studies, among which those worth mentioning are the ones proposed by Labov [49], and Lakov [50, 511, it is already recognized that many meanings are characterized by.fuzq boundaries. An utterance is fuzzy if it has the property of referential opacity. For instance, a lecture could be not bad. a girl may be rather preth', and John may have manji,friends.In these cases quantifiers and qualifiers allow us to carve the meaning of an expression more or less intensively and with a certain semantic shape. Within this perspective, according to Zadeh [52] and Zhang 1531 among others, meaning consists of a fic=i*set. that is a class of communicative units with a continuum of grades of membership. The same linguistic process is reduplicated with hedges which render a fuzzy reading. For instance, three o'clock is precise, but it becomes fuzzy when combined with around. as in: around three o'clock. Hedges like about, almost. or so behave in the same way as around. Also in the attributive clauses such as John is cleiper. Moq- is tall. the hedges \-en. and sonteit,hat modify the degree of fui~ziness.In fact. \*en.in John I F i m 7 . c-ie\vr- pushes

L. Anolli /MaCHT

11

the degree of meaning upwards; while somewhat in John is somewhat clever pushes the semantic value of clever downwards. Likewise, a word can have a meaning with fuzzy boundaries. In a classic study by Labov [49], based on the analysis of words referring to material artifacts like bowl, cup, mug, and dish, which compose the semantic field of “domestic containers for food and drink”, the results shown that the meaning of these words varied considerably in relation to the presence and relevance of some perceptual features like depth, breadth, height, presence of a handle and the like, as shown in figure 1.1. As things deviate progressively from a standard (or prototypical) type, we enter a semantic vagueness zone, where the same object could be, in turn, a bowl, a mug or a glass. The borderline between them is not clear-cut, but is undetermined and graded. It seems like a continuum more than a fair delimitation.

Figure 1.1. Series of cup-like objects. Adapted from Labov [49]

Semantic fuzziness phenomena were studied in depth with the prototypical fi-amework of meaning, proposed, among other scholars, by Givon [54], Kleiber [48], Posner [55], and Rosch [56, 571. Lakoff [50, 511, for instance, had already underlined the fuzzy nature of the semantic boundaries of meaning, considering the different position of a few members along the ranking of a category membership like bird-likeness, starting from robins and eagles to chickens and ducks and further on to penguins, ostriches, and pelicans. According to the prototype theory, even in its recently “extended” version advanced by Kleiber [48], and Geeraerts [47, 581, meaning in many circumstances does not symbolize a discrete and unitary category with clear-cut and closed boundaries, but represents a salient features pattern on the base of which we operate by inference and by partial similarity judgements. In the field of semantic vagueness it is worth mentioning the family resemblance phenomenon, underlined by Wittgenstein [59]. For instance, the meaning of game is not univocal but it spreads over a multiplicity of semantic subsets,, not always related to each other, as in Wittgenstein’s words: “Considerfor example the proceedings that we call “games”. I mean board-games, card-games, ball-games, Olympic games, and so on. What is common to them all ?--Don’tsay: “There must be

12

L. .4ndli / MaCHT

something common, or thev would not be called 'games'"-but look and see whether there is anvthing common to all.-For lfyou look at them you will not see something that is common to all, but similarities, relationships, and a whole series o f them at that. To repeat: don 't think, but look!--Lookfor example at board-games, with their multifarious relationships. Now pass to cardgames; here you .find many correspondences with the,first group, but many common .features drop out, and others appear. when we pass next to ball-games, much that is common is retained, but much is lost.-Are t h q all 'amusing'? Compare chess with noughts and crosses. Or is there always winning and losing, or competition between plajiers P Think o f patience. In ball-games there is winning and losing: but when a child throws his hafl at the wall and catches it again, this features has disappeared. Look at the parts played by skill and luck: and at the difference between ykill in chess and skill in tennis. Think now of games like ring-a-ring-a-roses: here is the element of amusement, but how man>*other characteristics features haw disappeared! And we can go through the man?', many other.groups oftgames in the same \-c.ai.;\tse can see how similarities c ~ o p up and disappear. .4nd the result of this examination is: we see a complicated network of similarities overlapping and criss-crossing: sometimes overall similarities. sometimes similarities of detail" [ p . 661

In this framework the meaning of game is not described by a complex of common and necessary traits but by a partial similarity shared by at least two or more game types. In this case, members of a specific category are connected to each other without the presence of common properties. 1.3. I . 3 Radial categories and semantic polisem!.

Referring to this perspective, Lakoff [38] has given an accurate analysis of the linguistic categorization system of Dyirbal. In this Australian indigenous language there are only four lexical categories: bajpi, balan, balam, bala, each one comprising totally heterogeneous meanings that cannot be described by referring to the classic categorization principles. For instance, the bayi category contains these items: men, kangaroos, possums, bats. most snakes and fishes, the moon, storms, rainbows, boomerangs etc. The balan category includes: women, dogs, most birds, some snakes and fishes, the sun and stars, fire. water etc. Lakoff [38] stresses that these categories are neither arbitrary nor random but they are regulated by a set of local similarities in terms of Dyirbal myths and beliefs. Therefore, we have to understand and manage complex categories made up by successive chains of items linked by local similarities. They can branch out along the so-called "radial categories". that is, categories "where there is a central case and conventionalized variations on it which cannot be predicted by general rules" [38, p. 841. Semantic fuzziness and vagueness phenomena like "family resemblance" and radial categories are neither isolated nor extraordinary, but they are related to the wider field of polisemjg. The polisemic word refers to a number of different meanings, semantically related to each other, each one, however, having its own autonomy. First of all. it is a matter of semantic polisem?, in which a word takes different levels of meaning. For instance, the Italian word .fresco (literally in English "fresh") implies three different meanings along three different levels of sense [45]: a ) temporal level of sense ("new. recent, just given": for instance. it is fresh news, it i.y a fresh vegeta Me). b ) positive state level of sense ("pure, uncontaminated, in optimal conditions": for instance. it is.freshf m i t - it is.fresh.fish,*freshmountain air), c ) thermal level of sense ("not warm": for instance. ihiy roont is ii.cvh. fhesh air t h i s nlor~rlirl
L. Anolli /MaCHT

13

As we can see, there is a partial overlapping area between a) and b) (fruit and vegetables) and between b) and c) (air). In this example, because of the plurality of different meanings the partially overlap, fresh constitutes a polisemic category and cannot have a single prototype, because a central case is not expected. In fact, polisemy provides for a basic meaning, which includes different courses of meaning. But there is also the analytic polisemy phenomenon, according to which a lexical category, still maintaining a basic intuitive unity, refers to many different referents [47]. For instance, the word bird, although constituting a univocal category with precise boundaries, refers to referents whereof each one is very different from the others (like sparrow, eagle, chicken, penguin, kiwi etc.). The great number of meaning routes of a polisemic word cannot be expected a priori, but can be explained only afterwards, with the connection between its basic meaning and secondary (derived) ones. 1.3.1.4 The gradability of the meaning

The assumption of parithetical value of the semantic traits expected by the NSC paradigm appeared to be unfounded too, according to a cluster of subsequent studies that have underlined the semantic gradability phenomenon. That is, the semantic traits of a word do not share the same semantic weight: there are essential properties (common to every member of a category) and typical properties (distinctive and specific of only some members of a category, which can increase their representativity and distinguish them from other members of the same category). The essential properties are criterion traits [44], that cannot be cancelled without negotiating the meaning of the word. For instance, for the bird category the essential features are only two: a) being oviparous; b) having a beak. All the other features are merely typical and can be cancelled without creating semantic modifications. A bird that cannot fly, like an ostrich or a penguin, is nevertheless a bird. The salience of typical properties is strictly connected to a given culture. Generally speaking, meanings are largely based on gradable or continuous criteria rather than all-or-nothing distinctions. Even in the case of words that, according to Sperber and Wilson [20], express “absolute” (or “well-defined”) concepts like bold or dead, they are gradable in English as in other natural languages. Thus completely dead, clearly dead, quite dead, nearly dead, almost dead, not quite dead, very dead, not very dead, hardly dead are scalar modifiers or intensifiers of dead. But, according to Thibault [60], the semantic variability of the word dead is also present in utterances like John is dead up top which does not mean that John is brain-dead but rather that he is stupid or dull. In this perspective dead can designate a biological condition, states of mind, moods, and attitudes towards someone or something. I .3.I . 5 The problem of literal meaning This semantic gradability allows us to overcome the distinction between the “literal” and the “nonliteral” meaning of a word or an utterance. According to the standard pragmatic theory, the former concerns the linguistic meaning as the combination of meanings of single words and it is the outcome of linguistic operations such as phonological, lexical, and syntactic operation. It is the basic (primary) meaning of a sentence, i.e. the plain, immediate and unproblematic meaning of a sentence (so-called “sentence meaning”). As a consequence, it has unconditional priority. In this perspective literal meaning is an abstraction, grounded on the hypothesis that words are like “meaning containers”, which are context-free (that is, they remain unchanged regardless of context of use). They are also sufficient to convey the sense of speech, that is, the “conduit metaphor”, according to

13

L. .3nolli MnCHT

which “ideas are conducted from one inner mind to another, transported in small compartments by the train of speech”, as Gibbs [61] has pointed out. On the contrary, the-figurative(or nonliteral) meaning is derived from the literal one and can be discovered by resorting to semantic substitution mechanisms (i.e. when nonliteral meaning takes the place of the literal one, as in the metaphoric expression, for example, surgeons are butchers). As a result, people can reach a translate and symbolic meaning, used in figurative speech by means of tropes of “speech figures”. such as the metaphor, metonymy, hyperbole, oxymora and the like. This standard theory is also shared by Grice [ 161 when he proposed the dual logic involved in conversation: the logic of language which concerns literal meaning (sentence meaning), and the logic qfconversation which applies to the pragmatic rules used by people to infer (“implicate”) what a speaker intends to convey (utterance meaning). “Conversational implicatures” require additional cognitive effort to go beyond the literal meaning of an utterance in order to grasp the speaker’s “intended meaning”. But such a distinction between literal and figurative meaning is not unproblematic and the concept of “literal meaning” has become a subject of theoretical revision upheld by the work of Gibbs [61] and Glucksberg [62]. The standard logic of the literal meaning follows a three-stage model: Derive the literal meaning of an utterance. h) Test the derived literal meaning against the conte-rt of the utterance. c) I f the literal meaning makes sense, accept that meaning as the utterance meaning, that is, the speaker’s intended meaning. I f it does not make sense, then seek an alternative. nonliteral meaning that does make sense in the conte-rt [62, p. 101. LI)

This standard model involves some testable implications. First, literal meaning is primary and basic, obvious and context-free (unproblematic). Second, literal meaning has unconditional priority, and it will always be derived first: only when it is “defective” can nonliteral meaning be attempted. Third, literal meaning is derived automatically, generated easily by the linguistic input (without conscious control). while nonliteral meaning is derived only optionally and requires additional cognitive effort. However. none of these statements has been verified empirically: on the contrary, the opposite has been verified. as Glucksberg [62] has pointed out. First of all, literal meaning is not exclusively the outcome of linguistic decoding. based on linguistic operations (i.e. phonological, lexical, and syntactic ones) but requires a semantic interpretation. The utterance cats are animals could be taken as granted and universally fixed in its literal meaning in any situation, independently from any context. However, the literal interpretation of this utterance is rather different if it is pronounced during a natural science lesson focusing on the distinction between animals and plants. or if it is said as an answer to the comment: “Our Pussy went on caterwauling last night“. As we already seen, context dependence especially concerns logical connectives. quantifiers (like some, j&*, many. and so on). deictic terms, qualitative adjectives. and pronouns. As Horman [63] underlines, there is a great semantic difference of literal meaning, for instance, of afew in: A .few people in the kitchen (four or five people) and in ‘4.few,people in the stadium (several thousand of people). As we have seen. the fact that words have a graded and gradable meaning inevitably requires reference to the context in order to be able to interpret correctly the semantic value of the literal meaning as well. A plain linguistic decoding is not sufficient to interpret the literal meaning of an utterance. Likewise, given that meaning (the literal one included) depends on the context. both literal and fi_eurati\.emeaning ha\.e the same t i m e of comprehension. as i t rewlts from the

L. Anolli /MaCHT

15

researches of Onishi and Murphy [64], and Pynte, Besson, Robichon, and Poli [65]. It is also unnecessary to understand literal meaning before proceeding to derive the nonliteral meaning. The interpretation of both these kinds of meaning is immediate, and follows the same time curve of computation, because they both show a parallel cognitive processing and activate the same inference mechanisms, as McElree and Nordlie [66] pointed out. As a consequence, the determination of so-called “literal meaning” is already the outcome of a choice of a certain interpretation among many other possible ones, all legitimate. The literal meaning of The cat is on the mat depends on the basic assumptions about such an utterance (for instance, in this case we do not refer to a floating mat). Such assumptions are not semantic in their nature but they belong to our knowledge encyclopedia, since they are derived from our experience within the culture of reference. Besides, these assumptions are not fixed, since each assumption involves other assumptions along an associative chain without end. Within this perspective, determining the literal meaning of every word or sentence is an impossible communicative operation. Each utterance offers an indefinite range of opportunities, among which the addressee has to choose the one that he/she considers more appropriate and effective to the context and communicative interaction (see below section 1.6.2). In this respect, so-called “literal meaning” is one of such possibilities. Actually, what people are interested in first of all is understanding the speaker’s communicative intention, which can be conveyed by any kind of utterance and language. They do not proceed to a philological study in order to discover the literal meaning, discriminating it from the figurative one. Instead, they are ready to grasp immediately what a speaker intends to communicate and, as a consequence, they are able to understand the variations, and also the nuances, of the meaning of an utterance. I .3.I . 6 The referring context Finally, the meaning of a word or an utterance or a gesture does not depend so much on a universal, abstract and fixed semantic system, isolated from the context, but it is strictly connected with the referring context. No meaning is totally foreseeable or definable a priori, because it depends on the context, defined by Searle [67] as background. Neither can meanings exist isolated from the context or without a context. In particular, the context calls attention to some specific properties of the meaning and, at the same time, it “narcotizes” other properties. Likewise, people’s interpretations of their own and other people’s expressions are not necessarily stable or constant over a period of time, but they can change as the context changes. For instance, the semantic features of book radically change according to the contexts, as in the following examples: (3) (4)

(5)

Peter is reading the book for the exam with attention (book as an object of study); Mary has lost the book that George gave her as a present (book as a souvenir in an inattentive and affective context); Paul, flown into a rage, has thrown the book against the glass of the window and broken it (book as a blunt tool in an anger context).

Within this perspective, the context performs a selection and determination of semantic traits, which can and must be activated in a specific communicative interaction, while it “narcotizes” other semantic traits, which can be activated in a different circumstance.

16

L. Anolli i MaCHT

The communicative phenomenon herein described is also involved in the so-called contextual risemanticization [45, p. 1491. In this case, a speaker can assign specific semantic traits to something that does not possess them of its own, but that obtains them thanks to a specific contingent situation. For instance, we can call chair a table, an empty

chest, a pile of books, if there are no free chairs, and we can also say: "don't take my seat". In this process the meaning of a word takes a double semantic valence: the table remains a table, but it performs the general function of a chair (that is, it permits the action of "sitting"). The contextual risemanticization stresses the great plasticity and high flexibility of the meaning, allowing a wide range in its use as the result of accurate processes of semantic adjustment. The whole of the semantic phenomena above mentioned enables us to criticize both the all-or-nothing dichotomous concept, and the notion of meaning as a fixed and objective entity, with clear-cut boundaries. Moreover, in the past this standard paradigm has deeply influenced the distinction and opposition between default communication and miscommunication. The latter was considered as an exception and a deviation from default values. On the contrary, the dynamic conception of meaning as a fuzzy set and gradable entity allows us to explain both communicative phenomena by default and those involved in the miscommunication domain. In particular, it encourages us to continue to view the miscommunication notion as a communicative chance. as MaCHT foresees. 1.3.2 Meaning as inferential outcome

Semantic instability and plasticity phenomena emphasize the fact that communicators, in order to understand each other, have necessarily to resort to some inferential processes. Meaning is not an immediate and fully evident datum, univocally correspondent to a real object or event, simply to be produced by a speaker and simply recovered by an addressee. Rather, it is created between the two communicators in a dynamic manner. Moreover, meaning is referred to a mental and cultural pattern that involves an inference process from the interlocutors, since it shows a specific point of view of reality. It does not only say something; it also points out and indexes how to intend what is said. In this way, meaning implies a semantic opaciry residue that needs to be elaborated and interpreted starting from some communicative cues and indices conveyed by the speaker. In fact, words, utterances, gestures and other non verbal signals are to be intended as communicative cues from which communicators can proceed to make suitable and opportune inferences through logical implication, analogy and similarity processes. Aristotle already defined the sign (ro qp.&'iov) as "a demonstrative premise which is necessary or generally accepted. That which coexist with something else, or before or after whose happening something else has happened, is a sign of that something's having happened or being (First analytic, 11, 27; 70a, 7). A classic example is given by the hypothetical enthymeme: "if she has milk, then she has given birth to a new-born child". Subsequently, the Stoics distinguished between "referring signs" (for instance, "where there is smoke, there is fire"), and "indicating signs" (for instance, "body movements are signs of the soul"). Within this perspective, meaning is not constituted by a fixed system of univocal correspondences between expression and content, but by a set of possible inferences that have different probability degrees of realization. In particular, from a realistic point of view. the meaning of a word, utterance or gesture is not to be considered as determined and established in a precise and definitive way by its necessary and sufficient components. because it does not refer to a metaphysical and abstract reality. In contrast. we ha\.e to

L. Anolli / MaCHT

17

conceive meaning as the semantic expression of our experience. Many different factors come into play here, since experience is a complex structure, made up at the same time of sensorial and perceptual, cognitive and emotional, inter-subjective and cultural processes, embedded and interconnected mutually with each other. Linguistic semantics and, more in general, communication are essentially based on our experience of the world, and, as a consequence, they will be influenced and guided by perceptual, cognitive and emotional constraints which control and manage the interaction with reality. Within this perspective, meaning involves a remarkable amount of instability and variability in order to fit and express in the right way the continuous flow and the great variety of our human experience. On one side, this semantic instability is at the basis of the meaning plasticity that allows speakers to use these meanings flexibly, according to their communicative intentions. In this way, we can observe a process of continuous remodeling of meanings in order to adapt them to referring contexts. Meaning is not a closed and univocal, fixed and hard pattern, but it refers to a mental and cultural pattern with a high degree of adaptability to contingent situations of interlocutors. Within fairly wide limits, meaning can be stretched, expanded or reduced according to the expectancies, beliefs and goals systems of the communicators. Essentially, it has a very wide degree of freedom and an opening value that allows a set of its possible and different expressions and interpretations, although such a set is neither undetermined nor chaotic nor random. In theory and in practice, it is consequently always possible to re-define and re-negotiate meanings in communicative interaction. On the other side, this semantic instability allows us to explain in a relevant way the phenomena and processes of default communication and miscommunication, reaching as far as processes of communicative distortion and pathology. Semantic instability concerns not only what is communicated but also what is said (Reboul, chapter 2; this volume). As a consequence, lexical and semantic ambiguities are common; sometimes, conversations are characterized by incomprehension and misunderstanding as well as confusion phenomena. Communicative games often appear and they make the conversation more interesting and intriguing.

1.4

Meaning stability as context regularity and a source of mutual understanding

The semantic instability and variability processes taken into consideration in the previous paragraph, if assumed exclusively, risk leading to a perspective of absence of communication and mutual comprehension. In themselves, taken singularly, they would generate a communicative disordered and chaotic situation. We would be compelled to live in a tower of Babel. Actually, semantic instability phenomena are compensated, completed and balanced by semantic stability processes, which make possible and explain the probabilities of order and regularity in meaning exchange. They are at the base of message intelligibility conditions and of mutual understanding between communicators. Competence is presumed of that meaning which is considered as shared inside a specific community of participants. In this sense we have to resume the notion of “code” but in a new perspective related to the traditional one. Semantic stability involves some kind of convention between the interlocutors, because of their sharing the same cultural identity. As culture is a mediation system which supplies people with a grid of categories, symbols, values, practices, and mental representations which enable them to read and interpret reality, it also provides the learning and sharing processes of signification and signaling systems. Such processes are to be considered as the

18

L. Anolli {'MaCHT

outcome of a long, complex and sometimes hard route to obtain consent and devise conventions between interlocutors (or conventionalizationprocess). On such a route the agreement on how something must be done or said is more significant than what is really done or said. I . 4.1 Context regularih' and production vs. re-productionprocesses A conventionalization process presupposes the active participation of the communicators,

as well as rules, practices, values, and meanings negotiation and sharing, although they may be local and temporary. It ends in working out a set of what Bruner calls communicative formats [68, 691, each of which is made of a structured sequence of interactive (verbal and non-verbal) exchanges, which allows communicators to reach a joint aim, follow the same procedures and rule systems, as well as share the meaning of what they are going to say or to do. Many communicative formats show a high and strong regularity structure, such as the greeting exchange, the call for apologies, a university exam, a political debate and the like. These cases can be called "standard (or default) formats", and are based on recognizing and accepting a shared system of rules and patterns. Usually, words and other communicative signs are "anchored" to a default format that makes their meaning foreseeable and definable. In particular, communicative formats oscillate from re-production processes to production processes. This oscillating motion is analogous to the one proposed by Bourdieu [70, 711 for cultural practice. On the one hand, given the nature of re-production processes, communicative formats tend towards repetition and recurrence in an almost stereotypical way, by creating proper "communicative routines" (obviously articulated in sub-routines), and also by establishing a continuity with semantic and communicative past conventions. These recurring and reduplicating processes are at the base of meaning stability and regularity. They are grounded on context regularip. since if it is true that contexts show a great deal of variability and unpredictability, then it is also true that in most cases contexts are structured and regular forms in our everyday experience of the world. On this platform. individuals build and share their scripts with reference to specific situations. The standard context is the context that presents a high routine regularity in the repetition of interactions, sequence of events and communicative exchanges. In such a way, we can assume that context regularity is equivalent to meaning regularih. No communicative act, like an utterance, a gesture and so on, exists that is context unrelated, since it is always indexed in a standard context of use. The meaning of such a communicative act is given by the mental representation of standard context regularity. On the other hand, thanks to production processes, communicative formats are neither totally constrained nor completely determined by the past and by context regularity, but they expect and produce variations and deviations as an effect of contingent conditions and novel unpredictable aspects that every communicative situation potentially brings on itself. Context regularity is the outcome of a historical and cultural process, not a logical necessity, and the past neither determines nor constrains the present, although the former steers the latter. In this way, contextual variations may always. in theory and in practice. take place related to the hic et nunc situation. These variations usually involve a signaling action and a communicative re-adjustment and re-negotiating process between the participants. Given this double process of re-production and production. regularity and variation are two essential constituents of meaning. reciprocallv presupposed and implied. Without

L. Anolli / MaCHT

19

regularity there is no awareness of variation and vice versa. That is, variation in itselfdoes not deny regularity, but demands it. As the standard communicative format is regular and stable, it contains in itself a range of unpredictability and interpretability conditions of meaning, as well as a set of applicability norms of communicative practices, since it provides the signification and indexing criteria. 1.4.2

Some more about code

Speaking about meaning regularity and stability as connected with context regularity and stability involves, as a consequence, the opportunity of speaking about the code notion under a new light. Some inferential frameworks of communication, in particular the relevance perspective proposed by Sperber and Wilson [20], are opposed to the communicative theory of code (see Reboul, chapter 2; this volume) and have concluded that it is also possible to communicate without resorting to the code. In its wide theoretic sense, a communicative code is to be intended as an organized set of rules, which allows communicators to associate in a biunivocal manner elements of one system with elements of another system. Referring to a classic example, the verbal code in language allows a competent speaker to associate biunivocally elements of the phonological system (specific sound combinations) with elements of the semantic system (specific mental images corresponding to these sound combinations). In a traditional standard perspective, referring, for instance, to the model of Shannon and Weaver [20], the code is conceived and formulated effectively as a mathematical construct and as a normative cognitive ideal, in which the biunivocal correspondence between the elements of the two systems is practically perfect and automatic. In this way the code, by itself, could to deal with and resolve every case of communicative indefiniteness and ambiguousness, as well as implicitness and semantic vagueness phenomena. According to such a framework, communication becomes a simple, immediate and automatic process of message encoding (ciphering) and decoding (deciphering). Concerning this strong version of code theory, Sperber and Wilson have correctly pointed out that it is untenable on a theoretical and practical level, since a perfect code simply does not exist and never will exist. No code is free of faults, interference, noises, and deviations. Nevertheless, it could be dangerous (perhaps, too dangerous!) to assume that communication can exist without any kind of code. Tommunication without code exists", according to the famous example of the two prisoners of different languages who, breaking stones back to back without seeing each other, succeed in communicating through the rhythm of their hammer strokes [20, p. 521. Actually, through the recognizable and shared rhythm of their hammer strokes, the two prisoners generate a local process of conventionalization, which allows them to convey intentions and meanings. Without some kind of communicative stability, created by sharing a set of regularities and a system of correspondences between the signs and the standard context of use, communication becomes impossible, since it would be totally unforeseeable, confusing and disordered. The scenario would be a chaotic tower of Babel. Therefore, we have to speak about a "code". However, the "code" herein underlined must intended as a limited communicative device that institutes general correspondences between one system of elements and another one. It is to be supposed a weak version of code theory, as a necessary but non-sufficient condition in order to carry out communicative exchanges. It needs to be integrated by the interlocutors with accurate inference processes of what is communicated. According to this perspective, the code

20

L . .4noIli / MaCHT

(whichever code) is not able to “disambiguate” and manage all the semantic combinations and correspondences enclosed in a communicative act. Such a weak theory of code seems to be particularly profitable, since it is consistent either with communicative processes (in the case of communicative exchange success) or with miscommunication phenomena (in the case of deviations and difficulties or even failure of communicative exchange). In fact, this weak theory of code does not involve a fixed semiotic system, but it presupposes a variable system in terms of power and eficacy. The more a “code” is powefil and “effective”, the better it can define a correspondence system and impose semantic constraints on the interlocutors. If they want to communicate something with other people, they are not totally free to invent words and expressions whenever and however they like, but they have to keep to a shared and sharable system of meanings, even if variations and deviations from it can be expected. This system is usually provided by the natural language created in each cultural community. But even in the extreme case of the two prisoners who use hammer strokes in order to communicate. meaning segmentation is connected with rhythm variations and intensity of strokes, that need to be created, recognized, and shared by the prisoners. However, in order to avoid possible ambiguities or misunderstandings. in this study we will use the expression signaling sjstern instead of code.

I .5

The plurality of signaling systems and semantic synchrony

According to MaCHT assumptions, outlined in this paper, miscommunication is neither a collection of abnormal communicative phenomena nor does it depend on the violation of an ideal model or of a standard system of communication. Miscommunication is a widespread phenomenon, and it belongs to the communicative exchanges of everyday life. It is enough to consider figurative and metaphoric speech, the ironic, deceptive. and seductive communication, as well as communicative games in general. In order to understand miscommunication, we should to bear in mind the continuous variation of meanings between stability and instability, as we said in previous section. We should also take into account the range of signaling systems, and the plurality of expressive means available to the communicators in order to express their communicative intention. In this paragraph we will focus our attention on the semantic sj*nchronj*process. intended as the speaker’s competence to organize, co-ordinate, and make the different communicative systems converge on each other in order to make explicit the meaning of onek own communicative intention in g unitary and coherent way. I .5.1 The plurality of‘communicativesystems The starting point is given by the observation that, according to Anolli and Ciceri [72], in order to communicate, the communicator has at his disposal a pluralih. qf signaling systems, verbal and non-verbal, as shown in figure 1.2. In the first places, language has a prominent position, as it remains the most powerful, flexible and stable communicative device and medium. exclusive to the human species. It allows an articulated and efficient expression of thoughts, beliefs, emotions. desires and the like through effective variable, and extensive semiotic and symbolic devices.

L. Anolli / MaCHT

1.5.I . I

21

The paralinguistic system

First of all, associated and concomitant with language, we have vocal communication, which includes both extra-linguistic characteristics and paralinguistic features. The first ones consist in the whole range of phonic qualities of the voice, which depend on the phonatory organ of the subject. The second ones consist in the supra-segmental organization of pitch variations, voice intensity and speed. As Sapir [73] already pointed out, it is not possible to pronounce even a simple word like /horse/ without giving it some form of interest and some prosodic variation. There are not clear and plain words or utterances, spoken in a neutral linguistic way, but they are always pronounced in a certain manner and with a specific prosodic configuration. The "pronounced word" as a synthesis of vocal verbal and vocal non-verbal aspects, is a powerful communicative device, that can regulate and manage the expression of one's own communicative intention flexibly and progressively. The paralinguistic (or supra-segmental) system allows us to organize in discrete patterns and categories the correspondences between elements of the content level (cognitive and emotive states, ironic intention etc.), and elements of the expression level (prosodic modulation of the pronounced utterance). The supra-segmental and prosodic pattern of joy is one thing; that of anger or contempt is another, as shown by Anolli and Ciceri [74], Scherer [75, 761, and van Bezooyen [77]. The paralinguistic system can create continuous variations of intensity, rhythm, and sound qualities inside the same category. For instance, inside the "family" of supra-segmental displays of "anger", it ranges from prosodic patterns of irritation and grudge to those of anger, increasing to an explosion of rage.

L. Anolli /MaCHT

LINGUISTIC PARALINGUISTIC

'

------------------_---. NON-LINGUISTIC

-

PROXEMIC posture 0 body touch interpersonal distance orientation, etc.

-

MOTOR-GESTUAL BEHAVIOR gestures: 1 . symbolic 2. iconic 3. illustrators 4. emblems. etc.

-

MIMIC VISUAL BEHAVIOR

-

VISUAL BEHAVIOR 0 gaze orientation 0 duration. etc.

Figure 1.2. The plurality of signaling systems. Adapted from Anolli and Ciceri 1721

The distinction between graded and discrete remains problematic. A repertoire is graded if signal variation is continuous, lacking clear acoustic boundaries for demarcating one signal category from another. Such acoustic continuity, however, can be broken up by a perceptual system, creating categorically distinctive classes of sound. According to Hauser. discrete repertoires consist of featurally bounded signals with no intermediates between signal types [78: p. 541. Human speech and acoustic elements of other species' repertoires represent systems that exhibit gradedness on the production end, but discreteness on the perceptual end, which serves as an efficient categorization device. In comparison with co-occumng linguistic elements the paralinguistic system is characterized by its own semantic independence, since it has the power of conveying emotional experiences, cognitive states (certainty, doubt, hesitation, etc.), irony, humor and the like autonomously. It may refer to the same utterance in order to express joy. love and hate, tenderness and anger, fear and sadness, shame and pride, conviction and doubt. tiredness and desire etc., as shown by Anolli and Ciceri [74: pp. 329-3371, Scherer [75. 761, and Pittam 11791.

L. Anolli /MaCHT

23

For example, let us take the sequence of sounds /go out/. According to the occasion, it can assume very different meanings, as in the following situations:

you go out, take an umbrella with you" (informative and descriptive utterance, which indicates the action of going out concomitantly with the risk of rain) ITS it necessary for you to go out just now?" (interrogative utterance of surprise in the case that the interlocutor should go out in an unexpected way) "GO OUT!" (imperative utterance, often as display of anger and with the meaning of "have done with it '7 "Doyou go out this evening?" (requestive utterance with a desire meaning of "do you leave me alone this evening? Please, stay with me!'? "Doyou go out this Sunday afternoon, too?" (regret utterance with a reproach meaning: "stay at home and do not leave me alone this Sunday afternoon, too '7. It is worth noticing that the linguistic input /go out/ remains unaltered, since the same phonemes are repeated in the same sequential order in each utterance. Changes refer especially to the prosodic and supra-segmental configuration whit which each utterance is pronounced, but these changes greatly alter and modify the meaning from one utterance to another. I . 5.I . 2 Theface system The face is also a powerfbl semiotic system that can convey simultaneously a large set of non-verbal signals, both static and dynamic. In particular, facial mimics constitutes an important kinetic system, which has been considered since ancient times and which has been experimentally investigated since the mid-nineteenth century firstly by Duchenne [80]. It is a variable system, clear, progressive, able not only to create unitary and discrete patterns of facial expressions as a whole, but also able to graduate their intensity and shading continuously. Among different techniques of analyzing facial mimics elaborated so far, the Facial Action Coding System (FACS) proposed by Ekman and Friesen [81] is the one that today has obtained a great scientific consensus. Built up using the components method, FACS is a system of observing and describing analytically every visible facial movement in relation to its anatomic-physiological corresponding components. It proceeds to partition the undifferentiated continuum of facial movements in 44 facial Action Units (AU), through which it is possible to investigate over 7000 facial movements and expressions in all their combinations. I .5.I . 3 The gestures system The whole range of gestures also represents another important communicative system of signaling, although organized in different subsystems like gesticulation, pantomime, emblems, and sign language, as Feyereisen [82, 831 and McNeill [84, 851 pointed out. Such subsystems are related with other communicative systems and properties in different way as reported in table 1 . 1 .

L. .4nolli / MaCHT

24

a) Relationship to speech Gesticulation obligatory presence of speech

Emblems optional presence of speech

Sign Language obligatory absence of speech

Emblems some linguistic properties present

Sign Language linguistic properties present

Pan tom im e not conventionalized

Emhlemy partly conventionalized

Sign Language

fully conventionalized

Pantomime global and analytic

Em hlems segmented and synthetic

Sign Language segmented and analytic

Pantomime obligatory absence of speech

b) Relationship to linguistic properties Gesticulation linguistic properties absent

Pantomime linguistic properties absent

c) Relations to conventions Geviriilntion not conventionalized

d) Character of semiosis Gesticulation global and synthetic

Table 1.1. Synoptic of relations between gestures and other communication aspects [ 8 5 ]

In particular, global property refers to the fact that “the determination of meaning in gesticulation proceeds in a downward direction. The meanings of the ‘parts’ are determined by the meaning of the whole. This contrasts the upward determination of the meanings of sentences” [85, p. 51. Likewise, synthetic feature refers to the fact that “a single gesticulation concentrates into one symbolic form distinct meanings that can be spread across the entire surface of the accompanyng sentence” (ibidem). While sign language is conventionalized, segmented, analytic, and shares language features, “gesticulation accompanies speech, is non-conventionalized. is global and synthetic in mode of expression, and lacks language-like properties of its own. The speech with which the gesticulation occurs, in contrast, is conventionalized, segmented, and analytic, and is hlly possessed of linguistic properties. These two contrasting modes of structuring meaning coexist in speech and gesture, a fact of profound importance for understanding the nature of thought and language in general, and how they function in communication” [ 8 5 , p. 61. I . 5.2

Some provisional conclirsions

At the end of this section, it is worth pointing out some consequences of this subject in order to better understand miscommunication processes and to outline the MaCHT more precisely. First, the different signaling systems herein mentioned have a different weight and significance of communicative eflicacy in relation to their power, flexibility, sharpness, stability and versatility in meaning-making. In particular, language appears to be prominent in comparison with every other communicative system. As Levinson [86. p. 81 suggests referring to Sadock’s and Jackendoff work. “phonolog. syntax, and semantics are areas

L. Anolli /MaCHT

25

each with their generative capacities, and there are significant mismatches between the structural strings in each representation that, in the end, come to be associated with one another”. Second, although they possess a different semantic strength, they show a relative degree of autonomy, since each of them is able to convey by itself meanings or pieces of meaning, and to refer to determinate aspects of reality. Third, these different systems of communication, although autonomous, are not disconnected from each other, but they are mutually connected by interdependence links. The outcomes of such interdependence are various: a) first of all, a great degree of freedom is offered to the speaker in order to manifest and calibrate hisker own communicative intention as related to the context; in this way, he/she can regulate and shade what he/she intends to communicate to the interlocutor, being able to resort to different communicative registers; b) the different signaling systems converge in a global and unitary communicative action, with a more or less high consistency degree; c) meaning is not connected with a unique and exclusive signaling system, but it is constituted by a network of semantic co-relations between different systems; d) the connection between different and distinctive signaling systems allows us to explain how possible conflicts and contrasts can appear between them, generating semantic loosening, and how, in extreme cases, semantic contradictions can arise when meaning portions conveyed by one specific signaling system are opposite to those transmitted by another signaling system, so that meaning tends to be confused or even null. Incidentally, concerning these phenomena, we may think about the antiphrastic communication in irony, wherein supra-segmental aspects communicate an exactly opposite meaning to the linguistic expression. Or also we may think of schizophrenic communication, in which nonverbal elements negate the linguistic contents, producing an unintelligible message. The plurality of signaling systems, herein mentioned, together with meaning stability and instability, contributes to explain the wide range of different communicative phenomena in a reasonable and viable way. In the case of default communication, we have communicative acts provided with coherence and unambiguousness, given that the meaning portions of a signaling system are consistently and univocally connected with the meaning portions of other systems. Instead, in the MaCHT here outlined, the existence of conflicts and contrasts between meaning portions conveyed by different signaling systems helps to explain different miscommunication forms, like ironic, seductive or deceptive communication. In the end, the open contradiction between different parts of meaning that are active in different signaling systems can involve a pathological communication when this contradiction brings about the cancellation of the meanings themselves.

I . 5.3 Semantic synchrony The assumption of a plurality of communicative systems involves the necessity of examining how they may be related to each other and how they can make up a unitary totality. In fact, meaning, whatever it may be, is unique and univocal in itself, although different meaning ascription is allowed, as we shall see in the next paragraph. What is needed is an attempt to explain how different signaling systems come to produce a communicative act that is coherent in itself, even if complex and articulated, since it displays a determinate communicative intention.

26

L. Anolli / M a C H T

1.5.3.I The "central communicativeprocessor *' h-vpothesis

Such a process of combining and unifymg these different signaling systems may be defined as semantic synchrony, since their non random combination and interdependence allow the production of a specific meaning of the communicative act. Therefore, we must explain this process of synchrony. With this aim, it could be useful to advance as a hypothesis the existence of a central communicative processor (CCP), similar to what many scientists have already supposed for the co-ordination and management of attentional resources. In this last field, an "attention supervisor system" has been hypothesized by Shalfice [87], also an "operative system" by Johnson-Laird [88], and a "central processor" by Mandler [S9], Posner [90],and Umilta [91] to explain unitary and focal action of attention and consciousness, although they are distributed in the central nervous system. First, it is essential to underline the fact that there is a strong liaison between communication and attention processes. In fact, communication is to be conceived as a flow of basically conscious acts, which can vary in complexity and in employment of mental resources (cognitive or emotional). In fact, an unconscious act, under the effect of drugs or in a particular biological condition (for instance, high fever), cannot be considered as communicative, even if it is speech. In this last case we speak of delirium, as we will see fiirther. Nevertheless, communicative acts are not always conscious in the same manner. We can range from routine and automated communicative acts (like greetings exchange or common questions) to really complex and demanding ones (like giving a report in fiont of a critical public, lying in an official context and so on). In the first case, an automatic processing of communication resources is enough, supported by an oriented attention condition; in the second one, a controlled processing of communication resources becomes necessary thanks to a kind of continuous focused attention. Variability and flexibility, as well as predictability of communicative acts involve - at least in a theoretical way - the action of a CCP as a distributed system, able to have a complete representation of the outside world and of the speaker's intentions. It can be hypothesized that such a CCP does not exert a direct, but an indirect control on different mental operations, taking part in the selection and activation of a communicative route that is privileged in respect of others. According to Umilta [92, p. 2451, most mental operations cannot have access to consciousness, since they are processed in an automatic way. Instead, it can be thought that a CCP exerts a conscious strategic control on the communication outcome, since it consents and favors the selection and activation of one specific sense route rather than another. That is. a CCP takes on the function of regulation between different communicative systems. Such a notion of CCP seems to be in line with the suggestions of Jackendoff [93] about the "architecture of language faculty": the associations between phonological, syntactic and semantic representations are ruled by a system of correspondences which are not deterministic in character, but such correspondences may be partial and not one to one. Moreover, the CCP hypothesis is supported by the notion of default reasoning-advanced by Bach [94J This kind of reasoning is governed by a "taking for granted principle" and proceeds to the first unchallenged alternative, i.e. to the first sense of the utterance that springs to the hearer's mind. Default reasoning is automatic; it relies on generalizations and stereotypes of everyday life and on the scripts of hisher own experience. People normally take it for granted that the interpretation that springs to mind is correct; no hrther reasoning is involved. According to Bach [94, p. 453, "I take for granted that the first thought that comes to mind is the right one - unless some reason against that or some alternative comes to mind".

L. Anolli / MaCHT

27

1.5.3.2 “Central communicative processor” hypothesis in respect of similarframeworh The CCP hypothesis is similar to other frameworks which are interested in explaining global processing of information and communication design. First, we may think that the CCP action is rooted in a parallel distributed processing (PDP) architecture of cognition, advanced by Rumelhart, McClelland, and the PDP Research Group [95]. PDP or connessionist model assumes that the mind is comprised of complex networks of interconnections that operate on parallel. We may represent a communicative goal as a pattern of activation over many units, where each unit represents the degree of presence or absence of a specific thought (or concept). Units are connected to one another by modifiable, weighted connections. Information is represented in the network as vectors of activation values across input and output units. The inputs effect the outputs by propagating their activation through the network of weighted connections. In this way the hypothesized model of the CCP has different properties, interesting for our purpose: a) while in a standard communicative and symbolic model an item of thought is represented by an atomic and discrete token (meaning) residing at a specific address in memory, in the CCP model an item is represented by a pattern of activation over many units, where each unit represents the degree of presence or absence of a certain meaning and thought; this processing allows items in memory to be content addressed, since one can recall a memory plainly by activating parts of the memory; b) while in a traditional communicative model a meaning is either present or absent in an all-or-nothing manner (see the NSC paradigm above mentioned), in the CCP model graded variations and continuous shadings of meaning are possible; c) while in a traditional cognitivist model the context of a meaning is provided by other meanings (or symbols), the context of a CCP meaning is a part of the internal structure, the distributed pattern of activation which is the representation of the meaning itself; d) in the end, in the CCP model learning is possible from single examples and a generalization process allows us to apply their meaning to novel inputs. Likewise, the CCP hypothesis can fit suitably with the global workspace model suggested by Baars [96, 971. According to such a perspective, conscious experience is the outcome of a nervous system activity in which multiple input processors compete for access to a transmitting capability: the winning processor can disseminate its information globally throughout the brain system. The main assumptions of the global workspace are basically the following: a) the existence of specialized semiautonomous processors; b) the global workspace, a memory that, like work memory, can be accessed by input processors; and c) goal contexts that influence the possibility for an input processor of accessing the global workspace. This architecture can be informally described as a “working theatre”, where focal consciousness acts as a “bright spot” on the stage, directed by the selective “spotlight” (or focusing) of attention. Because such an architecture involves both conscious and unconscious events, it deals directly with subjective experience. One dramatic contrast is between the vast number of unconscious neural processes happening in any given moment, compared to the very narrow bottleneck of conscious capacity. This narrow limit has a compensating advantage: consciousness seems to act as a gateway, creating access to essentially any part of the nervous system. Conscious experience creates access to the mental lexicon, to autobiographical memory, as well as to voluntary control over automatic action routines. The “contrastive analysis”, that is a set of paired comparisons between similar conscious and unconscious processes, provides evidence in supporting of the global workspace theory of Baars [96, 971.

28

L .4nolli 1MaCHT

The CCP perspective here mentioned is similar also, in some way, to the message generation model as local management of situated beliefs, proposed by O'Keefe and Lambert [98]. They followed the framework of speech production elaborated by Henmann [99]. As figure 1.3 shows, the representation of situation contains information about relevant situation parameters as well as declarative and procedural knowledge retrieved by the situation-interpreting program. This declarative and procedural knowledge creates the propositional base of the utterance, that is, the foundation of what is meant, including what is perceived, imagined. inferred, presumed, etc. by the speaker. In turn, the semantic input is the foundation of what is said, and it is a selection of what is meant: people verbalize according to pars pro toto principle, saying only a part of what they have in their mind, i.e., what they "mean". Generally speaking, when communicators say something, they find themselves committed to much more, by virtue of choices between all the ways they could have said it. In an ideal Gricean communication act, the semantic input is informative, clear, relevant, truthful and so on; in an actual communication exchange the semantic input generation depends on the specific and contingent features of a certain situation. According to the CCP model and MaCHT, in line with the global workspace model of Baars [96, 971, the "situationalist" approach of Herrmann 1991,and the local management model of O'Keefe and Lambert [98], the communicator has, moment by moment, the possibility of organizing, regulating, managing and monitoring hisher own communicative act related both to hisher intentions and the particular conditions of the context, thanks to semantic synchrony. He/she has the direction and strategic possibility of choice between different communicative routes in the single exchanges with the interlocutor. In this way he/she can orient hidher message even in terms of miscommunication, if he decides that it is suitable, advantageous and favorable for himher to take this line. Therefore, according to these principles of the MaCHT, miscommunication can no longer be intended as a simple deviation or exception related to fixed default patterns of communication. but it provides the communicative devices for the communicator to calibrate hisher own input and optimize the chances at hisher disposal.

L. Anolli /MaCHT

29

SITUATION

+ Situation interpreting program & declarative

Propositional

Selection program Semantic input SI: Subset of PB

Encoding program Syntactic, lexical, and

Produced utterance

Figure 1.3. Herrmann's Speech Production Model. Adapted from Herrmann [99]

1.6

The intention design and communicative synchrony

As is well known, since ancient times it has been accepted that intention plays a fundamental role in people's communicative exchanges. Describing communication in terms of intention and interpretation allows us to use another theoretical device to explain communication and miscommunication simultaneously, as well as to give further support to the MaCHT. As Searle [19] pointed out, intentionality is a property of certain mental states. If individuals had no mental states, they would have no intentionality and no intentions.

30

L . .4noNi / MaCHT

Dennett [ 1001 pointed out a general intentional stance as an attitude to interpret the behavior of an entity (newborn baby, animal, artifact) by treating it as if it were a rational agent whose actions are ruled and managed by considering its beliefs, desires and goals. People cannot ignore the fact that speech, gestures. texts and the like arise from human activities that have intentional purposes. This general attitude concerns not only the speaker in producing hisher communicative act but also the addressee in recovering and interpreting the meaning of the speaker’s message, attributing to it a specific intention. In this way, as, among other theorists, Anolli and Ciceri [72] have proposed, the communicative exchange is created and governed by a reciprocal game between the communicators: the display and ostension of a given intention by the speaker (‘‘intentionalization”process) and the ascription and attribution of a certain intention to h i d e r by the addressee (“re-intentionalization” process). The unavoidable gap between these two positions contributes effectively to explain both communication and miscommunication processes, as well as to sustain theoretically the MaCHT. 1.6.I

Intention.from the speaker’spoint of lierr.

The pragmatic approach in analyzing communication processes underlines the significance and essential role of communicative intention, so that we can speak about “communicative exchange”. If there is no intention, we are not in the presence of a communicative act, but only of an informative one. For instance, speech said in a delirium condition (cause by high fever or drug assumption) is informative but not communicative, since it does not communicate the speaker’s intention. 1.6.I.I

ToMVard an intention definition

Scientists usually make a distinction between two different meanings of intention: a) intention as a propositional character of a mental state (an action performed on purpose), and b) intention as an act concerning and directed at some state of affairs in the world. In the first sense, individuals deliberately perform an action in order to reach a goal. In this theoretical field some scholars - Anscombe [ 10 I], Searle [ 191, and Bratman [ 1021, among the others - make a hrther distinction between intention-in-actionand prior intention. The former means: “We do things intentionally”, while the latter means: “We intend to do things”. “All intentional actions have intentions but not all intentional actions have prior intentions. I can do something intentionally without having formed a prior intention to do it, and I can have a prior intention to do something and yet not act on that intention” [ 19. p. 851. The second sense of intention takes up the idea of meaning proposed by Brentano [ 1031 and, then, by Husserl [104, 1051 as an act of consciousness directed toward an object. Intention, from Latin etymon in-tendere, means to speak about something and it concerns the “aboutness” or “directedness” of mental states. In such a perspective any subjective, conscious experience - no matter how minimal - is an experience qf something. Intentionality is a basic feature of consciousness, which aims beyond itself at an object. Unlike desires, intentions are self-referential, since the satisfaction conditions of intentions refer to the same intentions [ 161. Unlike beliefs, intentions involve a willingness to change a certain state of affairs in the world. Both of these meanings of intention operate and are effective in communicative interaction. In particular, as we have already seen. Grice [I61 pointed out a certain “intentional clearness” in the sharing of the same communicative intention by the interlocutors thanks to the Co-operative Principle. He distinguishes on purpose between

L. Anolli / MaCHT

31

"what is said" and "what is meant"; the communicative distance between these two semantic layers is filled with implicatures. Even Sperber and Wilson [20] distinguished between informative intention (the content which is communicated) and communicative intention. The latter has been defined as making it "mutually manifest to both audience and communicator that the communicator has this informative intention" [20, p. 6 11. Nowadays, the topic of intention has become very important. Beyond the informative and communicative intention, Jaszczolt [ 106, p. 521 proposes the "Primary Intention Principle" (its role "is to secure the reference of the speaker's utterance") and deepens the notion of referential intention proposed by Bach [107, p. 1401: "what is said, to the extent that it is not fixed by linguistic meaning, is determined by the speaker's intention, which itself can include the intention to refer to what one is demonstrating". In this way intention seems to be a borderline and a cross-road in a synthesis process which is interposed between the inner world of the speaker (what he/she intends to say and display to the interlocutor), the outer world (the I'aboutness" above mentioned), and the produced message (what is said or expressed through the signaling systems).

I . 6.I .2 The intentional gradability Intention does not constitute by itself an "on-off' process, but it is characterized by an articulated graduation and differentiation within itself. First of all, in everyday life, intentionality (as an attitude to produce distinct and specific intentions), in the same way as consciousness and attention, is regulated by continuous variations of intensity and precision. This intentional gradability allows communicators to manage the focusing of different communicative acts during everyday life. Next to plain (common and ordinary) communicative acts, in which the intentional process is almost automatic, there are complex communicative acts (like seductive, ironic or deceptive communication) in which the communicator has to be attentive and conscious of the communicative intention that he conveys to the interlocutor. The strength of intention seems to be directly proportional to the informative content of the message [ 107, p. 681, to the interlocutor's significance, as well as to the nature of context (public or private). Moreover, a single communicative act can be governed by a plurality of intentions, embedded in each other and hierarchically organized. Such is the case of a prepared (packaged) lie, in which, as Anolli, Balconi, and Ciceri (chapter 6, this volume) have underlined, different layers of communicative intentions are at work: a) a hidden (covert) intention (the speaker intends to deceive the addressee by manipulating the information); b) an ostensive (overt) intention (the speaker intends to convey the information manipulation to the addressee). This second intentional layer is, in its turn,twofold: bl) informative intention (the speaker wants to give the addressee the manipulated information as if it were true); b2) "sincerity"intention (the speaker wants the addressee to believe that what he has said is true, in order to respect the Sincerity Rule of Searle [ 1081: "I want you believe that I believe what I am saying to you"). But without also resorting to deceptive communication, in everyday conversation a speaker has to select and choose an intentional layer to convey what he/she has in mind. Here we may come back to the pars pro toto principle, since in the production of a communicative act a speaker can only express a part of hisher mental content. Among other conditions, this process has a physical constraint, since phonetic articulation is a bottleneck in a communicative system that could otherwise run about four times faster [88]. As a consequence, the speaker is obliged to follow only one of the different meaning routes he/she has at hisher disposal for conveying what he/she has in mind.

32

L . Anolli i MaCHT

Such a choice and continuous gradation of communicative intentions make the communicative act particularly complex, since, on the one hand, it needs a precise cognitive and emotional direction; on the other, it can give rise to communicative uncertainties and difficulties. One intention is usually embedded in another which surpasses and includes the first one: in the case of the prepared lie, the hidden intention. for instance, subsumes and explains the manifest intention. This situation is more widely diflbsed more than is thought. After an episode of serious marital conflict, when one partner tries to make peace with the other and asks h i d e r . for instance, for a glass of water, this request carries at least two different intentions at the same time: a) to make the explicit (direct) request of obtaining a glass of water; b) to make the first step toward reconciliation in an indirect and implicit manner. A similar communicative condition is the ironic communication. as Anolli. Infantino and Ciceri have pointed out (chapter 6, this volume). The communicative outcome of this intentional articulation and embedment is the possibility of selection by the speaker of a certain route of sense instead of another one. It greatly increases the degree of communicative freedom at the disposal of the communicator. But in such a way, intention plurality and graduation entail. as a consequence, an intentional opacig?, since in many cases communicative intention. embodied in an utterance or gesture, is shaded and changeable. In the end, these considerations about intention plurality and gradability strengthen and shed some light on the continuity assumption between communication and miscommunication. If the intention field is organized and processed like this, it becomes impossible to decide where communication ends and where miscommunication begins. In the same way these considerations give further support to the MaCHT. 1.6.2 Intention according to the interloctrtork point qf‘view’ As we have already said, in the standard traditional framework communication is conceived as the transmission of a message fi-om a source to a receiver, and the communicative exchange is considered as having happened if this passage has been concluded. Within such a perspective the addressee is represented as a passive “terminal”.

I .6.2.I The recoven*of speaker’s intention In modem times, with the radical change produced by Grice’s [ 161 and Sperber and Wilson’s [20] contributions, the notion of reciprocity in the communication has been strengthened. In order to succeed, the communicative exchange has to be characterized not only by the producing of a communicative intention (m-intention) in the speaker, but also by its recovery by the addressee. The recovery act involves a precise and conscious activity and the participation of the receiver in the meaning elaboration, since meaning is defined only when the addressee recognizes the speaker’s communicative intention. In this sense meaning can be considered as the outcome of a two-person-construction, and the addressee’s intervention is as significant as that of the source. A basic assumption to explain the notion of intention reciprocity is given by social psychologists like Cooley [I091 and Mead [ I 101, according to whom the addressee’s understanding of the speaker’s intention rests on a special source of knowledge that is not available in other fields of knowing: the anaZoBj to the df.The implicit and practical reasoning of the comprehension of the other’s intention is based on the principle: “you are like me”. As a consequence. I am in a condition to grasp your intention.

L. Anolli /MaCHT

33

To seek understanding of the speaker’s intention is to assume that his/her action has a meaning, and make an effort to discover it. As a fundamental condition to reach a so defined communicative exchange a cognitive environment, mutually know, is requested where every manifest hypothesis is mutually manifest [20]. Such a condition allows a valid and reliable degree of co-ordination between individuals in order to communicate. The speaker’s communicative aim is that of modifylng the cognitive environment of the receiver. The idea of recovery as an act of recognizing the source’s communicative intention certainly assumes a bi-directional conception of communication, but its nature also entails an asymmetrical direction, like in a dance where one leads and the other is led. In this sense the speaker appears more significant than the hearer. The former takes a more important position as a source; the latter has only to carry out the recovery as accurately as possible. The more precise the hearer is in this activity, the more suitable he/she is in the communicative exchange and the better he/she does his/her communicative task. But in this perspective the intention recovery by the addressee presents a sort of semantic dependence in respect of the speaker. In fact, the former’s task consists of a “philological work” of semantic interpretation of the latter’s intention. Needless to say, in this task the addressee may make mistakes of over- or under-interpretation, since he/she may add or take away parts of meaning in the recovery of the speaker’s communicative intention. This semantic swinging in the addressee’s interpretation shows up the existence of a gap between the speaker’s intention and the hearer’s recovery of this intention. We are faced with an impossible task, given the intentional opacity of the speaker’s mind. In fact, even in the highest conditions of mutual awareness it is impossible to read directly into the other’s mind and to access directly the other’s intention. All “mind reading” is always indirect, mediated by a complex and fragmentary system of communicative devices, cues and expressions, both linguistic and extra-linguistic. Given the impossibility of the direct accessibility of an intention, any interpretation of the source’s intention is partial and limited, since it necessarily follows the totum ex parte principle. The addressee ascribes a complete and consistent intention to the speaker’s communicative act on the basis of a small and restricted set of communicative cues and hints. 1.6.2.2 The ascription of speaker j . intention

As we have already seen, the recovery act appears to be insufficient to explain the interpretation activity of the hearer in relation to the speaker’s intention. We need to add the idea of intention attribution (or ascription) to the former position with regard to the latter’s message. This means that the addressee attributes a certain intention to the source’s communicative act. This intention ascription in the hearer is the “other face” of the intentional stance and is the equivalent of the intention production in the speaker. Such a process is characterized by different features: a) it is an autonomous process, carried out only by the addressee; b) it is an active and subjective process, because it depends completely on the addressee’s abilities as well as on the perspective he/she chooses to follow. Needless to say, there is a great difference between intention recovery and intention attribution. In fact, the latter includes rules and manages the first one. By making an intention ascription, the addressee may recover the source’s intention in a reliable and precise way, but he/she may also ascribe a totally different intention to the speaker’s message. For instance, the addressee might perceive an ironic and sarcastic intention in the

34

L . Anolli / MnCHT

speaker’s utterance, while the latter had only a plain, narrative intention to say something in describing a given situation. It is worth remarking that we are faced with an essential condition of communicative interaction, because the speaker and the hearer have different points of view about the same events and objects as topics of communication, even if they share a mutual cognitive background and mutual, culturally defined knowledge. Such a difference in their points of view, that can be radical in some circumstances, involves and presupposes in any case different interpretations and meaning-making in the speaker and the hearer with reference to the same episodes and facts. They possess a different encyclopaedia of knowledge, because they have a different amount of experience in qualitative and quantitative terms. Moreover, they may refer to different beliefs and values systems. as well as being guided by different desires and goals. In such a perspective it is obvious to expect a pluraliw of interpretations in the intention attribution from the addressee to the source’s message. This plurality is not to be considered as a pathological accident, but it is the standard activity of the addressee. Given the plurality of interpretations as a matter of chance in the hands of the addressee, concepts like “literal meaning” or “authentic interpretation” have to be considered in a different light. These concepts seem to be a residual of an ideology based on the natural stance to communication. The literal or nonliteral (figurative) meaning of an utterance depends on the intention ascription operated by the addressee, and the so-called “literal interpretation” may be only one (and perhaps the weakest!) among many other possible solutions. Note that all these interpretations can be plausible and justified, and that the literal meaning, by its nature, does not possess an outstanding or privileged position. By analogy, the “authentic interpretation” is a matter of approaching the actual intention of the speaker through a gradated approximation. There are very often corrections and negotiations in this approaching and sharing of communicative intention. We actually understand each other, but this mutual understanding is not automatically guaranteed by the structure of linguistic or extra-linguistic tools or devices. Intention ascription is an open field, and the addressee is allowed a broad space to justify hisher interpretative choice. connected with some cues of the communicative act. 1.6.2.3 ‘* Taking.forgranted ”principle and the intention interpretations

In a standard condition, however, during the fast flow of communicative exchanges. intention ascription seems to be a default and immediate process. It clearly includes the “taking-for-granted” principle stated by Bach [943 and above mentioned, which says that one has to accept the first sense of a communicative act that springs to the interlocutor’s mind and that is not immediately refuted by another meaning. This default process of intention ascription is rooted in context regularity and in the routine exchanges we saw previously in meaning stability phenomena. A script-like structure of communicative interaction affords, in fact, the possibility of foreseeing the speaker’s communicative intention. In this sense Levinson [86] has emphasized the notion of “presumptive meaning” or preferred interpretation. But in many other circumstances intention attribution may be a significant and subtle communicative task for the addressee. In these cases the fundamental questions may be the following: b‘Whydoes he/she tell me such and such a thing? Why does heishe use this voice intonation?”, and the like. It is not only a paranoid problem. It is an issue for anybody who tries to understand the actual intention of the speaker. We are not interested in limiting our communicative understanding to the apparent, surface intention: we are usually interested in grasping the profound. actual intention of the speaker. If the addressee confines him-

L. Anolli /MaCHT

35

/herself to taking into consideration only the surface intention, he/she is quite likely to go wrong sooner or later; in any case, he/she does not seem a very competent communicator. However, the actual intention is often implicit, not directly said by the speaker, but made understandable by means of sufficient leakage cues. In this sense miscommunication is a borderline in relation to communication. Moreover, such an intention ascription involves a shift from “message philology” to “intention interpretation” in communicative interaction. The former is centered on the literal meaning and what is said; the latter is focused on the speaker’s thought, what he/she intends to communicate. Nevertheless, the discovering of the actual intention involves a precise, clever inferential process in the addressee. It is an issue of practical reasoning, often based on heuristics and biased by previous patterns of interaction with the interlocutor. Sometimes a few weak clues are sufficient to produce a large inducing process and to reach significant conclusions, which are not always correct. Such logical biases, driven by what Simon calls a “bounded rationality” [ l l l ] , are usually rooted in past experience with the partner, especially in close relationships, characterized by learning set and communicative patterns. In any case, these logical and psychological conditions allow the addressee to reach not the perfect (“Olympian”) interpretation but only the local and temporary best one. It is needless to say that, given the significance and the relevance of his/her intention ascription, the addressee is as important as the speaker in the communicative exchange. They are both on the same parithetic level and share the same responsibility in the construction of the communicative act. For this set of reasons intention ascription is indeterminate, since it can take different routes, none of which is either privileged or foreseeable. Such vagueness in intention attribution by the addressee adds to the opacity in intention production by the speaker. The connection between these two aspects effectively helps to explain the continuity between communication and miscommunication phenomena. There is not a fixed sharp borderline between them, but indefiniteness both in intention production and in intention ascription makes them fuzzy and graded by nature. As a consequence, communication seems to be an unforeseeable route, not deducible from previous exchanges. It is a game played by the participants, hand by hand and move by move. Within this route there is room, of course, for an extended range of miscommunication phenomena, as expected according to the MaCHT. I . 6.3 Communicative synchrony

In the processes of intention production and ascription, speaker and interlocutor are at the same level and share the same responsibility in communicative exchange-making. Communication presupposes and involves participation, since speech is the product of a collaboration between communicators. As Cohen and Levesque [ 1121 have pointed out, “the joint action of dialogue claims that both parties to a dialogue are responsible for sustaining it. Participating in a dialogue requires the conversations to have at least a joint commitment to understand one another”. Moreover, a participation framework implies a mutual process of co-constructing and sharing meaning in a continuous stream of communicative exchanges, as shown by Duranti [ 1131. When people communicate, according to Burgoon [114], they have to “adapt their interaction styles to one another. For example, they may match each other’s behavior, synchronize the timing behavior, or behave in opposite ways”. Actually, it is well understood in common observations as well as in the scientific field that participants in communicative interaction are usually engaged in a common rhythm. As Cappella [ 1151

36

L . .4no/li 1MaCHT

points out, “co-ordination in social interaction means that people adjust their actions to those of their partners”. The notion of interactive co-ordination and mutual adaptation in communication covers a broad range of processes, such as synchrony, mirroring, matching, reciprocity, compensation, convergence as well as divergence. Some of these phenomena have more to do with the temporal co-ordination of the interaction, whereas others focus on the reciprocal adaptation of communication styles in terms of converging vs. diverging from one’s partner’s style. In the MaCHT perspective we collect all these processes under the label of communicative synchron-v as a global and basic property of communication, that has been developing since birth in newborn babies through the interaction system with their caregivers. From a sociological point of view, different scientists have drawn special attention to time patterns in organizing communicative exchanges. The notions of “synchrony of rhythms” proposed by Goffman [ 1 161 and “temporal symmetry” advanced by Zerubavel [ 1171 are in this sense. Schutz [ 1 181 has widened the concept of a mutual tuning-in relationship as a theoretical extension of the notion of mutual sharing of Cooley [ 1091 and Mead [ I 101. The “mutual tuning-in relationship” involves, in a particular way. the organization of interaction sequences between partners in a temporal succession. Within the psychological field, among others scholars, Giles [ I 191, Giles and Powesland [ 1201, and Giles and Coupland [ 12 I ] deserve attention for their proposal of the Communication Accommodation Theo? . According to the CAT, attuning and accommodation strategies consist of a broad set of linguistic and extra-linguistic signals which enable us to adapt our communicative acts to those of our partners, shifting them along a convergent or divergent direction in the sequence of exchanges. In the first case. the communicative styles of the participants become more similar and assume a homogeneous shape; in the second one, the differences become greater, creating a process of schismogenesis. The CAT has been supported by experimental results, since it has been verified that pronunciation [ I 191, speech rate [ 1221, utterances and length of pauses [ 123], vocal intensity [ 124, 1251, as well as vocal fundamental frequency patterns in childish babbling [ 1261 vary in a sequence of communicative exchanges in conformity with mutual adjustment and attuning. More recently, Gregory [ 1271, Gregory and Webster [ 1283, and Gregory, Dagan, and Webster [ 1291 have analyzed the accommodation of vocal long-term features, showing that communicators are able to mutually co-ordinate their speech styles in a subtle way. by assuming and maintaining the same specific range of vocal pitch. The emergence of this synchronized and adapted pattern appears to be of crucial importance in ensuring efficacy and comprehensibility in communication, and in determining the perceived quality of the relationship, effectively managing relational and communicative distance. This topic has been stressed by Grammer [ 1301, and Grammer, Kruck, and Magnusson [131] in investigating the role of communicative synchrony in seduction interaction among people who meet for the first time. In fact, rhythmed coordination is fundamental in determining the development of interpersonal attraction and a higher degree of satisfaction about the relationship (see Ciceri, chapter 4. this volume: Mantovani, chapter 10, this volume). patient satisfaction in the physician-patient relationship is also highly related to interactive and communicative synchrony, as shown by Koss and Rosenthal [ 1321. This communicative synchrony constitutes another significant theoretical device in support of the unitary framework of communication and miscommunication phenomena. as well as the MaCHT. In fact. communicative attuning and synchrony allow communicator5 to decline their communicatix e competence in a practical wa\‘ for

L. Anolli /MaCHT

37

convergence or divergence, and create a broad set of chances for different solutions, ranging from direct and open communication patterns to indirect and cryptic ones.

1.7

Conclusions: toward a "re-definition" of miscommunication

At this point, it is worth proposing a definition of miscommunication in agreement with the MaCHT. According to the standard model of communication, miscommunication is associated with forms of misinterpretation that promote a wide spectrum of disagreements and misunderstandings [31. Within this traditional perspective, miscommunication includes communicative phenomena like disruption, relational instability and mutual misconstruction, misunderstanding, contradiction and the like. According to other theorists, miscommunication concerns the very elusive, enigmatic and inscrutable aspects of communication and the "maze of messages" [ 11. In this contribution, within the MaCHT perspective we propose a more optimistic notion of miscommunication. In fact, for us "say not to say" is not an empty say, that is, say nothing. Instead, say not to say is saying something in a different manner, that is, in an implicit (unsaid) and indirect manner. Miscommunication is an "oblique" communication with the purpose of optimizing interaction with the partner. In this light we can go back to the convenience principle, according to which ''say what is good for the relationship", as underlined by Turner, Edgley, and Olmstead [ 1331. This principle must not be intended in a utilitarian way, but as a criterion for optimizing the cognitive, emotional and relational resources of the communicator. The notion of miscommunication as a chance increases greatly the degree of freedom at the disposal of communicators. The communicative space created in such a way by the miscommunication process makes a concrete contribution to the framework of the local management of message design, advanced by O'Keefe and Lambert [98] and above mentioned. Such a local management model could risk indicating a general and suitable principle, but without examining the actual and operating mechanisms underlying it. According to the MaCHT perspective, communicators enjoy the opportunity of managing their communicative focus in the best possible way, given the contextual constraints and their respective encyclopaedia of knowledge. Communicative focus is concerned with how the speaker lets the addressee know what in particular he/she notices about the prominent aspect of the communicative act. Through hisher choice of specific linguistic and extra-linguistic hints and clues, like intonation, the word order in an utterance, an extra stress and so on, the speaker can convey different meanings, as in the following sentence: (1 1) (12)

YOU'RE not mad with me You're not mad with ME

Likewise, the order of the words in a sentence can change the communicative focus, like in: (13) ( 14)

It was John that went to the party It was the party that John went to

In (13) "John" is the focused information, while in (14) it is the "party". While the communicative topic presents the entity "about" which the predication affirms something in the given context, the communicative focus represents what is relatively the most

38

L. .4nolli / M a C H T

important and salient information in the given context. Focusing is an active, mutual process, engaged in by the participants, of concentrating attention and interest on a subset of their shared reality. It not only involves attention on what is said and meant, but is also a perspective on these aspects. A communicative act arises as the focus moves through the field of structures of beliefs, driven by the communicator’s goals and guided by thoughts and communicative devices. Within the MaCHT perspective, any communicative act can be given multiple characterizations, some of them mutually incompatible, and all of them equally justified. The theoretical fiamework herein outlined, grounded on the fundamental processes of communication such as meaning stability and regularity vs. meaning instability and flexibility, plurality of signaling systems and central communicative processing, communicative intention production (in the speaker) and ascription (in the addressee), intention gradation and articulation, semantic and communicative synchrony, can shed some light on communication and miscommunication phenomena. They are not to be thought of as discrete and separate categories, but as a unitary structure, although differentiated within itself. In such a way, the MaCHT provides some suggestions to overcome the Gricean distinction between “what is said” and “what is meant (or implicated)”. The essential thing is “what is communicated”, and this cannot be conceived simply as the sum of “what is said” plus “what is meant”. What is communicated subsumes both what is said and what is meant, but it is also more than that and different tiom that. We ought to overcome the distinction between sentence-meaning and utterance-meaning, since meaning, in any case. is a unitary totality of what is communicated, though it is neither monolithic nor rigid. but composite and flexibly organized within itself. The MaCHT is a preliminary attempt to sketch out a new perspective on the communication field, from a psychological point of view. Here is a challenge to develop this unitary and parsimonious perspective, by proceeding to its experimental control and theoretical expansion.

1.8

Acknowledgements

I would like to thank Fondazione “Piero, Pietro e Giovanni Ferrero” (Alba, Italy) for the support given to the present study. Rita Ciceri, Fabrizia Mantovani, and Giuseppe Riva made helpful comments and suggestions as my theoretical approach was evolving. Needless to say. they are not responsible for the uses I made of their suggestions. 1.9 [ 11

[2] [3] [3]

[5] [6]

References W. R. Cupach and B. M. Spitzberg. The dark side of’ interpersonal communication. Hillsdale: L. Erlbaum, 1994. B. Fraser, No conversation without misrepresentation. in: Pretending to communicate. H . Parret. Ed. Berlin: De Gruyter. 1994, pp. 143- 153. C. D. Mortensen, Miscommunication. Thousand Oaks, London: Sage Publications, 1997. H. Parret, Indirection, manipulation and seduction in discourse. in: Pretending to communicnte. H. Parret, Ed. Berlin: De Gruyter, 1994, pp. 223-238. L. W. L. Young, Cross talk and culture in Sino-American communication. New York: Cambridge University Press. 1994. I. Markova. C. F. Graumann. and K. Foppa. Eds. .Mlrtualitie.s in dialogue. New l‘ork: Cambridge

L. Anolli /MaCHT

[71

1131 ~ 4

1241 [251

39

University Press, 1995. C. E. Shannon, and W. Weaver, The mathematical theory of communication. Urbana: Illinois University Press, 1949. A. V. Cicourel, Cognitive sociology. Harmondsworth: Penguin Books, 1973. H. Garfinkel, Study in ethnomethodology. Englewood Cliffs: Prentice Hall, 1967. H. Sacks, An analysis of the course of joke’s telling in conversation, in: Explorations in the ethnography ofspeaking, R. Baumann and J. Scherzer, Eds. Cambridge: Cambridge University Press, 1974, pp. 337-353. E. A. Schegloff, Sequencing in conversational openings, American Anthropologist 70 (1968) 10751095. E. A. Schegloff, On some questions and ambiguities in conversation, in: Structures of social action, J . M. Atkinson and J. Heritage, Eds. Cambridge: Cambridge University Press, 1984, pp. 28-52. E. Goffman, Strategic interaction. Philadelphia: University of Pennsylvania, 1969. 1E. Goffman, Frame analysis: An essay on the organization of experience. New York: Harper and Row, 1975. E. Goffman, Forms of talk. Philadelphia: University of Philadelphia, 1981. H. P. Grice, Logic and Conversation, in: Syntax and semantics (vol. 3), P. Cole and J. Morgan, Eds. New York: Academic Press, pp. 41-58. J.L. Austin, How to do things with words. Oxford: Clarendon Press, 1962. J. R. Searle, Speech acts: An essay in the philosophy of language. New York: Cambridge University Press, 1969. J. R. Searle, Intentionality: An essay in the philosophy of mind. Cambridge: Cambridge University Press, 1983. D. Sperber and D. Wilson, Relevance: Communication and cognition. Oxford: Oxford University Press, 1986. S. A. McCornack, Information manipulation theory, Communication Monographs 59 (1992) 1- 16. S. Jacobs, D. Brashers, and E. J. Dawson, Truth and deception. Communication Monographs 63 (I) ( 1996) 98- 103. S. Jacobs, E. J. Dawson, and D. Brashers, Information manipulation theory: A replication and assessment. Communication Monographs 63 (1) (1996) 70-82. G. Bateson, Steps to an ecology of mind: Collected essays in anthropology, psychiatry, evolution, and epistemology. Northvale: Jason Aronson, Inc., 1987. G. Bateson, The cybernetics of “self’: A theory of alcoholism, in: The dynamics and treatment of alcoholism: Essential papers, J. D. Levin and H. R. Weiss, Eds. Northvale: Jason Aronson, Inc., 1994, pp. 3 13-322. G. Bateson, D. D. Jackson, J. Haley, and J. H. Weakland, Toward a theory of schizophrenia, in: Double bind: The foundation of the communication approach to the family, C. E. Sluzki and D. C. Ransom, Eds. New York: Grune & Stratton, 1976, pp. 3-22. S. Minuchin, The leap to complexity: Supervision in family therapy, in: The evolution in psychotherapy: The third conference, J . K . Zeig, Ed. Philadelphia: BrunnedMazel, Inc., 1997, pp. 27 1-28 1. S. Minuchin and M. P. Nichols, Structural family therapy, in: Case studies in couple and family therapy: Systemic and cognitive perspectives. The Guilford family therapy series, F . M. Dattilio, Ed. New York: The Guilford Press, 1998, pp. 108-131. M. Palazzoli Selvini, A systemic course of family therapy, in: The work of Mara Selvini Palazzoli, M Palazzoli Selvini and M. Selvini. Northvale: Jason Aronson, Inc., 1988, pp. 323-359. M. Palazzoli Selvini, S. Cirillo, M. Selvini, and A. M. Sorrentino, Family games: General models of psychoticprocesses in the family. New York: W. W. Norton and Co, Inc., 1989. V. M. Satir, Conjoint family therapy: A guide to theory and technique. Palo Alto: Science and Behavior Books, 1964. H. H. Clark, Using language. Cambridge: Cambridge University Press, 1996. R. W. Gibbs, Intentions in the experience ofmeaning. Cambridge: Cambridge University Press, 1999. G. Frege, Ueber Begriff und Gegenstand [On concept and object]. VierteGahrsschrqt .fir wissenschaftliche Philosophie 16 (1892) 192-205. G. Frege, Ueber Sinn und Bedeutung [On sense and meaning]. Zeitschrift fur Philosophie und philosophische Kritik 12 (1 892) 25-50. R. Carnap, Der logische Aufbau der Welt [The logical structure of the world]. Hamburg: Meiner, 1928. R. Camap, Meaning and necessity. Chicago: University of Chicago Press, 1947.

40

[471 1481 1491

[553

1641

1671

L . .4nolli I' MaCHT

G. Lakoff, Women,fire, and dangerous things. What categories rer.eal about the mind. Chicago: The University of Chicago Press, 1987. H. H. Clark and E. V. Clark, psycho log?^ and language: An introduction to psvcholinguistic. New York: Harcourt Brace Jovanovich, 1997. J. C. Gomez, Mutual awareness in primate communication: A Gricean approach, in: Self-awareness in animals and humans: Developmental perspectives, S. T. Parker. R. W. Mitchell. and M. L. Bwcia (Eds.). New York: Cambridge University Press, 1994, pp. 6 1-80. H. P. Grice, Meaning. Philosophical Revien. 66 (1957) 377-388. J. J. Katz and J. A. Fodor, Structure of a semantic theory. Language 38 ( 1963) 170-210. L. Hjemslev. Prolegomena to a theon oflanguage. Madison: University of Wisconsin. 1943. D. A. Cruse, Lexical semantics. Cambridge: Cambridge University Press. 1986. P. Violi, Signifcato ed esperienza. [Meaning and experience]. Milano: Bompiani, 1997. H. Putnam, Mind, language and realin?: Philosophical papers (vol. 2 ). Cambridge: Cambridge University Press, 1975. D. Geeraerts, Prospects and problems of prototype theory. Linguistics 27 (4) ( 1989) 587-61 2. G. Kleiber. La semantique du prototype. Categories et sens lexical [The semantic of prototype: Categories and lexical meaning]. Paris: PUF, 1990. W . Labov. The boundaries of words and their meaning, in: Neiv ways Of'anal~~sing variation in English, C. N. Bailey and R. Shuy. Eds. Georgetown: Georgetown University Press. 1973. pp. 340373. G. Lakoff. Linguistics and natural logic. in: Semantic's of' natural language. D. Davidson and 6. Harman, Eds. Dordrecht: Reidel, 1972, pp. 545-665. G. Lakoff. Hedges. a study in meaning criteria and the logic of fuzzy concepts. Journul of Philosophical Logic 2 (1973) 458-508. L. A. Zadeh, Fuzzy sets. Informationand control 8 (1965) 338-353. Q. Zhang, Fuzziness - vagueness - generality - ambiguity. Journal qffragmatics 29 ( 1998) 13-31 . T. Givon, Prototypes: Between Plato and Wittgenstein. in: .Voun classes and categori:afiori. C . Craig. Ed. Amsterdam: John Benjamin, 1986. pp. 77- 102. M. Posner, Empirical studies of prototypes. in: .Voun c1as.se.s and categorization. C. Craig. Ed. Amsterdam: John Benjamin. 1986, pp. 53-6 I . E. Rosch, Principles of categorization. in: Cognition und categorization. E. Rosch and B. B. Lloyd. Eds. Hillsdale: Erlbaum, 1978. pp. 37-48. E. Rosch, Linguistic relativity, in: Human i~ommunication: Theoretical perspectii.es. A. Silverstein. New York: Halsted Press. 1974. pp. 95-121. D. Geeraerts, Polisemy and prototypicality. Cognitive Linguistics 3 ( 2 ) ( 1992) 2 19-13. L. Wittgenstein, Philosophical investigations. Oxford: Blackwell. 1958. P. J. Thibault, Communicating and interpreting relevance through discourse negotiation: An alternative to relevance theory-A reply to Franken. Journal offragmatics 31 (4) (1999) 557-594. R. W. Gibbs, Jr.. Intentions in the experience of. meaning. Cambridge: Cambridge University Press. 1999. S. Glucksberg, Crnderstanding figurafitv Iunqtcuge: Front nrutaphora to idioms. Cambridge: Oxford University Press. 200 1. H. Horman. The calculating listener or how many are einige, mehrere. and ein paar (some. several. and a few). in: Meaning. use, and interpretation oflanguage, R. Bauerke. C. Schwarze. and A. \.an Strechan, Eds. Berlin: de Gruyter, 1983, pp. 22 1-234. K. H. Onishi, and G. L. Murphy. Metaphoric reference: When metaphors are not understood as easily as literal expressions. Memo? and Cognition 21 (1993) 763-772. J. Pynte, M. Besson, F. H. Robichon. and J. Poli. The time-course of metaphor comprehension: An event-related potential study. Brain and Language 55 (1996) 293-3 16. B. McElree, and J . Nordlie. Literal and figurative interpretations are computed in parallel. P.s>rhonomicBulletin and Retien' 6 ( 1999) 486-494. J. R. Searle. indirect speech acts. in: Slwtar and semantics 2: Speech acts. P. Cole and J . Morgan. Eds. New York: Academic Press. 1975, pp. 59-83. J. Bruner. .4ctual minds. possible i+*orlds.Cambridge: Harvard University Press. 1986. J. Bruner, .4cZs ofmeaning. Cambridge: Harvard University Press. 1990. P. Bourdieu, Outline c?f a theor?. af'practice.Cambridge: Cambridge University Press. 1977. P. Bourdieu, The logic ofpractice. Stanford: Stanford University Press. 1990. L. Anolli, and R. Ciceri. I1 gioco delle intenzioni nell'atto molare: Per un modello psicologico della comunicazione [Intentional game in the molar act: ToLvards a psvchological model of communication].

L. Anolli / MaCHT

41

in: Elementi di psicologia della comunicazione: Processi cognitivi e aspetti strategici [Elements of communication psychology: Cognitive processes and strategical aspects] . Milano: LED, 1995, pp. 25114. [73] E. Sapir, Language. A n introduction to the study of speech. New York: Harcourt Brace, 1921. [74] L. Anolli, and R. Ciceri, La voce delle emozioni. Verso una semiosi della comunicazione vocale nonverbale delle emozioni [The voice of emotion. Towards a semiosis of nonverbal vocal communication of emotions]. Milano: F. Angeli, 1992. [75] K. Scherer, Nonlinguistic vocal indicators of emotion and psychopathology, in: Emotion in personality andpsychopathology, C. E. Izard, Eds. New York: Plenum Press, 1979, pp.189-220. [76] K. Scherer, Vocal affect expression: A review and a model for future research. Psychological Bulletin 99 (1986) 143-165. [77] R. van Bezooyen, Characteristics and recognizability of vocal expression of emotion. Dordrecht: Foris Publication, 1984. [78] M. D. Hauser, The evolution of communication. Cambridge: MIT Press, 1996. [79] J. Pittam, Voice in social interaction: An interdisciplinary approach. Thousand Oaks: Sage Publications, 1994. [80] C. B. Duchenne de Boulogne, The mechanism of human facial expression. Paris: Jules Renard, 1862. (Ed. and trans. by R. Andrew Cuthbertson, Cambridge: Cambridge University Press, 1990). [81] P. Ekman and W. V. Friesen, Manual for Facial Action Coding System. Palo Alto: Consulting Psychology Press, 1978. [82] P. Feyereisen, and J. D. delannoy, Gesture and speech: Psychological investigations. New York: Cambridge University Press, 1991. [83] P. Feyereisen, Neuropsychology of communicative movements, in: Gesture, speech, and sign, L. S. Messing and R. Campbell, Eds. New York: Oxford University Press, 1999, pp. 3-25. [84] D. McNeill, Hand and mind: What gestures reveal about thought. Chicago: The University of Chicago Press, 1992. [851 D. McNeill, Language and gesture. Cambridge: Cambridge University Press, 2000. [861 S. C. Levinson, Presumptive meanings: The theory of generalized conversational implicature. Cambridge: MIT Press, 2000. [87] T. Shallice, Specialization within the semantic system. Cognitive Neuropsychology 5 (1) (1988) 133142. [88] P. N. Johnson-Laird, Mental models. Cambridge: Cambridge University Press, 1983. [89] J. M. Mandler, Stories, scripts, and scenes: Aspects of schema theory. Hillsdale: Erlbaum, 1984. [90] M. I. Posner, Chronometric explorations of mind. Hillsdale: Erlbaum, 1978. [91] C. Umilta, The control operation of consciousness, in: Consciousness in contemporary science, A. S. Marcel and E. Bisiach, Eds. Oxford: Oxford University Press, 1988, pp. 334-356. [92] C. Umilta, Attenzione e coscienza [Attention and consciousness], in: Manuale di psicologia generale [Handbook of general psychology], P. Legrenzi, Ed. Bologna: I1 Mulino, 1994, pp. 185-250. [93] R. Jackendoff, Architecture ofthe languagefaculty. Cambridge: MIT Press, 1997. [94] K. Bach, Default reasoning: Jumping to conclusions and knowing when to think twice. Paczjic Philosophical Quarterly 65 (1984) 37-58. [95] J. L. McClelland, D. E. Rumelhart, and the PDP Research Group (Eds.), Parallel distributed processing. Cambridge: MIT Press, 1986. [96] B. J. Baars, A cognitive theory of consciousness. New York: Cambridge University Press, 1993. [97] B. J. Baars, In the theatre of consciousness: Global workspace theory, a rigorous scientific theory of consciousness. Journal of Consciousness Studies 4 (4) ( 1997) 292-309. [98] B. J. O’Keefe, and B. L. Lambert, Managing the flow of ideas: A local management approach to message design, in: Communication Yearbook 18, B. Burleson, Ed. Newbury Park: Sage Publications, 1994, pp. 54-82. [99] T. Herrmann, Nuetzliche Fiktionen. Anmerkungen zur Funktion kognitionspsychologischer Theoriebildungen [Useful pretence. Considerations about the function of cognitive psychology in theory making] . Sprache und Kognition 2 (2) (1983) 88-99. [ 1001 D. C. Dennett, The intentional stance. Cambridge: MIT Press, 1987. [ 1011 G. E. M. Anscombe, Intention. Oxford: Basil Blackwell, 1957. [ 1021 M. E. Bratman, Intention, plans, andpractical reason. Cambridge: Harvard University Press, 1987. [ 1031 F. Brentano, Psychologie vom empirischen Standpunkt [Psychology from an empirical point of view]. Leipzig: Dunker & Humblot, 1874. [ 1041 E. Husserl, Logische Untersuchungen, 1. Halle 1900-1901. Repr. in: Husserliana, vol. XVIII, The Hague 1975; vol. XIX, The Hague 1984. Engl. transl. (from the I1 ed.): Logical investigations.

42

L . Anolli / MaCHT

Routledge & P. Kegan. London, 1970. [ 1051 E. Husserl, Ideen zu einer reinen Phhomenologie und phihomenologischen Philosophie. Buch 1 [Ideas pertaining to a pure phenomenology and to a phenomenological philosophy. Vol. I]. Allgemeine Einfu'hrung in die reine Phanomenologie [A general introduction in the pure phenomenology]. Halle. 1913. [ 1 061 K. Jaszczolt, Discourse, beliefi and intentions: Semantic defaults and propositional attitude ascription. Amsterdam: Elsevier, 1999. [ 1071 K. Bach, Paving the road to reference. Philosophical Studies 67 ( 1992) 1 15- 137. [ 108J J. R. Searle, Expression and meaning. Cambridge: Cambridge University Press, 1979. [ 1091 C. H. Cooley, Social organization. New York: Pergamon Press. 1920. [ 1 101 G. H. Mead, Mind, se!f;and sociep. Chicago: University of Chicago Press, 1937 [ 1 1 11 H. A. Simon, Models uf bounded rationality. Cambridge: MIT Press, 1982. [ 1121 P. R. Cohen. and H. J. Levesque, Preliminaries to a collaborative model of dialogue. Speech Communication 15 (4) (1994) 265-274. [ 1 131 A. Duranti, Linguistic anthropology. Cambridge: Cambridge University Press, 1997. [ 1 141 J. K. Burgoon, Cross-cultural and intercultural applications of expectancy violations theory, in: Intercultural communication rheor?, (International and Intercultural Communication Annual. Vol. 19. pp. 194-214). R. L. Wiseman, Ed. Thousand Oaks: Sage Publications, 1995. [ 1 151 J. N. Cappella, Behavioral and judged coordination in adult informal social interaction: Vocal and kinesic indicators. Journal of Personali@ and Social Psjdtologr. 7 2 (1) ( 1 997) I 19- 1 3 1 . [ 1 161 E. Goffman, interaction ritual: Essay on the,face-ro;fucebehavior. New York: Doubleday. 1967. [ 11 71 E. Zerubavel, Hidden rhythms: Schedules and ca1endar.s in social life. Chicago: The University of Chicago Press, 198 1. [ 1181 A. Schutz. GesammelteAufsdtze. Volume I: Das Problem der sozialen Wirklichkeit (Collected papers. Vol. I: The problem of social reality]. Den Haag: Martinus Nijhoff. 197 1. [ 1 191 H. Giles, Accent mobility: A model and some data. Anthropological Linguisrics 15 ( 2 ) ( 1973) 87-105. [ 1201 H. Giles, and P. F. Powesland. Speech style and social evaluation. London: Academic Press, 1975. [ 12 1 ] H. Giles, N. Coupland, and J. Coupland, Eds. Contexts ofaccommodafion:De\~elopments in applied sociolinguistics. Cambridge: Cambridge University Press. 199 1 . [122] J. T Webb, Interview synchrony. An investigation of two speech rate measures in an automated standardized interview, in: Studies in diadic communication, A. W. Siegrnan and B. Pope. Eds. New York: Pergamon, 1972, pp. 115-133. [ 1231 J. Jaffe, and S. Feldstein, Rhythms of dialog. New York: Academic. 1970. [ 1241 M. Natale, Social desirability as related to convergence of temporal speech patterns. Perceptual and Motor Skills 40 ( 1975a) 827-830. [I251 M. Natale, Convergence of mean vocal intensity in dyadic communication as a function of social desirability. Journal of Personalin)and Social Psychology 32 (1975b) 790-804. [ 1261 P. Lieberman, Intonation, perception, and language. Cambridge: MIT Press, 1970. [ 1271 S. W. Jr. Gregory, A quantitative analysis of temporal symmetry in microsocial relations. .Jmerican Sociological Review 48 (1983) 129-135. [I281 S. W. Jr. Gregory, and S. Webster, A non-verbal signal in voices of interview partners effectively predicts communication accommodation and social status perceptions. Journal qf' Personalrot and Social Psychologv 70 ( 6 ) ( 1996) 123 1 - 1240. 11291 S. W. Jr. Gregory. K. Dagan, and S. Webster, Evaluating the relation of vocal accommodation in conversation partners' fundamental frequencies to perception of communication quality. Journal of' h'onverbal Behavior 21 (1997) 23-43. [ I301 K. Grammer, Human courtship: Biological bases and cognitive processing, in: The sociobioloe of se.rua1 and reproductive strategies. A. Rasa. C . Vopel. and E. Volland. Eds. London: Chapman and Hall. 1989. pp. 147-169. [ 13 1 ] K. Grammer. K. Kruck. M. S. Magnusson, The courtship dance: Patterns of nonverbal synchronization in opposite-sex encounters. Journal ofllronverbal Behavior 22 (1) ( 1998) 3-29. [ 1321 T. Koss, and R. Rosenthal, Interactional synchrony, positivity and patient satisfaction in the physicianpatient relationship. Medical Care 35 (1997) 1158-1 163. [ 1331 R. E. Turner. C. Edgley. and G. Olmstead, Information control in conversations: Honesty is not always the best policy. Kansas Journal ofSocinlo'sp-1 I (1) f 1975) 69-89.

43

Say Not to Say L. Anolli et al. (Eds.) IOS Press, 2002

2

Semantic Transparency, Semantic Opacity, States of Affairs, Mental States and Speech Acts Anne REBOUL IAGO: Men should be what they seem, Or those that be not, would they might seem none. OTHELLO: Certainly, men should be what they seem. Shakespeare, Othello

If things were always what they seemed, how impoverished would be the imagination of man. Pursewarden, quoted by Durrell, Balthazar.

Abstract: There are two well-known views of linguistic communication: the code model and its counterpart, the hypothesis of semantic transparency. If both of these views were correct, then there would be only one possible type of mishap in communication, that due to noise in the communication channel. However, none of these views is correct. I will sketch a quick history of pragmatics relative both to the code model and to the hypothesis of semantic transparency. As we will see, the most recent pragmatic theories have discarded them. This enables them to account for the whole range of communication mishaps, from mere misunderstandings to sophisticated deception strategies. Incidentally it also enables them to propose a solution to Moore's paradox.

Contents 44

2.3 2.4 2.5 2.6

Introduction.. ....................................................... The code model of communication and the birth of pragmatics ........................................................... Going beyond the code model: relevance theory.. ............. Lies ................................................................... Moore's paradox.. .................................................. Conclusion.. ........................................................

2.7

References.. .........................................................

68

2.1 2.2

45 50 60 64 68

44

2.1

.4 Rehoul ’Semantic Transparens\.

Introduction

It is often considered that language is - and has evolved as - a tool for communication and that it is semantically transparent relative to what is communicated by a given sentence. In other words, there would be nothing more to linguistic communication than encoding and decoding. In the scientific formulation of that hypothesis, which strongly relies on Shannon and Weaver’s theory of communication [ 11, the only possible ground for misunderstanding is the noise which may interfere with the transmission of the signal. There are two sides to the hypothesis of the semantic transparency of utterances: the first one, according to which utterances are semantically transparent relative to the states of affairs they represent - which was called by Austin [2] the descriptive .fallacv -, has been largely discredited after the advent of speech acts theory; the other, according to which utterances are semantically transparent relative to the mental states of the speaker is still very active and speech acts theory has been instrumental in making it so. Whatever its version, semantic transparency is wrong not so much because language cannot represent reality, whether by reality one designates public states of affairs or private mental states. but because language is semantically underdetermined as has been shown by. notably. Grice [3]. The semantic underdeterminacy of language entails a corresponding semantic opacity of utterances and this, in its turn, can be a rich ground for miscommunication. be it mere (involuntary) misunderstanding or willful miscommunication. The semantic opacity of utterances has a main consequence: the interpretation of utterances cannot be a simple matter of decoding and, more generally, the process of linguistic communication is not only a matter of encoding and decoding either. Rather the whole process rests on a complicated “system” of communicative and informative intentions on the part of the speaker and the (more or less correct) recovery of these intentions on the part of the hearer. Going from the purely linguistic interpretation of the utterance to its “complete” interpretation entails passing through a pragmatic process of “free enrichment”, relying on a piecemeal context. The present paper will be squarely in the theoretical framework of Relevance Theory [4]. and will, among other things, assume its distinction between informative and communicative intentions. From that vantage point, I will try to give a general panorama of misunderstandings and various kinds of miscommunication, going from the case where the speaker is misleading his hearer unwillingly, because he himself is mistaken about the state of affairs he is describing, to the case where the speaker is willfullymiscommunicating. leading his hearer to believe that he (the speaker) is lying and not telling the truth. when. in fact, he is telling the truth but merely exploiting the mistrust of his hearer and leading him to the (false) belief that the propositional content of his utterance is false when it is in fact true. I will try to elucidate how those kinds of more and more Machiavellian and complicated miscommunication rely on the human ability to metarepresent one’s own and others’ representations or metarepresentations. I shall show that this is the basis of what Dennett [ 5 ] has called the interpretive stance and that it exploits the so-called principle of charity which is generally operative in human communication. Finally I shall draw the consequences of that for the Gricean and neo-Gricean accounts of communication and for the strong version of speech acts theory (Vanderveken’s [6]) where some types of speech acts are supposed to anah~ricalh.imply some types of mental states on the speaker‘s part.

A . Reboul /Semantic Transparency

2.2

45

The code model of communication and the birth o f pragmatics

Suppose that you have two communicating devices, for instance human beings. These communicating devices are able to entertain internal representations, which we will call messages. They also are able to send signals, which, according to Sperber and Wilson [4], can be defined as ' I . . . a modification of the external environment which can be produced by one device and recognized by the other" (p. 4). In other words, the message which is an internal representation of communicating device CD 1 is encoded as a signal, transmitted through a channel to communication device CD2, which then decodes that signal to uncover the message. Communication is successful if, at the end of the process, CDl and CD2 both entertain the message which, initially, was an internal representation of only CDl. Under the code model of communication, the whole process, following a diagram by Shannon and Weaver [ 11, can be represented as follows [4, p. 51: received acoustic \signal

/Thougm\

thought

\ I thought processes

II I

decoder

encoder

thought processes

f SPEAKER

noise

Figure 2.1 The code model of communication

According to this view, there are two important hypotheses which must be taken into account in any analysis of communication: 0 The only possibility for miscommunication is a misunderstanding due to noise in the communication channel; 0 Communication is successful in as much as the message recovered is the same as the message sent, i.e. if both communicative devices share the same internal representation as a consequence of the communication process. As we shall see, both of these hypotheses, which are still widely accepted in some linguistic circles, are mistaken.

2.2.2 Two versions of the hypothesis of the semantic transparency of linguistic communication

The hypothesis of the semantic transparency of linguistic communication is obviously dependent on the code model of linguistic communication. Indeed, it is hard to see how semantic transparency could be secured outside of the code model. However, depending on the nature of the message which is communicated, it can be said that the hypothesis of the

.4. Rehoul /Semantic Transparent\*

46

semantic transparency of linguistic communication has two sides: the first one is relative to the case in which the message being communicated is the representation of a state of affairs in the world; the second is relative to the case in which the message being communicated is not the representation of a state of affairs but is itself a state of mind of the communicative device which entertains it. A state of mind, in this instance, can be a belief, a desire, etc. The hypothesis of the semantic transparency of linguistic communication in the case in which the message is a representation of a state of affairs in the world claims that a given signal can be decoded in such a way that it will yield a full proposition, i.e. a representation which can be evaluated in terms of its truth-value. The hypothesis of the semantic transparency of linguistic communication in the case in which the message is itself a state of mind claims that a given sentence is, so to speak, a window to the speaker's mind. For a long time, philosophers considered that the aim of linguistic communication, and indeed the function of language was the representation of reality. They adopted the hypothesis of semantic transparency. A sentence was thus considered as a means of conveying the description of a state of affairs and its interpretation was considered as yielding a proposition, in the truth-conditional sense. In keeping with the code model. communication would be seen as successful in as much as the proposition recovered by the hearer is identical with the proposition that was communicated. Thus, until the fifties, the code model of linguistic communication as well as the semantic transparency hypothesis were concentrated on linguistic descriptions of states of affairs in the world. However, some obvious problems were pointed out, notably regarding the use ofdeixis. Indeed, if John says sentence ( 1 ) in London in 1897. the proposition he expresses is quite different from the proposition that Giovani expresses when Giovani says the same sentence in Roma in 200 1 : ( 1)

I am here

OM'.

This encouraged philosophers to distinguish between the fype (or sentence) and the token (or utterance). The sentence or type is an abstract entity, which is realized on different occasions as different tokens or utterances. A given utterance of a sentence is interpreted relative to a set of parameters (i.e. speaker. addressee, time and place of utterance). There is no need to go beyond the code model. Sentences such as (1) can still be considered as expressing propositions when they are uttered, though the proposition expressed by a given utterance of a sentence can differ with that expressed by another utterance of the same sentence, if deictics are involved. Thus, the propositions expressed by John and by Giovani are respectively represented (informally) as (2) and (3): (2) (31

John is in London in 1897. Giovani is in Romn in 2001

2.2.2 Austin and the notion of speech act This was the situation in 1955 when John Austin, an Oxford philosopher, gave the William James Lectures at Harvard [2], roughly a year before a symposium at MIT which involved Allen Newell, Herbert Simon, Noam Chomsky and George Miller and which is considered as marking the birth of cognitive sciences. Austin's William James Lectures are the true origin of a discipline, which at that time did not see itself as a cognitive science. i.e. pragmatics. Austin described the position adopted by previous analytic philosophers as the descriptive fallnq-. He attacked the notion of language as a tool for the description of

A. Reboul /Semantic Transparency

47

reality, by pointing out that, even if only affirmative sentences are considered, some - if not most - of them cannot properly be analyzed in terms of truth-conditions and as describing reality. Hence, evaluating them for truth-value was just ineffective, most of them not yielding to an analysis according to which they were true or false. Indeed, properly considered, without any a priori prejudice, affirmative sentences could be ranged in two classes depending on whether they actually were, or were not, describing reality: the first class is the class of constative utterances, while the second is the class ofperformative utterances, so called because such non-descriptive affirmative utterances were used to act on the world rather than to describe it. Below, examples (4)-(5) and (6)-(7) are respectively examples of constative and of performative utterances: (4) (5) (6) (7)

The cat is on the mat. It's raining. Ipromise that I will come tomorrow. I baptize you in the name of the Father, the Son and the Holy Spirit.

What is more, constative and performative utterances differed not only in their functions but also in the way they could be evaluated. Constative utterances express propositions which can be evaluated for truth-conditions, while performative utterances are evaluated relative to their success, i.e. they are not true or false but felicitous or nonfelicitous. This was Austin's position in the first few lectures he gave in the William James Lectures cycle. However, in the following lectures, his position changed in as much as he realized that constative and performative utterances could not be contrasted relative either to their function or to the mode of their evaluation. Indeed, constative utterances are used to modifL the hearer's states of mind, i.e. to act and can be evaluated for felicity. And performative utterances do express propositions - for instance, (6) express the proposition that the speaker will come tomorrow - which can be evaluated for truth-value and their felicity may depend on the truth or falsity of the proposition expressed. This led Austin to a modification of his view, according to which any grammatical sentence uttered in "normal" conditions was ips0 fact0 the accomplishment of a speech act of a particular kind, i.e. an illocutionary act. Indeed, this new version of speech act theory introduced a triple distinction between kinds of speech acts: The locutionary act: the act of saying something; The illocutionary act: the act accomplished in saying something; 0 The pdocutionary act: the act accomplished by saying something. In other words, locutionary acts are the very acts of uttering something; illocutionary acts are, for instance, promises, orders, assertions; perlocutionary acts are, for instance, persuasion. Utterances which in the first lectures would have been classified as constative were classified as illocutionary acts of assertion in the following lectures. Austin's William James Lectures were the foundation of a new discipline in language studies, i.e. pragmatics. However, as we will see below, the Lectures were posthumously published and Austin's work was pursued by John Searle, an American philosopher. 2.2.3 Grice and the notion of non-natural meaning

Before we turn to Searle, it is interesting to note that another milestone for pragmatics occurred when another philosopher, Paul Grice, published a seminal paper in 1957, soberly entitled Meaning [7]. Ten years later, he also gave the William James Lectures in Harvard

48

'4. Rehod/ Semantic Transparenel.

[3], thus securing as important a place for himself in the history of pragmatics as that of Austin. Grice's definition of meaning was formulated as follows: "'[the speaker] meant something by x' is (roughly) equivalent to '[the speaker] intended the utterance of x to produce some effect in an audience by means of the recognition of this intention"' [7, p. 3771. This definition, in itself, does not indicate that the recognition of the speaker's intention should be done through a decoding process, neither does it indicate that any linguistic convention is involved. Thus, the Gricean definition of meaning can be considered as a first step out of the hypothesis of semantic transparency and the code model of linguistic communication. This was reinforced when Grice pronounced the William James Lectures [3], describing the communication processes involved in conversation and linguistic communication. He made it obvious that, indeed, there was more to communication than mere encoding and decoding. Grice introduced a distinction between what is said and what is communicated. What is said may indeed be recovered through an encoding-decoding process and thus may be subject to the hypothesis of semantic transparency. However, what is communicated is a different matter. While what is said is a proposition, what is communicated is both what is said (i.e. the proposition expressed) and other things which can be recovered from what is said but which are not part of the explicit content of the utterance. These are called by Grice implicatures and there is no question that the recovery of implicatures quite simply is not done through an encoding-decoding process. Grice's hypotheses on the processes through which implicatures are recovered are well known: according to him, conversation is a cooperative process in which the speaker and the hearer obey the cooperatitie principle. The cooperative principle is articulated in nine conversational maxims. which play a role in the communication process in as much as they give rise toimplicatures. Let us take the first maxim of quantiht: Make your contribution ns informative as i s required. Suppose that you are told: (8)

Anne has,four children.

On a strictly semantic understanding, (8) can be interpreted as Anne has four children and maybe more. This, however, is clearly not the meaning which the speaker intended to convey. That meaning is: Anne has not more than four children. It is recovered through an inferential process in which the hearer, relyng on the first maxim of quantity, makes the hypothesis that if the speaker had intended to communicate that Anne has at least four children, he or she would have said it explicitly and that his having said (8) is a good indication that he intended to communicate that Anne has no more than four children. Thus, Grice's account of linguistic communication goes beyond the code model in that what is said does not exhaust what is meant or communicated. Thus, though the code model applies to the recovery of what is said, it does not apply to the recovery of implicatures. Hence, the Gricean model of communication is mixed: it is codic concerning the semantic interpretation of what is said. and inferential concerning the recovery of implicatures. What is more, the communicative intentions of the speaker, or hypotheses on these, play a role in this inferential process. In this sense, Grice clearly anticipated Dennett's notion of intentional stance [S].which can be informally defined as follows: The intentional stance is adopted whenever one interprets another's behavior on the basis of two general principles: 1. Other individuals are rational agents. They have beliefs. desires and other mental states. -. 7

A. Reboul /Semantic Transparency

49

Though, as we will see, Grice's position is not entirely satisfactory, the intentional stance is very much to the fore of contemporary pragmatics. 2.2.4 Searle and the conventionalist trend in speech acts theory

Austin died prematurely in 1960 and his William James Lectures had a posthumous publication in 1962. This was not the end of speech acts theory as an American philosopher, John Searle, carried on Austin's work and gave what became for the next fifteen to twenty years the foundation of pragmatics, i.e. a basic theory of speech acts [8]. Searle's main contribution was in articulating precisely much of what was left implicit or insufficiently developed in Austin's work. His first task was to make explicit the fact that speech act theory is based on two notions: intention and convention. In other words, a speech act implies an intention of the speaker which is recovered by the hearer through linguistic convention. In order to do this, Searle concentrated on only one of the three categories of speech acts distinguished by Austin, illocutionary acts. He then borrowed Grice's definition of meaning, though he gave it a twist. First, he redefined the recognition of the speaker's intention in such a way that it would depend exclusively on the linguistic conventions governing the sentence uttered, i.e. on its semantics. Second, given that illocutionary acts are supposed to be linguistically encoded, he replaced the notion of the effect that the speaker intended to produce on the hearer through his utterance - which, in Grice's definition, could be interpreted as either perlocutionary (i.e. non semantic) or illocutionary (i.e. semantic) - by the notion of the illocutionary effect which the speaker intended to produce in his hearer through his utterance. Hence, Searle's main goal in his book seems to be to make pragmatics - given that speech acts theory pertains to pragmatics - a part of semantics. Another contribution of Searle's which will be of interest here is that he tried to make explicit the conditions under which a given illocutionary act is felicitous. Predictably, felicity conditions, though they partly differ from one type of illocutionary act to the next, include conditions relative to absence of noise in the communication channel, community of linguistic code and a sincerity condition which I will now detail. The sincerity condition is to be found in felicity conditions regardless of the type of speech acts considered. Basically, what it says is that the illocutionary act in question can be felicitous iff the speaker is in the required state of mind: for instance, if the illocutionary act is an assertion, the speaker must believe in the truth of the proposition expressed, while if it is a promise, the speaker must have the intention to fulfill his promise, etc. The sincerity condition is the source of a variety of difficulties for speech acts theory as I pointed out in previous works ([9], [ 101, [ 1 11, [ 121, [ 131). I will have more to say about it below. Finally, Searle introduced a distinction between the marker of illocutionary force and the marker of propositional content. In a utterance such as (9), the marker of illocutionary force is Ipromise that and the marker of propositional content is I will come tomorrow: (9)

Ipromise that I will come tomorrow.

Thus, both illocutionary force and propositional content are linguistically marked and both are subject to the code model and are considered as semantically transparent. 2.2.5

Conclusion

What then of the code model in the first decades of pragmatics? As we have seen, though what pragmatics called the "descriptive fallacy" was discredited by Austin in his William

so

'4. Rehoul Semantic Trunsparencj,

James Lectures, nevertheless the code model was preserved in the recovery of the proposition expressed by the utterance, as well, in the later Searlian theory of speech acts, as the recovery of the illocutionary force. It is interesting to note that on Searle's view, interpreting a utterance is, just as it is for Grice, recovering the intention of the speaker. However, and this is where Gricean pragmatics and Searlian pragmatics part, the Searlian model is just the code model applied both to the recovery of propositions and to the recovery of illocutionary force, exhausting the intention of the speaker, while the Gricean model limits the code model to the recovery of what is said and defends an inferential model for the recovery of further implicatures. This inferential model heavily relies on the notion of speaker's intentions as well as on the cooperative principle and conversational maxims. By contrast, there is nothing of the kind in the Searlian model. Thus, though both models rely on a double-barreled intentional definition, where the speaker has first the intention to communicate a given content through his or her utterance and the further intention that the first intention be recognized as such, that further intention plays a major role in the interpretation process according to Grice and plays no role in the interpretation process according to Searle. Hence, Grice's pragmatics can be said to be authentically a model of double-intentionality whereas Searle's pragmatics cannot be described in such a way. And, as we will see, double-intentionality in this authentic form is exactly what is needed to go beyond the code model of linguistic communication.

2.3

Going beyond the code model: relevance theory

Before going beyond the code model, let us just go back to the semantic transparency hypothesis. According to Searle, utterance are semantically transparent - in the sense that they can be interpreted along lines suggested by the code model - both relative to their propositional content and relative to their illocutionary force. According to Grice, they are semantically transparent - in the same sense - relative to what is said. Nevertheless, the code model does not suffice to determine the implicatures of a given utterance and utterances are thus not semantically transparent relative to what is being communicated in excess of what is said. This, however, might be a simple matter of a terminological disagreement, where Searle and Grice would have two different uses (and understandings) of the phrase what is communicated. In other words, it might be that they agree on the definition of what is said as being the sum of illocutionary force andpropositional content, but disagree in as much as Searle would see what is communicated as equivalent to what is said while Grice would see what is communicated as the sum of what is said and the implicatures of the utterance. Thus, the question would be, not whether what is said is semantically transparent, but whether Grice is right to include implicatures in what is communicated, i.e. to make them a part of what the speaker intended to communicate. After all, it might be argued that the recovery of what is said is quite enough to recognize the speaker's intention and that the recovery of anything further is not necessary to it. Again, this would mean that all of what is necessary for the analysis of linguistic communication is given by the code model with its two corollaries: 0 The only possibility for miscommunication is misunderstanding due to noise in the communication channel; Communication is successful in as much as the message recovered is the same as the message sent. According to what is said above, a possible definition for miscommunication would be that it occurs whenever communication is not successhl. i.e. when there is a discrepancy

A. Reboul /Semantic Transparency

51

between the message recovered and the message sent. This would certainly be true according to the code model. However, it would be interesting to have a more intuitive clue that miscommunication has happened. I propose that miscommunication can be said to occur when the participants in the interaction seem to be talking at cross-purposes, as, for instance, when each participant is saying the same thing over and over again. This, it should be clear, can happen when there is noise in the communication channel. People would repeat themselves in the hope of being heard or read by their audience. Whether it only happens in such a case is however a good test of the validity of both the code model and Searle's view. If, indeed, there is nothing more to what is communicated than what is said, given that what is said is supposed to be semantically transparent, then without noise in the communication channel, no miscommunication should occur. 2.3.I Misunderstandings without a noise in the communication channel

However, regrettably for the code model of linguistic communication, such miscommunication can occur, and in fact frequently does. Some time ago, the transcription of a radio conversation at sea between Americans and Canadians was widely circulated on the web because of its comic impact: (1 0)

AMERICANS: Please reroute your ship by I5 degrees North to avoid a collision. Over. CANADIANS: Please reroute YOUR own ship by 15 degrees South to avoid a collision. Over. AMERICANS: Here the captain of a US navy ship is speaking. I repeat: please reroute YOUR ship. Over. No, please reroute YOUR ship. CANADIANS: AMERICANS:THISIS THE uss AIRCRAFT CARRIER LINCOLN,THE SECOND SHIP IN

us

NAVY. W E ARE ACCOMPANIED BY THREE DESTROYERS AND A GREAT NUMBER THE OF OTHER SHIPS. I ASK YOU TO REROUTE YOUR SHIP BY 15 DEGREES NORTHOR RESTRICTING MEASURES WILL BE TAKEN TO ENSURE THE SECURITY OF OUR SHIP .

OVER. CANADIANS: This is a lighthouse.

Though it is very far from certain that example (10) actually is the transcription of an authentic radio exchange between the American and Canadian navies, it is nevertheless interesting in as much as it is a good example of a misunderstanding - most of it represents Americans and Canadians as speaking at cross-purposes - and there does not seem to be anything like noise involved. In fact, what is happening in example (10) is something over and above the predictions of the code model. Every sentence seems to be perfectly well understood by its addressee but no common ground for communication is found. The conversation, indeed, consists mainly in each speaker in his or her turn repeating his or her previous utterance, adding, in the case of the last American intervention, a clear menace. The final Canadian intervention clears the misunderstanding and gives a clue as to its origin. In (10), the miscommunication arises through a discrepancy between the informations accessible to the Americans and those accessible to the Canadians. The Americans know the spatial position of the Canadians and the Canadians know the spatial position of the Americans. Both the Canadians and the Americans know that their respective addressee belongs to the opposite navy. However, the Canadians have an information that the Americans do not. The Canadians know that they are not on a ship (i.e. a mobile object)

52

A. Rehod/ Semantic Transparent,,

but in a lighthouse (i.e. a non-mobile object). The American ignorance of this fact explains the insistence on their part that the Canadians reroute. The Canadian knowledge of this fact equally explains their own insistence on the Americans rerouting. It is only when the Canadians, in their last intervention, finally say that they are in a lighthouse that the misunderstanding is cleared. Thus in this specific case, there is a clear miscommunication without any noise in the channel of communication. What is more, it is not only the code model of communication that is weakened by such an example. It is also the Searlian position as outlined above, that is, the equivalence between what is said and what is communicated. Clearly, what is communicated, notably in the last Canadian intervention is more than what is said. What is said, on a Searlian analysis, is a propositional content This is a lighthouse with an illocutionary force of assertion. What is communicated is however much more than that. What is communicated is a refusal to obey the previous American order as well as an explanation of this refusal. The interpretation of the last Canadian intervention appeals not only to Gricean maxims of conversation but also to encyclopedic knowledge about what lighthouses are. This raises a few questions: though the Gricean interpretation mechanism is inferential and. hence, part of the interpretation process is inferential as well, it is not clear that conversational maxims are sufficient to account for these inferential interpretation processes. For instance, how does linguistic interpretation interact with encyclopedic knowledge in the above instance? What is more, it is very far from certain that Grice would have endorsed a contextualist pragmatics (see, e.g., Recanati [ 141) and it is hard to see how encyclopedic knowledge can interfere in the inferential interpretation process on an anti-contextualist view of pragmatic interpretation. This is where yet another pragmatic theory. which might be called post-Gvirean rather than neo-Gricean, comes into play. 2.3.2 Relevance Theon,

Relevance Theory [4] can be said to be postGricean, rather than merely neoGricean, in that, though it adopts some of the major tenets of the Gricean theory of communication. it does not adopt all of them. For instance, as Grice does, Sperber and Wilson see the communication process as a mixture of both encoding-decoding and inference. However, the frontier between what belongs to the code domain and what belongs to the inference domain does not run in the same place in Grice's theory an8 in Relevance Theory. This. relative to the question of semantic transparency. may be Relevance most important departure from Gricean theory. What, in fact, Sperber and Wilson say is that semantic underdetermination is more widespread than what was previously believed and that it infects not only what is communicated over and above what is said but also what is said. They point out that it is not so much that utterances are incomplete but that language does not semantically determinate the proposition expressed by a given utterance, even if a parametric approach (see 6 2.2.1) is adopted. They point out the existence of frequent lexical or syntactic ambiguities as well as the semantic underdetermination of referring expressions and the even more frequent use of loose talk. In other words, though encoding-decoding processes play a part in linguistic communication, they are far from the whole story even if the proposition expressed is concerned. What is more. and this is one point of agreement with Grice, there is much more to what is communicated than what is said. However, there is no need to postulate a cooperative principle between communicators, nor to deal with nine maxims. A single cognitive principle can do the whole job.

A . Reboul /Semantic Transparency

53

Let me briefly outline the theory. The first thing to note is that it may well be the first pragmatic theory to explicitly claim a place for itself among cognitive sciences. What is more, it is clearly chomskyan in its view of both what goes on in linguistics and what goes on in pragmatics, placing pragmatics squarely out of the linguistic domain. In its first guise, that is, on the first publication of [4], in 1986, Relevance Theory adopted the modular view of mind advocated by Fodor [ 151. According to Fodor, the mind is organized in a hierarchical way, with vertical modules specialized in the interpretation of stimuli in a given modality and a horizontal, non specialized central system. Examples of the vertical modules would be, for instance, a visual module, an olfactory module, an auditory module, etc. Modules would output a first interpretation of a stimulus that would then serve as input to the central system. Incidentally, on this view, the binding problem - i.e. the problem of how different stimuli all produced by the same source can be bound together to yield a recognition of this source - would be solved by the central system where inputs from the relevant modules could be considered simultaneously. Translated in the domain of linguistic communication, which was the main interest of Sperber and Wilson, the modular view of mind amounts to the following: there is a linguistic module specialized in the interpretation of linguistic stimuli, i.e. utterances. This linguistic module deals with the problems traditionally considered as pertaining to phonology, syntax and semantics. In keeping with Sperber and Wilson's view about the underdetermination of language, the output of the linguistic system is not a full propositional form, which can be evaluated for truth-value, but a logical form in which some elements (reference assignation for instance) are not hlly determined. This logical form consists in a structured sequence of concepts. It is the input to the pragmatic processes which take place at the level of the central system. Pragmatic interpretation, as is the case in Grice's theory, is largely inferential and nondemonstrative. It operates not from the logical form of the utterance and from a set of Gricean maxims but on the logical form of the utterance and from a set of premises, i.e. a context. This, apart from the pragmatic intrusion in the determination of what is said, is the second major departure of Relevance Theory from Gricean pragmatics. Relevance Theory is and Gricean pragmatics is not contextualist. According to Relevance Theory, utterances are interpreted relative to a context, and the context is not given once and for all: it is rebuilt for each new utterance. As said above, the context is a set of premises, i.e. of propositions which the hearer entertains as true or probably true. These propositions come from different sources: the interpretation of previous utterances in the same conversation or discourse; the situation in which the communication occurs, i.e. fiom the physical environment; encyclopedic knowledge, accessed through the concepts in the logical form. However, it seems clear that these sources yield an enormous amount of potential premises and that a selection process has to apply in order to yield a manageable context. Similarly, applying deductive rules - which Sperber and Wilson describe as operating both on the premises in the context and on the logical form of the utterance - could be a non-finite process yielding ever more conclusions. Again some sort of restraining influence must be brought to bear on the interpretation processes if they are not to run away. Relevance solution to these problems goes through a single and general cognitive process: the Principle of Relevance, to which the whole theory owes its name. The name of the principle is borrowed from Grice, but it has two main differences from its Gricean counterpart: it relies on a definition of Relevance; it subsumes all other Gricean maxims, hence its singularity. Relevance is defined (roughly) as follows:

54

A. Rehoul /Semantic Transparency

Relevance: The more effects a given stimulus has, the more relevant it is. a) The less efforts a given stimulus demands for its processing, the more b) relevant it is. Hence, relevance is a matter of equilibrium, of balance, between processing efforts and effects. In the last version of Relevance Theory [4], Sperber and Wilson distinguish between two principles of Relevance, a general cognitive principle which applies throughout cognitive processes to the effect that human cognition is geared to achieve maximum relevance and a communicative principle of relevance, limited to stimuli produced in the course of a communicative exchange. It is the second principle that we will be interested in here. It states that "Every act of ostensive communication communicates a presumption of its own optimal relevance" (p. 260). Before I articulate more precisely how the communicative Principle of Relevance should be interpreted, let me just say that it operates both on the selection of the context fiom the different sources indicated above and on the limitation of the inference process of interpretation. I would like to begin the exegesis of the communicative Principle of Relevance by explaining what ostensive communication is. As indicated above, Relevance Theory, just as Gricean pragmatics, is an authentic doubleintentionality theory and this manifests itself in that acts of communication are described as implying both an informative intention and a communicative intention. The informative intention is defined as the intention of "[making] manifest or more manifest to the addressee a [given] set of assumptions" (p. 58). The communicative intention is defined as the intention "to make mutually manifest to the communicator and the addressee that the communicator has this informative intention" (p. 61). "A fact is man$est to an individual at a given time if he is capable at that time of representing it mentally and accepting its representation as true or probably true" (p. 39). The cognitive environment of an individual at a given time is the set of facts which are manifest to him at that time. Two individuals can share a mutual cognitive environment i.e. a cognitive environment in which it is manifest which people share it. In a mutual cognitive environment, every manifest assumption is mutuallv manifest. In other words, an act of communication is ostensive in as much as the communicator's informative intention is mutually manifest both to him and to his addressee. As ostensive acts, communicative acts have a special claim to relevance: given that the communicator is claiming a non-negligible share of her addressee's attention and cognitive efforts, her act of communication must yield enough effects to be worth the addressee's attention. What this means is that communicative acts are interpreted in such a way that the interpretation obtained is consistent with the communicative Principle of Relevance. i.e. in such a way that the contextual effects are enough to offset the processing efforts. Very briefly, contextual effects are cognitive effects, i.e. they exist in as much as they modify the representation which an individual entertains. at a given time, of the world. They can be of three sorts: 0 a new information is added to the existing representation of the world; 0 a pre-existing information becomes either more or less plausible, i.e. the degree of confidence with which it is entertained is modified either positively or negatively; 0 a pre-existing information is contradicted and either it or the new contradictory information drawn from the interpretation of the stimulus is erased, depending on the degree of confidence with which each is entertained.

A . Reboul /Semantic Transparency

55

Thus, ceteris paribus, the (right) interpretation will be the most accessible and, more generally, both the full propositional form and the implicatures of a given utterance will be considered in the order of their accessibility. Finally, let us turn to speech acts, which were the basis of pragmatics at its birth. Sperber and Wilson very clearly part ways there with previous pragmatic accounts. According to them, two types of illocutionary acts should be clearly distinguished: those that are socially defined (baptism, most administrative declarations, legislative acts, etc.) and which rely on non-Iinguistic conventions; those that rely on linguistic conventions and that are not socially determined. Though the first class may have a sociological interest, it does not pertain to pragmatics. The second class does pertain to pragmatics but, concerning it, Sperber and Wilson question one of the main assumptions of Searlian speech acts theory, that is, the hypothesis that each utterance should be classified as performing one or another illocutionary act, such as menace or prediction. This, according to Sperber and Wilson, is not the case, and the same sentence (e.g. I will come tomorrow) can be interpreted, depending on the context, as a promise, a menace, a prevision, etc. in defiance of the Searlian conventionalist view. Thus, it is more realistic to distinguish three classes of non-social illocutionary acts: acts of saying that, acts of telling to and acts of asking whether. The main thing to say is that success of one of these acts does not depend on the satisfaction of a sincerity condition as was the case in Searlian pragmatics. The central speech act for the present paper is the act ofsaying that, that Sperber and Wilson define as follows: "Let us define saying that P , where P is the propositional form of the utterance, as communicating that the thought interpreted byP is entertained as a description of an actual state of affairs" (p. 237). As we shall see below, this definition does not meet with the difficulties that plague accounts with sincerity conditions.

2.3.3 An extension of Relevance Theory to discourse interpretation Relevance Theory is, properly speaking, a theory of utterance interpretation, given that utterances are considered as a special case of ostensive communication. It can, however, also deal with all manners of ostensive communication from non-verbal communication to verbal communication extended over the limits of an utterance. In other words, it can account not only for utterance interpretation, but also for discourse interpretation (for a detailed defense of this claim, see [16], [17], [18], [19], and [20]). The first thing any theory of discourse should presumably do is to offer a definition of it. My own would run as follows: A discourse is a non-arbitrary sequence of utterances. This much should not be controversial. Controversy begins, however, with the seemingly inoffensive expression non-arbitrary, which can be interpreted in two quite different ways: as implying that the non-arbitrariness of discourse is due to the fact that there are linguistic rules governing the production and interpretation of discourse; as implying that discourse is non-arbitrary in a pragmatic sense, i.e. there are pragmatic rules governing the production and interpretation of discourse. These two different interpretations have widely different implications: in the first case, discourse can be - and has often been - seen as another linguistic unit, to be added to phoneme, morpheme and sentence; in the second, discourse can be seen as the product of pragmatic rules which would be specific to it or as the product of the same rules that govern the production and interpretation of utterances and it should be noted that the choice between the two depends on which version of pragmatic theory one favors. If one adopts a conventionalist view of pragmatics, then one will probably choose the first interpretation; if one adopts, as I do, a postgricean theoretical framework, such as Relevance Theory, one will probably defend the second interpretation. The choice between those two interpretations is not trivial: choosing the first one would presumably lead to

56

.4. Rebod/ Semantic Transparenci

considering discourse as a pragmatic unit, whereas choosing the second would not (discourse would merely be considered as just another act of ostensive-inferential communication). The preceding distinction between a linguistic or conventional pragmatic approach and a radical pragmatic approach has a corollary: if there are rules - linguistic or conventionally pragmatic - specific to discourse, those rules would govern theform of the discourse, not its content, just as syntactic rules govern the form of the sentence. not its content. Non conventional pragmatic rules would not only not be specific to discourse; they would just relate to its content, not its form. This is the view briefly outlined below. In a linguistic theory of discourse, there are linguistic rules relative to discourse; in a conventionalist pragmatic theory of discourse, there are pragmatic rules relative to discourse. In both cases, the non-arbitrariness of discourse can be accounted through the existence of a set of rules which would be specific to it and whose application would guarantee its coherence, just as the application of syntactic rules guarantees the grammaticality of a sentence. In a pragmatic framework which does not hypothesize the existence of such a set of rules, how can one account for the non-arbitrariness of discourse, and - even more crucially according to some currents in discourse analysis - account for its coherence or incoherence? An answer to this question can be found when one considers that, indeed, discourse is a case of ostensive communication, and, as such, implies the same double intentionality as do utterances. What is more, just as the Principle of Relevance and non-demonstrative inference processes can account for the recovery of intentions behind utterance production. they can account for the recovery of the intentions behind discourse production. In effect, this means that the two types of intentions in utterance, i.e. informative intention and communicative intention, are also to be found in discourse. Thus, the speaker of a discourse has a global informative intention and a global communicative intention, concerning the whole of his discourse, just as she has a local informative intention and a local communicative intention for each utterance in her discourse. These global intentions are recovered through the interpretation of the utterances composing the discourse, though the process is not purely incremental (i.e. the product of the interpretation of a discourse is not a sequence - or a sum - of the products of the interpretation of the successive utterances in that discourse) and this is where non-arbitrariness creeps back in. Why do we need global intentions in addition to local intentions? Well, this of course hinges on the fact that local intentions in Relevance are taken to concern utterances, not sequences of utterances and though they are enough (together with the rest of Relevance) to account for the fact that utterances are interpreted relative to context (which includes much more than just the interpretation of previous utterances), they are not quite enough to explain why, in a discourse, a given utterance occurs when it does. This is, among other things, because speaker’s intentions regarding discourse are not always straightforward: for instance. in literature, an author may willfully mislead her audience, that is, some utterances may lead the reader in a direction which is not consistent with the whole of the text. before surprising him with the conclusion (e.g. A. Christie in The Murder qfRoger ,4ckroyd). In this specific case, the reader will eventually recover as the global informative intention the culpability of the doctor, whereas he might legitimately have recovered most local informative intentions as indicating the doctor’s innocence. By the way, this is also why the interpretation of the discourse is not equivalent to the sum of the interpretation of the successive utterances, though each of these local interpretations is done relative to a context. And, again, this is a matter of contents and relations between them and not of form. The non-arbitrariness of discourse is due to two main factors: the whole of the discourse falls under the global communicative intention (i.e. it i s an act of \t*illfiil communication)

A . Reboul /Semantic Transparency

57

and the local informative intention for each utterance is determined at least in part by the global informative intention. This is also what accounts for coherence and here is how it does it. Coherence has been traditionally regarded in linguistic discourse analysis as the discourse equivalent of grammaticality in syntax: it was thought that just as grammaticality is defined in syntax as the property of sentences, coherence can be defined as the property of discourses. The threat of circularity is to be avoided in both cases through the fact that (syntactic or discursive) rules are made explicit and are exhaustively described. One can criticize syntax for not being exhaustive; however, and that is much worse, one can criticize linguistic discourse analysis for not being explicit at all, i.e. linguistic discourse analysis does not seem to have produced any set of rules. Given that, it seems that coherence is at best a pre-theoretical notion which cannot account for anything (and certainly not for the non-arbitrariness of discourse), but which itself is in need of an explanation. I think that the framework proposed above can do precisely that without in any way presupposing that coherence is a theoretical notion and without using it theoretically. In saying that coherence is a pre-theoretic notion, I mean that one should not try to give it a definition which gets beyond its vernacular use. There is no way to deny that people do have spontaneous judgment of coherence about discourses, just as they have spontaneous judgment of grammaticality about sentences. However, grammaticality can attain theoretical status through the production of an explicit set of rules aiming at exhaustivity (Generative Grammar comes to mind). As said above, it is doubtful that this is the case for coherence. Hence, coherence is still in need of an explanation. The explanation which can be offered in a relevance oriented pragmatics of discourse such as that proposed above is the following: a discourse will be judged as more or less coherent by its hearer depending on the ease with which the global informative intention can be recovered and depending on the richness and complexity of the information transmitted. That is, depending on the effort necessary for the interpretation of the discourse and on its effects, i.e depending on its overall relevance. Note that all discourses will not provoke the same expectations of coherence (i.e. of relevance): the coherence expected of a scientific text is presumably much higher than the coherence expected of a discussion between friends in a restaurant. But that does not detract for the ability of a relevance oriented pragmatics of discourse to both account for the non-arbitrariness of discourse and explain coherence judgements. Here is a short, authentic and unabridged example: (1 1) Dare I tell the story which was confided to me while taking the shade under the shadow of a graveyard wall in a delightfully green square of alfafa? Why not? I am already dishonored as saying truths which shock the fashion of 1838: The curate was not old; the maid was pretty; people were talking which did not discourage a young man from a near village to court the maid. One day, he hides the kitchen tongs in the maid’s bed. When he came back a week later, the maid said: “Tell me where you put the tongs which I have looked for everywhere since your departure. This is a very bad joke” The lover kissed her, with tears in his eyes, and went away. Stendhal, Voyage dans le midi, My translation. I will take it for granted that no one would find this text incoherent and that all judgments of coherence would be favorable. The question is why is it (judged to be) coherent? The first thing to note is that there are no connectives in this text and very few pronouns. In fact, almost the only expressions which could related to the judgment of

58

A. Rehoul/ Semantic Transparencv

coherence are temporal (e.g. a week later), though they are certainly not sufficient to account for it. I think that the best way to answer the question is in terms of the hypotheses which Stendhal sows in the reader's mind as to what he intends to tell (i.e. these hypotheses concern not so much the local informative intention of each successive utterance than the global informative intention of the whole text) and this begins with the first utterance of the text. He then exploits the reader's encyclopedic knowledge with the first utterances of the story itself, i.e. curates are male andmaids female and young people of both sexes can fall in love. The following utterances of the text mainly confirm these hypotheses. Why is the extension of Relevance Theory to discourse important for our present concern? As indicated in the introduction, I will be interested in what follows both in falsity masquerading as truth and in truth masquerading in falsity, i.e. in lies. It is a commonplace to say that the main difficulty involved in lying is not so much on the inventive side (i.e. in the content of the lie), but in keeping that content consistent with the rest of one's discourse, i.e. the main difficulty occurs not at the level of the utterance, but at the level of discourse. Before turning to lies of both kinds, I, however, want to outline the relations between pragmatics (specifically the interpretation of utterances) and metarepresentation. 2.3.4 Pragmatics and metarepresentation

Before turning to lies of all hues, I would like to outline the potentially complex relations between pragmatic interpretation and metarepresentation. First of all, let me say what I mean by metarepresentation,relying on the very sophisticated discussion given in Perner's book [21]. Perner squarely belongs to the representational tradition in cognitive sciences adopted here. He distinguishes three kinds of representations: 0 primary representations, representations of an object, i.e. a photograph of myself is a primary representation of myself; 0 secondary representations, voluntary misrepresentations of an object, i.e. representations in which a property is falsely attributed to an object. as. for instance, in hypothetical reasoning; 0 metarepresentations, "represent[ing] that something (another organism) is representing something" (p. 7), and how it is representing it. This may be clearer fiom figure 2.2.

A . Reboul /Semantic Transparency

59

Figure 2.2 The different sorts of representation

Let us forget secondary representations and misrepresentations for the moment and concentrate just on representation and metarepresentation. On a representational view of cognition, mental states are, or at least involve, representations. This is the case, for instance, for intentions. Thus the intentions which are recovered in utterance interpretation are representations. However, given that what is recovered is an attributed intention, the form of utterance interpretation will be a representation (on the hearer's part) of the intention of the speaker, hence a metarepresentation. This much is the common ground behind all pragmatic accounts, whether conventionalist or not, and whether contextualist or not. Can we go farther than that? On the Searlian conventionalist account, as said above (see 5 2.2.5), it does not seem to be the case. What, however, can be said of the Gricean and post-Gricean accounts? On a Gricean account, it seems obvious that metarepresentation has a role to play in the interpretative process itself, notably when conversational maxims are flouted. Then, the hearer must go through a complex and rather cumbersome reasoning process as to what the intentions of the speaker could be, etc. On a post-Gricean account, however, things are more simple: no such reasoning process is necessary though metarepresentation is the reason why the communicative Principle of Relevance comes into play. However, it does not necessarily play a role in the communication process, as long as everything goes well. The story, however, may become different when communication is not straightforward, i.e. in case of miscommunication, or in case of the detection of lies.

60

2.4

.3. Rehoull Semantic Transparenq.

Lies

As said above, I will be concerned here with two kinds of lies which I will designate. rather ornately, asfalsity masquerading as truth and as truth masquerading asfalsity. Both have a common aim: leading the hearer to a false belief or, more precisely, leading the hearer to believe something which the speaker believes to be false. I will have more to say about lies below, but before I turn to them, I would like to borrow a few reflections on metarepresentation and utterance interpretation proposed by Dan Sperber [22 J. 2.4.I

Three interpretation strategies

Not surprisingly, Sperber bases his paper on the assumption that language is semantically underdetermined. The consequence of this is that utterances are semantically opaque, that is, no amount of decoding is going to yield the interpretation of a given utterance. Hence. the linguistic decoding interpretation process has to be supplemented by a pragmatic inferential interpretation process. This implies a modicum of metarepresentation. However how much metarepresentation is involved in the pragmatic interpretation process is going to depend on the degree of sophistication of the interpretation strategy adopted by the hearer. On lines reminiscent of the three little pigs of the children tale, Sperber describes three strategies ranging from the less sophisticated to the more sophisticated. Interestingly. the less sophisticated here means not only the less complex as far as processes go, but also the more likely to lead the hearer in the clutches of a less than benevolent speaker, i.e. to lead the hearer to believe any lie which the speaker might intend him to believe. By contrast, the more sophisticated the strategy. the less risks there is that the hearer will be led to false beliefs. As Sperber points out, given Relevance Theory, there are two ways in which the communication process can flounder, i.e. not recover the speaker's intentions: the speaker may be incompetent, thus leading the interpretation process astray by not producing an utterance which can be interpreted so as to yield an interpretation consistent both with the communicative Principle of Relevance and with the speaker's informative intention; 0 the speaker may be less than benevolent, i.e. she can want to deceive the hearer. Given those two ways of producing miscommunication, the three stratepes are distinguished by whether they take them into account or not. The first strategy is naive optimism. In naive optimism, the hearer (let us call him John, following Sperber [22]) considers the speaker (let us call her Carol, still following Sperber) as both competent and benevolent. The only thing he has to do to interpret Carol's utterance is to follow the communicative Principle of Relevance and accept the first interpretation available to him which is consistent with it. The second strategy is cautious optimism. In cautious optimism. Carol is presumed to be benevolent, but not necessarily competent. In this strategy, John will again follow the principle of least effort, but he will not adopt the first interpretation consistent with the communicative Principle of Relevance. Rather, he will adapt to what he believes true of Carol, and, notably, what he believes that Carol believes about him. In other words, he will adopt not the first interpretation consistent with the communicative Principle of Relevance but the first interpretation that Carol might have believed to be relevant to him, John. This entails a good amount ofmetarepresentation on John's part. The third strategy is sophisticated understanding. Under sophisticated understanding, the speaker is believed to be neither competent nor benevolent. She is assumed to semi competent and benevolent.

A . Reboul /Semantic Transparency

61

The three strategies differ not only in the amount of metarepresentation involved, but also in the level of metarepresentation involved. Let us look at the following examples where different degrees of metarepresentation are involved: (1 2) ( 13) (14) ( 15)

(16)

Primary representation: It is raining. First-order metarepresentations: John says/believes/intends to communicate [that it is rainingJ. Second-order metarepresentation: John intends [me to believe [that it is rainind] . Third-order metarepresentation: John intends [me to believe [that he believedintends to communicate [that it is raining]]]. etc.

There are, in principle, no logical limits to the orders of metarepresentation, though cognitive limits do exist. The naive optimism strategy implies the capacity in the speaker and the hearer of forming first-order metarepresentations. The cautious optimism strategy implies a capacity to form second-order metarepresentations, while the sophisticated understanding strategy implies the capacity of forming at least third-order metarepresentations. It should also be noted that each strategy is more costly in terms of processing efforts than is the previous one. Thus, ceteris paribus, in most circumstances, we will tend to adopt the naive optimism strategy. It is only in specific cases, for instance when we know that the speaker is likely to be incompetent (if, for instance, it is a small child), or when the stakes are particularly important to us that we resort either to cautious optimism or to sophisticated understanding. This leaves a loophole in the communication process, which unscrupulous speakers can and frequently do exploit. 2.4.2 Lies and perlocutionary effects

Lies are one of the stumbling blocks of Searlian speech acts theory. The difficulties it raises for it mainly depend on the redefinition of meaning by Searle and on the place of the sincerity condition in the felicity rules for assertion. In order to see why this should be so, let me begin by a definition of a lie. It is generally thought that lying is saying something false. However, as was pointed out by Davidson [23], this is incorrect: "It is sometimes said that telling a lie entails what is false; but this is wrong. Telling a lie requires not that what you say be false but that you think it false" (p. 258). I will however not define lying through the nature of what is said: whether we take lies to entail saying what is false or saying what we believe to be false, this does not make lies different from most metaphors. Hence it is hardly a defining characteristic of lies. I will rather define a lie through a general characteristic of the speakerk intention. The speaker of a lie has the intention to produce a specific effect: that his hearer should believe something which he, the speaker, takes, rightly or wrongly, to be false. What is more, he intends to obtain this effect through his utterance(s). This, it should be clear, is not an illocutionary effect. It is aperlocutionary effect. As such, it does not tally with Searle's redefinition of meaning along (what Searle takes to be) Gricean lines. In that definition, Searle insist that the effect intended cannot be anything but illocutionary. I want to defend the idea that there are two ways of achieving the intended effect of a lie: you can say something which you believe to be false and get your hearer to think that you have said something which you believe to be true on good grounds; you can say something which you believe to be true and get your hearer to think that you have said

62

.4.Rehoull Semantic Transparencjq

something that you believe to be false on good grounds, leading him to adopt a contradictory belief (which you think false). There are thus two varieties of lies though the first one is the one generally considered. To avoid misunderstandings, I will reserve the word lie to the first variety, which I will characterize as falsity masquerading as truth. I will call the other variety, truth masquerading as falsity, truthfir1 deception. There is a third kind of deception, apart from lies and truthful deceptions: it occurs when you lead your hearer to derive a false conclusion. I will call it Machiavellian deception. Why, apart from the fact that deception in general aims at aperlocutionary rather than illocutionary effect, do lies make any difficulty for Searle's theory of speech acts? It seems impossible to describe lies in terms of speech acts and yet it seems rather weird to deny that lying is accomplishing a speech act. Let us suppose that lies are indeed a type of speech act. It seems clear, given that deception is involved, that lies cannot be illocutionary acts. This, however, does not mean that they cannot be another type of speech act and that they cannot have felicity conditions. Let us suppose that this is the case. Lies are perlocutionary acts and their felicity condition is that the intention of the Iiar should be satisfied, i.e. the hearer believes that the propositional content of the lie is true. Under speech acts theory, any utterance must correspond to the accomplishment of an illocutionary act. I will here adopt the simplifjmg assumption that all lies correspond to the performance of an illocutionary act of assertion. However, this cannot be the case, given that a defining condition of assertion is the sincerity condition, which is obviously not satisfied in lies. However, it does seem that if a lie is to be successful, the corresponding assertion also should be successful. But this cannot be the case if the sincerity condition is not satisfied. Thus, it does not seem to be the case that lies can be given a satisfying description in Searlian speech acts theory. It should be noted, however, that, on Sperber and Wilson's definition (see 4 2.3.3) of the act of sa,ving that, lies are just a variety of saying that. Indeed, saying that P is merely described "as communicating that the thought interpreted by P is entertained as a description of an actual state of affairs". which does not imply any kind of sincerity condition. 2.4.3 Deception on Iago 's mind

Iago is well known as the epitome example of a deceitful man. Othello will be my main illustration in the remaining part of this paper. Its scenario is well known, but I will briefly recapitulate it. Iago has been bypassed by Othello, a Moor employed by Venice in its fight against the Turks, who has chosen Cassio, of whom he is fond, as his officer, leaving to Iago the less prestigious role of ensign. Iago, to revenge himself, decides to make Othello believe that his wife, Desdemona, is Cassio's mistress. To this aim, he gives Othello a few hints and tops his work by a "proof", an handkerchief of Desdemona, given to her by Othello, which he steals and plants in Cassio's room. Othello, maddened by jealousy, kills Desdemona, understands his error and commits suicide on his wife's body. The turning point in the play occurs in the third scene of the third act. It begins whencassio has been disgraced through a plot of Iago and has asked Desdemona to intercede with Othello on his behalf. This is when Iago begins to lead Othello slowly but surely to the belief that his wife is unfaithfid. Scene 111, 3 begins with Othello secure in his marital bliss and ends with Othello sure of his cuckold status. The question is how does Iago manage to bring about this change? Iago's deceptions cover the three strategies outlined above: lies, truthful deception and Machiavellian deception. Let us, however, recognize that Otheilo is the ideal victim. only stirring out of naive optimism when Iago leads him to it.

A . Reboul /Semantic Transparency

63

Let us try and describe the whole process through the categories defined above. What then is Iago's global informative intention? His global informative intention is to make it manifest to Othello that Desdemona is Cassio's lover. He begins by subtly drawing Othello's attention to the fact that Cassio, who was speaking with Desdemona and asking her help, is leaving subrepticely. To do this, Iago mutters: "Ha! I like not that". His aim is to be heard, but not understood by Othello and to induce in Othello the belief that his utterance was an involuntary exclamation. His local informative intention is to make Othello believe that he, Iago, uttered an exclamation of dismay. Questioned by Othello on what he says, he reinforces Othello's initial interpretation by saying: "Nothing my lord. Or if, I know not what". In this exchange, Iago adopts the Machiavellian deception strategy. It is not the content of his utterances which Othello should believe: it is the inference that Othello draws from his interpretation of Iago's attitude which leads him to a dim kind of disquiet. The next episode where Iago again is active consists in Iago asking Othello whether Cassio was involved in Othello's wooing of Desdemona. This leads Othello to question him on his opinion of Cassio's honesty. This is where Iago resorts to truthful deception. His answers affirm Cassio's honesty, and he does believe Cassio to be honest. The whole exchange between Othello and Iago at this point deserves to be quoted in full: (17)

For Michal Cassio, I dare be sworn I think that he is honest. OTHELLO I think so too. IAGO Men should be what they seem, Or those that be not, would they might seem none. OTHELLO Certain, men should be what they seem. hGO why then, I think Cassio's an honest man. Nay, yet there's more in this. OTHELLO Iprithee speak to me as to thy thinkings, As thou dost ruminate, and give thy worst thoughts The worst of words. hG0

The situation in this exchange is the following. Iago knows Cassio to be honest. He says so. But he says it in such a way that Othello is led to the belief that Iago thinks Cassio to be dishonest. In terms of metarepresentation, this could be formulated as follows: ( 18)

Othello believes [that Iago thinks [that Cassio is dishonest]]

Iago's intention is the following: (19)

Iago intends [that Othello believes [that Iago thinks [that Cassio is dishonest]]]

His strategy can be articulated as: (20)

a. Iago intends [that Othello believes [that Iago says [that Cassio is honest]]] b. Iago intends [that Othello believes [that Iago does not believe [that Cassio is honestll] c. To achieve b., Iago intends that [Othello believes [that Iago lies [in saying that Cassio is honest]] ] d. To achieve c, Iago intend that [Othello believes [that Iago does not want [Othello to worry about Cassio's and Desdemona's relationship]]]

A. Reboul /Semantic Transparenet

64

This, in effect, is exactly what happens. Let us now turn to a lie. One occurs later in the scene, after Iago has adjured Othello not to fall into suspicion regarding Cassio and Desdemona. Othello assures him, falsely, that he has not. Iago then says: (21)

IAGo

Should you do so, my lord, My speech should fall into such vile success Which mt,thoughts aimed not.

Here, what Iago says is that Othello succumbing to jealousy was not his (lago's) intention. It is of course a blatant lie. What is Iago's intention? (22)

Iago intends [that Othello believes [that /ago did not intend [Othello to be jealous]]]

All of this should now be assessed relative to the notions of semantic transparency and semantic opacity. Linguistic underdetermination has a corollary: utterances are semantically transparent neither relative to their propositional content nor relative to the speaker's state of mind. This is why Machiavellian deception is possible: it relies on exactly the same inferential processes as utterances where no Machiavellian deception applies, as for instance, in the interpretation of example ( I 1). This is also why truthful deception is possible: utterances are not always taken to be sincere or truthful and they are not considered as semantically transparent relative to the speaker's state of mind. So far so good, but if this is the case, how are lies possible? Why does the sophisticated understanding strategy not apply across the board? To answer this question. I propose to quickly examine a well known philosophical puzzle, Moore's paradox. Moore's paradox

2.5

Moore's paradox was proposed by the Cambridge philosopher George Moore and can be enunciated in the following way:

(23)

It rains but I don't believe it rains.

The first thing to note is that, strictly speaking, Moore's paradox is not a contradiction does not contravene the law of excluded middle -. neither is it a paradox in the sense in which. for instance, the liar paradox is:

- it

(24)

I'm lving.

Thus the paradox in Moore's paradox resides in the fact that we consider such a utterance as (23) as weird in some mysterious sense, though it does not seem to have any of the anomalies listed in catalogues of semantic teratology. So why is Moore's paradox intuitively bizarre and what is its relationship with lies? The present section will be devoted to an answer to the first part of this question and this answer will allow us to deal with its second part.

A. Reboul /Semantic Transparency

2.5.1

65

A logical solution to Moore's paradox?

A possible solution to Moore's paradox would be to show that, despite appearances, it in fact is a contradiction. This seems to be the position adopted, notably, byvanderveken [6] and Levinson [24]. In other words, the solution would consist in considering the utterance in (23) as semantically equivalent with either (25) or (26):

(25) (26)

I believe it is raining and I don't believe it is raining. It is raining and it is not raining.

The first thing to note is that there is no problem in replacing but in (23) by and in (25)(26), given that but is generally considered as logically equivalent to and. However for (25) to be equivalent to (23), it should be considered that sayingIt's raining is logically equivalent to saying I believe it's raining. Similarly, for (26) to be equivalent to (23), it should be considered that saying I don't believe it is raining is logically equivalent to saying It is not raining. How likely is that? Saying that two linguistic expressions are semantically or logically equivalent is saying that they make the same contribution to the truth-conditions of any sentence where they might occur. In other words, in a given sentence, replacing one of two semantically equivalent expressions by the other will not change the truth-value of the whole sentence. When the utterance being considered is a compound of two sentences, each of which expressing a proposition, saying that one of the sentence in the compound is semantically equivalent with another sentence is just saying that replacing the sentence in the compound with the semantically equivalent sentence will not change the truth value of the compound. It is on this reasoning that advocates of the logical solution toMoore's paradox base their argumentation. As, for instance, It is raining is semantically equivalent with I believe that it is raining, (25) is semantically equivalent with (23) and thus both are necessarily false because they are contradictions. This is good reasoning in itself apart from the fact that it begs the main question which is that of the logical equivalence of either it is raining and I believe it is raining or I don't believe it is raining and It is not raining. What ground is there for thinking that it is so? Logical equivalence between sentences means quite simply that the two sentences have the same truth-values. One way of ensuring that is to make sure that they have the same truthconditions, i.e. that the truth of either of them implies the truth of the other. In other words, for It is raining and I believe it is raining to be equivalent, both (27) and (28) -where the sign is the logical sign for material implication -would have to be true: (27) (28)

It is raining a I believe it is raining I believe it is raining a It is raining

For (27) to be true, semantic transparency relative to the speakerk state of mind would have to be true and, as we have seen before, this is just not plausible. For (28) to be true, an even less likely condition should be met, i.e. that the speaker is infallible. Let us now turn to the second possible equivalence. Again, for I don't believe that it is raining to be semantically equivalent to It is not raining, both (29) and (30) should be true: (29) (30)

I don't believe that it is raining =jit is not raining It is not raining 3 I don't believe it is raining

.4. Reboul/ Semantic Transparenciq

66

(30) corresponds to an operation called Negation raising and is seen as legitimate. However, no such legitimacy is attached to (29), which would depend on two operations. The first one would be to consider that I don't believe that it is ruining is semantically equivalent to I believe that it is not raining. The second one would consist in an elimination of the preface I believe. None of these two operations are legitimate: I can not believe that it is raining without simultaneously believing that it is not raining. For instance, I may just not know whether it is or it is not raining. And, as pointed out before, unless I am infallible, my beliefs do not necessarily come true and, hence, it may well be the case both that I believe that P and not-P. This, incidentally, is why Moore's paradox is not a contradiction. Thus, it does not seem to be the case that there is a logical solution toMoore's paradox and neither can it be seen as contradiction. As a last argument against the logical solution, let me borrow (with modifications) two examples from Tsohatzidis [25]: (3 1) (32)

I don't believe that he feels sick but as he says he does and as his words would be accepted as the best evidence of his feeling sick, I assert that he feels sick. *I don't believe that he feels sick but as he says he does and as his words would be accepted as the best evidence o f his .feeling sick, I believe that he .feels sick.

Examples (3 1) and (32) are identical except for the replacement of1 assert in (3 1 ) by I believe in (32). (32) is definitely contradictory while (31) is perfectly acceptable. If semantic transparency relative to the speaker's state of mind was a sound hypothesis. assertion - as in (3 1) - and belief - as in (32) - would be equivalent. The comparison between (3 1) and (32) shows that this is not the case. 2.5.2 A pragmatic solution to Moore'sparadox

Thus, Moore's paradox is not a contradiction. What should be done, then, is explaining why (23) is, nonetheless, weird. The solution outlined below is a pragmatic solution, where pragmatic implies cognitive considerations. The first thing I would like to insist upon here is the indexicality of Moore's paradox. Let us compare (23) and (33) and (34): (23) (33) (34)

It is raining but I don't believe it is raining. Mary says it is raining, but I don't believe it is raining. It is raining in Caracas, hilt I don't believe it i s raining in Lvons.

The weirdness of (23) disappears in all of (33) and (34). In other words, it is essential to Moore's paradox that the person who says that it is raining should be the same as the person who says that he/she does not believe that it is raining, and it is just as essential that the location where it rains should be the same in the first and the second clauses of the sentence. This should. on the face of it. be mysterious: there is no location explicitly designated in (23). An answer to that problem has been suggested by American philosopher John Perry from example (35): (35 )

It is raining.

As Perry remarks, the truth-value of this utterance can only be evaluated relative to a precise place, though no place is mentioned in (35). He introduces the notion of

A . Reboul /Semantic Transparency

67

unarticulated constituent (see Perry [26]). An unarticulated constituent is a constituent of truth-conditions which is solved directly on the situation in which the communication takes place, but which is not mentioned explicitly in the utterance. Thus unarticulated constituents share with indexicals their deictic aspect. What is more, Perry introduced another notion, the notion that some indexicals are essential. Presumably, the best way of explaining Perry’s ideas on the matter is to quote his example ([27], p. 33): (36)

“I once followed a trail of sugar on a supermarket jloor, pushing my cart down the aisle on one side of a tall counter and back the aisle on the other, seeking the shopper with the torn sack to tell him he was making a mess. With each trip around the counter, the trail became thicker. But I seemed unable to catch up. Finally it dawned on me. I was the shopper I was trying to catch. I believed at the outset that the shopper with a torn sack was making a mess. And I was right. But I did not believe that I was making a mess. That seems to be something I came to believe. And when I came to believe that, I stopped following the trail around the counter and rearranged the torn sack in my cart. My change in beliefs seems to explain my change in behavior. ”

The notion of essential indexical applies to those indexicals which are indispensable to explain the behavior or the changes in behavior. In the above example, replace the indexical by a co-referring expression (e.g. John Perry) and you will loose the explanatory weight of I am the shopper with a torn sack. In this case, indexicality is essential in that it is impossible to eliminate it. Not all indexicals are essential however. Only indexicals which both are interpreted relative to the situation in which the communication takes place and which refer to components of that situation are essential in the above sense. This means that essential indexicals are I, here, now and the unarticulated constituents. Unarticulated constituents intervene to fix the location where it is raining though the speaker does not believe it in (23). However, they are not indispensable to the weirdness of Moore’s paradox: identity of location, however it is ensured, is. Let us extend the notion of unarticulated constituent out of truth-conditions and include in it a non-vericonditional preface any utterance: in other words, let us suppose that each utterance is accompanied by the unarticulated constituent . Thus, (35) - which is identical with the first part of (23) - should be understood (non-vericonditionally)as:

(37)

it rains

This allows the following (not semantically equivalent to (35)) deduction:

(38)

( it rains)

I say that it rains

=j

Let us now resort to the intentional stance (see 52.2.3). According to it, the interpretation of behavior relies on the attribution to other people of states of mind, through the hypothesis that people are rational. Suppose that we combine the intentional stance with a naive optimism strategy regarding utterance interpretation. Then from the fact represented in (38), we will deduce that indeed the speaker of (35) believes that it is raining. However, when we apply the intentional stance and the naive optimism strategy to Moore’s paradox, we find ourselves faced by a contradiction between the state of mind we attribute to the speaker from the interpretation of the first sentence of the compound and the state of mind he describes himself as being in the second sentence of the compound. There is a contradiction, but that contradiction is not between the two sentences of the

A. Rehoul I' Semantic Transparencr.

68

compound. It is between a deduction which the first sentence seems to allow and the propositional content of the second sentence. That, in itself would not be sufficient to explain the weirdness because, after all, we only have to change our strategy and abandon naive optimism in favor of sophisticated understanding. This would leave us with a difficulty, however, in that the intentional stance would still apply but its second requisite would not be complied with. Indeed, it seems that there is no way to attribute to the speaker of (23) one or several states of mind consistent with the rule that the speaker is considered as a rational individual. Let me explain shortly why this is so. Let us suppose that the speaker of (23) is sincere: thus, when he says It is raining he believes that it is raining; similarly when he says I don? believe it is raining he does not believe that it is raining; however, it is not possible to believe both that it is raining and that it is not raining of the same place; hence, the speaker of (23) cannot be rational. Let us now suppose that the speaker of (23) is not sincere, i.e. he is lying: when he says It is raining, he intends his hearer to believe that he (the speaker) believes that it is raining; similarly, when he says I don't believe that it is raining he wants to make his hearer believe that he (the speaker) does not believe that it is raining; it is impossible to have and to satisfy simultaneously two contradictory intentions: hence. the speaker of (23) is not rational. In other words, Moore's paradox draws its weirdness from the fact that it is impossible to attribute to its hypothetical speaker a state of mind consistent with the rationality requirement of the intentional stance. 2.5.3 Moore's paradox and lies We are free to adopt any interpretative strategy we want, as long as it yields results consistent with the intentional stance. As we saw before, though the most reliable strategy presumably is sophisticated understanding, it also is, unsurprisingly, the most costly strategy. This is why we generally adopt naive optimism unless we have good reasons to believe that the speaker is neither competent nor benevolent. This also is why we tend to believe the lies we are told, as long as the liar is competent enough not to get himself into obvious incoherence. This is also why we tend to believe that it is raining when in fact it is not and why there is a tendency to believe that Moore's paradox is a contradiction. However. it is not.

2.6

Conclusion

Semantic transparency is just a myth. The code model of linguistic communication is far from satisfying. However, there are models that can do justice to the intrinsic complexity of human communication. Relevance Theory is such a model as I hope to have shown above. What is more, a plausible model of communication should cover all the avatars of communication, from perfect functioning to gross imperfection. This, I think, is just what the code model and the hypothesis of semantic transparency cannot do. Relevance Theory can and, hence, it is. now, the most promising model for communication.

2.7 [I]

123

References C E Shannon and W Weaver, The Mathematical Theon of Communication Urbana Unnersih of Illinois Press, 1949 i 4uqtin. HOMto do ~hings\+ i r h itordq Oxford Clarendon. 1%:

A. Reboul /Semantic Transparency

69

P. Grice, Studies in the way of words. Cambridge, MA: Harvard University Press, 1989. D. Sperber and D. Wilson, Relevance: Communication and Cognition. Oxford: Basil Blackwell, 1996 (2ndedition). D. Dennett, The intentional stance. Cambridge, MA: The MIT Press, 1987. D. Vanderveken, Les actes de discours [Discourse acts]. Bruxelles: Mardaga, 1988. P. Grice, Meaning, Philosophical Review 66 (1957) 377-388. J.R. Searle, Speech Acts. Cambridge: Cambridge University Press, 1969. A. Reboul,: The logical status of fictional discourse: what Searle's speaker can't say to his hearer, in: Speech Acts, Meaning and Intentions: critical Approaches to the philosophy of John R. Searle, A. Burkhardt, Ed. Ber1in:de Gruyter, 1990, pp. 336-363. A.Rebou1, The description of lies in speech acts theory, inpretending to communicate, H. Parret, Ed. Berlin: de Gruyter, 1993, pp.292-298. A. Reboul, La fiction et le mensonge: les "parasites " dans la theorie des actes de langage [Pretence and lie: The "parasites" within the speech act theory], in L'assertion en dkbat. La description du monde dans la conversation [Assertion in discourse: World description in conversation], Ch. Brassac, Ed. Psychologie de 1'Interaction 5-6, (1997) 87-125. A. Reboul, Le paradoxe du mensonge dans la theorie des actes de langage [The paradox of lie in the speech act theory], Cahiers de linguistique franqaise 13 (1992) 125-147. A.Rebou1, Represented speech and thought and auctorial irony: ambiguity andmetarepresentation in literature, in: Hommages h Ronald Landheer, J. Rooryck, P. Smith and P. Bogaarts, Eds. Amsterdam: Rodopi, to appear. F. Recanati, Contextualism and anticontextualism in the philosophy of language, in Foundations of Speech Act Theory: Philosophical and Linguistic Perspectives, S. Tsohatzidis, Ed. London: Routledge, 1994, pp. 156-166. J. Fodor, The modularity of mind. An essay onfacultypsychology. Cambridge: MIT Press, 1983. A. Reboul and J. Moeschler, Le dialogue n'est pas une categorie naturelle scientifiquement pertinente [Dialogue is not a natural category scientifically relevant], Cahiers de linguistiquefranqaise 17 (1995) 229-248. A. Reboul and J. Moeschler, Faut-il continuer a faire de l'analyse de discours?[Do we still need discourse analysis?], Hermes, Journal of Linguistics 16 (1996) 61-92. A. Reboul and J. Moeschler, Reduction and contextualization in pragmatics and discourse analysis, in Pragmatik, B. Eckard Ed. Linguistische Berichte, Sonderheft 8 (1 997) 283-295. A. Reboul and J. Moeschler, La pragmatique aujourd 'hui. Une nouvelle science de la communication [Pragmatics nowadays: A new communication approach].Paris: Le Seuil, 1998. A. Reboul and J. Moeschler, Pragmatique du discours. De I 'interprktation de I 'enonck a 1 'interprktation du discours [Discourse pragmatic: utterance interpretation and discourse interpretation]. Paris: Armand Colin, 1998. J. Perner, Understanding the representational mind. Cambridge, MA: The MIT Press, 1991. D. Sperber, Understanding verbal understanding, in What is intelligence?, J. Khalfa, Ed. Cambridge: Cambridge University Press, 1994, pp. 179-198. D. Davidson, Inquiries into Truth and Interpretation. Oxford: Oxford University Press, 1984. S. C . Levinson, Presumptive Meanings: The Theory of' Generalized Conversational Implicatures. Cambridge, MA: The MIT Press, 2000. S. L. Tsohatzidis, Speaker meaning, sentence meaning and metaphor, in Foundations of speech act theory. Philosophical and linguistic perspectives, S. L. Tsohatzidis, Ed. London: Routledge, 1994, pp. 365-373, [261 J. Perry, Indexicals, Contexts and Unarticulated Constituents, in Proceedings of the 1995 CSLZAmsterdam Logic, Language and Computation Conference. Stanford: CSLI Publications, 1998. z271 J. Perry, The problem of the essential indexical, in The problem of the essential indexical and other essays. Oxford: Oxford University Press, 1993, pp. 32-52. [31 [41

This page intentionally left blank

SECTION I1 PRETENDING TO COMMUNICATE: DECEPTION, SEDUCTION AND EQUIVOCATION The ability to deceive involves hiding information known to oneself from another person. Successful deception requires taking into account the knowledge available to another person and thus indicates appreciation of mental states and the ability to predict the behaviors of others on the basis of their mental states.

Yirmiya, Solomonica-Levi and Shulrnan, I996

This page intentionally left blank

Say Not to Say L. Anolli et al. (Eds.) IOS Press, 2002

3

73

Deceptive Miscommunication Theory (DeMiT): A New Model for the Analysis of Deceptive Communication Luigi ANOLLI, Michela BALCONI, Rita CICERI

Abstract: Deceptive communication has been recently studied by many scholars, both in naturalistic studies in the field and in experimental research programs carried out in the laboratory, but this scientific domain remains devoid of a viable theory. These theoretical perspectives risk creating a sort of “mythology” about deception as a separate communicative domain, characterized by distinct and specific verbal and nonverbal features. The present chapter intends to overcome this deception mythology and offer a viable model which can explicate the main characteristics of deceptive communication and the local management of the deceptive message in its different expressions. The framework in which we move is the miscommunication as a chance theory (MaCHT) proposed by Anolli. According to the theory herein followed, deception is a kind of miscommunication and a chance in communication terms, since deceptive miscommunication greatly enhances the degrees of freedom available to the communicator. Within this perspective we mean to propose a new model, called the Deceptive Miscommunication Theory (DeMiT). The main theoretical points of this model will be sketched out in the present chapter.

Contents

3.8 3.9

Introduction............................................................ The “family” of deceptive acts.. ................................... Intentional framework in deceptive miscommunication. ....... The cognitive design of deceptive miscommunication.......... Deception as communicative interact. ............................. Modes of deceptive saying.. ........................................ Machiavellian attitude and self-deception as devices for a skilled deceptive miscommunication.............................. Conclusions. ........................................................... Acknowledgements...................................................

92 95 96

3.10

References.. ...........................................................

96

3.1 3.2 3.3 3.4 3.5 3.6 3.7

74 76 77 80 85 87

74

3.1

L . Anolli et al. /Deceptive Miscommunication Theon. (DeMiT)

Introduction

Deceptive communication has been the object of much study and research in these last 30 years, but, as McCornack [ l ] clearly points out, this scientific domain remains devoid of viable theory. Two main lines of research have been developed: naturalistic studies in the field to analyze spontaneous lying behavior and experimental research program carried out in the laboratory. In this second line, among the various theoretical models formulated by scholars, we can mention the Information Manipulation Theory (IMT)proposed by Lapinski [2], McCornack [3], McCornack, Levine, Solowczuk, Torres, and Campbell [4]; the Interpersonal Deception Theory (IDT)advanced by Buller and Burgoon [S], and Burgoon, Buller, Guerrero, Afifi, and Feldman [6]; the social model of lying presented by Bradac, Friedman, and Giles [7]; studies about the nonverbal aspects of deceptive speech act suggested by Buller, Burgoon, White, and Ebesu [8], and Ekman [9]. Differing and often conflicting between each other, these models have highlighted some distinct aspects of deceptive communication. According to IMT, which reverses Grice’s maxims (Quality, Quantity, Relevance, Manner) regulating conversational exchanges, the deceptive message derives from speakers “transforming” relevant information, but it fails to explicate this purported transformation process and, as Jacobs, Dawson, and Brashers [ 101 pointed out, it confuses three distinct issues: a) deceptive message production, b) deceptive message features. c) interlocutor interpretation of such a message. In its turn, IDT is grounded on a large set of formal statements concerning deception variables, such as suspicion, behavioral leakage, communicative strateges, relational intimacy and the like. Deceivers strategically control the information in their deceptive messages with the purpose of presenting themselves behaviorally as expressing truthfblness and veracity, since they are engaged “in greater strategic activity to manage information, behavior, and image [than truth-tellers]” [5, Proposition 31. However, non strategic unconscious leakage cues (like arousal and nervousness, negative affect and incompetent communication performance) can reveal deception [8, 61. According to DePaulo, Ansfield, and Bell [ 1 11, despite extensive research and numerous theoretical assumptions, IDT also fails to explain the production mechanisms responsible for deceptive message encoding, and the cognitive processes implicated in deceptive message interpretation. Firstly, we have to question the ecological validity of these current theories and ask ourselves how much their findings resemble naturalistic processes in everyday deceptive communication. Secondly, both IMT and IDT risk building a sort of “mythology” about deception as a separate communicative domain, characterized by distinct and specific verbal and non verbal features. According to McCornack [ 11, these “hopeful myths” are: the encoding of deceptive messages entails active, strategic, and detailed cognitive processing, the encoding of deceptive messages requires greater cognitive load than the encoding of truthful messages, the encoding of deceptive messages is more ph.vsiologicallr7 arousing than that the encoding of truthful messages, there is an identifiable and consistent set of deception-arousal-basedbehavioral cues that deceivers “leak” when encoding deceptive messages, human beings are innate!i* capable of deception detection. and deceptive messages halye specifiable characteristics that render them distinct from t n t thfii 1 messages.

L. Anolli et al. /Deceptive Miscommunication Theory (DeMiT)

75

The aim of the present chapter is to overcome this deception mythology and to offer a viable model which can explicate the main characteristics of deceptive communication and the local management of the deceptive message in its different expressions. The theoretical domain in which we move is the miscommunication as a chance theory (MaCHT), as sketched out by Anolli (Chapter 1, this volume). According to the perspective herein followed, deception is a kind of miscommunication and a chance in communication terms, since deceptive miscommunication greatly enhances the degrees of freedom at the communicator’s disposal. It represents another route to express the speaker’s sensations, thoughts, beliefs, emotions, and desires; likewise, at communication level, it may be really advantageous to have the chance of hiding, omitting, concealing or, simply, blurring information. After all, truth is not a matter of black or white, and deceptive communication may be an opportunity both in a Machiavellian, opportunistic sense and in an everyday relational situation. Within this perspective we mean to propose a new model, called the Deceptive Miscommunication Theory (DeMiT). The main theoretical points are the following: a) a deceptive miscommunication theory should be included in a general framework capable of explaining the default communication; therefore, deceptive miscommunication is neither an alternative to truthful communication, nor an exception nor yet a violation regarding a standard paradigm of default communication; b) deceptive miscommunication is not a homogeneous but a heterogeneous communication field, with different kinds of deception and deceptive message features; c) deceptive miscommunication is managed by an intentional stance characterized by an internal gradation; these degrees of intention serve to arrange and monitor a deceptive message design suitable in a given situation and a contingent relational web; d) the deceptive message follows the same mechanisms and processes of mental planning and execution as the default communication message; e) deceptive miscommunication is context-bound, then requests a “local management” of conversational exchanges, because every context is constructed in situ; here a distinction may emerge between naive (“bad”) deceivers and skilled (“good”) ones; f) deceptive miscommunication, like default communication, uses different kinds of linguistic and nonverbal expressions, although they may be qualified by distinctive communication patterns in certain circumstances; g) a useful explicative key to understand deception mechanisms is given by the link between deception and self-deception. The aim of the present chapter is to sketch the DeMiT, starting with the deception “family” and the “resemblance of family” in deceptive phenomena, and following with the intentional framework of deceptive miscommunication in order to point out the intentional gradation involved in it. In this section we will also see some motives implicated in deception and how children manage to understand deceptive actions. In the next session we will take into consideration the cognitive design of deceptive miscommunication, examining the “cognitive load hypothesis”, making the distinction between the cognitive processes implied in low-content and high-content deceptive acts, and searching a contextual management conception of the deceptive act. After looking at deceptive miscommunication as a relational game and communicative inter-act (section 5 ) , we will take into consideration the different modes of expressing deception, examining both IMT and IDT, as well as examining closely the linguistic styles of deceptive miscommunication. The Machiavellian attitude and self-deception seen as devices for skilled deceptive miscommunication are the topics of the last section. In this chapter we will only allude to the emotional aspects of deception.

76

3.2

L . .4nolli et al. /Deceptive Miscommunication Theon?(DeMiTl

The “family” of deceptive acts

Deception as an articulated and complex miscommunication act is an emblematic pattern of adaptive behavior in interpersonal relational management, with the aim of influencing others’ beliefs, as it is outlined by Bond, Omar, Pitre, Lashley, Skaggs, and Kirk [ 121, and by Buller and Burgoon [13]. In this way the deceptive act, as it occurs within everyday conversations, is both universal and casual. In the often quoted naturalistic study by Turner, Edgley, and Olmstead [ 141 61.5% of utterances reported by the 130 participants were, somehow, deceptive, even if less than one sixth were outright lies. Also DePaulo, Kashy, Kirkendol, Wyer, and Epstein [ 151 found in a naturalistic study that college students during a week’s observation told lies “in approximately one out of every three of their social interactions, and people from the community lied in one out of every five social interactions” (p. 16). These lies were not very serious and did not imply a large cognitive effort in planning and telling; subjects also reported that they expected to be believed and did not worry too much if they would be caught out by the deceived. In the everyday context the majority of deceptive messages involve subtle and complex combinations of truthful and deceptive information [3] and are told in an unplanned and simple manner ~51. In contrast to naturalistic studies, laboratory assessments tackled the prepared deception. Especially in the lie domain, strategic moves and countermoves by deceivers and deceived were discovered in experimental studies, and a self-conscious deliberative process has needed to lie [5]. The lie is considered as an intentional and conscious act to deceive another person who is unaware and unwilling. To this end, the need to arrange and execute a complex communicative plan is linked to the complexity of an intentional plan: that is, the speaker’s intention to make the interlocutor deliberately believe what heishe knows not to be true. Instead of considering the contraposition between field and laboratory research as exclusively a theoretical one, in the DeMiT, in line with MaCHT perspective (Anolli, Chapter 1, this volume), we prefer to deal with deception as a ‘:fami/r*”of miscommunication phenomena and processes. In this “family” we can consider selfdeception and pathological frequent deceptiveness; prepared lies (mainly managed to avoid sanctions) and unprepared lies (to face up to an embarrassing situation), pedagogic lies (for example, told to a child for reassurance) and white lies (aimed at managing a threateningface situation). For this reason truth and falsehood are not always separated by a dichotomous line and in many situations they are mixed together. In particular, crossing the broad distinction of “having the interlocutor believe the false“ and “not having the interlocutor believe the true”, we obtain four deceptive “subfamilies”: a) omission (the speaker omits to give the addressee some information that he/she thinks or knows is relevant to the addressee’s goals); b) concealment (the speaker withholds and hides some information by giving the addressee some other divergent/diversionary information that is true but not relevant. in order to perpetuate false assumptions in him/her); c) .ful.sification (the speaker deliberately conveys to the addressee some information that he/she believes false); d) masking (the speaker withholds some information by giving the addressee some other false information). We can continue this analysis of the deception ”family” and see we can lie telling the tnrth. as in the case of half-truth (the speaker only refers part of the truth or the truth is minimized) and exaggeration (the speaker provides more information than the truth called for). Moreover, truthful messages can be used to deceive by causing deceptive implicatures on the grounds of false presuppositions. as i t is outlined by Jacobs. Dawson. and Brashers [lo]. and by Castelfranchi and Vincent [I61 I n general. the different kinds of deceptke

L . Anolli et al. /Deceptive Miscommunication Theory (DeMiT)

77

forms here mentioned are covered by the main distinction proposed by Chisholm and Feehan [171: deception by commission and deception by omission. It is worth pointing out that deceptive “subfamilies” and other related phenomena do not constitute “deception types” in the DeMiT. Speaking of “deception type” means accepting a priori taxonomy, in which the deception categories would be discrete and characterized by clear, definite and inflexible boundaries. Instead, according to DeMiT, there is a continuum among the different kinds of deceptive “subfamilies”: as Wittgenstein [ 181 says, there is “a resemblance of family”, in which the boundaries are fuzzy and vague, even though this “resemblance” does not deny the presence of great differences in the deception system. Now we must ask which, if any, could be common denominators in this range of deceptive forms. We think that DeMiT could be useful in this task.

3.3

Intentional framework in deceptive miscommunication

According to the prototypical perspective proposed by Coleman and Kay [19], the lie is defined by three basic features: a) the falsehood of the content of the utterance; b) the awareness of such falsehood; and c) the intention to deceive the addressee. But we realized that falsehood is not always necessary to produce a deceptive message, as you can tell a lie telling the true. Moreover, you can tell a falsehood by lying, or by making a mistake. He/she who mistakenly makes a false statement does so unknowingly (thinking of he/she is saying the truth) and successively learns the truth; instead, he/she deliberately tells a lie, first knows the truth and then tells the lie. As a consequence, the intention to lie is the essential feature of deceptive communication, as it is shown by Sweetser [20], and Anolli, Balconi, and Ciceri [21]. Anolli and Balconi [22] suggested that, whether or not a deceptive act is successful (the addressee is actually deceived) does not alter the fact that a deceptive act taken place; in the same way, the fact that the addressee does not believe what the deceiver has said, does not modify the fact that the speaker has deceived. According to Grice’s [23] speech acts theory, the performative effect does not alter the illocutory force of the deceptive speech act. Speaking of intention is a delicate subject. As Anolli (Chapter 1, this volume) pointed out, we have to consider different kinds of intention: referential, informative, and communicative, as well as the intentional gradation. Intentional stance is a matter of degree in the communicative process, since intentions are directly proportional in their strength to the informative content of the message, as it is outlined by Jaczczolt [24, p. 681. In deceptive miscommunication, according to the DeMiT, we theoretically have different layers of intention: a) covert (hidden) intention (the speaker intends to deceive the addressee by manipulating the information, but this intention must not be revealed); b) overt (ostensive) intention (the speaker intends to convey the manipulation of information to the addressee). This second intentional layer is, in its turn, twofold: bl) informative intention (the speaker wants to give the addressee the manipulated information as if it were true); b2) “sincerity intention (the speaker wants the addressee to believe that what he has said is true, in order to respect the Sincerity Rule of Searle [25]: “I want you believe that I believe what I am saying to YOU”). Therefore, deceptive communication appears to require at least a second-order intentional system and in certain cases (especially in prepared lies) a third-order intentional system. In this second-order intentional layer we have to further distinguish between the deception famiZy and the joke family (such as teasing, irony, pretence, parody, sarcasm, banter, etc.). In the former the speaker intends to deceive the addressee, whereas in the ”

78

L. Anolli et al. [Deceptive Miscommunication Theon (DeMiT)

latter the speaker has the intention not to deceive but to be disbelieved. The speaker’s intention to deceive is therefore defined in terms of what the speaker wants the addressee to think. The deceitful speaker wants the listener to think that the statement is true, while the joking speaker wants the listener to know that the statement is false (see figure 3.1 ). Declarative statement

Truth value is not relevant

Politeness, greetings, etc.

Truth value is relevant

I Jokes, teasing. Pretence. wit, etc. parody, etc.

True

Knowledge is not relevant

Knowledge is relevant

(the communicator

(the communicator has

lacks knowledge)

good knowledyet

False

True Telliny the truth

False Telling a Ire

not justified

justified

genuine mistake

careless mistake unprepared lie

prepared lie

Figure 3.1. Route of miscommunicativeacts of deception

But the second-order intentional layer is a common and ubiquitous feature in human communication, grounded on what Dennett [26] defines as the intentional stance and theory of mind. In everyday conversations people normally attribute beliefs, motives, desires, and intentions to the mind of others, whatever they are saying or doing. They have a mental representation of the others’ mental representations. that is, they have metarepresentation. This mindreading implies intersubjectivity (like the sharing of subjective states and common reference), perception of intentionality (the idea of others as animate, self-directed, and goal-oriented), and attribution of mental states to others. As Whiten [27, p. 631 pointed out, also mindreading is graded, and this gradation corresponds ta the depth of penetration that the mindreader makes into the mind of the other. starting from epistemic seeing to arriving at false belief ascription. According to Perner [28]. in

L. Anolli et al. /Deceptive Miscommunication Theory (DeMiT)

79

this last condition children achieve a truly “representational theory of mind” around four years of age. Intentional stance and metarepresentation are inherently recursive processes, tied to belief-embedding: a representation (or intention) can refer to another representation (or intention) that, in turn,refers to another, and so forth ad infinitum. According to the DeMiT, deceptive miscommunication is not the psychological counterpart of truthful communication, and is not a separate communicative type needing a distinct message-production mechanism. Intentional gradation, requested by any communicative act, can also manage different kinds of deceptive message, from white lies to prepared and bold-faced ones. The framework of deception as a miscommunicative act implies an increase of the degree of freedom on the part of the speaker to choose a definite path of message design according to context expectancies and cultural standards. Moreover, intentional system and gradation operating in deceptive miscommunication produce a sort of intentional opacity in the speaker: except in bold-faced lies, he can hide hisher deceptive act as a result of lack of information, a mistake, poor judgement, a moment of bewilderment, etc. In this way he/she can “save hisher face”. Likewise, intentional stance can also be useful in understanding both deception motives and deception comprehension in children. Concerning the former, the deceptive message, like any communicative or miscommunicative act, has the intent to influence another person in line with the adaptive perspective of human behavior. Here we are not interested in building a taxonomy of motives for deception, as recently proposed by O’Hair and Cody [29]; what we are interested in is to point out the combination of intentional gradation and motivational range. The deceptive act is not gratuitous, suggested by the Evil One or caused by madness; it is the outcome of an evaluation of the contingent situation in order to optimize the possibilities and reach a convenient and preferable solution in terms of costs and benefits. In any case, the deceptive message is generated by rational people, affected, however, not by an Olympian but by a bounded rationality, as it is called by Simon [30]. As the study of heuristics and judgement under uncertainty proposed by Kahneman, Slovic, and Tversky [31] has extensively shown, they can only reach the so-called “local best”, that is, the solution that maximizes opportunities and minimizes risks at a certain time. In this perspective one can resort to a deceptive message in order to acquire or protect one’s own resources (money, time, possessions, privacy, etc.), or to manage the relationship with a partner (initiate, continue, or avoid interaction; avoid conflict and embarrassment, avoid punishment, leave take, avoid self-disclosure, etc.), or else to maintain or enhance self-esteem (save one’s face, improve one’s social position, increase one’s social desirability, etc.). These motives are included in the benign fabrication of deceptive acts, so called by Goffman [32]. On the contrary, there are other motives defined by Goffman as exploitative fabrication. Among them, we can mention the deceptive message with the purpose of gaining at the expense of other people, and manipulating or harming them. Revenge, vindictiveness, retaliation, sabotage, and hatred can serve as examples of exploitation and malevolent deception. These and other indefinite numbers of motives underlying deception confirm the flexibility and variability of deceptive communication as context-bound and dependent on the contingent relational web. In a naturalistic study DePaulo and Kashy [33] found that individuals would tell fewer lies per social interaction to the people to whom they felt close, and they would feel more uncomfortable when they lied to those people. Because altruistic lies can communicate caring, DePaulo and Kashy [33] found also that relatively more of the lies told to friends would be altruistic rather than self-serving, whereas the reverse would be true of lies told to acquaintances and strangers.

80

L. h o l l i et al. 1 Deceptive Miscommunication Theov (DeMiT)

Moreover, intentional stance and belief representation appear useful in explaining the development of the comprehension of deceptive acts in children. Most scholars agree with Sodian’s [341 perspective, that four-year-old children are capable of distinguishing lies from jokes, since at this age they already possess the concept of belief and can understand deception as manipulation of others’ behavior by influencing their beliefs a b u t reality. Early on they can distinguish deception from teasing in terms of the speaker’s intention to influence the addressee’s behavior. However, important precursors of a theory of mind are found earlier than four years of age. As Cadinu and Kiesner [35] pointed out, three-year-old children, engaged in pretence and deception situations, demonstrated a significant understanding of other people’s mental states. Likewise, Chandler and Hala [36], and Hala and Chandler [37] found that actively involving three-year-old children in planning deception facilitated performance on false-belief questions. In fact, when children simply watched the execution of deceptive planning (where to relocate an object), they did not differentiate whether the object was moved for deceptive or practical reasons. In contrast, children who had strategically planned a deception were markedly better at answering questions about another’s false beliefs than those who simply witnessed the object transfer. Moreover, it is worth noting that, according to Leekam [38], children of four years understand desires and intentions before they understand beliefs. With regard to ability in active deception, in a recent naturalistic study it was observed that three-year-old children (and perhaps even younger ones) resorted to everyday deceptions in a varied, flexible, and context appropriate manner, too complex to be “blind” learned strategies. Children’s deceptive skills are more likely to develop fkom pragmatic need and situational exigencies than from conceptual development; as Newton, Reddy, and Bull [39] pointed out, they may learn to lie in the same way as they learn to speak. This learning continues till adolescence. To sum up, intentional gradation and metarepresentation (theory of mind) are basic features which explain not only standard communication, but also deceptive miscommunication both in everyday conversations (unprepared lies) and in significant situations (prepared lies). Higher order intention, belief state, and message forms are strictly intertwined between them to form the center of deceptive miscommunication.

3.4

The cognitive design of deceptive miscommunication

3.4.I

The “cognitiveload h-vpothesis

I’

Buller, Burgoon, White, and Ebesu [8], DePaulo, Ansfield, and Bell [ 1 I], and Vrij, Semin, and Bull [40] have pointed out that the deceptive message is more cognitively difficult and demanding than the truthhl message, since the fabrication of a plausible and convincing lie that is consistent with the contingent situation and everything the addressee knows is a laborious cognitive task. Referring to this topic, McCornack [I ] advances the cognitive load hypothesis, of which he criticizes two basic assumptions: a) according to Zuckerman and Driver [41], deceivers “must construct a message from scratch, and the content of the message must be both internally consistent and compatible with what the listeners already know”; b) deceptive communication is cognitively more difficult because of the number of constraints placed upon the formulation of the lie, as Greene, O’Hair, Cody, and Yen [42] sketched out. In his criticisms McCornack [ 11 is absolutely right, showing the theoretical fragility of this dichotomous distinction between deceptive and t r u t h h l communication. Saying that

L. Anolli et al. /Deceptive Miscommunicatiori Theory (DeMiT)

81

deceptive messages are constructed “from scratch” means that truthful messages are not built from scratch and so they are the result of a preformed message design; but this statement is untenable from the point of view of any contemporary model of message production, as it is shown by O’Keefe & Lambert [43]. In the same way, assuming that deceptive messages are internally consistent presupposes that truthful messages could lack internal consistency. This statement is also untenable. Likewise, stressing the high number of constraints in deceptive communication means a failure to recognize the mental convenience and opportunity to resort to deception in certain complicated situations, in which truth appears much more cognitively laborious than a lie. However, on a more advanced level, the problem of the cognitive demand of deceptive miscommunication remains in its entirety. According to DeMiT, we can assume that: a) the cognitive load of the deceiver arises as a function of the entity and gravity of the deceptive contents; b) the cognitive load of the deceiver depends on the context significance. To develop these assumptions, first of all we have to distinguish between prepared and unprepared deception. The former is cognitively planned in advance and examined by the deceiver at least in its main aspects; the latter is spontaneously said, often as the answer to an unexpected question by the interlocutor, without any mental planning. This distinction is not dichotomous; rather, there are different degrees of planning and preparation in telling a lie. Secondly, we have to introduce a further distinction between high- and low-content deception. The former concerns a serious topic, is said in an important context, and is characterized by the presence of notable consequences and effects for the deceiver or for the addressee or even for other people. The latter regards a minor topic, can be said in any kind of context, and does not have consequences or, if any, they are of scarce importance. As for the previous distinction, high- and low-content deception is a continuum, and not a dichotomous matter. We have to investigate this subject further. 3.4.2 High- and low-content deceptive acts

In the DeMiT we hypothesize that low-content deceptive acts can be said almost automatically, without any anticipatory planning or careful, conscious thought, especially in everyday conversation and between intimates. In this domain white and pedagogic lies are included, as well as different kinds of concealment, omission, and evasive message. Oflen they are unprepared and are justified by the deceiver in the name of cultural standards and the addressee’s (or others’) interests (altruistic lies). In these situations the deceptive intents are not really to be deceitful, but they are benign and benevolent in order to promote or maintain the relational web. Concerning this aspect, Lindskold and Walters [44] carried out a research on college students’ perception of the ethical evaluation of deception and found a hierarchy of “acceptability” for deceptive acts in a balancing of the costshenefits of the deceiver with the costshenefits for the deceived target. Moreover, low-content deceptive messages are likely to take place in contexts characterized by nonface-threatening (or low face-threatening) condition. The deceiver can feel at hisher ease in these contexts and does not need particular cognitive demands in generating this kind of deceptive message. It is something taken for granted. Instead, another type of deception consists of high-content deceptive acts, since they have serious effects and consequences for both the deceiver and the deceived. The deceiver runs the risk of “losing face” and being treated as dishonest and untrustworthy, if hidher deceptive act is uncovered or detected. The detection of a lie intrinsically implies a threat to the deceiver’s image and an attack to hidher self-esteem, generating in himher negative self-conscious emotions such as guilt, shame, and embarrassment. He can be openly

82

L .4nolli et al

Dec'eptisr Miscommunication Theon, /DeMiT)

accused by the addressee, who can have feelings of aggression and retaliation, bewilderment, and lingering suspicion. The deceived, in hisiher turn, runs the risk of being harmed or hurt in hisher interests to deceiver's advantage. Because high-content deceptive acts bear high costs both for deceiver and deceived. there must be good reasons for producing them. Generally, they are likely to happen in complicated relational situations and in conflicting or face-threatening contexts, in which a problem of disclosure arises, with a dilemma like this: is it better tell the truth and the salient information, recognize the risks connected with sincere and honest disclosure and choose a way in which the truth can be presented in a acceptable manner. or tell a deceptive message, recognize the penalties associated with insincere and dishonest disclosure and prepare information suited for contextual constraints? Or choose a middle course'? In any case. the speaker has to face up to high cognitive demands in both alternatives, since, telling the truth or telling a lie, he has to fabricate a message with the lowest risk and penalty. For this basic reason, most people resort to a kind of "in-between" message, which involves subtle and complex packaging of both false and tmthfbl elements [3]. True and false are meshed each other, so that it becomes impossible to tell them apart. Focusing our attention on cognitive demands for high-content deceptive acts. we can see that telling a high-content lie requires considerable cognitive complexity, since manipulating information necessitates careful mental planning. This cognitive complexity lies in the discrepancy between private knowledge and public statement: according to Anolli and Ciceri [4S] the deceiver k n o w the truth (which heishe does not tell). but publicly tells a lie (which he/she does not believe. but has to make the hearer think that helshe does believe it). How can this cognitive demand for a high-content deceptive message appear'? As we have already seen in section 3.2, there are no fixed and universal signals of deceptive communication. It would be so easy to detect all the liars in the world. with no need to resort to the lie-detector! Each speaker has learned hislher own communicative style in telling lies, as helshe has learned other communicative styles of persuading people. defending himiherself. bargaining, seducing, etc. If one looks for these signals as compelling evidence of the cognitive load in deceptive miscommunication. one is on the wrong track. Instead, there are different and flexible communicative modes for deception; they can also be the opposite. Take, for instance, hand movements: some people show a decrease in hand movement during deception due to the cognitive demand. while others show an increase, as it is shown by Ekman, O'Sullivan, and Matsumoto [46], and by Vrij. Akehurst. and Morris [47]. It is not a matter of a universal and reflexive-like signal. but it depends on various factors. Among others, individuals with high public self-consciousness (i.e. the ability to become aware of another's perspective and to act from that perspective) and high self-control make fewer hand movements during deception compared to truth-telling [47]. Likewise, in high-content deception there is a rise in the number of shorter and more recurrent pauses than in low-content deception or truth [4S]. This phenomenon has been already observed in purely linguistic tasks in which frequent interruptions in the utterance by filled or unfilled pauses are the sign of a complex linguistic task, as it is outlined by Goldman-Eisler [48], and Rochester [49]. In the same way, studies by Berger and Jordan [50] on the association between planning difficulties and verbal fluency highlight a significant correlation between the number of pauses and the level of difficulty in generating a speech that will fulfill one's aim. The packaging of personal cognitive processes and the systems of signaling combined with hi_gh-contentdeceptive acts are learned during one's ou.n learning contexts and arc.

L. Anolli et al. /Deceptive Miscommunication Theory (DeMiT)

83

the result of one’s own experience. It is a matter of micro-culture (or idioculture), given by the primary (i.e. family) and secondary (i.e. school, peer group) social systems of socialization. One can learn to stare at the addressee while deceiving; another can learn to avoid the mutual gaze. One is prompt in answering an unexpected question; another has to prepare hisher speech taking care over details. It is likely that within a high-content deceptive miscommunication the tense relationship with the partner could generate an emotional arousal, produced by the fear of being detected and of “losing face”. According to Tagney and Fischer [Sl], this arousal especially concerns negative self-conscious emotions, like guilt, shame and embarrassment. Shame and embarrassment are “toxic” emotions, aroused by the risk of being disapproved or condemned by the addressee, as well as by being inadequate by cultural standards, whereas guilt, due to consciousness of the transgression of moral and conventional norms, can lead, in some cases, to the spontaneous confession of one’s own deception and the wish to make amends. All in all, the segmentation of the deception “family” in subgroups has enabled us to shed some light on cognitive demand. According to DeMiT, as for truthful speech, deception requires an extensive range of cognitive effort and the speaker has to choose the more suited communicative path as the occasion demands. 3.4.3 Deceptive message design

For the analysis of deceptive message design as a communicative act we can start from the Q Principle and the R Principle which stress the dialectic opposition between “speaker’s economy” (use as few words for concepts as necessary) and “auditor’s economy” (use as many expressions as messages to make the communication clear). Horn [52] defines them in the following manner: The Q Principle: Make your contribution sufficient. Say as much as you can (given R). The R Principle: Make your contribution necessary. Say no more than you must (given Q). Between these two principles, which replace Grice’s maxims, the communicator has to plan hisher message, which could be truthful or deceptive. This communicative space is not to be intended as the transmission of a message from a sender to a receiver governed by discrete state communication systems (that is, sender, receiver, message), but it is to be intended as a co-regulated communication in which each partner may be influenced by the other, as Fogel [53] sketched out. Along this theoretical line communication is a continuous process not ritualized or over-controlled by either one or the other partner but created by the dynamics of mutual action, so that the results are the emergent qualities of the relationship. They are neither foreseeable nor deducible from the previous moves of each partner. Therefore, the information is not ‘in’ the situation nor ‘in’ the communicator, but it is created when he/she engages in an active discourse with the partner. From this point of view we do not see message generation (including the deceptive one) as the result of a planning system in a holistic and functional way. In classical planning systems we have three alternatives: a) a top-down (or “prescriptive”) planning system which invests most of the resources in plan construction and involves selection from functionally indexed high-level forms; b) a bottom-up (or “reactive”) planning system which pays more attention to execution and selects the concrete linguistic utterance from fimctionally indexed low-level forms; c) an interleaved (or “limited-commitment”)

84

L Anolli et al

1 Deceptive Miscommunication

Theon. (DeMiT!

planning system which passes control back and forth between the plan construction and the plan execution. But, as O’Keefe and Lambert [43] pointed out. each of these systems: “depends on there being a decontextualized relationship between .form and &function. But, to put it bluntly, deconte-xtualized linguistic.forms have no .functionalsignificance. It is a truism to sa-v that the meaning of a .form depends on the context of its use. At the level qf’

discourse acts or message-features,the evidence shows that the formTfunction relationship is mediated b?. reasoning from context-specificbeliefs“ (p. 57). As a consequence, messages as communicative acts are not planned and executed according to an abstract and universal rule set, but according to the contextual conditions. In fact, “message structure and function are not holistic, but rather reflect the grounding of messages in an ongoing stream of thought and action” [43, p. 591. This theoretical line has brought a new perspective on message generation, governed by a local management [431. According to DeMit, considering communication situations as organized fields of thoughts and intentions, and messages as the outcome of thought and intention selective expression, message structure arises as the focus moves through the field of thoughts. Focus is an active process of concentrating attention. shared with the partner, on a certain element of reality (event, object, etc.); it involves not only attention to cognitive aspects. but also a perspective on those aspects. Focus is driven by intentions and guided by the route that the communicator formulates to move through the field of thoughts. The resulting message, rather than being a functionally packaged and unified act, is a collation of thought, each of which can have distinctive effects. The diversity of messages in communication situations depends on the fact that the communicators can have different intentions and different routes combined with those intentions, as it is suggested by Anolli (Chapter 1, this volume). This diversity in focusing leads to a diversity in the thoughts selected for expression and differences in the message forms. This model of situated and context-bound message planning and production can help us in the understanding of deceptive message design. Most low-content deceptive acts may be produced in the flow of communicative exchanges with the partner by the activation of relevant thoughts. They are unforeseen and unprepared, and are the result of the selection of a route suitable to bring the speaker to a “local best” in that situation, or. at least, to an acceptable end state. They may also be parsimonious in communicative terms (enabling the speaker to choose the most convenient and unproblematic message in that situation). and they may be advantageous in relational terms (allowing the speaker to avoid painful or harmful disclosure and to maintain the relationship). On the contrary, high-content deceptive acts may request previous planning. since they are generally foreseen and prepared. In this case the deceiver has to elaborate hisher best communicative way to convince the partner. First of all. heishe has to be careful in the deceptive message planning, paying attention to its internal consistency and compatibility with the partner’s knowledge, as if it might be true, or, at least, likely. Secondly. helshe has to be as spontaneous as possible in the deceptive message execution in order to be believable. Therefore, in the high-content deceptive message execution the speaker has to manage as well as possible the interaction with the addressee, monitoring hisher speech and non verbal systems of signaling, but in this task heishe may fail. In fact, in the local management of deceptive utterance production the speaker can either exceed in hisher self-control or show a lack of control. In the first case. in the over-control condition deceivers produce, for instance. an average deeper tone of voice. without variations. whereas in the lack of control condition they have a higher tone associated \vith man?

L, Anolli et al. /Deceptive Miscommunication Theor?, (DeMiT)

85

vocal variations. Compared to these naive (“bad”) deceivers, the skilled (“good”) ones show the same vocal intonation and profile both in truthful and deceptive message production [45]. According to Riggio, Tucker, and Throckmorton [54]. It could be that naive deceivers are people who are characterized by a high self-monitoring level, and are then able to adapt themselves to the interaction process with the partner. Summing up, in the DeMiT we have to underline three aspects of the cognitive processes implied in deceptive miscommunication. First, deceptive and truthful communication have the same cognitive mechanisms both for planning, production and execution of the messages, consistently with what has been said concerning communication and miscommunication, as Anolli sketched out (Chapter 1, this volume). Second, the cognitive demand for deceptive communication is not a homogeneous and constant factor, but there is an increasing of cognitive request in function of the deception subgroup. For low-content deceptive acts the default cognitive processes seem to be sufficient, since they can be planned, produced and managed in a straightforward manner, almost automatically. In fact, their risk and possible penalty are rather low. Instead, for high-content deceptive acts a notable cognitive effort is needed to package a consistent message and to be credible, possibly in a spontaneous manner. Third, the cognitive demand of deceptive message production is associated with the necessary activity of monitoring and controlling one’s own performance. In this activity naive deceivers may exhibit either an excess or a lack of self-control, producing leakage cues (even if minimal), whereas skilled deceivers show an efficient and effective performance maintaining a flexible self-control.

3.5

Deception as communicative interact

Some studies on deception have followed a unidirectional approach to deceptive communication, considering that deceivers convey signals which receivers absorb passively: as Berger and Bradac [55], and Surra and Ridley [56] outlined, deceiver and deceived seem to have little agency when they come to communicate. But, as we have seen in the previous section, deceptive miscommunication, like default (truthful) communication, is not a unidirectional transmission of message from the deceiver to the deceived, but an interactive and interpersonal act by nature. It is a dynamic and coregulated activity, since the interaction between deceiver and deceived, on the one hand, is a complex system, which is continuously modified by the actions of both the partners as they occur in a given situation. According to Anolli (Chapter 1, this volume). Such an interaction is governed by the communicative synchrony principles. On the other hand, it is a social process by which individuals locally alter their actions according to the ongoing and anticipated actions of the partner. In deceptive miscommunication, if one hides, it naturally means there is someone to hide from. Hence, the deceptive act, like all communicative acts, is a relational game between two individuals: on one side one finds the speaker playing the role of the more or less able deceiver; on the other side the other individual in the role of the victim or detector [45].In this psychological game there is not one active participant (i.e. the deceiver) and another passive one (i.e. the deceived), given that deceptive miscommunication, like default communication, is not like dancing where one partner leads and the other follows. Deceptive (and truthful) assumptions are not simply created and recovered but rather created between the two in a dynamic manner [24, p. 821. As a consequence, in deceptive interaction the success of the lie depends on the balance between the communicative ability of the deceiver and the acquiescent or shallow attitude

86

L ..inolli et nl

Deceptive Miscommunication Theon, fDeMiT)

of the deceived; alternatively, its.failur-e is determined by the reverse (the inability of the deceiver to deceive and the ability of the deceived to detect himher). In the relational game of deception a significant position is taken by the addressee. He can assume an acquiescent attitude toward what is said (truthfiil or deceptive) by the speaker. In this situation the deceptive message is successful: but it is likely to be a matter of low-content deceptive acts. On the contrary, facing high-content deceptive messages the addressee may have a suspicious and unmasking attitude. inquiring and questioning. In this perspective Stiff, Corman, Knzek and Snider [57] have verified that deceivers modified their response latencies in function of the type of questions asked by the interviewer: when the questions were direct and pointed, the deceivers lengthened their response latency more than when the questions were simple, especially in the middle segment of the conversation. This process has been interpreted as an “impulse decay”, that is, deceivers take more time to reflect, arrange and monitor their messages in a suitable manner. compatible with the conditions of the contingent contextual interaction. Likewise, many studies on the probing efect have pointed out the significance of the addressee’s attitude towards the deceiver’s behavior. Probing consists of the direct questioning of a message source regarding the veracity of the information presented, and requesting additional information to be given. According to Buller, Comstock, Aune, and Strzyzewski [58]. Buller, Strzyzewski, and Comstock [59], and Stiff and Miller [60]. In general, probing produces an increase of honesty and credibility attribution to the message source (i.e. the potential deceiver), as probed sources are rated more “honest” than non probed sources. The Behavioral Adaptation Explanation (BAE) has been considered as a plausible account of the probing effect. According to BAE. the perception of suspicion by the message source (the deceiver) makes himher more careful, so that hekhe can modify hidher behavior to appear more believable: as a consequence, the addressee becomes more convinced of the source’s honesty. In other words. probing causes the addressee to attribute truthfulness to the speaker. In fact. the perception of suspicion appears to be correlated with four main cues in the probe deceiver’s behavior: more speech errors, more speech hesitations. longer response latencies. and more eye contact. But. as Levine and McComack [61] point out. these behavioral changes in probed deceivers, instead of being evidence for BAE validity, show that they display cues indicative of deceptiveness when the sources are probed. The subsequent dispute between Buller, Stiff and Burgoon [62] and Levine and McCornack [63] has pointed out that behavioral adaptation is certainly required in deceptive interaction, but that it does not account for the probing effect. Moreover, when the addressee is suspicious and inquiring, the naive deceiver is likely to increase hidher inadequate emotional responses and to become either ”over-emotional“ or “over-controlled”. In the former case he/she can get nervous, with more blinking, more self- object manipulation, frequent postural shifts and general body activity. In the latter case he/she shows an overall exaggerated control over his/her behavior. with postural rigidity. lack of spontaneity and overdone performance. The effects of the “deception-detection game’’ (as a sequence of interactions distributed along a certain time period) can arrive at a fie-bias or a truth-bias, as it is shown by Levine and McCornack [64], and by McCornack and Parks [65]. In the former the tendency to view one’s relational partner as deceptive becomes prevailing in many instances. since he/she is willing to distort the truth without much hesitation. In a few cases, this relational pattern could become aberrant, with frequent explicit actions of deceptiveness. In the latter the partners develop a truthful relationship since in most instances they are telling the truth. in this honest relationship relational involvement. confidence, truth-bias. and accuracy form a causal chain. Obviouslv. a relationship can change from tnith bias. to suspicion. to

L. Anolli et al. /Deceptive Miscommunication Theory (DeMiT)

87

lie bias rather quickly when a partner has been detected telling lies. Vice versa, the opposite path - from lie-bias to truth-bias - is much more complicated and difficult, because, once an individual has been labeled as a “liar”, he/she must prove hisher honesty and believability over and over again before that he/she can be believed, and in certain circumstances he/she is in no state to prove the truth of what he/she says. In this way he/she falls into the liar paradox: a liar is not believed, even when he/she tells the truth. Summing up, according to DeMiT, high-content deceptive miscommunication involves an interactive process in which deceiver and deceived have an apparent co-operative position but, beneath the surface, they have opposite and competing agendas. In this communicative field each partner may use various degrees of freedom to simulate or detect a certain state of affairs. In any case, deceptive interaction, like default communication, is a matter of negotiation between the partners in a recursive manner, because utterances can be interpreted in many ways, and are not always planned in advance, as Anolli sketched out (Chapter 1, this volume). In communicating, the speaker (for instance, the deceiver) does not convey to the addressee what is objective for himher: it becomes objective only through communication, by changing the state of the speaker’s knowledge. In this way he/she can not presume what interpretation the receiver will assign to his/her utterance. Interpretation is product between the partners, and it is not a plain assumption recovery, since in the communicative act interpretation is incremental and context is constructed in a contingent and local manner.

3.6

Modes of deceptive saying

3.6.I

Basicjhdings

In more than 30 years of study about deception and lyng most of the research and the theoretical effort of scholars has been devoted to analyzing communicative processes and performances underlying deceptive messages, even if there is not a global and viable theory on deceptive communication at the present, as we already said in section 1 of this chapter. However, it is worth noting that great advances in the right direction have been made in this field, in the attempt to understand the distinctiveness of deceptive communication processes and also, more recently, the commonalties with default communication. McCornack [3], taking his distance from the idea of deceptive communication types as something already preformed, has suggested that individuals manipulate information simultaneously along different dimensions when they deceive others, the result being a potentially indefinite range of specific deceptive message forms. Information Manipulation Theory (MT) is a framework for describing the different ways that information can be manipulated to accomplish deceit [4]. According to this model, the observable variations in deceptive message design reflect a continuum of covert to overt misrepresentation of information, in which relevant information serves as one anchor and explicit falsification of truthful information as the other, as it is shown by Metts [66]. A significant conceptual problem concerns the assumption of the continuum between covert/overt misrepresentation. Verbal deception, by its nature, is covert; linguistic acts that openly misrepresent information are processed and perceived as intentional “flouts” of conversational principles (like sarcasm, parody, pretence and the like) rather than deception [23]. IMT, following Grice’s Co-operation principle and the four maxims (Quality, Quantity, Relevance, and Manner), views deception as arising from covert violations of one or more

88

L . .4nolli et 01. i Deceptive .Miscommrrnication Theon. IDeMiTi

of these maxims governing conversational exchanges. According to IMT, individuals, in everyday conversations. monitor the information that they disclose along four different primary dimensions: amount, veracin,, relevance, and clarity. These dimensions form the platform for inferential processing underlying conversational sense-making. Deceptive messages are “deceptive” in that, although they deviate from the principles of conversational understanding, these departures remain veiled. In particular, research proposed by Ekman [67] and Metts [66] has showed that deception can be fabricated by manipulating the amount of information conveyed, the veracity, the relevance. as well as. according to Bavelas, Black, Chovil, and Mullet [68], McCornack, Levine. Solowczuk. Torres, and Campbell [4], and Yeung, Levine. and Nishiyama 1691. the clarity of information. Jacobs, Dawson and Brashers [ 10, p. 73) claim that IMT conceptualization “does not center on the contribution of the deceiver’s message to the hearer’s substantive beliefs. Rather, it centers simply on failures to detect [gricean] maxim violations”. Moreover, IMT is inconsistent with Grice’s theory of conversational maxims. and the idea that messages are deceptive only when they violate the Quality maxim. On the other side, scholars like Buller and Burgoon [ 131, Burgoon and Buller [70]. Burgoon, Buller, Guerrero, Afifi, and Feldman [6], Burgoon, Buller, Dillman, and Walther [71], Burgoon, Buller, and Guerrero [72], Burgoon, Buller. Floyd, and Grandpre [73]. and Burgoon, Buller, White, Afifi. and Aileen-Buslig [741 have developed the Interpersonal Deception Theov (IDT), in which a great deal of study has been devoted to the analysis of the communicative patterns of deception. IDT follows a strategic approach to deception comprehension. Buller, Strtyzewski. and Hunsaker [75] say: “The [deceptive] conversation is characterized b?* a series of’ moves and countermoves. the skill of.both communicators and deception ’s ultimate success or failure is afiected to see through the multiple laJ?ersof meaning and react in way.c.Mqhich fiwther their goals” (p. 28).

Within this perspective Buller and Burgoon [ 131 distinguished between straregic and non-strategiccommunication in deception. Strategc (or intentional) patterns to deceive are those manipulated by deceivers with the goal of presenting themselves in a truthful and believable manner. Since deceivers know that deception can produce detection cues. they will attempt a strategy to give an honest impression through communicative signals thought to be indicative of veracity. According to IDT, the main strategic communicative patterns in deception are the following: a) uncertainoy (or vagueness) giving rise to ambiguity by sending intentionally mixed messages through irrelevant information, fewer references to self-experiences, more frequent hand gestures, fewer absolute verbs. etc.: b) reticence and withdrawal (or non immediacy) signals used in creating a distance from the partner, or withdrawing oneself from direct interaction by means of shorter responses. verbal non immediacy. longer response latencies, less mutual gaze. less forward lean. greater proxemic distance, etc.; c) disassociation signals used to divert responsibility for a previous message or an attempt to remove oneself from the deceptive act by means of fewer self-references, fewer self-interest statements. more other references. verbal non immediacy, etc.; d) image and relationship protection signals designed to present oneself in a favorable light by nodding. smiling. refraining from interruption. etc. and thus hoping to avoid deception scrutiny The non-strategic communication in deception is formed by unintentional leakage cues. and deceivers are unaware that they are revealing deception. or are unable to control them in preventing deception detection. According IDT. there are three categories of

L. Anolli et al. /Deceptive Miscommunication Theory (DeMiT)

89

unintentional non-strategic leakage cues: a) revealing arousal and nervousness with more blinking, a higher pitch of voice, vocal nervousness, more speech errors, longer response latencies, less gesturing, more leg and foot movements; b) revealing negative affect by means of less pleasant facial expressions, more frequent negative feedback, reduced eye contact, less pleasant vocal profile, and more negative statements; c) incompetent communication performance with more speech errors, hesitation, word repetitions, rigidity, brief utterances, channel discrepancies, dissynchrony, exaggerated performance, and lack of spontaneous behavior. Therefore, IDT sees the deceiver’s communication as a balance between intentional and unintentional (involuntary) signals: if the former prevails, the deception succeeds, whereas, if the latter takes priority, failure is likely. As a consequence, IDT seems to conceive the deceptive exchange as a battle field, in which the deceiver aims to make strategic communicative signals in an intentional and consistent way to convince the addressee, who, however, aims to detect the deceiver by scrutinizing hidher behavior and communicative style. 3.6.2 Linguistic styles of deceptive miscommunication According to van Dijk [76], we can define linguistic style as an articulated dynamic pattern of micro- and macro-components of language, intentionally adopted by the speaker to produce specific context effects on hidher addressee, as it is proposed by Sperber and Wilson [771. In deceptive conversation, like in default conversation, the connection between text and context encourages the production of specific linguistic styles during the flow of exchanges. These styles are to be thought of as a kaleidoscopic processes, calibrated in relation to the necessities of the speaker-addressee system, as Sandig and Selting [78] sketched out. By combining different linguistic elements, it is possible to represent a consistent and systematic textual production, diachronically organized with regard to the dynamics of communicative exchanges. The diachronic character of stylistic profiles is caused by the need to introduce opportune changes in the linguistic plan during interaction, adapting one’s own message to the context variations, moving the narrative focus as the occasion demands. The diachronic model of the verbal style, proposed by Sinclair [791, foresees continuous narrative changes, called “postural changes”, which include the ongoing definition of the conversational positions and the relational boundaries between the partners in order to adapt the local intent to a contingent communicative exchange. In this stream of conversational sequences the dynamic of postural changes involves the selection of verbal styles, defined both by textual and trans-textual (contextual) properties. Within this perspective, DeMiT assumes that in deceptive miscommunication deceivers may resort to a considerable variety of linguistic styles, which may be very different one from another and, in certain circumstances, also contrastive. In particular, according to Anolli and Balconi [80], three main linguistic styles can be recognized in deceptive conversation, as shown in figure 3.2. First, deceivers can have a linguistic style characterized by ambiguousness and prolixity, especially when facing an acquiescent and silent addressee. Within this linguistic profile they use many modifiers in the semantic value of words (with a prevalence of dubitative forms), a high number of “levelers” terms (like “all”, “nobody”, etc.), complex and long utterances, a scant number of factual utterances, high variation in the choice of voice, a wide use of irrelevant information. In particular, in telling lies, deceivers refer less contextual and sensory information in favor of subjective experience, as well as accounts of imagined memories (equivalent to the deception condition) which differ from accounts based on perceived memories (equivalent

90

L .4nolli et al. Decepti1.e Miscommirnication Theon. IDeMiT)

to the truth condition), as it is suggested by Alonso-Quecuty [8l]. Aspects of this vague. noncommittal and unverifiable linguistic style have also been found in many previous studies, as in Buller and Burgoon [ S ] , Carpenter [82], Dulaney [83], Knapp, Hart, and Dennis [84], Kuiken [85], Miller & Stiff [86], and Todd-Mancillas and Kibler [87]. By means of this speech style deceivers try to “dilute” the falsehood and make it less identifiable. They use semantically rich and syntactically well-constructed utterances. which, at the same time, are neither clear nor relevant. In this way they appear to “neutralize” and “narcotize” the falsehood, without disconcerting the addressee. It is a matter of “gray shades”, as Baumeister [88] says. In everyday life, no event can be described simply as being true or false, black or white; rather, it has to be represented by different gradations of gray, nearer or farther from one of these two extremes. With ambiguousness and vagueness deceivers are able to lie without running the risk of exposing themselves excessively and of being openly believed wrong if they should be detected, as Buller and Burgoon [5], and McCornack. Levine, Morrison. and Lapinski [89] sketched out. This line of representing reality, however, is very common, because ubiquitous cultural artifacts take on a regular mediation function between the subject and the environment. In particular, secondary (or symbolic) artifacts provide people with cognitive devices by means of which they can adapt and arrange a sequence of events, as it is outlined by Cole [90]. According to Geertz 191, p. 51, in this perspective the human being may be thought of as “an animal suspended in webs of significance he himself has spun”. Second, deceivers may resort to a linguistic style that is characterized by concise nssertiveness and elliptic avoidment, especially when the addressee appears suspicious and inquiring. Within this linguistic style they adopt a reticent attitude, prefemng to say the bare minimum; the messages are frequently short, without a subject and with an implicit predicate; the utterances are brief, with a reduced number of words, many long pauses. as well as long response latencies. Elements of this concise, reticent and elusive speech style have also been found in previous studies, as in b a p p and Comadena [92], Kuiken [85]. Stiff. Corman, Krizek. and Snider [57]. Stiff and Miller [60], Todd-Mancillas and Kibler [87], and Vrij [93]. By means of the reduction of information and the great simplification of speech. deceivers choose to shirk saying, since they can decide not to tell their own lie and to define its boundaries as little as possible. In this way their assertiveness can appear as a linguistic device intended to a guarantee the truthfblness of the contents and corroborate the reliability of their speech. Third. deceivers can resort to a linguistic style that is characterized by depersonalization. Their speech includes a small number of self-references, frequent recourse to third parties, frequent factual utterances. great use of impersonal conditions (the impersonal “one” and the plural “we”). This linguistic phenomenon was also found in previous research [ 13, 67, 661. By using the depersonalization style deceivers aim to avoid taking responsibility for their own statements and to dissociate themselves fiom their messages. shifting the focus to aspects of the external context. The variety of speech styles suitable for deceptive miscommunication allows the speaker to adopt and to follow the most effective linguistic route in view of the addressee‘s attitude. likewise the speech styles in default communication. It worth noting that deceivers use words to “transform” reality, so as to make the falsehood appear reliable and believable. Deception asks for a continuous and iocal check of the esectiveness of message design: one who deceives has to be sensitive to the interlocutor‘s changes of attitude. It is a question of tuning not only to avoid loss of reliability. but also to appear honeqt and trust).

<

L. Anolli et al. /Deceptive Miscommunication Theory (DeMiT)

I

(A): Say not to say

)

( # c ~ - - -1 @

Argumentativity Index (+)

(,

/

0 -

0

construction (+)

of arguments (+) NO Noofwords(+) ofpauses (+) NO

No of semantic modifiers (+) No of "levelers" Eloquency index (+) Complex utterance (+)

pronouns (+)

- 1 -

91

--

I \

\

\

\

Noof variations of the

\

\ / I 0

\

I

\

standard phrase (+) Noof dubitative semantic modifiers (+) No of softening modifications (+) No of softening equivalents (+)

/

PROLIXITY

0

0

- d @

AMBIGUOUSNESS

(B): Elude saying e@-----I

Personal pronouns (+) No of dubitative semantic modifiers (-) No of emphasizing semantic

I

of arguments (-) Nominal construction (+) NO

Disfluency index (+) Content substitutions (+)

\

I I / 0 0

0

I

0

CONCISE ASSERTIVITY

-I

\

*r

\ Third parties (+) Impersonal condition (+) Factual utterance (+)

/

/

\

of softening modifications (++) NO

\

\

No of semantic modifiers (+)

I I / 0 0 d

DEPERSONALIZATION

AMBIGUITY / FUZZINESS

Figure 3.2. The main linguistic styles in deceptive miscommunication

/

I

0

92

L . .3noili et al. 1 Deceptive Miscommunication Theon.lDeMiT)

Machiavellian attitude and self-deception as devices for a skilled deceptive 3.7 miscommunication In this section we will tackle the question of some psychological features involved in a skilled deceptive miscommunication. Until now the routes of study followed in this field have been mainly two: the structural route concerning the personality of the skilled deceiver (so called “Machiavellian personality” or Machiavellianism), and the functional route regarding the self-deception mechanism as a condition to tell the falsehood. 3.7.I

The Machiavellian personalih construct and d4ceptit.e miscommtrnication

Riggio and Friedman [94], Riggio, Tucker, and Throckmorton [54], and Zuckerman. DePaulo, and Rosenthal [95] have considered that extroverted and dominant people are generally more successful than introverted at controlling leakage cues. In particular. individuals with high self-monitoring ability and less anxiety are more able to increase facial animation and head movements, verbal fluency, eye contact and use of personal pronouns. These people are good actors, appearing as skilled communicators, untroubled by scrutiny. As a consequence, some people may look more “honest” or “dishonest” both in their behavior and their physical features, regardless if they are telling the truth or a lie. This ”demeanor bias” shows how low the competence of human beings is in deception detection and how much depends on unreliable factors, as it is shown by Ekman [9]. Frank and Ekman [96], and Hess and Kleck [97], even though professionals with special interest or skill in deception detection (like officers, judges, clinical psychologists, etc.) are fairly accurate in judging people who are lying or telling the truth about their opinions, as Ekman, O’Sullivan, and Frank [98] sketched out. The “Machiavellian personality” construct finds its justification in this theoretical platform. It can be described as an attitude used to manipulate and exploit others, with an adaptive function. In fact, according to Wilson, Near, and Miller [99] Machiavellian behavior can be regarded as a “defect” strategy, which possesses advantages (i.e. gains from exploiting others), as well as disadvantages (i.e. retaliation and avoidance by others). The Machiavellian worldview is essentially defined by three features: a) it involves using manipulative strategies such as deceit and flattery in interpersonal relations; b) it implies a cynical perception of others as weak and untrustworthy; c) it includes a basic indifference toward conventional rules of morality in thought and action. as it is suggested by Fehr. Samson, and Paulhus [ 1001. According to this construct, Ashton, Lee, and Son [ l o l l proposed that high Mach individuals would be more skilled in deceptive miscommunication, by increasing eye contact and fabricating plausible verbal lies. Moreover, Machiavellian people are characterized by a cool detachment. which makes them less emotionally involved with others and less worried about saving face in embarrassing situations. High Mach, more than low Mach individuals, believe that others are under their influence, and are ready to plot and proceed to the manipulation of these people. According to Shepperd and Socherman [ 1021, they customarily act in a domineering style and prefer a show of strength in front of others, avoiding the use of weakness tactics like sandbagging (a manipulative strategy in which people f e i p incompetence and display low ability to induce an opponent to reduce effort or lower hidher guard). Summing up, the structural route sees competent deceptive miscommunication as the outcome of a personality protile. Machiavellian people resort to deception in a skilful manner as a means. among many others. to obtain their goal of potver and dominance.

L. Anolli et al. /Deceptive Miscommunication Theory (DeMiT)

93

since, as Machiavelli said, “the end justifies the means” 3.7.2

What better deceiver than a sincere deceiver?

In the analysis of skilled deceptive miscommunication, instead of following a structural route which considers deception ability as a personality trait, it is possible to proceed to a functional one and see self-deception as a means to tell other-deception in a believable manner. If non- true becomes true for myself, why should it not become true for others? Self-deception especially concerns complex situations such as interpersonal games, affective experiences, social competitive connivance, relational negotiation, and so on. It is not an easy matter, because self-deception has been and is a field of contrasting perspectives. How indeed can falsehood become true for oneself? If one follows the classical definition of deception (A deceives B using p utterance if (i) A knows that p is false and non-p is true, (ii) A induces B to believe that p ) , then self-deception is submitted to a double paradox: a) static paradox, given that an agent deceives hindherself only if he/she believes that p and that non-p at the same time; b) dynamic paradox, given that the selfdeceiver must have the intent to fabricate a false belief in him-herself who is aware of the truth. To overcome these paradoxes, different explicative models have been proposed. First, according to the standard model, Davidson [ 1031 has proposed a self-deception view grounded on human rationality, in particular on the rational choice theory. Ideally, individuals would optimize the choice which assures the highest degree of expected utility; but really they follow a “bounded rationality” which barely satisfies the expected utility. Therefore, self-deception results from the weakness of the warrant, since the self-deceiver, among the reasons used to formulate a judgment, does not consider the best possible reasons but the most functional ones for hisher aims and desires. It is a matter of practical reasoning, in which he/she prefers in some way (conscious or unconscious) to avoid an unpleasant condition (misleading information, like that p ) , rather than accept correct knowledge (that non-p). To reach this aim, the self-deceiver has to remove attention from the reasons contrasting the falsehood, consider only the useful alternatives, and resort to a particular biased system to process information and transform hisher beliefs. According to the words of Kunda [104], the motivation can influence “the generation and evaluation of hypotheses, of inference rules, and of evidence”, and “motivationally biased memory search will result in the formation of additional biased beliefs and theories that are constructed so as to justifjr desired conclusions” (p. 483). Second, to overcome self-deception paradoxes other scholars have proposed a mental partition model, according to which, given the distinction between “primary” consciousness (i.e., the immediate, plain and almost automatic knowing activity) and “secondary” one (i-e., the metaknowledge as knowing to know), self-deception is due to an automatic “unconscious” process, since the self-deceived is not conscious of professing one of the two belief p or belief not-p for motivational reasons, as it is suggested by Gur and Sackeim [ 1051. Also Davidson [ 1061 has recently claimed that “obviously two beliefs could coexist only if they were somehow kept separate” (p. 8). In particular, the threatening belief assumes a central position in the consciousness of the self-deceiver, while the deceptive belief takes an avowed position in hisher consciousness, as Gozzano [lo71 proposed. A stronger mental partition model has been advanced by the modular perspective of Pears [lo81 and Heil [109], according to which the mind is divided into different sub-systems. In this way the self-deception sub-system not only possesses its own “rationality” but also can follow different beliefs with reference to the main system (figure

3.3). Third, Mele [ I 10, 1 1 1, 1 121 has devised a “deflationary” model of self-deception. which is conceived as the avoidment or re-interpretation of a belief perceived as “threatening” by the individual. First of all, Mele proposed to overcome the static paradox. The self-deceived, driven by the desire that p , may believe that p . without needing a previous belief of that non-p. This process is based on cognitive biases typically linked with “cold” cognition, like vividness of information, confirmation bias, availability heuristics, tendency to search for causal explanations. etc. In particular, negative and positive misinterpretation, selective focusingattending and selective evidence-gathering are the main four ways by which a person’s desiring that p can contribute to his,her believing that p . In addition, Mele intended to tackle the dynamic paradox. To overcome it, he stated that. crudely put, self-deception is not an intentional subjective activity; rather, it is a motivationally biased information processing. According to Trope and Lieberman [ 1 131, in his “anti-agency view” Mele made reference to the confidence threshold in order to accept (or reject) a hypothesis in natural reasoning. For each hypothesis this threshold is defined by two values: a) an acceptation threshold (that is, the lowest level of “trust” in the truth which is necessary in order to accept that truth), and b) a reject one (that is. the lowest degree of trust in the hypothesis falsehood which is necessary in order to reject it).

Figure 3.3. The model of mental partition according to the modular perspective. Adapted from Heil [ 1091

The acceptation value threshold depends on the ratio between the acceptation costs of a false hypothesis and the information costs. The former consists in the subjective significance of avoiding a wrong belief of that p (the false hypothesis); the latter concern the cognitive resources and effort of information processing relevant to the hypothesis. The reject value threshold depends on the ratio between the reject costs of a true hypothesis and the information costs. So the individual’s desires may orientate the verifjmg or falsifying process of a hypothesis in two steps: a) they guide the choice of the hypothesis to be verified; b) they appraise the hypothesis according to the confidence thresholds in order to avoid the mistake costs (i.e. the subjective importance of wrongly avoiding belief of that p). The higher the costs of the mistake. the higher the acceptation value threshold and the IoLver the reject value threshold. Castelfranchi [ 1 141 has stressed this anti-intentional point

L. Anolli et al. / Deceptive Miscommunication Theory (DeMiT)

95

of view and has proposed a goal-oriented model of self-deception. Self-deception seems to be not a “goal-governed” behavior (i.e. intentional and subjectively finalized), but a “goaloriented” one, since it is a way of behavior finalized to reach a goal without an anticipatory representation of the goal itself by the agent. In this sense self-deception is an unconscious defensive mechanism like animal mimicry; in order to avoid psychological disease, the self-deceiver applies a cognitive operation on negative beliefs, such as to make them false or incomplete. In this way he/she acts “as if” he/she were governed by that goal. Self-deception calls for an incertitude condition, so that individuals can behave as if they could believe they may influence something that is already determined and not modifiable, by means of a sort of “almost magic thought or action”, as Shafir and Tversky [ 1151 say. Given the complexity of the situation, they may be self-persuaded because of a shift in their belief system: some undesired hypothesis can be weakened in favor of alternative beliefs that are more compatible with their desires, more convenient and, finally, more “true”. Certainly, we can assume that nobody has the intentional aim of deceiving him-herself, and that nobody is able to believe something just because he/she wants to, and so decides that it is indeed true. Two conditions seem to be essential: a) the plausibility of the belief of that p (in fact, if p is plausible - i.e. it respects the “reason constraints” - we have plenty of space in which to manoeuvre with the motivationally biased reasoning); b) the threatening value of belief to be inhibited or to be re-interpreted. In fact, Feingold [ 1161 found a positive correlation between self-deception and selfesteem. Although these two psychological constructs are completely distinct, they both account for subjective looks-personality aspects, and represent a similar self-orientation: to have a positive subjective view of self. These are basic aspects of the construction of selfperceived attractiveness and social desirability, which allow the individual to occupy a satisfying position in hidher own social world. In this sense self-deception is not considered as a failure of human rationality; rather, it seems to be and to work as a “defense mechanism” to reduce the psychic disease generated by a threat against the aims and self-image of individuals. The self-deceiver, like the illusionist, is able to “transform” reality in hisher own mind, and a false belief becomes true in hisher mind. As a consequence, self-deception could be an effective way to fabricate other-deception. In fact, the self-deceiver has to do only with the truth, and he/she is “sincere” in believing what he/she believes, even if it is the outcome of a disbelief. In this way a self-deceiver is “sincere” when he/she tells a lie to others, because he/she tells them what he/she believes. Therefore, what better deceiver than a sincere deceiver? Otherwise said, the self-deceiver is the best other-deceiver.

3.8

Conclusions

Deception seen as a family of subjective and interactive processes which take place in our everyday life in different contexts of use, need a global model to explain the range of phenomena included in this family. The miscommunication perspective, as outlined in the MaCHT proposed by Anolli (Chapter 1, this volume) has seemed to us an interesting and promising route to follow in order to illustrate and understand the different shades of deceptive interaction. In fact, deception may be interpreted essentially as a communicative act aimed at improving environment resources, protecting self-image and self-esteem, respecting social standards of relational interaction, as well as avoiding rudeness and discourtesy. In this field high- and low-content deceptive act may be an useful device. Generally speaking, in our theoretical perspective, we have not followed a Cartesian paradigm, based on the dichotomous distinction between mind and body, that is, between

96

L . .4nolli et al. /Deceptive Miscommunication Theoor?. (DeMiT)

“pure” rationality and “impulsive” or “perverse” irrationality. Truth-tellers would belong to the first sphere; false-tellers would belong to the second one. Rather, in the DeMiT perspective we have been inspired by a baconian paradigm, according to which mistakes and other wrong things such as lies and deceptive acts pertain to human nature in its information and processing and in its knowledge activity. We assume that we can not tell truth and falsehood apart as if they were black or white boxes. The ambiguousness and disease aspects of deception can not be thought of simply as an intentional decision to protect oneself or damage others, nor as a Machiavellian personality effect. nor yet as a rationality paradox and an interference of emotionally and motivationally biased reasoning. Instead, this ambiguousness can be thought of as a global aspect of operating and interacting with other people, because in this way our mind organizes the network of social experience according to our cultural environment. In this sense, in an everyday relational situation, we can consider deception as a route to express the communicator‘s sensations, thoughts, beliefs, emotions, and desires; likewise, at communication level, it may be useful to have the chance of hiding, omitting, concealing or. simply, blurring information.

3.9

Acknowledgements

We would like to thank Fondazione “Piero, Pietro e Giovanni Ferrero” (Alba. Italy) for the support given to the present study.

3.10

References S. McCornack, The generation of deceptive messages: Laying the groundwork for a viable theory of interpersonal deception, in: Message production: .4dvance.s of Communication theon.. J. 0. Greene. Ed. Mahwah: Erlbaum, 1997, pp. 9 1 - 1 16. M. K. Lapinski, Deception and the selJ! .4 cultural examination of information manipulation theor?.. Unpublished master’s thesis. Manoa: University of Hawaii. 1995. S. A. McCornack, Information manipulation theory. Communication Monographs 59 ( 1992) 1 - 16. S. A. McCornack, T. R. Levine, K. A. Solowczuk, H. 1. Torres, and D. M. Campbell. When the alteration of information is viewed as deception: An empirical test of information manipulation theoy. Communication Monographs 59 (1) (1992) 17-30. D. B. Buller. and J. K. Burgoon, Interpersonal deception theory. Communication Theor?. 6 C 1996) 203-242. J. K. Burgoon. D. B. Buller. L. K. Guerrero. W. Afifi. and C. Feldman. Interpersonal deception: XII. Information management dimensions underlying deceptive and truthful messages. Commrrnic~urinn .Monographs 63 (1996) 50-69. J. Bradac. E. Friedman. and H. Giles. A social approach to propositional communication: Speakers lie to hearers, in: Languuge,for hearers. G. McGregor, Ed. Oxford: Pergamon Press. 1986. pp. 127- 15 1 . D. B. Buller. J. K. Burgoon, C. White, and A. S. EtKsu, Interpersonal deception: VII. Behavioral profiles of falsification. concealment. and equivocation. Journal of‘ Language und Social P.q>*choloq?. 13 (1994) 366-395. P. Ekman, Deception, lying. and demeanor. In: Stares @’mind: American and post-Soviet perspectiiv nn contemporar?. issues in psycholog., D. Halpern and A.E. Voiskounsky. Eds. New York: Oxford University Press. 1997, pp. 93- 105. S. Jacobs, E. J. Dawson. and D. Brashers. Information manipulation theory: A replication and assessment. Communication Monographs 63 ( 1996) 70-82. B. M. DePaulo. M. E. Ansfield, and K. L. Bell. Theories about deception and paradigms for studying it. C’nmmunicarionTheon,6 ( 1996) 297-3 1 1. C. F. Jr. Bond. A. Omar. U. Pitre. B. R. Lashley. L. M. Skaggs, and C. T. Kirk. Fishy-looking liars: Deception judgment from expectancy violation. .JOUI-IIU~of Pcruonulit~~ unil Soc~icilP c i . c h l o , ~63~ (6) ( 1992) 969-977. D R Hiillrr. and .I K R u r m o n . Decention: Stratepic and nonstrategic communication. in: .Tmrft3eic

L. Anolli et al. /Deceptive Miscommunication Theory (DeMiT)

97

interpersonal communication, A. Daly and J. M. Wiemann, Eds. Hillsdale: Erlbaum, 1994, pp. 191223. R. E. Turner, C. Edgley, and G. Olmstead, Information control in conversations: Honesty is not always the best policy. Kansas Journal ofPsychology 11 (1) (1975) 69-89. B. M. DePaulo, D. A. Kashy, S. E. Kirkendol, M. M. Wyer, and 3. A. Epstein, Lying in everyday life. Journal of Personality and Social Psychology 70 (1996) 979-995. C. Castelfranchi, and J. Vincent, L’arte dell’inganno [The art of deception]. In: Menzogna e simulazione [Deception and pretence], M.A. Bonfantini, C. Castelfranchi, A. Martone, I. Poggi, and J. Vincent, Eds. Naples: Edizioni Scientifiche Italiane, 1997, pp. 155-201. R. M. Chisholm, and T. D. Feehan, The intent to deceive. Journal ofPhilosophy 74 (1997) 143-159. L. Wittgenstein, Philosophical investigations. Oxford: Blackwell, 1958. L. Coleman, and P. Kay, Prototype semantics: The English word lie. Language 57 (1981) 26-44. E. Sweetser, The definition of lie: An examination of the folk models underlying a semantic prototype, in: Cultural model of language and thought, D. Holland and N. Quinn, Eds. Cambridge: Cambridge University Press, 1987, pp. 43-66. L. Anolli, M. Balconi, R. Ciceri, Fenomenologia del mentire: aspetti semantici e psicologici della menzogna [Phenomenology of deceiving: Semantic and psychological aspects of deception]. Archivio di Psicologia, Neurologia e Psichiatria 45 (1994) 268-296. L. Anolli, and M. Balconi, Le parole della menzogna: per un’analisi delle strategie verbali del mentitore [Deception words: Towards an analysis of the verbal strategies of the deceiver]. Rassegna Italiana di Linguistica Applicata 2 ( 3 ) (1999) 15-27. H. P. Grice, Studies on the way of words. Cambridge: Harvard University Press, 1989. K. Jaszczolt, Discourse, beliefi and intentions: Semantic defaults and propositional attitude ascription. Amsterdam: Elsevier, 1999. J. R. Searle, Expression and meaning. Cambridge: Cambridge University Press, 1979. D. C. Dennett, The intentional stance. Cambridge: MIT Press, 1987. A. Whiten, Grades of mindreading. In: Children ’s early understanding of mind: Origins and development, C. Lewis and P. Mitchell, Eds. Hove: Lawrence Erlbaum, 1994, pp. 47-70. J. Perner, Understanding the representational mind. Cambridge: MIT Press, 1991. H. D. O’Hair, and M. J. Cody, Everyday deception, in: The dark side of interpersonal communication, W. R. Cupach and B. Spitzberg, Eds. Hillsdale: Erlbaum, 1994, pp. 181-213. H. A. Simon, Models of bounded rationality. Cambridge: MIT Press, 1982. D. Kahneman, P. Slovic, and A. Tversky, Judgment under uncertainty: Heuristics and biases. Cambridge: Cambridge University Press, 1982. E. Goffman, Frame analysis. An essay on the organization of experience. Harmondsworth: Penguin, 1974. B. M. DePaulo, D. A. Kashy, Everyday lies in close and casual relationships. Journal of Personality and Social Psychology 74 ( 1 998) 63-79. B. Sodian, Early deception and the conceptual continuity claim, in: Children ’s early understanding qf mind: Origins and development, C. Lewis and P. Mitchell, Eds. Hove: Lawrence Erlbaum, 1994, pp. 385-401. M. R. Cadinu, and J. Kiesner, Children’s development of a theory of mind. European Journal qf Psychology of Education 15 (2000) 93-1 1 1. M. J. Chandler, and S. Hala, The role of personal involvement in the assessment of early false-belief skills, in: Children ’s early understanding of’ mind: Origins and development, C. Lewis and P. Mitchell, Eds. Hillsdale: Erlbaum, 1994, pp. 403-425. S. Hala, and M. J. Chandler, The role of strategic planning in accessing false-belief understanding. Child Development 67 (1996) 2948-2966. S. R. Leekam, Jokes and lies: Children’s understanding of intentional falsehood. In: Natural theories of mind. Evolution, development and simulation of everyday mindreading, A. Whiten, Ed. Oxford: Basil Blackwell, 1991, pp. 159-174. P. Newton, V. Reddy, and R. Bull, Children’s everyday deception and performance on false-belief tasks. British Journal of Developmental Psychology 18 (2000) 297-3 17. A. Vrij, G. R. Semin, and R. Bull, Insight into behavior displayed during deception. Humnn Communication Research 22 (4) (1996) 544-562. M. Zuckerman, and R. E. Driver, Telling lies: Verbal and nonverbal correlates of deception, in: Multichannel integrations of nonverbal behaviour, A. W. Siegman and S. Feldstein, Eds. Hillsdale: Erlbaum, 1985, pp. 129-148. J. 0. Greene, H. D. O’Hair, M. J. Cody, and C. Yen, Planning and control of behavior during deception. Human Communication Research 11 (1985) 335-364.

,

98

i

.4t?o//iet a1

Decepti1.e Miocommunication Theon. fDeMiT)

~ 4 3 1 B. J. O'Keefe, and B. L. Lambert. Managing the flow of ideas: A local management approach to message design. in: Commirrrication Yearbook 18. B. Burleson. Ed. Newbury Park: Sage Publications. 1994, pp. 54-82. S. Lindskold, and P. S. Walters. Categories for acceptability of lies. Journal of'Social f.v?.cholog\.120 (1983) 129-136. L. Anolli, and R. Ciceri. The voice of deception: Vocal strategies of naive and able liars. Jorrrnul ol ,Vont-erbalBehavior 21 ( 1997) 259-284. P. Ekman. M. O'Sullivan. and D. Matsumoto, Confusions about content in the judgment of facial expression: A reply to contempt and the relativity thesis. Moti\*arion and Emotion 15 ( I99 1 a 1 169-I 76. A. Vrij, L. Akehurst, and P. Morris. Individual differences in hand movements during deception. Joitrnal qf Nont.erbal Behavior 21 ( 2 ) ( 1997) 87-102. F. Goldman-Eisler. Ps\~cIin/in~Qui.otic~s: Expt~rirncvits i n spontarieotr,~ ~peei'li.New York: Academic Press. 1968. S. R. Rochester. The significance of pauses in spontaneous speech. .lournu1 of' P.s~~c~holiiipui.stic Research 2 ( 1973) 5 1-8 1 . C. R. Berger. and J. M. Jordan. Planning sources. planning ditficulty and verbal fluency. Communication Monographs 59 ( 1992) 1 30- 149. J. P. Tagney. and K . W. Fisher, Self' conscious emotions Tlw p.y\dio/o
L. Anolli et al. /Deceptive Miscommunication Theoly (DeMiT)

99

suspicion on perceived communication and nonverbal behavior dynamics. Human Communication Research 22 (1 995) 163-196. J. K. Burgoon, D. B. Buller, and L. K. Guerrero, Interpersonal deception: IX. Effects of social skill and nonverbal communication on deception success and detection accuracy. Journal ofLanguage and Social Psychology 14 (1995) 289-3 11. J. K. Burgoon, D. B. Buller, K. Floyd, and J. Grandpre, Deceptive realities: Sender, receiver, and observer perspectives in deceptive conversations. Communication Research 23 ( 1996) 724-748. J. K. Burgoon, D. B. Buller, C. H. White, W. Afifi, and L. S. Aileen-Buslig, The role of conversational involvement in deceptive interpersonal interactions. Personality and Social Psychologv Bulletin 25 (6) (1999) 669-685. D. B. Buller, K. D. Strzyzewski, and F. G. Hunsaker, Interpersonal deception: 11. The inferiority of conversational participants as deception detectors. Communication Monographs 58 (1) (199 1) 25-40. T. A. van Dijk, Context models in discourse processing, in: The construction of mental representations during reading, H. van Oostendorp, S. R. Goldman, Eds. Mahwah: Erlbaum, 1999, pp. 123-148. D. Sperber, and D. Wilson, Relevance: Communication and cognition. Oxford: Oxford University Press, 1986. B. Sandig, and M. Selting, Sprache und Gespriichsstile [Speech and styles of conversation]. Berlin: de Gruyter, 1997. J. M. Sinclair, On the integration of linguistic description, in: Handbook of discourse analysis (vol. 2), T. A. van Dijk, Ed. London: Academic Press, 1985, pp. 13-28. L. Anolli, and M. Balconi, Mentire con le parole: Analisi degli indicatori verbali micro- e macrostrutturali [Lying with words: An analysis of micro- and macrostructural verbal indexes]. Ricerche di Psicologia 22 (2) (1 998) 121 - 149. M. L. Alonso-Quecuty, Deception detection and reality monitoring: A new answer to an old question'? In: Psycholou and law: Znternational perspectives, F. Losel, D. Bender, and T. Bliesener, Eds. Berlin: de Gruyter, 1992, pp. 328-334. R. H. Carpenter, Stylistic analysis for law enforcement purposes: A case study of a language variable as an index of a suspect's caution in phrasing answers. Communication Quarterly 29 (1991) 32-39. 1831 E. R. Dulaney, Changes in language behavior as function of veracy. Human Communication Research 9 (1982) 75-82. M. L. Knapp, R. P. Hart, and H. S. Dennis, An exploration of deception as communication construct. Human Communication Research (1974) 15-29. D. Kuiken, Nonimmediate language style and inconsistency between private and expressed evaluations. Journal of Experimental Social Psychology 17 ( 1981) 183-196. G. R. Miller, and J. B. Stiff. Deceptive Communication. Thousand Oaks: Sage, 1993 W. R. Todd-Mancillas, and R. J. Kibler, A test of concurrent validity for linguistic indexes of deception. Western Journal of Speech Communication 43 (1979) 108-122. R. F. Baumeister, Lying to yourself: The enigma of self-deception, in: Lying and deception in evelyday life, M. Lewis and C. Saarni, Eds. New York: The Guilford Press, 1993, pp. 166-183. 1993 S. A. McCornack, T. Levine, K. Morrison, and M. Lapinski, Speaking of information manipulation: A critical rejoinder. Communication Monographs 63 (1) (1996) 83-92. M. Cole, Interacting minds in a life-span perspective: A cultural-historical approach to culture and cognitive development, in: Interactive minds: Lfe-span perspectives on the social foundation of cognition. P. B. Baltes and U. M. Staudinger, Eds. New York: Cambridge University Press, 1996, pp.59-87. C. Geertz, The interpretation of cultures. New York: Basic Books, 1973. M. L. Knapp, and M. A. Comadena, Telling it like it isn't: A review of theory and research on deceptive communication. Human Communication Research 5 (1 979) 270-285. A. Vrij, The impact of information and setting on detection of deception by police detectives. Journal of Nonverbal Behavior 18 (1995) 117-136. R. E. Riggio, and H. S. Friedman, Individual differences and cues to deception. Journal of Personali8 and Social Psychology 45 (1983) 899-915 . M. Zuckerman. B. M. DePaulo, and R. Rosenthal, Verbal and nonverbal communication of deception, in: Advances in experimental social psycholog-v (vol. 14). L. Berkowitz, Ed. New York: Academic Press, 1981, pp. 1-59. M. G. Frank, and P. Ekman, The ability to detect deceit generalizes across different types of high-stake lies. Journal of Personality and Social Psychology 72 ( 1 997) 1429-1439. U. Hess, and R. E. Kleck, The cues decoders use in attempting to differentiate emotion-elicited and posed facial expressions. European Journal of Social Psychology 24 (1 994) 367-38 1.

f 00

L. Anolli et a/. !Deceptive Miscommirnication Theon, (DeMiTj

[98] P. Ekman. M. O’Sullivan, and M. G. Frank. A few catch a liar. P.yvchological Science 10 ( 1999) 263266. [99] D. S. Wilson, D. Near, and R. R. Miller, Machiavellianism: A synthesis of the evolutionarl; and psychological literatures. fsvchological Bullerin 1 19 ( 1996) 285-299. [ 1001 B. Fehr. D. Samson, and D. Paulhus, The construct of Machiavellianism: Twenty years later, in: .4dvances in fer.sonalig*.4ssessment. C. D. Spielberger and J. M. Butcher, Eds. Hillsdale: Erlbaum. 1992, pp. 77-1 16. f l O l ] M. C. Ashton, K. Lee, and C. Son. Honesty as the sixth factor of personality: Correlations with Machiavellianism, Primary Psvchopaty. and Social Adroitness. European Journal nf Personalie. 14 (2000) 359-368. [ 1021 J. A. Shepperd, and R. E. Socherman, On the manipulative behavior of low Machiavellians: Feigning incompetence to “sandbag” an opponent. Journal of‘ Personalih, and Social f.yJ.chologj. 7 2 ( 19971 1448-1459. [ 1031 D. Davidson, Deception and division. in: Actions and events: Perspectives on the philosophj. of’ Donald Davidson, E. LePore and B. McLaughlin, Eds. Oxford: Blackwell. 1985. pp. 138-148. [ I041 Z. Kunda, The case for motivated reasoning. fsschological Bulletin 108 ( 1990) 480-498. [IOS] R. C. Gur, and H. A. Sackeim, Self-deception: A concept in search of a phenomenon. Joumul I?/ Personaliy and Social Pslrholog*37 ( 1979) 147-I 69. [ 1061 D. Davidson, Who is fooled? In: Se/f-deception a n d parado.res of ratiortalig.. J . P. Dupuy. Ed. Cambridge: Cambridge University Press, 1998. [ 1071 S. Gozzano, Condizioni per l’autoinganno [Conditions for self-deception]. Sisremi Inrelligmfi I 1 ( 3 ) ( 1999) 469-487. [I087 D. Pears, The goals and strategies of self-deception. in: The multiple .self,. J. Elster. Ed. Cambridge: Cambridge University Press. 1986, pp. 59-77. [I091 J. Heil, Going to pieces, in: Philosophical p.sj*choparholokgj.,G. Graham and L. Stephens. Eds. Cambridge: MIT Press. 1994. pp. 1 1 I - 134. [ 1 101 A. R. Mele. Real self-deception. Behavioral and Brain Science 20 (1) ( 1997) 9 1 - 1 36. [ 1 1 I J A. R. Mele. Motivated belief and agency. Philosophical Ps?r.holom. I 1 ( 3 )( 1998) 353-369. [ 1 133 A. R. Mele. Twisted self-deception. Philosophical f.ychologr. 12 ( 2 ) ( 1999) 1 17-137. [ 1 131 k’. Trope, and A. Lieberman. Social hypothesis testing: Cognitive and motivational mechanisms. in: SOCILIIps?rhologr: Handbook of basic principles. E. Higgins and A . Kruglanski. Eds. New York: Guilford Press, 1996, pp. 239-260. [ I 141 C . Castelfranchi, Come studiare la mente (per quello che P ) [How to studv the mind (for what it is)]. Sistenti Intelligunfi 12 ( I ) ( 1999) 39-66. [ 1 IS] E. Shafir. and A. Tversky, Thinking through uncertainty: Nonconsequential reasoning and choice. Coyniti1.e Psrchologv 24 (4) ( 1992) 449-474. [ 1 161 A. Feingold. Good-looking people are not what we think. P.yJrhological Bulletin 11 1 ( 1992) 304-341.

101

Say Not to Sa-v L. Anolli et al. (Eds.) IOS Press, 2002

4

Seductive Communication: Paradoxical Exhibition, Obliquity and Non Verbal Synchronization Rita CICERI

Abstract: Seductive interaction is here analyzed as a flexible plurality of behavioral patterns, corresponding to the variety of communicative intentions: exhibition, approaching the partner, deepening reciprocal knowledge, and reaching of a level of intimacy. The attention is moved from a structural analysis in which seduction is described as a timed flow of interactions characterized by different “steps”, to a complementary approach which analyzes the communicative seductive behavior in each phase and shows some of the timerelated dimensions, such as sequence, frequency, synchronization and simultaneity, which are required to describe seductive communication behavioral patterns. It particularly makes it possible to analyze the connections between different systems of expression (verbal and nonverbal) and to describe several seductive strategies of obliquity and to disguise tactical communication, which are defined as miscommunication forms. These communicative strategies paradoxical exhibition, forms of discursive obliquity, the multimodal message and nonverbal synchronization - are based on the “undefined content” of the message. It is because of this that seductive communication is tantalizing, leaving much to the partner’s imagination and promising her more than she has already seen. Moreover, the fact that the content is “undefined” allows the content itself to be adapted and modified to best suit the situation, thus lessening the risk of being too invasive and of being rejected. Secondly, it is shown how these strategies are carried out concentrating on the “undefined form” of the message, or rather, on the synergy between different expressive signals which make up the message.

Contents 102 102 103 105

4.7 4.8 4.9

Introduction. ......................................................... Seductive communication: a matter of heart or head?. ........ The bio-social models and the steps of seductive interaction. Paradoxical exhibition: getting out of anonymity.. ............. The gradually reciprocal disarmament and the strategies of obliquity.. ......................................................... Multimodality: the seductive power of signals and componentiality of the seductive message.. ..................... Reciprocity and synchronization strategies. ..................... Conclusions. ......................................................... Acknowledgements .................................................

4.10

References. ...........................................................

114

4.1 4.2 4.3 4.4 4.5 4.6

107 108 110 113 114

I02

4.1

R. Cicer-i Seductib’e Commrrnrcation

Introduction

During the courting stage even the cockerel uses astuteness to attract the attention of the hen. To catch her attention and to entice her towards him, he shows off the food he has in his beak. This is an extremely simple sequence of actions. which has been consolidated throughout the centuries. Courting, both in the animal world and in human interaction, involves resorting to masked exhibitionism, trickery and, more generally, a form of indirect and obiique communication. The act of seducing is a subtle and enticing game, which requires using the right amounts of exhibition or disguise, flaunting or revealing slightly, saying something but not sayng too much. This alternating between one and the other of the above plays a big part in seductive interaction in which the intention to reduce the interpersonal distance with hopes to increase intimacy is closely intertwined with the wish to save one’s face in case of rejection. It also concerns the choice of communicative behavior. which is strategically allusive: to be efficient, it must involve giving just a hint of what’s on offer without revealing too much. Clarity and referential information are certainly not the distinctive features of seductive communication; in fact, what seems to count most is “how the game is played.” rather than the contents of the game itself. For these characteristics, seductive communication stands out in its own right as a point of observation in communicative behavior. Particularly it makes it possible to analyze the connections between different systems of expression (verbal and nonverbal) and to describe several strategies of obliquity and disguise techniques. which are defined as miscommunication forms. Having defined this phenomenon which we are referring as “seduction” we hereby propose to describe some of these communicative strategies.

4.2

Seductive communication:a matter of heart or head?

The etymology of the Latin verb “seducere” (composed of sed “separation” and ducere “to lead,” which in German comes out as VerTfihrung)shows that the word has taken on various meanings throughout the course of the centuries and points out different and complementary aspects of seduction. One negative meaning of it is “to lead astray,” therefore trick the other person somehow. In literature, from the classical tragedies through to decadentism, seduction was considered to be an unstoppable force, a trap that was capable of leading one onto the wrong path. The separation of the image of the seducer, a hunter, cold and indifferent to the feelings of others, and the seduced, prey and victim of his or her own feelings, is highlighted. The former is driven by his head, the latter by his heart. Seductive communication is the place and the means, which are used to lay the snare. Another, but quite different meaning of’keducere” is “to be led along.” Here. he who is seduced is attracted and led, quite simply. “somewhere else.” In terms of relationships, it is stressed that seduction plans, or rather, creates of suitable settings and the conditions for a more intimate encounter and the getting to know of one another. The set borders between the seducer and the seduced seemingly break down since seductive roles coexist in both partners. In romance literature and in much contemporary literature seductive communication comes under this meaning being a prelude to relationships and special encounters. and made up of siLgnals and e\pident s i p s of choosing Heart eyer mind and

R. Ciceri /Seductive Communication

103

conversely, or, better still, passion bound with knowing one another. In a broader sense, nowadays the word “seduction” indicates a form of behavioral and communicative strategy. Leaving aside the connotation, which is most associated with the relationship between the two sexes, today the term “seduction” means a mixture of attitudes and devices, which is used to enable oneself to come out of anonymity and to get noticed and chosen. In this sense seduction can be seen at work in all social spheres and contexts, from private relationships, to the public economy to politics to the media and press. Seductive communication then has much to do with creating and keeping up an image that must be suitably attractive to obtain what is desired. Seduction could therefore be an advertisement, a picture or an electoral campaign. In other words, in what Baudrillard [ 13 defines as “appearance strategies,” seductive communication can combine an informative capacity with the capacity of pleasure and “sensual involvement .” The “art of pleasure” requires a careful plan of behavior, which has communicative relevance, and simultaneously must be able to elicit an emotional response. If it can provoke not only an interest, but also attraction, then it can create or increase a feeling of desire or need. For this prerogative it has been defined by Umiker-Sebeok [2], and by Taylor, Hoy, and Haley [3], quite paradoxically, “cold seduction”. This type of seduction, aiming itself at several people at once though making each one feel exclusive, is of a contractual nature and is played consciously and knowingly by both partners. In this last interpretation the strategic aspect of seductive behavior assumes the form of a conquering technique in which emotional involvement and interaction with the partner are not a relevant part of the game, but is the actual means to an end. Seductive communication is a complex matter for the heart and for the mind since it is used as a planned, rational strategy and is also called into action when there is emotional involvement on behalf of the seducer. The different amount of these two components will depend on the seducer, what his aims are, how far he wants to go and how much is at stake for him. Moreover, as we’ll see in the next paragraph, these components vary also depending on the actual stages of the seductive game

4.3

The bio-social models and the steps of seductive interaction

In scientific literature there have been two main approaches to analyzing seduction: the ethological-evolutionist research and the studies which examine the social and psychological dimensions. Both highlight an exchange of resources as a part of the courting ritual. The evolutionist models highlight the reproductive function of the exchange as being largely due to sexual attraction, while the socio-psychological models point to the function of culture in establishing patterns of exchange. Heart/mind can be weighed up against instinctheason or nature/culture. The most recent bio-social seduction models aim to bind the biological aspect, concerning animal behavior, with the interpersonal aspect, this last with all the various influences bestowed upon it by culture. Because of the most recent developments in psychology and intimate behavior outlined, among other scholars, by Hewes, [4],and Dindia [ 5 ] , the Steps Theory of Givens [6] and the bio-social model proposed by Kendrick and Trost [7], seduction could be defined as a strategic and intentional sequence of moves in which the primary motive is to attract (usually sexually) another person (usually of the opposite sex). The main goal of seduction

R Ciwri Sedtrctilv Commrrnication

I 0.1

is to build an intriguing bond with the partner with the aim of reaching an intimate relationship. Within this perspective, we could describe seduction as a timed flow of interactions characterized by different "steps" or unfolding phases. as it is outlined by Morris [8]. .4s shows figure 4.1, the starting point of the seductive process is the identificationchoice of a charming partner based on attraction and interests (first step). Then the subject aims at establishing contact with the potential partner through the strategies of exhibition and of catching hisher attention (second step). The subject then tries to assess more realistically the degree of interest and attraction he feels towards the partner (third step). Finally, the seductive process develops into a phase of reciprocal approach, with the aim of establishing an intimate relationship between the partners through the progressive reduction of the uncertainty level (fourth step). The whole sequence ends with the decision to maintain a stable bond (fifth step).

Step 1 THE SEARCH FOR

STIMVLLS _------------__--

Is the person attracti\ e''

Step 2 MAKING CONTACT

_----__--_--------

Is this person interested in me:'

Is he she

Step 5 -intelligent" -friendly'

Can I flirt with him her'?

4

I

Can 1 trust him her'?

KEEPING UP THE BOND

Can I open up IO him'her''

Can the relationship po on !' Could we Ine together '

\ I I

__-_--------_--_--

I

Figure 4.1. The steps of seduction

Figure 4.1, above, shows the procedural dimension of the seductive interaction. time related and structurally divided into progressive stages, as it is shown by Moms [9]. Such segmentation allows us to distinguish and analyze intentions. strategies and communicative behavior at each stage and as far as the description and analysis of seductive communication are concerned; two limits are considered which could lead to a simplification of the phenomenon. Above all, the model is primarily interested in the actions of the seducer. showing a rigid distinction between the roles of the partners. Secondly, only the sequential dimension of the process is described, thus neglecting not only the aspects of simultaneity and concurrence, but also of, and specifically, the communicative aspect. A complementary approach is outlined in Figure 4.2. It shows some of the time-related dimensions which are required to describe seducti\.e communication. The communicatiLe

R. Ciceri /Seductive Communication

105

gestures of both partners show how these gestures are essentially interrelated in terms of sequence, frequency, synchronization and simultaneity. According to Watt and Van Lear, [lo], Greene [ 111, and Ciceri and Mistri [ 121, this increases the grades of freedom which the partners have in making their choices and their actions. The attention is moved from a structural description of seductive interaction (its phases) to the analysis of the communicative behavior which characterizes each phase. Bearing these aspects in mind, in the following pages we will focus our attention on some of the communicative strategies of seduction.

Figure 4.2. Types of behavioral coordination during seductive interaction: movement synchrony, simultaneity, frequence and sequence

4.4

Paradoxical exhibition: getting out of anonymity

The first step of the seductive sequence is to draw attention to oneself to be chosen. In fact, the aim of the seducer is to come out and “change status: from being anyone to being someone” [l]. This primary aim of the would be seducer can be seen nearly everywhere, from the greatest Don Juan to the world of advertising, their greatest wish being that of coming across as exclusive and unique. Seductive behavior must be broken down into its various subcomponents such as temptation, charm and attraction. It is characterized by communicative performances in which the esthetical aspect is enhanced and heightened whereas the referential and informative aspect is lessened [3]. It’s necessary to come out and show oneself, get noticed and subsequently go on to be selected and to conquer. In this process both partners are compelled to enhance and stress their qualities and their

R Ciceri Seductive Communicatron

106

strong points, as well as to hide their defects and mask their limitations to be attractive. The goal is to become the other person’s object of sexual desire. So, during the courtship. each partner tries to appear better than he/she really is, and, in some way, the enticing communication shows several points of contact with deceptive communication. This “basic” intention in the first phase of seduction is also the prospective of the evolutionist models. Here, sexual attraction is defined as the primary biological dnve involved in courtship. Several ethological studies concerning different species of animals highlight the presence of courtship behavioral forms which have reproduction and pleasure as their primary aim. What we’re dealing with is specific and articulated forms of exhibitive behavior which are used to attract the partner, in the same way as aggressive behavior (contention, struggles, etc.) is used to conquer directly. The analysis of these behavioral aspects shows just how relevant they are in this interactive game of exhibition. The “game“ itself ends with the complement (union of the two partners) and it is repeated on successive occasions, whenever the reproductive cycle allows it. According to this perspective. the characteristics involved in determining the choice of partner are directly connected to the possibility of bearing an offspring. It is through this that gender differences can be explained (Coolidge effect). Females, who play a bigger role in the rearing of the offspring. are more selective in their choice of partner. They show a preference for males who have attributes of strength and dominance. Males. on the other hand. are less selective. Since their primary aim is reproduction. their strategy is the number of times they can copulate and the number of partners they can copulate with [7]. Socially and culturally influenced examples of the Coolidge effect can be seen in some partner selection criteria in western culture. These examples have been pointed out in recent research. During the courtship period men tend to emphasize their social status and the resources they have to offer such as dominance and the guarantee of looking after and protecting the offspring. Women. on the other hand, are viewed as being competitive when it comes to conquering a partner, drawing attention to her morphological indicators of youth and health which are symbols of greater fertility. as Tooke and Camire [ 131. Barber [ 131. and Hirsch and Paul [ 151 sketched out. However, these aspects constitute only one of the factors taken into consideration by the seducer in hidher evaluation and selection process in the first phase. The criteria for choosing a partner don’t involve these characteristics since it must also be taken into account ( 1 ) what exactly one wants from the relationship ( o p e of’ relationship, according to Kendrick. Sadalla. Groth and Trost [ 163: Klohnen and Mendelsohn [ 171) and (2) each individual’s beliefs, wishes and expectations (.dfcomponents, according to Fletcher. Simpson, Thomas, and Giles [ 181). For example, for a simple date much emphasis is put on physical attraction whereas in a more stable relationship other criteria are involved such as trust. social status and the complementary, likeness-to-oneself criteria. This exhibition is a strong relational strategy because it implies being directly involved. Also. it entails an implicit declaration of willingness to begin an adventure, as well as settling a commitment and a relational responsibility towards a potential partner [5]. Moreover. this strategic behavior must be carefully weighed out according to the time the partner takes to react and respond. The exhibition is played out in a relatively narrow psychological space defined, on the one hand. by the desire to attract the partner and. on the other. by the need to respect hisher freedom in order to become. in turn. the object of desire. Consequently. to reach the goal of enticement. it i s necessay to leave the partner some degree of freedom and room *

R. Ciceri /Seductive Communication

107

for movement. This initial phase of “approach,” nonetheless, cannot last too long, because a prolonged exhibitive behavior risks becoming boring or irritating since it can come across as having a hint of arrogance and dominance. Alternatively, it may have the opposite effect, as an indication of insecurity and weakness. What we are up against is a paradoxical exhibition because, though explicit, it is not declared. It is evident, but not formalized. Since during seduction the individual himself is at stake, he cannot run the risk of a flat refusal by the potential partner. If this occurred, it could heavily damage the self-image, and lower the level of self-esteem. So, the seductive exhibition should not go beyond certain limits and should not become so excessive as to avoid the intrusiveness of the partner’s subjectivity. Here, he/she could protest and oppose, generally, decisive forms of rejection.

The gradually reciprocal disarmament and the strategies of obliquity

4.5

Once the contact between the partners has been set up, there follows the phase of reciprocal approach. In fact, in the seductive process the partners are inclined to “withdraw,” as a basic condition for a growing intimacy, as Duck [ 191 outlined. In this step, they encourage each other to reveal themselves gradually through a progressive and selective exchange of information about personal history. A seductive intention must be shown gradually, bit by bit, because of the impossibility of knowing with certainty the intentions or feelings of the other. It is this impossibility of knowing the other’s intention which makes the situation risky, and, being over-explicit in showing an interest could lead to rejection, this in turn causing grave damage to how one feels about himself. It’s therefore necessary to adopt a type of communication which is “oblique” or indirect [l], a strategy of “saying a little but not saying too much.” It’s necessary to be able to get the right balance between allowing oneself yet denying oneself, putting oneself forward yet restraining oneself. It’s a question of stating a little but not too much and simultaneously being able to observe how the other is reacting. Let us consider the different possibilities of forwardness and risk in the following examples of allusions to a hrther encounter. (1) (2) (3) (4) (5) (6)

Do you come here often? I wonder ifwe’ll meet again I wonder ifwe could meet again It would be nice to see each other again I would like to see you again I would like to see you again. What about you?

We can observe the change in the level of indirectness: from the somewhat “cautious” request for information in (1) which doesn’t quite reveal the intentions of either party, to the rather latent declaration of interest and request for direct response in (6). The level of obliquity, or indirectness, depends on the amount of feedback the seducer is getting from the partner. He is trying to calculate how much intimacy is gained from the questions he has posed to the partner and the response he has been given [19]. Explicitly personal questions are usually avoided as they could lead to a violation of intimacy. The seducer knows that there are risks involved in being too forward and open, and he doesn’t want to jeopardize the possibility of strengthening the bond with the partner and so he must try to use the right amount of forwardness. Oblique communication therefore allows him to combine openness and pleasure whilst allowing him at the same time the reciprocal

1 ox

R Ciceri Srductiiv Communicuriori

"disarmament," regarding the content, timing and method involved in the communication. According to Derlega, Metts, Petronio, and Margulis [20], Dindia [2 11, and Berger [ 2 2 ] , this process of sharing reciprocal experiences allows a gradual reduction of the degree of uncertainty, limiting the risk of becoming vulnerable too early, as pointed out by the uncertainty reduction theory. Sometimes an exhibition of vulnerability is used as an oblique strategy. It's frequently adopted to make a first move, and it aims at getting the other person to drop hidher defenses or "arms". According to this. the strategy allows the seducer to increase the level of intimacy and then lessen or even neutralize the partner's resistance. He does this by adopting a submissive and somewhat childlike behavior. Child-like in the sense that the behavior conveys a need to be looked after similar to that found in the mother-child relationship. Hass 1231 stated that also animals use signals of subordination and mothering or "fosterage" to encourage or invite the partner to approach. This facilitates mating and curbs the instinct of the mate to protect hisher body and living space. It is this instinct of self-protection that causes aggressive and defensive behavior in many vertebrates. Courtship rituals in some species of birds see one bird approaching its mate and behaving as if it were a baby bird still in the nest. In the same way, doves, for example, press their beaks together prior to mating. This is reminiscent of the beak contact involved when a parent dove gives food to the young. Research on the vocal qualities of the seducer shows that a "babyish" voice is associated with less power yet more warmth and approachability. In other words. it seems that an infantile voice is a clear sign of willingness to get to know the other person, and it is a signal which is sent out during the first stages of courtship. Berry [24] noticed that an attractive male voice is associated with the characteristics of strength. invulnerability and dominance whereas an attractive female voice is reminiscent of warmth, honesty and kindness. However, all these characteristics can still be detected even when either sex adopts a "babyish" voice. Such information is confirmed in studies by Givens [6] in which he describes the type of voice used by the seducer to assure the partner that heishe can approach without any fear. What we are dealing with here is the tone of voice used by adults when they are speaking with children: i.e., a high-pitched yet low-intensity \.oice tone. When the seducer is smiling, looking at and facing the other person. he seems to take on a high-pitched quickened voice. This signals an interest in the partner and a desire to interact with her. In the same way. a submissive attitude, especially where the male is concerned. is a sign of desire to get closer to the partner. as Givens 16). and Kendon and Ferber 1253 sketched out.

Multimodality: the seductive power of signals and componentiulity of the 4.6 seductive message As Barthes [26] reminds us, the cautious seducer always moves about with a "reserve disguise" which he points to in subtle and allusive way. He needs to be able to express things in such a way that gives to understand something without to take upon oneself any liability for what he has been said. Therefore, the meaning of the seductive message is "transdiscorsive." it goes beyond what has actually been said, as it is shown by Burgoon and Dillman [27]. The form of seductive message is multimodal, depending on the relation between signals belonging to different systems of expression (vocal. verbal. visual. kinetic). which in[-olves the [%.hole person.

R. Ciceri /Seductive Communication

109

During the course of seductive interactive game, communication is influenced by verbal and nonverbal components. Unlike persuasion, seduction has not only to convince the partner, but also to attract himher. Furthermore, the seducer must be able to act in an implicit and intriguing manner because enticing the other person needs an attentive and sensitive process of negotiation. The research which has been carried out in this field up till now has aimed on the one hand to isolate and define nonverbal traits, or signals charged with seductive power. On the other hand, there are experimental studies that try to analyze the role of strategic nonverbal behavior during the management of seductive interaction and to describe the procedures of componential (verbal and nonverbal) message production. To begin with, more emphasis has been given to the role of eye-contact. The dilation of the pupils is a clear signal of attraction, and a precise psychophysiological signal of interest [9] charged with seductive power. Besides, the time and modality of eye-contact typical in the seductive game have also been described in Western culture (e.g., glimpsing, or looking askance at someone; Givens [6]; Simmel [28]). The higher the tendency to exchange glances and prolong the period of eye-contact, the higher the chances are of putting into action a process of reducing interpersonal distance. Specifically, according to Cupach and Metts [29], women in the West seek and cast more glances than men and, in an alluring interaction, rely more heavily on visual feedback. One example of this type of communicative behavior is the position that Simmel [28] describes as being typical of a flirt, be it a man or a woman. Here, it takes the form of a sidelong glance so that while the eyes are staring at the partner, the face is slightly turned in a different direction. This method of eye-contact is suitable for conveying ambiguous messages of “interest yet disinterest,” for example, when one doesn’t want to risk revealing too much. The suitability of using such a strategy, it being made up of various components, or, multimodal signals, can be seen when one acts in a certain way and yet conveys a totally different message which would appear contradictory to hisher intentions. Grammer [30, 311 confirms that in courtship rituals subjects try to attract the attention of the partner using signals composed of an indirect and often contradictory nature rather than via explicit verbal communication. In this way there’s no actual declaration of the “longing” felt by the seducer who is therefore able to avoid the possibility of rejection and getting hurt. Other than eye-contact there are certain other behavioral strategies which are used to show interest or indeed, to show the opposite (for example, to get away from or to get out of an unpleasant relationship). In the first case behavior patterns well noted in literature are gestures which involve touching oneself (stroking, touching one’s hair etc.), smiling a lot and assuming the posture of a person who is interested (chest sticking out, facing the other person and standing really close to himher). In the second case (wishing to show disinterest) there is less self-touching and smiles, and posture is stationary. Even the eyecontact is different, glances flitting here and there but not necessarily lowered or directed at the partner, as it is outlined by De Weerth and Kalma [32]. According to the studies by Koeppel, Montagne-Miller, O’Hair, and Cody [33], signals of interest in the first phase of interactive seduction constitute expressive patterns which make it possible to recognize what’s going on between two people; whether there it is flirting or simply, friendliness, this latter considered to have inferior emotional involvement. There has been less research done on nonverbal vocal aspects which are a way for the seducer to introduce himself, and it is from the actual voice that the partner can guess or get an idea of some of his physical, social and psychological characteristics such as sensitivity, intelligence, strength, maturity and beauty, as Mallory and Miller [34],

110

R C‘icrri Sedurri\.e Communication

Addington [35], Laver [36], and Scherer [37. 381 sketched out. More specifically, some scholars, i. e., among the others, Mallory and Miller [34]. Scherer [39], and Zuckerman, Hodgins, and Miyake [40], have highlighted just how important and relevant the voice is in interpersonal attraction thus confirming the notion of the “vocal attraction stereotype.“ If the seducer’s voice comes across as being attractive. then the partner detects an overall desirable feeling about him. The voice is also considered to be a very trustworthy means of finding out what type of interaction is going on between two partners [33]. In an interaction based solely on friendship the tone of voice is neutral. warmth of the tone is limited. there is not much laughing and there are lengthy silences. In an interaction involving flirting on the other hand, there’s an increase in giggling, warmth and interest. the tone of voice is more intimate, speech is livelier and flows more freely and silence periods become shorter.

4.7

Reciprocity and synchronization strategies

Many seductive strategies and techniques show a high level of reciprocity, as it is shown by Tooke and Camire [ 131, Greer and Buss [41]. When the seducer introduces himiherself he or she seemingly adapts to the expectations of the other. Both sexes tend to emphasize the physical and psychological characteristics which he/she knows the other would appreciate [30]. Studies by Net0 [42] and De Weerth and Kalma [32] point out the most widely used seduction strategy. One such technique is a communication strategy which involves sending out signals of interest and willingness to look after or take care of the partner. Such a strategy sees the seducer taking a back seat and focusing his attention on his partner. Further studies by Crook [43] which aimed to outline the differences between the male and female approach to seductive strategies highlighted how the selection of different strategies is connected to reciprocity. The woman activates a series of “body signals“ which ignite the curiosity of the man thus capturing his attention. Since the reaction in the man constitutes his response. the frequency of the man’s advances is highly correlated to the frequency which the woman sends out these signals. much of this sewing to reduce the man’s sense of insecurity. Of particular interest are the researches by Grammer [30, 3 I]. and Grammer. Kruck. and Magnusson [44] which highlighted how in an encounter between a woman and a man who don’t know each other the degree of reciprocal interest determines the increase in the degree of them synchronizing their behavior. More synchronization corresponds to a high degree of satisfaction which the man and the woman get from the relationship. Such synchronization is noway a “mirror-version.” but it has gotten rhythm. something of a courtship “dance”. Results reveal that opposite sex interaction differs substantially from same-sex interaction. In the first there is more complex pattern organization and more stereotypical. longer patterns. Communication in the early stages of courtship should take place on a level where intentions are hidden and unveiled only slowly. If allusion and deception play a role in male-female initial interactions. communication is forced to level of high ambiguity. The fact that the rhythmic patterns are hidden in “noise“ makes it difficult for the seducer to perceive actual partner interest directly. and rhythmic pattern synchronization might thus be a versatile tool to communicate without running into danger of deception. According to LaFrance and Ickes [45]. this can also explain the findings about negative relation between rapport and posture mirroring in the first step of interaction between males and females and about the irrelc\ ant role that echo and iinitation

R. Ciceri /Seductive Communication

111

behavior play in courtship [44]. These patterns of behavioral synchronization are obvious signals, too. Cappella [46] found that smiles, mutual gaze and illustrators correlated positively with synchronization ratings. This underlines the fact that interpersonal coordination is not perceivable directly from single stimuli, because the whole body is involved in the rhythmic structuring of seductive interaction. Thus making it an ideal tool for manipulation. Research by Singh [47] shows that how attractive a person is correlated to how nice the other person finds him to be. Fondness increases when two people observe a likeness or similarities in their attitudes and behavior. Grammer’s studies show that fondness is born out of the ability to coordinates one’s attitudes with those of the other person. This last case is confirmed by recent research carried out by Anolli and Ciceri [48]. Our aim was to contribute to the analysis of suprasegmental transitory characteristics (variations in rhythm, duration, pitch and intensity of speech). Having done this, our aim was then to describe any variations undergone by the persons involved in the seductive interaction from the first phase of selection (unit A) to the phase of self-disclosure (unit B) to the phase of approaching the partner (unit C). We then went on to highlight the differences between successful and unsuccessful seductive language. Generally, successful seducers (those who succeeded in obtaining a subsequent meeting from the partner) differ, from those who failed to do this, in the ability to modulate the prosodic aspects of their voice in a more flexible way during seductive interaction with their female partner (Figure 4.3). In fact, as regards pitch, successful seducers generally show more substantial changes in passing from a high modulated profile in the initial phase to a low monotone profile in the final phase of the seductive interaction, whereas in unsuccessful seducers there constantly prevails a medium value of pitch in every phase of courtship. A similar difference is also observed with the intensity of voice, because the successful seducers appear more competent than the unsuccessful ones in varymg the gradations of the suprasegmental aspects. In fact, the first ones start with a loud, variable volume but then lower the intensity of their voice remarkably in the following moments of the seductive game. Considering the indications of variability, they could be considered “variable subjects,” because they are more skilled in modifying the volume of their voice, not only in the gradual succession of the interactive moments, but also within the same moment. They can, therefore, rely more on an “orotund’ voice when the condition requires competence, confidence and enthusiasm. They are also able, however, to use a medium and low intensity of voice to transmit tenderness and warmth in the subsequent phases of the seductive interaction. On the other hand, the unsuccessful seducers appear as “stable subjects” because they consistently resort to a weak volume of voice in all the phases of courtship.

112

R ('iceri Sedirctiiv Communication

Figure 4.3. Comparison between successhi \ s. unsuccesshl seducers' \ ocai protiles

Even in the rate of articulation the successfd seducers show a higher variability than the unsuccessfbl ones, because the former are remarkably more capable of varying the rhythm of their speech in the seductive interaction, compared with the latter. Generally. the successful seducers have greater and more sensitive attention than the

R. Ciceri /Seductive Communication

113

unsuccessful ones in modulating the prosodic aspects of their voice during seductive interaction. It is not only a matter of making one’s voice attractive and warm, but also a matter of interactive competence. What is at stake here, is one of the central aspects of seductive communication: the display of unconditional attention toward the partner, making her feel exclusive in the game of role reversion, in which the one who appears to be seduced is the one who seduces. We can interpret the result of the successful seducers according to the theory of the local management of communication proposed by O’Keefe and Lambert [49]. In fact, they show they possess in all situations a stronger ability to regulate the features of their voice while interacting with the partner. It is not only a matter of planning and programming the communication of a message. It is also necessary to know how to choose the most effective move at that moment, optimize the resources available thereat, as well as turn the hints offered by the partner into opportunities. On the contrary, the unsuccessful seducers are less variable during the seductive interaction. Their quite stable voice, characterized by a generally low pitch and weak intensity, can be defined “flat” voice, with a more limited number of oscillations and variations. This type of voice seems to belong to shy and depressed people, or to people who prefer to be ignored rather than to let themselves be heard and noticed. Furthermore, the fixed nature of their vocal profiles does not appear very suitable for the seductive game, in which it is necessary “to yield” to the affective mood of the partner.

4.8

Conclusions

In conclusion, seductive interaction does not appear as a uniform and homogeneous phenomenon, but as a flexible communicative process which expresses itself in a wide variety and plurality of behavioral patterns, corresponding to the variety and plurality of communicative intentions (exhibition, approaching the partner, deepening reciprocal knowledge, and reaching of a level of intimacy). Such flexibility does not depend only on the seducer’s ability to display a wide range of communicative behavior charged with seductive power, or rather, vibes of exhibitionism and invitation such as an attractive voice, good looks, dilation of the pupils, etc. Above all, if the seductive game is to be effective then it must involve a degree of flexibility that of being able to harmonize and synchronize with the partner. Every move must be carried out and evaluated according to the partner’s response. The communicative strategies which we have described - paradoxical exhibition, obliquity, synchronization - show at least two things in common; the “saying a little but not too much” aspect and some forms of miscommunication which will both have to be looked into in more detail. First, these strategies are based on the “undefined content” of the seductive message. As we have already seen (in the descriptions of paradoxical exhibition, the forms of discursive obliquity and the multimodal message) the content of the seductive message is allusive. Its potential efficiency is because of the message is somewhat “unclear” and incomplete. It is because of this that seductive communication is tantalizing, leaving much to the partner’s imagination and promising her more than she has already seen. Moreover, the fact that the content is “undefined” allows the content itself to be adapted and modified to best suit the situation, thus lessening the risk of being too invasive and of being rejected. Secondly, it is shown how these strategies are carried out concentrating on the “undefined form” of the message, or rather, on the synergy between different expressive forms which make up the message. This allows the seducer to maintain a certain distance

R C'icrt-i Seductive Cnmmunicatioti

I I4

with the way he speaks while also enticing the partner by using body movements. Or it can allow him to emphasize and call on the emotional involvement of the other through the seductive power of nonverbal signals. Indeed, he can also convey his pleasure and interest in the partner by synchronizing and "mirroring" his behavior with hers in this complex communicative game of words. gestures and movements which involve the heart and the mind.

4.9

Acknowledgements

I would like to thank Fondazione "Piero, Pietro e Giovanni Ferrero" (Alba. Italy) for the support given to the present study. The data and theoretical framework presented in this chapter are taken from the paper: "Analysis of the vocal profiles of male seduction: From exhibition to self-disclosure". by Luigi Anolli and Rita Ciceri, to be published in The Jotrrnal of General P.s>-cho/qq*. References J. Baudrillard, De la seduction [About seduction]. Paris: Galilee. 1979. J. Umiker-Sebeok. Marketing and semiotics. Berlin: de Gruyter. 1987. R. E. Taylor. M. G. Hoy. and E. Haley. How French advertising professionals develop crzatiLc. 25 ( 1996) I - 14. strategy. Joiirnal r?/'..idr,erti.sing D. E. Hewes, The cognititv bases ofinterpersonal comntunication. Hillsdale: Erlbaum. 1995. K. Dindia, Sex differences in self-disclosure. reciprocity of selfdisclosure. and self-disclosure and liking: Three meta-analyses reviewed, in: Balancing the secrets of yr-irate di.sclo.sure.s. L E.4 't commrrnication series, S. Petronio, Ed. Mahwah: Lawrence Erlbaum Associates. 2OOO. pp. 2 1-35. D. B. Givens, The nonverbal basis of attraction: Flirtation. courtship. and seduction. Ps\rlriarc. 41 ( 1978) 346-359. D. T. Kendrick, and M. R. Trost, A biosocial theory of heterosexual relationships. in: Females. niu1e.s and sexualig.: Theories and reseurch. K. Kelley. Ed. Albany: State University of New I-ork Press. D. Morris, Manu*arching:,4 ,fieldto ltuman behavior. London: Triad Panther. 1978. D. Moms, The human seses: .-I natural histon. of'man and woman. London: Network Books. 1997. C. A. Watt. and J. H. VanLear. Eds.. i)\-namic.pa~ternin rommunic.atioti poce.sst's. Thousand Oaks: Sage Publications. 1996. J. 0. Greene. Ed.. Message producriori: ..fd1wnces i n ~~ot?it?iutiic.atic~ti t h c ~ o ~ ? .Mahwah: . Laurence Erlbaum Associates. 1997. R. Ciceri. and M . Mistri, Conoscenza procedurale [Procedural knowledge]. in: Chmunicurc~il pen.sier.o [Communicate thought]. R. Ciceri. Ed. Turin: Omega. 2001. u'. Tooke, and L. Camire. Patterns of deception in intersexual mating strategies. Erholop. untl Sociohiologt*12 ( 199 1 ) 345-364. N. Barber, The evolutionary psychology of physical attractiveness: Sexual selection and human morphology. Etholo
R. Ciceri / Seductive Communication

115

Publications, 1994, pp. 27-57. C. R. Berger, Producing messages under uncertainty, in: Message production: Advances in communication theory, J. 0. Greene, Ed. Mahwah: Lawrence Erlbaum Associates, 1997, pp. 22 1-245. ~ 3 1H. Hass, The human animal. New York: Putnam’s, 1970. 1241 D. S. Berry, Vocal attractiveness and vocal babyishness: Effects on stranger, self, and friend impressions. Journal of Nonverbal Behavior 14 ( 1990) 141- 153. A. Kendon, and A. Ferber, A description of some human greetings, in: Comparative ecology and behavior of primates: Proceedings of a conference held at the Zoological Society, R. P. Michael and J. H. Crook, Eds. London: Academic Press, 1973, pp. 34-49. R. Barthes, Fragments d ’un discours amoreoux [Fragments of a lover’s discourse]. Paris: Edition du Seuil, 1977. J. K. Burgoon, and L. Dillman, Gender immediacy and nonverbal communication, in: Gender, power and communication in human relationships, P. J. Kalbfeish and M. J. Cody, Eds. Hillsdale: Lawrence Erlbaum Associates, 1995, pp. 63-81. G. Simmel, Die Koketterie [The flirtation]. Berlin: Wagenbach, 1909/1986. W. Cupach, and S. Metts, Facework. Thousand Oaks: Sage Series on Close Relationships, 1994. K. Grammer, Human courtship behavior: Biological basis and cognitive processing, in: The sociobiology of sexual and reproductive strategies, C. Vogel and E. Voland, Eds. London: Chapman & Hall, 1989, pp. 147-169. K. Grammer, Strangers meet: Laughter and nonverbal signs of interest in opposite-sex encounters. Journal of Nonverbal Behavior 14 (1 990) 209-236. B. De Weerth, and A. Kalma, Gender differences in awareness of courtship initiation tactics. Sex Roles 32 (12) (1995) 717-734. L. B. Koeppel, Y. Montagne-Miller, D. O’Hair, and M. J. Cody, Friendly? Flirting? Wrong? In: Interpersonal communication: Evolving interpersonal relationships, P. J. Kalbfeisch, Ed. Hillsdale: Lawrence Erlbaum Associates, 1993, pp. 13-32. E. B. Mallory, and V. R. Miller, A possible basis for the association of voice characteristics and personality traits. Speech Monographs 25 (1 958) 255-260. D. W. Addington, The relationship of selected vocal characteristics to personality perception. Speech Monographs 35 (1968) 492-503. J. Laver, The phonetic description of voice quality. Cambridge: Cambridge University Press, 1980. K. R. Scherer, Inference rules in personality attribution from voice quality: The loud voice of extroversion. European Journal of Social Psychology 8 (1978) 467-487. K. R. Scherer, Vocal correlates of emotional arousal and effective disturbance, in: Handbook of’social psychophysiology, H. Wagner and A. Manstead, Eds. K. R. Scherer, Judging personality from voice: A cross-cultural approach to an old issue in interpersonal perception. Journal ofPersonafity 40 (1972) 191-2 10. M. Zuckerman, H. S. Hodgins, and K. Miyake, The vocal attractiveness stereotype: Replication and elaboration. Journal of Nonverbal Behavior 14 (1990) 97- 1 12. A. Greer, and D. Buss, Tactics for promoting sexual encounters. Journal qf Sex Research 31 (3) (1994) 185-201. F. Neto, Romantic acts in Portugal. Psychological Reports 81 (1997) 147-151. J. Crook, Sexual selection in Primates, in: Sexual selectiori and the descent of man, B. Campbell, Ed. New York: Aldine, 1972, pp. 1871-1971. K. Grammer, B. K. Kruck, and M. S. Magnusson, The courtship dance: Patterns of nonverbal synchronization in opposite-sex encounters. Journal ofNonverba1 Behavior 22 (1 998) 3-29. [451 M. LaFrance, and W. Ickes, Posture mirroring and interactional involvement: Sex and sex typing effects. Journal of Nonverbal Behavior 5 ( 1981) 139-154. J. N. Cappella, Behavioral and judged coordination in adult informal social interactions: Vocal and kinesic indicators. Journal OfPersonality and Social Psychology 72 (1997) 119-131. D. Singh, Adaptive significance of female physical attractiveness: Role of waist-tohip ratio. Journal of’ Personality und Social Psychology 65 (1973) 293-307. L. Anolli, and R. Ciceri, Analysis of the vocal profiles of male seduction: From exhibition to selfdisclosure. Journal of’Genera1Psychology (in press). B. J. O’Keefe, and B. L. Lambert, Managing the flow of ideas: A local management approach to message design, in: Communication Yearbook 18, B. R. Burleson, Ed. Thousand Oaks: Sage Publications, 1995, pp. 54-82.

This page intentionally left blank

117

Say Not to Say L. Anolli et al. (Eds.) IOS Press, 2002

5

The Meaning of Expression: Views from Art and Other Sources Jose-Miguel FERNANDEZ-DOLS, Pilar CARRERA, Cristina CASADO

Abstract. Contemporary views on facial expression take for granted that there are some basic emotions that are manifested by a few typical facial expressions. During the last twenty years psychologists have tried to demonstrate the link between basic emotion and facial expression by carrying out experiments on recognition, in which people attribute a particular emotion to a particular facial expression. Studies on recognition bear a striking similarity to those approaches to art, particularly painting and sculpture, in which the artist’s main goal consists in trying to convey a particular emotion to the observer. Interestingly enough, the patterns of representation of emotional facial expression in most of the art available in museums, from antiquity up to the beginning of this century, do not follow the patterns of emotional facial expression outlined by contemporary studies on recognition of emotions. We believe that a considerable part of this systematic inconsistency can only be explained as the result of some important and profound misunderstandings about the concept of emotional expression in current psychology. In this chapter we address these misunderstandings by discussing current research in the psychology of emotion and examples from art and scientific research.

Contents

5.5 5.6

Introduction.. .......................................................... Neo-Idealism in the study of the facial expression of emotion. The Facial Expression Program.. .................................. An alternative to the Facial Expression Program.. ............... Conclusion: Expression and lack of expression.. ................ Acknowledgements.. .................................................

118 119 121 128 130 131

5.7

References. ............................................................

131

5.1 5.2 5.3 5.4

J -.W Femandez-Doly et al

118

5.1

The Meaning of~Expression

Introduction

Since the beginnings of history, human beings have been concerned about the communication of intangible entities. An important set of intangible entities is constituted by psychological processes. Are psychological processes entities with their own independent existence? Psychologists would say that they are, but they have to come to terms with the lack of direct evidence of psychological events. Take, for example, “aggression“: in psychologcal terms, aggression cannot be defined as behavior that causes either moral or physical harm. The indiscreet but innocent remarks of a child can cause moral harm: someone who faints can cause physical harm to another person and so on. Psychologically. “aggression” is defined not by the physical, observable consequences of a particular behavior. but by the subject’s intention. People are aggressive or commit aggression if their intention is to harm others, even if they are unsuccessful. Unfortunately. “intention” is not an observable phenomenon. We can infer others’ intentions, and societies have developed complex institutions (e.g., penal systems) in order to detect and punish intentions when harm is caused to others. No psychological process is a physical, observable event, and psychologists have to guess. on the basis of physical, observable events. the shape of non-physical. unobsewable psychological phenomena. However, if psychological phenomena cannot be observed, how can psychologists become aware of the existence of a particular phenomenon? The key to this problem is language. Psychologists’ starting point for the development of a particular psychological concept is everyday language. Psychologists study “thought”, “belief’ or “desire“ (hiding these common terms behind more respectable, academic labels. such as “cognition“. “representation” or “motivation”) because people, in their own cultures. claim that they have thoughts. beliefs or desires. This starting point is fine as long as we are prepared to pay the price it bears. Building a scientific discipline following everyday language creates important tensions and problems. One of these problems is the cultural relativity of a scientific discipline whose main concepts are taken from a cultural, arbitrary system of representation. Can psychologists aspire to developing universal laws? Or are their findings restricted to the language in which they think and talk about psychological “reality”? Research on emotion is one of the fields of research in which this problem is most salient. Emotion is an elusive concept that even defies a consensual definition. For more than one hundred years [ l , 21 researchers have tried to isolate the essence of emotions. their components and their behavioral manifestations. but we are still far from attaining agreement in this respect. Why is emotion even more problematic than other psychological concepts? Because emotion belongs to a realm of phenomena -such as mood, affect. feeling-that are particularly opaque to language. People can talk about their (true or false) memories or they can describe (with varying degrees of success) what they are perceiving, but they find it difficult to describe the emotion they are feeling without recourse to a more or less clear label. Emotion challenges language, and human beings have had to craf? allusive. connotative Lvays of describing the indescribable. However, the opacity of emotions would not be such a problem were it not for a certain troubling feature: fkom a psychological point of view. “emotion” is synonymous with “important”. As Silvan Tomkins points out, important things cause emotion and emotion confers importance on things. Art, and particularly music and poetry. are important because they are. somehow, describing or manipulating emotions. Other common linguistic resources. such as insults. habe --apart from their pragmatic functions the semantic function of

J.-M. Fernandez-Dols et al. / The Meaning of Expression

119

connotatively expressing the importance of our own (negative) emotions while playing on others' emotions. Psychologists have tried to deal with the opacity of emotion language in many ways. Interestingly, their search for a consensual definition of emotion bears a telling resemblance to the old philosophical discussion about the relationship between language and reality. As in philosophy, alternative solutions to the relationship between everyday talk about emotion and the concept of emotion are based on varying assumptions about the ontological status of emotion. All of these assumptions can be described as lying along a continuum between two extreme ontological perspectives. At one extreme, which could be labeled as Idealism (or Platonic realism), emotions are universal, real entities that can shape what people say about emotion but are independent of language. At the other end, which could be labeled as Nominalist, emotions are mere labels, i.e., words that lack a universal, real denotation, and summarize the sameness of some psychological events [3]. Psychological events labeled as emotion are specific and idiosyncratic; there are no universal realities such as emotions. The study of the facial expression of emotion has been linked, over the last thirty years, to a rather radical, extreme Idealist position. Its assumptions have blinded psychologists to many aspects of facial expression that are extremely interesting and should be urgently considered. In this chapter we describe the evolution of facial expression within this Idealist approach, the main problems that arise from this view, and an alternative view on the facial expression of emotion that emphasizes (a) that facial expression is an icon, and (b) that facial behavior is an event. This approach to facial expression concludes that actual facial demonstrations of emotion are, in fact, better defined as the lack of those expressions usually attributed to emotion by current mainstream psychology.

5.2

Neo-Idealism in the study of the facial expression of emotion

From a philosophical point of view, the belief in universals is a basic feature of Idealism. Modem Idealism is also connected to Romanticism, a movement that rejects rationality and claims a role for emotions and the transcendental in the understanding of human beings. In our view, the search for universals in emotion must be discussed in relation to these intellectual doctrines. The common use of words such as "Idealism" or "Romanticism" suggests an abstract, somewhat impractical way of approaching human actions. Nevertheless, Idealism and Romanticism have been, and probably still are, two of the most influential approaches to practical dilemmas of the last one hundred years, competing with the social sciences and, particularly, disciplines such as social psychology. Whereas social psychology emphasizes Situationism, as an a-historical tension between immediate situational constrains and cognitive construals [4], an Idealist and Romantic psychology has emphasized universalism as a dialectical relationship between transcendental collective or individual objective trends and subjective manifestations of these transcendental collective or supraindividual trends. The Idealist and Romantic approaches have been a source of inspiration for remarkable achievements of modem art, but also for most of the political regimes that, throughout the twentieth century, considered human beings as disposable pieces in the attainment of ''historical'' destiny. In fact, for most of the twentieth century, art and literary criticism tried to separate emotion from psychology in the public mind; one of the legacies of Romanticism was the popular assumption that psychologists should not study such intangible phenomena as love, happiness or hate. This attitude contrasts with the lasting

I20

J - bt Fernlindez-Doly et al

The Meaning of Erprewon

prevalence of emotions in key social phenomena, such as politics (e.g., nationalist doctrines, terrorism, the Cold War), the mass media (e.g.. the commercial movie industry), religon, and even economics (e.g., through key concepts such as “investor confidence” or “stock-market panic”). The Romantic view of emotions was led by Silvan Tomkins, a psychiatrist with a wide range of interests and a formidable intellectual curiosity. Tomkins’ style was Romantic in form and content; his approach to problems was a sort of large-scale Fresco on which his theoretical reviews could span not just hundreds but thousands of years [5]. Tomkins‘ approach to emotion is reminiscent of Hegelian dialectics: subjective affects or emotions are innate, supra-historical phenomena whose material counterparts are objective human faces: the dialectical synthesis is absolute, “basic” facial expressions. In Tomkins’ words, “the affects are the primary motives of man, and if, as I also believe, the face is the primary site of the affects, then the face is the man” [ 5 ] . Tornkins’ disciples. Carroll Izard [6] and Paul Ekman [7], developed an empirical version of Tomkins’ approach by linking up the idea of universal subjective affects with Darwin‘s work on facial expression. In 1872 Darwin published The Expression qf the Emotions in Man and Animals, a minor work aimed at showing that not only morphological features but even behavior is a result of evolution and hereditary transmission. Tomkins’ disciples read Darwin’s approach as a psychological theory IS]. but neither the context nor the content of Darwin’s book suggests that this was the case. Darwin’s book belongs to a nineteenth-century tradition of books on human and animal anatomy that extended the descriptive field of this medical discipline; this tradition was a sort of scientific, medical version of the seventeenth and eighteenth-century tradition of academic treatises on the pictorial representations of passions in painting. exemplified by Le Brun’s Confirence stir I’expression [9]. Le Brun designed his expressions in a deductive way. without empirical support, from a set of theories -based particular on Descartes’ ideas. but also on those of Aristotle and Quintilian- about the relationship between body and soul. Le Brun’s approach gave rise to a set of rules about the “right” expressions that later inspired Darwin’s psychological approach to expression Whereas Le Brun’s method was rationalist, and his goal was to organize as a standardized set of norms the old conventions on the symbolic representation of human passions, Darwin’s method was empirical, and his aim was to describe the rich anatomical bases of some muscular movements vaguely described, in colloquial English. as emotional. Darwin also had another immediate goal: to criticize the philosophical assumptions of another contemporary medical book on the anatomy of facial expression -Charles Bell’s .4natont\* and Philosophv of Expression. Bell believed that facial muscles were created by God exclusively for expression [lo]. Darwin [ l l ] tried to criticize Bell’s views, which underpinned his contemporaries’ mainstream approach, arguing that “man is derived from some lower animal form” because facial expressions, as muscular movements, can be observed in all human beings, and there is “to a certain limited extent” a physiognomic continuity between the expressions of animals and human beings. For Darwin. God should be substituted by Evolution. Darwin developed no psychological theory of emotion in his book. “Emotion” was used as a synonym of “expression”, and his careful and insightful description of the expressions of animals and humans contrasts with his vague and conventional list of feelings such as meditation, hunger, determination or love; his empirical meticulousness in the study of expression contrasts with his list of emotions drawn from Le Brun and other rationalist sources. Darwin’s definitions of his famous three Principles of Expression make no mention of emotion. and he concludes his book by inviting “further attention“ from an^ able p h i*c.ioloqivf”.

J.-M.Fernandez-Dois et al. / The Meaning of Expression

121

Apparently, Psychology was not on Darwin’s agenda, but the Romantic approach to his work gave a new emphasis to emotion over expression, turning his observations into a quest for transcendent, universal emotions. Darwin’s finding of similar expressions across cultures, ages, and even species became a finding of similar emotions across cultures, ages and species, while his studies on the recognition of expressions became studies on the recognition of emotions. Darwin’s empiricist, materialist approach became the Romantic search for universal, intangible entities (basic emotions) in human faces. 5.3 The Facial Expression Program

The search for universal emotions as facial expressions has shaped contemporary research on emotion. In 1971 Carroll E. Izard published the first version of his Differential Emotions Theory [ 121. In 1972 Paul Ekman established the bases of the Neurocultural Theory [ 131. According to these models, developed over the last thirty years [ 141, facial expressions are part of a broader, universal phenomenon known as “basic emotions”. The main features of these influential models, and the assumptions they have inspired since then, have been summarized by Russell and Fernandez-Dols as the “Facial Expression Program” (FEP) [ 151. Some of the FEP assumptions are as follows: There are a small number (seven plus or minus two) of basic emotions. Each basic emotion is genetically determined, universal and discrete. Each basic emotion is a coherent pattern of facial behavior, experience, physiology and instrumental actions. The production and recognition of distinct facial expressions of basic emotion is the result of evolution. Any state lacking its own facial expression is not a basic emotion. There is consensus on the existence of six basic emotions: happiness, surprise, fear, anger, disgust and sadness. Some authors have also proposed contempt, shame and interest. All emotions other than the basic ones are subcategories or mixtures of the basic emotions. Expressions of emotion are spontaneous. Voluntary facial expressions can simulate spontaneous ones. Different cultures establish different display rules. Display rules inhibit, exaggerate or mask spontaneous expressions. Any facial expression that deviates from the universal expressions of basic emotion is a mixture of universal expressions or a voluntary display. Emotional state is revealed by facial measurement. Observation or electromyography can detect the muscular movements linked to facial expressions of basic emotion. If emotional states can be described by facial measurement, verbal reports of emotion can be bypassed. The true criteria for the existence of a basic emotion are to be found in the subjects’ facial movements. The expressions of basic emotion are easily recognized by all human beings. The ability to recognize an expression of basic emotion is innate rather than culturally determined. The mental categories by means of which recognition occurs are genetically rather than culturally determined. The meaning of a facial expression of basic emotion is invariant across changes in

J.-M. Femandez-Dols et 01.

:The

Meaning o f E.rpr-ession

the context in which it is produced. The “recognition of emotions” in contemporary psychology bears an unfortunate similarity to the Renaissance enterprise of the Neo-Platonic painters that tried to elicit the immediate “recognition” of some intangible entities (e.g., virtues) in observers of their art. Gombrich [ I61 has suggested, for example, that Botticelli’s mythological paintings were educational programs that included a kind of “moral agenda” drawn up by a humanist intellectual, probably Ficino. Venus becomes, in this view, a tangible expression of humanitus, a moral virtue that includes all the desirable traits of a Renaissance gentleman. The contemplation of Venus was a direct contemplation of humunitus. so that paintings became an active and decisive educational device. Like the Neo-Platonic Renaissance painters and scholars, the FEP followers have assumed that some facial expressions ure universal emotions that will be recognized by human beings irrespective of their culture, epoch or previous experience. This fascinating parallelism is even more evident if we consider the problems the FEP has encountered. What is the evidence for the universality of facial expressions of basic emotion? Testing the universality hypothesis requires tests of at least three different propositions [IS]: -

The same patterns of facial movement occur in all human beings. Observers in different societies attribute the same specific emotions to these universal expressions. These expressions are manifestations of those same emotions in all human beings.

5.3.1 Patterns of’facial movenienr The first proposition (the same patterns of facial movement occur in all human beings) has always been considered obvious, even though empirical research on the matter has been scarce or non-existent among psychologists. In fact, there are no empirical data on the distribution of particular facial movements across different populations. Eibl-Eibesfeldt [ 171 is probably the author that has come closest to a description of facial movements across cultures. He has found some striking similarities across cultures. such as the ”eye-brow flash”. as an expression of social contact readiness, or the smile, as an expression of friendly contact readiness. Nevertheless, there is no “incontestable classification of facial expression either on the basis of motivation or. alternatively. on the basis of function” [p. 4461. and some data raise interesting questions for the hypothesis of universal -facial moi*ements: facial musculature shows considerable ethnic and individual variation [p. 46 1 1. Furthermore, there is also some intriguing data on the prevalence of other patterns of facial movement in particular circumstances. substituting the patterns of expression assumed by the FEP. For example, Garcia-Higuera [ 181 found that Spanish bullfighters, while in the ring with the bull, display a strikingly uniform pattern of facial movements that have nothing to do with the expressive movements described by the FEP, even though the situation is obviously intense in emotional terms: a man and an animal fighting for their lives according to an extremely primitive, brutal script. Garcia-Higuera found a specific and idiosyncratic pattern of expressions of intense happiness that has nothing to do with smiling: when the bullfighters were “in the flow”. absorbed in the fight, having been successful in deceiving the animal, they raised the chin while funneling the lips, an expression that corresponded to their ou’n reports of intense. unique happiness (see Figure 5 . 2 )

J.-M. Fernandez-Dols et al. / The Meaning of Expression

123

On the other hand, the FEP prototypical expressions of happiness (see Figure 5.1) were not particularly frequent in the sample of amateur and professional bullfighters he studied (see Figure 1). Garcia-Higuera’s findings suggests that the prevalence of some particular facial movements may be less universal than it seems to Western psychologists; the “universal expressions” are anatomically possible for any human being, but they are not necessarily displayed, and we do not know whether they were even the most prevalent at any time in human evolution.

Figure 5.1 Source: A. Bauerfeind (1982) Mimische Schemata Munich: Max Plank Institut f i r Psychiatrie

For example, the bullfighters’ expressions perhaps were, and still are, more prevalent forms of expression for those that have to take risks in a physical fight; in fact, Grant [ 191, Blurton-Jones [20] and Camras [21] observed similar expressions in children in agonistic contexts. Finally, a cursory analysis of Western culture throughout history suggests that expressions of basic emotion have not constituted such a common pattern of facial movement as the FEP suggests. We systematically codified all the paintings in the catalogue of the Prado Museum in Madrid (Spain) [22] that depicted events. After excluding 716 paintings (portraits, still lifes, etc.) whose subject had no obvious emotional denotations, we selected 135 works painted between the fourteenth and nineteenth centuries- that clearly had an intensely emotional subject and portrayed visible faces. The emotional content and facial action of the pictures were rated by three judges, two of them experts in FACS [23], an observational system that identifies muscular actions through changes in facial appearance. Interestingly, 75 (56%) of these paintings displayed neutral expressions. The remaining 60 (44%)were coded by the judges in terms of the emotion linked to the painting’s theme according to the standards of contemporary observers. In these 60 paintings, 25 were categorized as portraying

124

J -M.Femandez-Dols et a/. ’ The Meaning ofExpression

a situation in which happiness or joy was felt, 15 as representing a situation in which sadness was felt, 14 as corresponding to a situation in which fear was felt, and 6 as portraying a situation in which anger was felt. Out of these 60 “emotional paintings” we found only six facial configurations that could be described as prototypical expressions of emotion according to the FEP; 56% of the sample of these 135 emotional-content paintings show neutral faces.

Figure 5.2 Source: A. Bauerfeind ( 1982) Mimische Schemata Munich: Max Plank Institut f i r Psychiatrie

The distribution of neutral and expressive faces across the centuries did not show significant differences (Chi-square(5) = 4.44, p>.05). With respect to the six paintings that included prototypical expressions, one fiom the sixteenth century (Patinir’s “The Temptations of St. Anthony”) and four from the seventeenth century (Velhquez’s “Vulcan’s Forge”; Rubens’ “Saturn devouring one of his sons”; Jordaens’ “The Fall of the Giants”; and Teniers’ “The Jolly Soldier”) all showed prototypical expressions of surprise. The exception was an expression of happiness with Duchenne smile (in Velkquez’s “Triumph of Bacchus”). In summary, we may ask ourselves why almost no artist, over six centuries, was capable of perceiving the facial movements linked to basic emotions, even though, according to the FEP, they are the universal, explicit and visible signals of emotion, and people can naturally recognize these expressions because they correspond to certain innate perceptual categories ~41.

5.3.2 Attribution of emotion The FEP claims that observers in different societies attribute the same specific emotions to

J.-M. Fernandez-Dols et al. / The Meaning of Expression

125

expressions of basic emotion. Figure 3 summarizes a large number of cross-cultural studies on the attribution of basic emotion to prototypical expressions. The standard method of these studies, usually called “recognition studies”, consists in showing subjects a set of pictures of prototypical expressions (such as those in Figure 5.1) and asking them to assign a category of basic emotion (e.g., happiness, sadness, anger, fear, disgust or surprise). Russell [25] has raised some methodological questions concerning such recognition studies. In Figure 5.3 the bars show the percentage of subjects that chose the expected category (e.g., happiness for a smiling face). The set of bars on the left comes from Western literate societies, the second set from non-Western literate societies, and the third set from non-Western illiterate societies. Even though the ratings of recognition are significantly high across samples, they are proportional to the subjects’ exposure to Western influence. Furthermore, they may be inflated due to the method of these experiments (e.g., the experimenters did not usually speak the subjects’ language in crosscultural studies; most of the studies used forced-choice response format; etc.) [25].

Figure 5.3 Recognition rates for six facial expressions of emotion [25]

The recognition of basic emotion in facial expressions can also be criticized from other theoretical points of view [26]. “Recognition” of emotion in the FEP experiments is not necessarily a Neo-Platonic re-representation of universal entities already known by the individual, but rather a mere process of attribution, in which people link some emotions to some facial patterns using lay explanations that lack any necessary or sufficient relationship to the actual experience of emotion and its behavioral consequences. According to this attributional hypothesis, people do not recognize emotions; they simply make a kind of illusory correlation: salient facial patterns and emotions, not

126

J -.MFemdndeez-Dols et a1

The Meaning of E.rpression

necessarily produced simultaneously but observed within the same situation, are lumped together into an ideal and unrealistically consistent representation of “‘emotional expression”. For example, people may frown when they threaten someone, and most people are angry bejore they threaten someone. A realistic description of the situation would say that a person is angry and, afterwards, s h e frowns when s h e threatens another (not necessarily the person that elicited hidher anger). An ideal description would say that people fiown when they are angry. Researchers have carried out experiments in which they modified the usual procedures of recognition studies, finding dramatic changes in the universal pattern of attribution of emotions. For example, Fernandez-Dols and Russell [27] explored this hypothesis by giving their judges a chance to perceive the relationship between emotion and expression as embedded in an event in which prototypical expressions and emotional experiences could happen either sequential(i7 or simultaneousli>. Stimuli were eleven pairs of faces (a neutral and a prototypical expression of basic emotion in each pair) and eleven pairs of sentences. Each pair of faces was displayed by the same model and included a prototypical expression of emotion. The two corresponding sentences represented the model’s experience and interaction at two different times of the event in which the expressions were displayed: a first point in time. at which the model was having an emotional experience (e.g.. “So, he harmed my son!”), and a second point at which the model was talking to someone as a consequence of the emotional experience (e.g., “You harmed my son!”). Using the graphic conventions of comics, the sentences were shown inside balloons, the first one represented as a thought and the second one as speech (see Figure 5.4).

Figure 5.4 The first balloon represents thnrrght and the second one represents YpeecIi

Judges were asked to assign the sentences to the neutral and the prototypical expression in the “most plausible” or “most natural” way. The prototypical facial expression was not associated with the point in time at which the models were having the emotional experience, but with the time in which the models were talking to someone afterwards (8 1Yo). In another experiment, Fernandez-Dols and Carrera [28] camed out a more direct test of the hypothesis about people generating an illusory correlation of expression and emotional experience within an ideal expression of emotion. In two studies, adults and children were asked to judge videos and slides of children undergoing a frightening experience (having a vaccination). Even thouzh none of the children shown in the videos and slides displayed

J.-M. Fernandez-Dols et al. / The Meaning of Expression

127

prototypical expressions of fear, adults were particularly prone to “recognize” prototypical expressions in these stimuli. For example, in the first study, 80% of the adults claimed they saw prototypical expressions of fear, whereas only 40% of the 6 and 7-year-old children reported seeing prototypical expressions of fear. These studies show that “recognition of emotion’’ may be explained as the result of an over-generalization based on schemas which, like those of other intuitive theories (e.g., stereotypes about gender), do not necessarily provide an accurate description of a person’s behavior (in this case behavior when feeling an intense emotion). These differences between adults and young children are reminiscent of the above-mentioned differences in recognition between literate and illiterate cultures shown in Figure 5.3. Attribution with regard to facial expressions is not, according to these and other data [29], a mere recognition of emotions, but rather a much more complex process in which lack of facial expression can, in particular contexts, be as informative as intense facial displays. In the study with paintings referred to above, more than one third of the Prado paintings with an intense emotional subject displayed neutral faces. Carrera and Fernandez-Dols [30] carried out an experimental test of the emotional meaning of neutral faces in the context of recognition tasks. They found that neutral faces can be strong sources of information when they interact with particular kinds of emotional contextual information: for example, neutral faces were more influential than prototypical expressions of emotion on judgments about the message of a facial display combined with an unusual, but clear, emotional situation. In other words, faces are intrinsically influential, irrespective of their emotional content, and interact in complex and as yet unknown ways with the context in which they are displayed [29]. 5.3.3 Actual expressions of emotion

The FEP claims that prototypical expressions are manifestations of the same emotions in all human beings. Fernandez-Dols and Ruiz-Belda [3 11 reviewed the evidence supporting this claim, in addition to that provided by the already-discussed studies on recognition. The first conclusion of this review was that studies on the actual spontaneous expression of adults during intense emotional episodes is surprisingly scarce. The most cited study in support of the FEP assumption was carried out by Ekman [ 131 and Friesen [32] at the beginning of the 1970s. They asked American and Japanese students to watch one neutral and three stressful films. Samples were taken of the facial behavior of each subject during the neutral and the final stressful film. Researchers discussed their findings in terms of a global correspondence between the American and Japanese students [32]-, but they provided no data of the precise correspondence between the specific emotion experienced by the students and their facial expression at the time. The experimental studies that have tried to code facial behavior while obtaining reports of subjects’ moment-by-moment emotional feelings have not provided evidence of a clear correspondence between emotional experience and facial expression. Rosenberg and Ekman [33] found a correspondence only between facial expressions of disgust and reports of intense disgust, and only at those moments at which the most intense expressions and the most intense reports coincided; no data were provided on which other faces coincided with intense reports of disgust, or on how many expressions of disgust coincided with no report of disgust. Fernandez-Dols, Sanchez, Carrera, and Ruiz-Belda [34] did not find even a partial correspondence between facial expressions and reports of negative emotion from people watching a clip from a horror film (see also Reisenzein) [35]. Field studies on the actual behavior of people in natural encounters in which researchers can infer their emotional state have not obtained better results for the hypothesis of a

128

J.-M. Fernandez-Doh et al. The Meaning of‘Expression

correspondence between facial behavior and emotional experience. For example, Kraut and Johnston [36] and Fernandez-Dols and Ruiz-Belda [37] found that happy people (e.g.. hockey fans after an event favorable for their team, gold medal winners at the Olympic Games) did not necessarily smile at happy times, but rather when they were interacting with others. The above-mentioned sample of Prado paintings produced only six prototypical expressions of emotion (five of surprise and one of enjoyment) as described by the FEP, from 60 paintings. Even these few prototypical expressions did not necessarily represent -in the painter’s mind- the basic emotion predicted by the FEP. Contemporary judges judged only three paintings as portraying a situation congruent with the basic emotion attributed by the FEP to their prototypical expressions (“Vulcan’s Forge” as surprise and “Triumph of Bacchus” as enjoyment); the other paintings were judged as representing situations in which fear (“Saturn devouring one of his sons”; “The Temptations of St. Anthony”: “The Fall of the Giants”), or happiness (“The Jolly Soldier”) were felt. Of course, judgments by contemporary observers would not be expected to coincide with the intentions and conventions of the artists and their original public (is martyrdom a happy, frightening or sad event for the martyr him/herself?), but the lack of any prototypical expression over several centuries cannot be explained as simply a consequence of lack of technical skill on the part of all the painters included in this sample. Movement and perspective may involve difficult technical problems, but the static representation of at least some of the facial expressions (e.g., open smiling) cannot be considered as too great a challenge for most of the artists whose work is included in the sample. In sum, there is a surprising lack of definitive data on the actual correspondence between the experience of some particular emotions (e.g.. happiness. sadness. etc.) and their corresponding facial behavior as predicted by the FEP.

5.4

An alternative to the Facial Expression Program

The above-mentioned review suggests that the FEP has been only partially successful in the Idealistic task of establishing a link between intangble, universal entities -the basic emotions- and some patterns of facial behavior. A careful description of the patterns of facial movement and attributions about these movements suggests that the relationship between facial behavior and emotion is still an unsolved mystery; nevertheless, we believe that history, and particularly the history of art. could again provide the key to understanding. Artists, over many centuries, have not resorted to the prototypical expressions claimed by FEP because painting emotion consists not in painting a particular, specific expression, but rather a pattern of movement in which no isolated gesture has a particular meaning outside of its context and sequence. In this vein, neutral faces may be much more readable as emotion than any expression. Expressions can become grotesque, unnatural p m a c e s rather than balanced representations of facial behavior linked to an intense emotion; on the other hand, neutral expressions can serve as representations of particular sequences of facial behavior. For example, the painter chooses the neutral face as the departure point for all the future actions entrained by a particular context. Interestingly enough. most of the smiles found in the works from the Prado are displayed by vulgar, drunk or crazy models. or by children. Smiling was not linked, as it is today, with beauty: a fixed. open smile was a sign, not of happiness. but of simpleness. The lesson we learn from these paintings is that expression is an interactive action ”entrained” by the context. As Ginsbur2 and Hamngton 1381 have pointed out. emotion is

J.-M. Fernandez-Dols et al. / The Meaning of Expression

129

action in context. The context “feeds forward” the sequence of facial action, establishing the path for the future expressive action. The link between expression and emotion is only indirect, and mediated by the motives and strategies linked to a particular sequence of actions. Thus, the emotional expression takes place within the context of a broader system of events and social relationships that are necessary for giving emotional meaning to any facial behavior. In fact, authors are beginning to discuss their findings, as painters did for centuries. Some authors [39] conclude that expressions of emotion are mediated by social or interactive factors, while others [40, 411 have suggested that facial expressions are produced not by emotion but for social reasons, or as a result of other psychological processes not necessarily linked to emotion. In any case, the future of the study of facial expression should take into account that facial expressions are not objects or concepts but events. Research on facial expression of emotion, like research on many other psychological processes, should pay more attention to the difference between objects and events [42]. Objects persist across time, whereas events last for a specific period of time. In our view, any facial expression is a unique event. People can categorize events but they cannot recognize a particular instance of an event because a given event can only be experienced once. Therefore, people process events guided by concepts. Facial expressions, then, are episodes that are experienced only once; concepts help to process these episodes in terms of emotions, motives or other sources of information. Nevertheless, the concepts that people use for making these categorizations do not necessarily provide an objective explanation of the causes and dynamics of facial expressions. In other words, people may attribute some emotions to some facial expressions because, for centuries, a particular culture has linked certain facial expressions to certain emotions, or because people like to explain behavior as a consequence of emotion, or for other reasons. Nevertheless, this consensus on the attribution of expressions to emotions does not necessarily mean that facial expressions are expressions of emotion. Again, an example from art can help us to understand this point: the failure of the neo-Platonic program -in which Renaissance painters tried to represent the actual expression of particular virtues or passions- gave rise to an alternative approach to the representation of abstract or invisible objects in art. This alternative view, of Aristotelian inspiration, approached expression in terms of “allegory”, using a complex language in which expressions might be considered not as indices but as icons of emotions. Following Peirce’s classic typology [43] of signs, there are codes made of symbols, indices and icons. A symbol bears an arbitrary relationship to its object (e.g., the word “gato” is the Spanish symbol for a small domestic feline we might see sleeping in our backyard), an icon has a relationship of similarity to its object (e.g., a painting of a woman, as an icon of woman), and an index is physically connected to its object (e.g., smoke as an index of fire). The FEP claims that a facial expression is an index of emotion, i.e., a substantive aspect of the emotion itself. Our hypothesis is that facial expressions are icons of emotion. As iconic allegories of emotion, facial expressions are neither causally nor hnctionally linked to the experience of intense emotion, but they can convey the experience of emotion with greater or lesser success, depending on their “readability” for observers. An expression can “allude” to some emotions for different reasons, even though it is not produced by the sender’s emotional state. Furthermore, an iconic view of facial expressions interprets facial behavior as a movement that cannot be frozen into a particular prototypical expression: “expression” cannot be summarized as a particular facial configuration. Static expressions are just

130

J - M Fernantic-Dols et ul

The Meuning of Erpression

conventional devices to suggest movement or, in psychological terms, interaction. This view allows us to understand the above-mentioned results of Camera and Fernandez-Dols, in which neutral faces were as communicative as static prototypical expressions, as well as the lack of expression described for most of the paintings analyzed in the Prado catalogue. The representation of expression is a representation of a sequence of movements. and m y part of the sequence can be as representative (or not) as the others. depending on the context. Representing movement is a complex task that requires more than just a short. invariable list of icons.

5.5 Conclusion: Expression and lack of expression To summarize, facial expression is (a) an icon and (b) an event. As an icon. facial expressions bear a relationship of similarity to emotions but they are not necessarily emotions. Goya's Mujus are icons of a woman but they are in no way a woman or part of a woman. The similarity of an icon to its object does not depend on the similarity between icon and object; in fact. in many cases a convincing icon has nothing to do with its object. The canvas, the colors, the drawn lines, etc. have practically nothing to do with the flesh. shape and features of a human being; the similarity between icon and woman comes from a conceptual process in the observer who is capable of reading, in a particular culture. a particular kind of bi-dimensional representation know as "a painting". The similarity of an expression to an emotion is also conceptual, and involves a complex process in which an event is processed as a static grimace that is not even necessarily present at the time at which the actor is feeling a particular emotion. Fridlund [41]. for example, claims that facial expressions usually attributed to emotions are probably linked to social motives associated with episodes in which people feel particular emotions. For example. smiling is an expression of affiliation but not of happiness: the link between smiling and happiness may be due to the fact that happy people are usually more sociable: crying is a sign of helplessness. but helpless people are usually (but not necessarily) sad, and so on. As events, facial expressions are flows of behavior. Expression is movement, and no particular pattern of muscular movements can be described as the "true" expression of emotion. The recognition of emotion through static expressions might be an experimental artifact which. as we described above. could easily be deconstructed by giving subjects the conceptual tools for approaching facial behavior as a sequence of movements in context. In this sense, plain faces are as informative as grimaces when someone -e.g.. a skilful painter- has to translate the sequence of movements into a single, static frame. Our research program is accumulating data that suggest that "expressions" of emotion. as far as their current mainstream definition in psychology textbooks is concerned. are mostly cultural products. The sixteenth-century endeavor of Le Brun and others established a shared code for reading facial expression. followed first by the intellectual elite (e.g., Darwin), and subsequently by most of Western culture (e.g.. smiles did not begin to appear in portrait pictures. as a necessary rule of self-expression. until u.ell into the twentieth century). On the other hand, actual facial behavior linked to the experience of intense emotion. and carefully analyzed frame-by-frame, appears as a disorganized, complex and fascinating event. in which large numbers of facial actions -usually not including the prototypical expressions claimed by FEP- are displayed. This facial behavior is the true expression of emotion. but it cannot he summari7td w.ithin a single facial contipration. a s the

J.-M. Fernandez-Dols et al. / The Meaning of Expression

131

proponents of FEP would argue. “Facial expression of emotion” is usually understood as a single, static, and precise facial configuration, but we would suggest that the real, true “facial expression of emotion” is neither single, nor static, nor precise. Indeed, actual facial expressions of emotion should be defined as lack of the expression. 5.6 Acknowledgements

This chapter was partially supported by a grant (PB98-071) from the Spanish DGICyT. Correspondence should be addressed to J. M. Fernandez-Dols, Facultad de Psicologia, Universidad Autonoma de Madrid, 28049 Madrid, Spain or, via Internet, to [email protected]. We thank David Weston for his help in the preparation of this manuscript.

References W. James, Reprint of The Physical Basis of Emotion, Psychological Review, 101 (1994) 205-210. P.C. Ellsworth, William James and Emotion: Is a Century of Fame Worth a Century of Misunderstanding?, Psychological Review, 101 (1994) 222-229. J.M. Fernandez-Dols and J.A. Russell, The Universal and the Cultural in Emotion and Affect, in Comprehensive Handbook of Psychology, vol. 9: Personality and Social, T. Millon and M.J. Lerner Eds., manuscript in preparation. L. Ross and R.E. Nisbett, The person and the situation. Perspectives of Social Psychology. New York: McGraw Hill, 1991. S.S. Tomkins, The Phantasy Behind the Face, Journal of Personality Assessment, 39 (1975) 551-562. C.E. Izard, Innate and universal facial expressions: Evidence from developmental and cross-cultural research, Psychological Bulletin, 115 (1994) 288-299. P. Ekman, Facial expressions of emotion: New findings, new questions, Psychological Science, 3 (1992) 34-38. P. Ekman, Biological and cultural contributions to body and facial movement, in Anthropo1oc.y qfthe body, J. Blacking, Ed. London: Academic Press, 1977. J. Montagu, The expression of the passions, Yale University, 1994. C. Bell, Anatomy and Philosophy of Expression, 1844. C . Darwin, The Expression of the Emotions in Man and Animals, Chicago: Chicago Press, 1965. C.E. Izard, Theface of emotion, New York: Appleton-Century-Crofts, 1971. P. Ekman, Universal and cultural differences in facial expressions of emotions, in Nebraska Symposium on Motivation, Lincoln: University of Nebraska Press, 1972. P. Ekman, An argument for basic emotions, Cognition and Emotion, 6 (1992)169-200. J.A. Russell and J.M. Fernandez-Dols, What does a facial expression mean? in The psychology ?f jacial expression. Cambridge UK: Cambridge University Press, 1997, pp. 3-30. E.H. Grombrich, Symbolic Images. Studies in the Art of the Renaissance II, London: Phaidon Press, 1972. I. Eibl-Eibesfeldt, Human Ethology, New York: Aldine de Gruyter, 1989 J.A. Garcia-Higuera, Conductas faciales en situaciones de dominancia, unpublished doctoral dissertation, Universidad Autonoma de Madrid, 1996. E.C. Grant, Human facial expression, Man. 4 (1969) 525-536. N.G. Blurton-Jones, Criteria for use in describing facial expressions of children, Human Bioloxv, 43 (1971) 365-413. L. Camras, Facial expressions used by children in a conflict situation, Child development, 48 (1977) 1431-1 435. Fundacion amigos del Museo del Prado y Fonds Mercator Paribas, Catalogo, 1996 P. Ekman and W.V. Friesen, Facial Action Coding System (FACS): A technique.for the measurement offacial action, Palo Alto, CA: Consulting Psychologists Press, 1978. P. Ekman. ExDression and the nature of emotion. in Amroaches to emotion. K.R. Scherer and P.

132

J.-M. Fernandez-Dols et al. The Meaning of Expression i

Ekman, Ed. Hillsdale, NJ: Lawrence Erlbaum. 1984. J.A. Russell, Facial Expressions of Emotion: What Lies Beyond Minimal Universaliq. P.yvchologica1 Bulletin, 118 (1995) 379-39 1. J.M. Femandez-hls, Facial expression and emotion: A situational view. in The social context of' nonverbal behavior, P. Philippot. R.S. Feldman and E.J. Coats, Eds. Cambridge UK: Cambridge University Press, 1999. J.M. Fernandez-Dols and J.A. Russell. Are facial displays social or emotional'?. unpublished manuscript. J.M. Fernandez-Dois and P. Carrera, The recognition of emotion "error", poster presented Annual Meeting of the Society for Personality and Social Psychology, 200 1 . J.M. Femandez-Dols and J.M. Carroll, Context and meaning, in The psvchologv of.facia1e.rpressiort. J.A. Russell and J.M. Fernandez-Dols. Eds. Cambridge UK: Cambridge University Press. 1997. pp. 2 75-294. P. Carrera and J.M. Fernandez-Dols, Neutral faces in context: Their emotional meaning and their function, Journal ofhronverbal Behavior. 18 ( 1994) 281 -299. J.M. Femandez-Dols and M.A. Ruiz-Belda. Spontaneous facial behavior during intense emotional episodes: artistic truth and optical truth, in The psycholog.l*qfjucial espression. J.A. Russell and J.M. Fernandez-Dols, Eds. Cambridge UK: Cambridge University Press. 1997. pp. 255-274. ~ 3 2 1 W.V. Friesen, Cultural differences in facial expression in a social situation: An experimental test of the concept of display rules. Unpublished doctoral dissertation, University of California. 1972. 1331 E.L. Rosenberg and P. Ekman, Coherence between expressive and experiential systems in emotions. Cognition and Emotion, 8 ( 1994) 20 1-229. ~ 3 4 1 J.M. Fernandez-Dols. F. Sanchez. P. Carrera and M.A. Ruiz-Belda, Are spontaneous expressions and emotions linked?. An experimental test of coherence. Journal of.Vonrwha1 Behai.ior, 21 ( 1997) 163177. R. Reisenzein. Exploring the strength of association between the components of emotion syndromes: The case of surprise. Cognition and Emotion, 14 (2000) 1-38. R.E. Kraut and R.E. Johnston, Social and emotional messages of smiling: An ethological approach. Joiimal qfPersonalih and Social P . y h o l o a . 37 ( 1979) 1539- 1553. J M . Femandez-Dols and M.A. Ruiz-Belda, Are smiles a sign of happiness:' Gold medal winners at the Olympic Games. Journal qfPersonali0, and Social Psychologv. 69 ( 1995) 1 1 13- I 1 19. G.P. Ginsburg and M.E. Harrington, Bodily States and Context in situated lines of action. in The emotions. Sociul. Ciiltirral and Biological Dinten.sion.s, R. Harre and W.G. Parrot. Eds. London: Sase. 1 996. H. Wagner and V. Lee, Facial Behavior Alone and in the Presence of Others in The social conre.rt rtomw-bal hehavior. P. Philippot. R.S. Feldman and E.J. Coats. Eds. Cambridge UK: Cambridge University Press. 1999. N. Chovil, Facing others: A social communicative perspective on facial displays in The ps?u.holog\.qf' facial e.rpression. J.A. Russell and J.M. Fernandez-Dols. Eds. Cambridge UK: Cambridge University Press, 1997. pp. 32 1-333. A. Fridlund, The new ethology of human facial expressions in The psychologs q1,facial expression. J . A . Russell and J.M. Femandez-Dols. Eds. Cambridge UK: Cambridge University Press. 1997. pp. 103-132. "421 J.M. Zacks and B. Tversky. Event structure in perception and conception. P.y*r.hologicalBulletin. 127 (2001) 3-21. [331 C . S. Peirce, Ohrn 1igic.o .retnicitic~n.Madrid: Taurus Comunicacibn. 1987.

SECTION I11 COMMUNICATE TO PRETEND: IRONY AND HUMOR Humor is a fundamental ingredient of social communication. It is a rare conversation in which at least one participant does not try to elicit laughter at some point or does not respond with amusement to something another has said or done. Jokes, witticisms, and other humorous verbal and nonverbal behaviors are commonplace in social interaction situations and can have a major impact on the quality of the interactions. For example, one's interpretation of a stranger's remarks as humorous can influence the impression one forms of that person. In addition, humor is often used strategically to decrease tension in a heated discussion or to enliven a boring one. Clearly the transmission and comprehension of humor are central features of social interaction. Wyer and Collins, I992

This page intentionally left blank

135

Say Not to Say L. Anolli et al. (Eds.) IOS Press, 2002

6

‘‘You’re a Real Genius!”: Irony as a Miscommunication Design Luigi ANOLLI, Maria Giaele INFANTINO, Rita CICERI

Abstract: In a standard (rhetoric) perspective ironic communication is considered a semantic inversion between the literal (primary) meaning and the nonliteral (implicated) one. Nevertheless, in common use irony is not necessarily bound to the rhetoric concept of semantic inversion. Rather, it can be defined as an alteration of a reference aiming at stressing the reality of a fact by means of the apparent dissimulation of its true nature. In a communicative perspective, irony aims to get an effective protection of interpersonal relationships, so as to leave wide ranges of fieedom for managing both meanings and interaction. The point of view herein followed concerns irony as a paradigmatic expression of miscommunication, since it is a chance in widening the psychological space available to the communicators. Irony as miscommunication fits rather well the MaCHT framework proposed by Anolli. In fact, by means of an ironic remark, the speaker can lay the responsibility of the ironic value of the utterance on the intention ascription of the interlocutor. In particular, the present chapter intends to sketch out a new theoretical perspective of ironic communication as miscommunication, by proposing the .fencing game (or irony situation) model. Here irony is not considered only as a comment or remark at a linguistic level but also as a complex communicative interaction between interlocutors, depending on contextual constraints and opportunities.

Contents 6.1 6.2 6.3 6.4 6.5 6.6 6.7

Introduction.. ........................................................ The irony family.. ................................................... Theoretical perspectives.. .......................................... What purpose does irony serve?. .................................. The “fencing game“ model of irony.. ............................. Conclusions.. ......................................................... Acknowledgements.. ................................................

136 137 137 141 145 152 154

6.8

References.. ..........................................................

154

6.1

Introduction

According to the standard (rhetoric) perspective, proposed, among others, by Booth [ I], Lausberg [2], Olbrechts-Tyteca [3], and Perelman and Olbrechts-Tyteca [4], irony is an antiphrasis, that is, an utterance expressing the opposite of its literal meaning. In fact. ironic communication consists in a semantic inversion between the literal (primary) meaning and the nonliteral (implicated) one. It is the classic witticism toward someone who has failed an exam: “You’re a real genius!”. Usually. the witticism is uttered in a particular intonation, as we will see below. The Greek etymology of the word “irony”. Etpwvtxa. means pretence. and it comes from ctpwv (the one who makes a question pretending to be naive or less knowing than meaning “true”, “overt”. The Semitic root of he/she is). which is opposite to CIAT~OEUT~KOS, the Greek word is derived from the Accadic term erewwm, “covering”, by means of which irony appears as a device to avoid the direct impact of an explicit word. In this sense. in common use irony is not necessarily bound to the rhetoric concept of semantic inversion. Rather, it can be defined as an alteration of a reference aiming at stressing the reality of a fact by means of the apparent dissimulation of its true nature. Anyway, to reduce irony to a mere rhetoric figure or a linguistic ruse involves not seizing its communicative significance due to the psychological web of its implicit meanings. In fact, in a communicative perspective, irony springs out as a strategic “QS if’. both by escaping the alternative of truth vs. falsehood, and by suspending the subsequent parameters of judgement. It aims to get an effective protection of interpersonal relationships, so as to leave wide ranges of freedom for managing both meanings and interaction. The point of view herein followed concerns irony as a paradigmatic expression of miscommunication, since it is a chance in widening the psychological space available to the communicators. Irony as miscommunication fits rather well the MaCHT framework advanced by Anolli (Chapter I , this volume). In fact, by means of an ironic remark. the speaker may say and not say; he/she may shield him-herself behind the opaque meaning of the ironic utterance: he/she can always withdraw the ironic meaning by taking refuge in the linguistic expression. In any case, heishe can lay the responsibility of the ironic value of the utterance on the intention ascription of the interlocutor. Along this line of thought, the present chapter intends to sketch out a new theoretical perspective of ironic communication as miscommunication, by proposing the .fencinggame (or iron!. situation) model. Here irony is not considered only as a comment or remark at a linguistic level but also as a complex communicative interaction between interlocutors. depending on contextual constraints and opportunities. In the ironic script the two opponents will be considered as using not a massive sword but a sharp foil. In order to make plain our assumptions. in the first section of the chapter we will propose some considerations about the irony family and most recent opinions of the present theoretical perspectives about irony: in the second section we will examine the main psychological functions of speakers resorting to irony communication; in the last one. we will tackle the “fencing game” model according to the miscommunication framework. underlining some possible advantages of this proposal.

L. Anolli et al. / Iron.v as a Miscommunication Design

6.2

137

The irony family

First of all, it is useful to consider that irony is not a fixed and narrow phenomenon, but a family of communicative processes. On one side, there is sarcastic irony, through which the speaker blames his/her interlocutor by means of literally praising words (blame by praise). In this way, he/she does not usually intend to soften a heavy expression of criticism, but to condemn the interlocutor without emoting him-/herself, by humiliating h i d e r through sarcasm and coldness. On the other side, there is kind irony, which consists in praising the other one by means of literally critical and offensive utterance (praise by blame). In such a way the speaker may mitigate the excess and euphoria of direct and explicit praise that may sometimes produce embarrassment in the interlocutor. The more frequent use of sarcastic rather than kind irony in Western culture can be explained resorting to the asymmetry of the afect hypothesis, advanced by Kreuz and Roberts [ 5 ] , according to which the essential nature of irony would consist in the intention to attack. Kind irony, as an affiliation strategy, should not have the same relevance as sarcastic irony. Other forms of irony should be explained. The so-called “Socratic irony”, as an elegant, ingenious, polite, urban way of communication, is convenient for discussing, and debating fashions and dogmas without unbalancing nor compromising. “Bantering irony”, instead, is like wit and serves to reduce the drama of a potentially tense or conflicting situation. Among these forms there exists a certain “air of family”, as Wittgenstein [6] should say: in fact, they are based in a similar way on common communicative ground. As Muecke [7, 81 has pointed out trenchantly, irony is the art of being clear without being evident. The clearness of the ironic utterance does not in any way involve communicative transparency. In this case “clearness” means that an ironic comment can be linguistically decoded by resorting to standard linguistic operations such as phonological, lexical, and syntactic operations. However, its communicative meaning remains opaque because of its semantic indeterminacy and obliqueness. In this sense irony is neither “evident” nor transparent. It is an obvious instance of miscommunication according to the MaCHT model proposed by Anolli (Chapter 1 , this volume). Ironic miscommunication consists in a mask that conceals what one thinks or feels, but it is a mask that, in some regards, reveals what it hides, and, in other regards, hides what it reveals. Therefore, an ironic speaker is not a deceitful one. Unlike the lie, where words and utterances are “deceptive”, irony is found underneath a disguise of pretence. In fact, while in deceptive communication a speaker deliberately omits or fabricates some significant conditions of truth and reality by concealing hidher intention, as, among others, Anolli, Balconi and Ciceri have pointed out (Chapter 4, this volume), pretence communication instead clearly cohabits with reality, and exhibits its “not being true”, by winking at what is hidden behind the mask of untruth, as it is shown by Mizzau [9].

6.3

Theoretical perspectives

Since ancient times irony has been a classical subject of study because of its intriguing communicative value and its theoretical challenge. In fact, irony is a paradigmatic instance of figurative language, since in a standard ironic exchange a speaker conveys (and the addressee interprets) a communicative meaning which is the opposite to what is said. In modern times several theoretical perspectives have aimed at explaining this complex communicative phenomenon.

138

6.3.I

L .4nolli et al

Iron\*us a Miscommunication De.sr
The rationalistic theon.

Grice [ 101 has proposed a rationalistic theory for explaining irony, by formalizing the standard fiamework of antiphrasis. The ironic utterance consists in “sa-vingp intending not-p”, and then involves a semantic gap between what is said and what is meant. Given these conditions, irony represents a violation of the maxim of Quality (“Do not say anything you believe to be false or for which you lack adequate evidence”). Therefore. it is the outcome of a conversational implicature. Such a hypothesis was shared also by Searle [ 111, who identified irony as a logical denial of the literal interpretation. The actor resorts to irony for producing perlocutory effects on the interlocutor, by not fulfilling hisher models of expectation. According to this theoretical perspective, irony is generated by the conflict between the actor’s subjective attitude (intention) and the attitude and expectations which are generated by the literal expression of an utterance. In this way, Fish [ 121 attributes particular importance to the ironic communication interpretation. Both literal and ironic interpretations of an ironic utterance are possible and come from a process of inference. There is neither univocal nor equivocal irony: the dilemma is loosed at the level of the interlocutor’s competence, because irony is an indirect but clear communicative strategy. Irony is neither allusion nor ambiguousness nor even equivocalness. The rationalistic perspective is still anchored to the linguistic design of ironic communication. In this way, it considers irony only as a violation (albeit apparent) of the Gricean maxim of Quality, based on truthfulness, without considering the interactive and contextual aspects of ironic communication. Crudely put, Grice [ 101 proposed a dual logic: the logic qf lungzruge which applies to literal meaning (sentence meaning), and the logic of’ conversation which applies to the rules that people use to infer (or “implicate”) what a speaker intends to convey (utterance meaning). In the ironic communication an interlocutor begins with the literal meaning of a sentence and ends with an utterance meaning by making an appropriate implicature. But such a distinction is not unproblematic, and the concept of “literal meaning” has recently become a subject of theoretical revision by a few scientists. like Anolli (Chapter 1. this volume), Gibbs [ 131, and Glucksberg [ 141. Literal meaning involves the idea of words as “meaning containers”. independent of any use and contextual constraints. But. as Gibbs [I31 has clearly pointed out. “the literal meaning of any word or sentence is almost impossible to determine” (p. 47). In the interpretation of an ironic comment we do not need to proceed in an additive way, analyzing first what is said (literal meaning or sentence meaning), and then what is meant (ironic meaning or utterance meaning). According to Gibbs, Buchalter, Moise, and Farrar [ 151. and Lakoff and Johnson [ 161, the interpretation of a comment as ironic is immediate. and it happens automatically, without any additional cognitive effort. since it does not require conscious control nor a complex computation. What people are interested in is the actual meaning of an ironic utterance pronounced by a speaker. They want to see and grasp hidher communicative intention at once, and, as Anolli has underlined (Chapter 1. this volume). communication and miscommunication design follow the same cognitive processes, and the utterance meaning depends on specific and contingent features of a certain situation.

L. Anolli et al. /Iron?, as a Miscommunication Design

139

6.3.2 The Machiavellian perspective Irony has also been considered from a Machiavellian perspective as a communicative form aiming at creating a set of effects on one’s interlocutor, without considering or respecting the formal rules of language or the truthfulness of the message. According to this perspective, attention moves from the analysis of the linguistic design to the interactive processes which are involved in ironic communication: in fact, irony can be an effective device for disorienting one’s interlocutors and for taking them unawares. A typical change of “pace” (footing, as Goffman should say) is required, by passing from seriousness to joking, from the descriptive utterance to the ironic one, so as to produce surprise and amazement. Within this framework irony should not consist in the violation of the Gricean maxim of Quality; rather, it should imply a transgression of one’s interlocutor’s contextual expectations, and should involve a careful interpretation to value shrewdly the “allowed” contradiction within the ironic comment, to use the expression proposed by Mizzau [ 171. According to the allusional pretence theory, advanced by Kumon-Nakamura, Glucksberg and Brown [ 181, ironic effects are generated by allusions to an unfulfilled expectation that was forecast at a conventional level (after a terrible blunder, one expects no appreciation but criticism). In particular, ironic communication is based on a condition of incongruity that is generated by a lack of consistency with an expected script. As Deckers and Buttram suggest [ 191, the understanding of irony takes place by means of a two-levels process: a) the perception of incongruity with the situation; b) the outcome of the ironic value of an utterance as a solution for the perceived inconsistency. In this sense, the effectiveness of ironic communication is inversely proportional to the presence of explicit signals, and the more subtle and implicit it is, the more effectively it hits the mark. According to this Machiavellian perspective, irony is a useful device for disorienting interlocutors, by avoiding one’s own responsibility for what is uttered. Other people are to be mocked, to be looked down upon, and to be subtly depreciated. In fact, because of its intrinsic ambiguity, irony as miscommunication can be a device for softening and confusing the borders of sense, and for scorning others. On one side, this feature is an advantage, because the ironist enjoys many more degrees of freedom than when he/she uses an explicit utterance, when the menace for one’s personal image is so high that people need a kind of relational shield for not losing face. On the other side, instead, irony can be used as a weapon to amaze one’s interlocutors and to exercise power on them, as Messman and Canary [20], and Mizzau [21] have underlined. But the Machiavellian viewpoint of irony stresses a pessimistic (and partial) conception of the human being, in line with Hobbes’s statement: homo homini lupus. As a consequence, irony becomes mainly (if not only) a matter of ruse and nastiness. It seems to be a reductive perspective which considers irony as a means to overwhelm and injure. On the contrary, it concerns a wide range of expressions and forms, as we have seen in the irony family. 6.3.3 The echoic theory

In turn, Sperber and Wilson [22, 23, 241 have underlined other psychological aspects of ironic communication as a way of re-proposing and evoking one’s interlocutor’s thoughts, actions, attitudes and feelings. Thanks to the ironic comment, the speaker produces an utterance echoing what hisher interlocutor previously said or did, as well as expressing hidher own either disparaging or humorous attitude. In particular, according to the echoic

130

L. Anolli et al. 1 Iron!. as a Miscommunication Design

theory, irony is a kind of indirect quotation used by people for echoing something in a misleading fashion. Consequently, on one side, ironic communication is an implicit mention (echo) of what one’s interlocutor said or did; on the other one, it is a way of showing one’s attitude toward the thought or the action that the ironic comment refers to, or even toward one’s interlocutor. In this sense, irony can refer to a specific subject, as well as to a category of people, or even to a social norm or a cultural trend. In ironic communication implicitness works as a substitution of literal sense. When someone says “Today is a gorgeous day!”. while outside there is pouring rain, the implicit meaning of “hideous day” substitutes the literal sense. According to Sperber and Wilson, ironic communication consists in “making people impose one idea on another idea”. If someone asks “Did you remember to water the flowers?,’, after days of rain, he/she echoes and comments the interlocutors’ anxiety to keep the flowers watered. In this perspective, the interlocutor realizes that the speaker produces a comment about the expression itself rather than the meaning of the expression. Within the echoic perspective the so-called tongue-in-cheekphenomenon is involved, as Almansi [ 2 5 , 261 has explicitly indicated. It consists of an ironic background comment to the interlocutor’s utterances. It is a continuous, shrewd and subtle comment, which is typical of the Anglo-Saxon culture. Similarly, within the perspective of irony as a mention. there is the quotational model too, proposed by Kerbrat-Orecchioni [27], Kreuz (281, and Kreuz and Glucksberg [29]. Irony should be a “quotation” of what happened. By means of such a mechanism, ironic communication implies an attitude of distance from the literal sense of an utterance, so as to provoke a doubling. The echoic theory, although it has pointed out the interactive aspects of the ironic exchange and has taken into consideration contextual conditions, runs the risk of accepting a dual nature of ironic communication, like the rationalistic theory proposed by Grice. In this framework ironic meaning seems to be an “extra-meaning”. added to the literal one. 6.3.4 The theatrical perspective

Final 1y, according to a theatrical perspective, ironic communication has even been interpreted as a pretence, since it is based on a “social pact” allowing the apparent untruth to “clearly cohabit with reality”. Within such a framework, pretence exhibits its “not being true” by winking at what is hidden behind the mask of untruth. Haverkate [30] speaks about “transparent pretence”, since the speaker says something in which he/she pretends to believe. and, at the same time, he/she signals such a pretence by means of paralinguistic indexes. Likewise, the concept of irony as simulation, proposed by Clark and Gerrig [3 11 in their “pretence theory”, can be included within this theatrical perspective. The ironist does not in fact resort to an utterance to reveal hisher inner contradictions; on the contrary. helshe pretends to use it, and creates a sort of complicity with hisher addressee. who acknowledges the presence of such a simulation. Let us take the subsequent situation. A says: “It was a wonderful dinner with a lot of Italian and French cheese!”. This utterance can be interpreted as serious praise by B (a stranger). but not by C, who. like A. knows that D. their friend, is allergic to cheese. Therefore, the utterance acquires an ironic and sarcastic value toward D. Within such a situation, C is an “initiate” so he can understand the whole sense of A’s ironic comment. In this sense. ironic communication is rather close to parody. as it involves the mechanism of antinomy. which is an antithesis between two terms so as to produce an ironic effect ac Hutcheon 132. 331 sketched out. .An essential constraint of parody is g i \ m

L. Anolli et al. /Irony as a Miscommunication Design

141

by understanding the action (or character) that is the object of the parody. The parodist reframes the antecedent (action or real subject) within a new picture, by a structural overlapping of texts. In this way, on one side, parody frees itself from the seriousness or authority of the situation (or character) which it mocks; on the other one, while ridiculing, parody revives it by referring to it. However, both irony and parody, like a theatrical performance, imply a sort of complicity between author and addressee, because they call for sharing communicative levels (primary vs. secondary). In this regard, an ironic comment includes the chance to select the addressee in accordance with hisher competence, the sharing of a specific background of knowledge, and reference to a defined context. Ironic communication consists in an enigma only for people who cannot understand it. In fact, it creates a condition of communicative trianguzation, by which participants are raised to a higher level of knowledge, so as to generate an inner circZe and banish people who cannot understand, as Attardo points out (Chapter 7, this volume). According to the theatrical perspective, underlined, among others, by Creusere [34] and Mizzau (171, ironic communication can be used as a strategy for excluding, that is, selecting those who should understand (as they know the assumptions) from the ones who should be prevented from understanding more than a literal sense. From the state-of-the-art knowledge, herein summed up, it emerges that the different theoretical models hitherto elaborated enable us to understand many deep aspects of ironic communication both at a linguistic layer and an interactive one. However, they have not been analyzed as a sequence of actions between the participants in a communicative exchange, since irony is patterned in a script-like structure. In particular, the ironic remark has not been examined as the outcome of an intentions game between communicators: on one side, the production of ironic intention by the speaker and, on the other one, the recovery and ascription of this intention to himher by the addressee. The unavoidable gap between these two positions helps to explain the success (or unsuccess) and effectiveness (or less) of the ironic exchange in the light of the miscommunication framework. These two aspects are basic in the “fencing game” model, as we will see in section 5 ; but first it is worth examining the psychological functions of irony as a fundamental premise to better understand ironic interaction.

6.4

What purpose does irony serve?

As we said above, ironic communication is miscommunication as an oblique communication. In fact, on one side, it shows what it hides, while, on the other one, it conceals what it says. In this sense, irony is “say in order not to say”. By means of an ironic comment, the ironist can remain “opaque” and impervious to the interlocutor on a relational level, though he/she is not silent. In fact, ironic miscommunication is a kind of semantic mask, by means of which it is possible to soften and fuzz the borders of meaning in order to improve the negotiation processes in a given situation. In this sense, from the perspective of Bakhtin [35], ironic comment may be seen as an emblematic instance of discursive dialogism, according to which the word is not semantically univocal (monosemic), but it possesses “more voices” (polysemic). Its interpretation assumes different forms in terms of both its “position” within the discourse and its relation to the “focusing” game, in which some of its features are in the foreground and others are veiled. Paradoxically, in irony the foregrounded mask plays a background role during the exchange between the ironist and hisher interlocutor, as it is shown by

I47

L . Anolli et al. Iron\. as u .Mi.scommunicalioti Design

Bakhtin [35] and Levinson [36]. As a consequence, irony carries out and satisfies different psychological functions, as Anolli, Ciceri and Infantino (in press a) have underlined. 6.4.I Ironic communication as a sign of respect fbr cornlentions (hair. to el’ade censure in a cuIturallr*correct ~ q * )

“Where the lion’s skin will not reach, it must be patched with the fox’s’’. the wise Greek strategist Lysander sentenced, according to Plutarch. A similar metaphor appears again in the works of Baltasar Gracian y Morales, a seventeenth century Spanish writer, as well as in Niccolo Machiavelli’s essays of diplomacy during the Italian Renaissance. This maxim describes the situation in which one realizes that a direct and impulsive expression is unsuitable for the interactive context, especially in face-threatening situations like conflict. Hence, it is not a matter of chance that its original version was uttered by Lysander, an expert in military planning. By extending this metaphor to everyday communication, an effective communicative interaction is enhanced by ironic comment in a subtle and diplomatic way, so that a speaker might achieve hisher aims in agreement with the “unwritten rules’’ of civilized behavior: the Latin concept of urbanitas, as Anolli and Infantino [37], and Infantino [38] have pointed out. Irony is a really effective strategy for “hitting the mark” in an indirect way. without violating the norms required by one‘s own cultural background. Irony as miscommunication arises from the need both to respect social standards. and to avoid other people’s censure, without abandoning, however, those topics that would otherwise be unacceptable. Ironic speakers accept the cultural norms and at the same time, violate them, remaining within the limits of social acceptability: they do not have to suppress their thoughts or their feelings. Therefore, ironic communication finds its edge in those cultures (like the Anglo-Saxon one) where self-control is very important and where it is thus a very positive thing to be able to keep coolly detached from events. without emotional arousal. In this way, a speaker can use irony to hide the expression of hidher emotions and safeguard hisher personal experience. In particular, English humor responds to these cultural expectations and standards, involving the ability to “stay in one‘s place”. In English culture, where one talks about emotions in preference to showing them. as Lutz [39] has observed, irony becomes not only a device to keep at a distance from emotions and “de-emotionalize” oneself. but also a way of showing consideration for the interlocutor’s feelings (without saying everything one feels or thinks about the other). in order to be polite and be cognizant of the situation. as Wierzbicka 1401 has pointed out. For instance, a speaker named Peter, if his colleague John arrives late at an important business meeting, may utter an ironic remark by saying: ”Congratulations. John! You are so punctual this morning!”. In fact, John’s delay irritated Peter. who decides to use sarcastic irony rather than direct blame, because he knows that the formal context forces him to address John by using indirect communication. without. howxever. avoiding criticism of his behavior, which is in contrast with social acceptability. By extending the metaphor of theatrical pretence to social life. and by applying the MaCHT perspective, the ironist can well be described as an able director of’his/her.O M N irnage. as Rutelli [41] has proposed, who can manage interpersonal relationships by veiling and unveiling him-iherself with moderation. In fact. according to Dolitsky [42. 431. human beings are like actors playing their part in a ”dramatic” and “carnival-like“ sociec. In this regard. a social interpretation of Diderot’s paradox of acting is feasible: the mask of irony as miscommunication allows one to express the duplicity between being and appearing. as \\.ell as their paradoxical unitication within the ironist‘c consistency mith the character he interprets

L. Anolli et al. /Irony as a Miscommunication Design

143

In this way, as interactionist scholars like Bakhtin [35],Ducrot [44],Goffinan [45,461, and Mizzau [21, 171 have underlined, the ironist succeeds in respecting the “unwritten laws” of polite behavior which call for both self-control and socially-correct moves. Irony is the resource used by the artful man while waiting for the right moment to carry out his performance on the stage of society. 6.4.2 Ironic communication as a border of reserve (how to safeguardpersonal space)

According to Allemann [47],Barthes [48],and Goffman [45,461, irony can be used not only as a device to evade social censure, but also as a planned action aiming at maintaining dignity, restraint, and demeanor, as well as one’s own privacy. An intriguing image is provided by Barthes’ [48]metaphor of the “dark glasses”, according to which in ironic communication attention is shifted from the informative function to the metacommunicative one. In fact, after crying, people do not wear their dark glasses to conceal the fact that they have cried. Rather, people put them on in order to disguise the distressing expression of pain, i.e., their swollen red eyes. The dark glasses are an “allusive mask”, aimed at preserving one’s own dignity and demeanor: they hint at the pain whose embarrassing effect they cover. People wearing their dark glasses metacommunicate: they intend to communicate that, though suffering, they do not want to exhibit their own pain. Therefore, the dark glasses, at the same time, make the ironist an actor and a witness of him-/herself and of others. They are useful for protecting both personal space and privacy. Irony as miscommunication can be described once again by means of a metaphor: that of the sacredfence, symbolic boundary, or even magic circle, which makes the ironist “intangible” and “unapproachable” in the interpersonal game. In fact, irony has often been considered as connected to the talented wise man, who succeeds in observing things from a distance, avoiding unbalancing and compromising himself. Just like Socrates, the prototypical ironist, people performing an ironic communication can calm passions as well as debating dogmas, as Bergson [49],JankelCvitch [50],Jouhandeau [5 11, Kierkegaard [52],and Pirandello [53]have pointed out. This intention not to be explicitly aggressive arises from the words Mary says ironically to Lawrence, her husband, who has tried to repair an old armchair, with the usual lack of success: “You’re so genial, Lawrence! Since our marriage, your cleverness has always fascinated me!”. In this case Mary chose to be ironic rather than direct because she does not want to start an open conflict with Lawrence. Conversely, Paula jokes with her brother, who has just got a high mark in his mathematics exam, saying “You were right, Andrew. As usual, you are an idiot!”. The girl comments ironically on Andrew’s success, because her brother was scared and pessimistic before the exam. In both the above-mentioned cases, irony is planned to reduce tension and defuse dramatic situations (both in conflict and in cooperation), mitigating their emotional arousal. In this regard, irony is a usehl device for maintaining or for re-establishing symmetry within relationships, in agreement with Brown and Levinson’s [54]model of politeness, as a strategy that interlocutors use for preserving their personal image mutually.

6.4.3 Ironic communication as a relational ambiguity (how to re-negotiate interaction)

The ironist, skilled in the art of skimming and lingering, just like the mythological Janus Bifrons, has two faces: one which laughs at the weeping of the other. In this way, it is possible to define irony as a “Janus-faced” communication. The paradox about ironic

113

1- 4nollr et nl

Irony

UF

a Miwommunication Deygn

communication is that, z f you want to be understood cIear(r7, you have to he misunderstood. In fact, the ironic comment is like a skin that alludes to the hidden content at the very moment in which it conceals that content. Therefore, ironic miscommunication can be used as an ambivalent strategy, a “tongue in cheek” producing puzzlement and disorientation in the interlocutors. as Almansi [25, 261 has underlined. In fact, the “Janus-faced” nature of ironic miscommunication allows people, on one side, to calm their passions, while on the other one, to shift in their own favor the fuzzy borders between the different possible (and legitimate) interpretations of their comment. The ironist benefits both from the “efectiveness of the word” and the “innocence qf silence”, to use the incisive expression of Ducrot [a]. For this reason, we can speak about pragmatic poZysemv in ironic communication. In fact, by means of a systematic process of “meaning negotiation”, in an ironic utterance speakers convey a communicative intention which allows the interlocutor to interpret it with different meanings. On this subject. some scholars, like Bakhtin [35], Dolitsky [42, 431, Ducrot [a], Merleau-Ponty [ 5 5 ] , and Mizzau [ 171, talk of ambiguity and a sort of “semantic three-dimensionality” in irony because of its indirect nature. irony as miscommunication is a complex communicative outcome in which different signaling systems interact at the same time. In particular, in the standard ironic comment, linguistic segments are combined in a specific paralinguistic (or supra-segmental) frame. The apparent opposition between these two signaling patterns generates the ironic meaning perceived by the addressee. On some other special occasions, when there are strong contextual constraints and clues, linguistic inputs are sufficient alone to create the ironic meaning. But, as Anolli has pointed out proposing the MaCHT perspective (Chapter 1. this volume), each signaling system has a different weight and significance in determining communicative efficacy of irony. Therefore, we can foresee an extended range of graded and fuzzy variation between linguistic units and paralinguistic patterns. so that. in some (exceptional) cases. the ironic value of an utterance is totally evident, and in other cases (the majority) it is less evident, to the point of arriving at an undecidable balance between a plain descriptive comment and a truly ironic one. The basic ambiguity of irony allows one to negotiate and re-negotiate the meanings of an ironic comment. In this way, the ironist is not constrained to undertake responsibilih. for hidher words. Consequently, according to Jorgensen [56] he has neither to compromise hidher image nor to lose face. Thus, responsibility for the appraisal of ironic value of an utterance moves from the ironist to the addressee. It is up to the latter to recover and attribute an ironic meaning and intention to the utterance of the former. In particular, the ironic sense of a comment is effective when it is not only intentionalized by the ironist, but also ”re-intentionalized’ by the addressee, as Anolli (Chapter I . this volume). and Anolli and Ciceri [57] have pointed out. This property of irony (being a skilful device to assure oneself of many more degrees of freedom than an explicit utterance does), arises from the ironic remark Anthony addresses to his friend Hillary. The lady goes to a cocktail party wearing a hideous dress and Anthony say: “Hillary! You’re so beautiful: just like Sharon Stone!”. “What do you mean’?!”, George. Hillary’s husband, intervenes, irritated by Anthony‘s sarcastic attitude. “Hillary is gorgeous, tonight! I was just paying your wife a compliment...“. Anthony pretends not to have been sarcastic, as if the meaning of his comment depended only on the linguistic input. In this dialogue the sense of the utterance is the subject of a skillful negotiation between them. because the semantic ambiguitv of the ironic remark allows ,4nthonv not to take full responsibility for his innuendo. 4s useful outcome of ironic miccommunication. speakers have the chance and

L. Anolli et al. /Irony as a Miscommunication Design

145

opportunity to calibrate the weight of the indirect meaning of their speech. An indirect expression of one’s thoughts, desires and feelings cannot only hide one’s real intention, but it can also define it and re-draw the limits of social interaction between interlocutors. First of all, according to the “tinge hypothesis”, proposed by Brown and Levinson [54], Dews, Kaplan and Winner [58], and Dews and Winner [59, 601, irony should express less condemnation and less approval than a direct utterance does (mitigation of the intended meaning). A criticism ironically made is apparently lighter and less offensive than an open insult; similarly, ironic praise is less positive than an explicit form. There is a kind of “regression to the center”, in which, on the one hand, the exultation is lessened, and, on the other, the aggressive charge is attenuated, as Giora [61], and Glucksberg [62] have clearly underlined. Conversely, according to other scholars, like Almansi [24], Colston [63], and Sperber and Wilson [23, 241, ironic criticism is used to emphasize condemnation rather than to dilute it: thanks to irony it should be possible to achieve one’s aims in a more pointed and controlled way, intensifylng the meaning of an utterance (enhancement of the intended meaning). Because of a cool detachment from emotions, irony may be used as a device for wounding someone in a much more cutting way than a direct criticism oriented in the same direction. In fact, an explicit insult can be produced in a moment of rage, as a consequence of the speaker’s mood in the contingent condition. Alternatively, an ironic insult can arise from a cold calculation, so as to express, besides blame, even the ironist’s intention of not losing hisher self-control in showing the interlocutor’s lack of success [56]. Similarly, within a context of praise, irony is feasible when the speakers know each other very well. Therefore, as Oring [64] points out, irony can arouse much more familiarity and fondness between them than direct praise can do. Such a phenomenon is evident in the apotropaic formula, e.g. the Italian “in bocca a1 lupo”(literally, “in the mouth of the wolf’), or the English “break your leg”, whose origin can be found in ancient Roman traditions, for instance in the so-called “Carmina Triumphalia” in which the soldiers joked about their leader’s victory in order to avoid any mischance which might be caused by arrogance [381. Actually, more recent studies, carried out by Anolli, Ciceri and Infantino [65, 661, Gibbs [13, 671, Lee and Katz [68] and Leggitt and Gibbs [69] go beyond this antithesis, and attribute manifold functions to ironic communication: it can be used now as a defense, then as a particularly stinging attack, or again simply to reduce the tension of emotionally heavy situations.

6.5

The “fencing game” model of irony

From these recent studies that intend to go beyond the antithesis between irony as a mitigation and irony as an enhancement, so as to illustrate the flexibility of irony in the process of the creation of meaning, the “fencing game” model of irony as miscommunication has been proposed by Anolli, Ciceri and Infantino [65, 66, 701. According to this framework, attention is focused not only on the linguistic level of irony but more especially on the irony situation, intended as a class of interactive episodes in which an ironic comment is generated as the best local solution between communicators given certain contextual constraints and opportunities. Within this perspective irony is seen as a suitable outcome embodied in the plot of communicative moves, since both speaker and addressee are engaged in the irony production as Bavelas, Black, Chovil, and Mullet [71] sketched out.

Both war and theater metaphors fit the “fencing game” model, as emerges from the two psychological perspectives on which it is based on. In fact, the interlocutors play the pan of two opponents holding not a massive and cumbersome sword, but a quick sharp foil. From a functional point of view, referring to the socio-interactionist .framework. proposed. among others scholars, by Goffinan [46], irony is a flexible phenomenon that people use like a mask in order to evade censure in a socially correct way. to safeguard their private space. and to re-negotiate meanings (theater metaphor). In this way it is easier for speakers to manage the interaction and irony may be included in what Dindia [72, 731 defines as “relational strategies”, that is, the creation and maintenance of interpersonal relationship by adapting one’s own repertoire of verbal and nonverbal actions to the situation with a definite aim, in order to be effective while respecting social standards. Planning relationship maintenance, trying hard to preserve it. and metacommunicating to get suitable interactive accommodation are the main strategies that promote the upkeep of desired relational situations. From a structural perspective, instead, referring to the planning communication theog., elaborated by Berger [74], Bratman 1753. Cohen and Levesque [76]. Greene [77. 781 and Monge and Kalman [79], irony can be described as a planned attitude. strategically aimed at achieving specific intentions (defined as mental dispositions which are oriented toward future actions), and as a direction of practical reasoning. Within this perspective, as Monge and Kalman [79] have pointed out, communicative interaction is ruled by sequentiality (i.e.. a chain of actions), simultaneity (i.e., a certain communicative pattern at a given time). and synchronicity (i.e., a simultaneous coordination of communicative processes). This chain of events embodies a sequence of actions aimed at striking the antagonist in an elegant but “stinging” fashion while observing social rules and cultural standards. It is organized with reference to a script composed of several phases, as shown in Figure 6.1. Among others, as Anolli (Chapter I , this volume) has pointed out. communication and miscommunication, ironic communication included. are regulated by a complex web of meanings and intentions in the speaker’s and addressee’s exchanges. This web is characterized by mutual interdependent moves calling for adjustments and progressive modifications in terms of tuning and interactive synchrony. In the irony situation the speaker’s intention of saying something that sound ironic, that could be sarcastic or “Socratic” or else witty, is as important as the addressee’s competence of recovery and ascription of an ironic meaning and intention to what the speaker has just said. As Grosz and Sidner [SO]. and Waldron 1811 say. we have to go beyond the bare recognition of intentions. In such a way, irony as a “fencing game”, rather than a simple utterance or comment. consists of a communicative situation which is ruled by a script that is predictable and well articulated. Four steps are foreseen: the asszrmptions, the,fixal event (the stirring up object of irony), the dialogic comment (the ironic utterance that the speaker expresses by alluding to the assumptions), and the ironic effect. depending on the addressee‘s feedback.

L. Anolli et al. /Irony as a Misconimunication Design

147

Figure 6.1. The articulation of the ironic script in the “fencing game” model of irony (it is a partially modified version of the figure that appeared in Anolli, Ciceri, & Infantino, 2000)

We will describe these four phases by explaining them with an example in everyday life. The frame of the example is given by an exam situation that takes place in a university, where three students (Cassie, Dan, and Bob) interact between each other. The exam is oral and is considered one of the most difficult. Both Cassie and Dan have to sit their exam, while Bob intends only to listen to his university mates’ exam. Before the test takes place, Cassie and Dan express their moods, while Bob, after his pals’ exam results, will express his comment. Assumptions - The assumptions which constitute the first phase of the irony situation, mainly consist of the mutual relational background, i.e., a store of shared systems of beliefs and meanings, interactive patterns, and cultural standards. In fact, in order to perform irony

148

L . .4noIli rl al. Iron\. as a Miscommunication Design ,I

effectively, interlocutors have to share a common knowledge which they should allude to in their ironic remark. Assumptions can be both textual and contextual. In fact, while speaking ironically, people can make reference (echo) to previously uttered expressions, as well as exclusively to social norms [23, 241. The textual assumptions concern previously uttered words which interlocutors know: substantially, they pertain to a precise utterance that will be the object of the ironic remark. The contextual assumptions, instead, refer to an extensive range of social rules and expectations, which are invariably positive. Of course. while an ironic remark grounded on textual assumptions will be more circumscribed and will imply a profound acquaintance between the interlocutors, contextual assumptions allow even a more generic mutual knowledge, because the allusion regards more widely shared know ledge. In the example, both textual and contextual assumptions can be individuated. Cassie is very nervous before she takes her oral exam of comparative literature, that is considered one of the most difficult. Therefore, she repeats to Bob: “1’11 fail. yeah. this time will be wrong, I feel it!”. Dan. instead, is arrogant as usual: he is sure of getting a high mark. although he did not study very much. “To get a high mark it’s enough to beat about the bush. as usual! It will be nothing for me to pass the exam!”. Dan exclaims. laughing. Both Cassie’s and Dan remarks might constitute the textual assumptions for developing the ironic script. Besides. in the above-mentioned example, there is a clear influence of social standards which are invariably positive. If a student gets a good remark. hekhe hlfills the social expectations of accomplishment and success. In this regard, it is worth mentioning the so-called “asymmetry of affect” phenomenon. according to which there is a divergence between the use of sarcastic irony (i.e.. blame by praise). and kind irony (i.e., praise by blame), as Kreuz and Glucksberg [29] haire suggested. In fact, sarcastic irony echoes the Western culture’s standards of success. which the irony target cannot fulfill. Conversely, kind irony arises from the opposite process. with a less obvious outcome. Therefore, in order to be better explained. it usually needs a textual reference, i.e., a mention of previously uttered words, or. at least, a private knowledge shared by interlocutors (in the specific case. Cassie’s habitual anxiety). Focal event - The assumptions constitute the background for the jocul ei*ent. ie.. the input that triggers the ironic comment. This event triggers the process of planning and preparation for the ironic utterance. It gives rise in the speaker to appropriate mental and ideative associations, such as suitable metaphors and rhetorical figures. idiomatic expressions and proverbs, contextual references, as well as cultural stereotypes. At the same time, he/she has to choose the prosodic intonation of voice. Together with the assumptions, the focal event forms the inductor element and the antecedent of ironic communication. The event can take place independently of the communicators: but it is important that they note it and make it the subject of their communicative interaction. In the above-mentioned example, the focal event is the exam itself when it takes place. involving the students’ results. Cassie, despite her fear. gets a high mark. while Dan. notwithstanding both his arrogance and pride, fails his exam. Dialogic comment - The third phase of the irony situation (or script) is the dialogic comment, i.e., the ironic utterance. as the expression of a definite communicative intention by the speaker. In fact. by means of the ironic comment. hekhe is able to blame. to praise. to mock, to amuse, to play the game of “hide-and-seek” and so on. The production of the speaker‘s intention requires an option among different kinds of ironic communication

h

L. Anolli et al. /Irony as a Miscommunication Design

149

(such as sarcastic irony, kind irony, bantering irony, Socratic irony and the like). At the same time, he/she has to proceed to a selection of linguistic unit, prosodic intonation of voice, face mimic, and gesture pattern. Within the “fencing game” perspective, the ironic utterance is a global act of communication and it is the unitary outcome of different signaling systems. In it linguistic (segmental) inputs are combined with paralinguistic (supra-segmental) patterns, face expressions, and gestures. Each of these aspects contributes effectively, each for its part, to produce the ironic meaning of an utterance said by a given speaker in a certain situation. They converge to produce an ironic comment that is coherent in itself, even if complex and articulated, since it displays a determinate communicative intention. Differently said, irony is a “multi-modular” communicative act. Therefore, as Anolli has pointed out (Chapter 1, this volume), irony, like other communication forms, is the result of a semantic synchrony process, in which the different signaling system are combined and unified in an interdependent way. It is likely that such a process might be regulated by a “central communicative processor”. In any case, for irony, we have a specific “contrastive semantic synchrony”, since the paralinguistic pattern appears to be in contrast to the linguistic input. With the linguistic unit speakers convey a certain meaning obtained by a pure linguistic encoding; with the paralinguistic pattern they convey an alternative (in most cases, opposite) meaning. The synchronic combination of these signs systems generates the ironic meaning, which is not to be grasped as an addition of “linguistic meaning plus paralinguistic (and nonverbal) meaning”, but it is to be intended as the unitary outcome of the simultaneous convergence of these communicative signs. As an effect of this semantic articulation, speakers can arrange and graduate in the best possible way their communicative intention in order to produce an ironic utterance in the manner most suitable to the contextual conditions. In this way irony cannot be considered as a homogeneous communicative process, but as a flexible and effective strategy, planned in a graded manner, referring to a certain interlocutor, given a determinate situation. The gradation of ironic intention by the speaker is a really useful device to manage locally semantic value and strength of speech. As a consequence, it is useful to foresee a famiZy of ironic voices, among which the speaker can choose the most appropriate one to express hisher communicative intention. In fact, experimental researches, carried out, among others, by Anolli, Ciceri and Infantino [66], Cutler [82], F6nagy [83], Haverkate [30], Kreuz and Roberts [5], have verified the existence of a range of supra-segmental profiles that qualify irony communication as opposed to normal speech. The irony voice is generally characterized by a combination of high and changeable pitch, strong energy and a slower rate of articulation, which provokes an emphatic lengthening of syllables. In irony there is a caricatured stress of the suprasegmental features, in which irony is a device to “play with the voice” [82]. It is employed not in a normal way but in a “studied” one, though spontaneous and natural. It is both a “premeditated” and “affected” voice by means of which the speaker intends to assign a special intentional value to the supra-segmental pattern. In fact, he/she can resort to different paralinguistic profiles and adapt his/her prosodic intonation according to hidher communicative intention and the interaction with the interlocutor in a given situation. As Gibbs [67] has pointed out, he/she can adjust the level of emphasis of the ironic comment according to the intensity with which he/she means to affect his/her interlocutor. Herein, as Infantino [84] suggests, it is worth taking into account two dimensions: context and empathy. In the first one the two endpoints consist of conflict and cooperation; in the other one, the two endpoints are represented by involvement and estrangement. In

1 so

L . .4nolli et al. i Irony as a Miscommirnication Design

both conflict and cooperation contexts according to the degree of empathic involvement with hisher partner, a speaker can appraise the weight of the indirect speech, choosing whether to strengthen hisher own intention or, instead, to weaken it. In the first case. when he/she wants to reinforce hisher intention indirectly. he/she can resort to a scornful voice to carry sharp blame or to a tender voice to express affectionate praise. In the second one. when he/she intends to minimize hisher comment, he/she can employ a bantering voice to express both criticism and commendation. Going into this matter thoroughly, the bantering voice, both kind and sarcastic. represents an effective device to cope with the inducing event, tingeing and mitigating the situation through an empathic estrangement. By means of a bantering voice a speaker can weaken hisher communicative intention in a compliant way without exposing himherself openly: for instance. he/she can censure a certain behavior of the interlocutor without attacking h i d e r , or, on the contrary, he/she can praise the addressee without exalting him/her and without putting himher in an awkward situation (attenuationcondition). However, a speaker can also strengthen hisher intention by means of an ironic comment (intenszjication condition). In a conflict context, for instance. heishe can profoundly offend the partner by resorting to a sarcastic comment pronounced in a scornful voice (that is, a low and monotone pitch, a slow rhythm and a stressed articulation of words) in a cold way, without showing an apparent emotional arousal [54. 181. Likewise. in a cooperation context a speaker can praise the partner by means of an ironic utterance pronounced with a tender voice in order not to exhibit verbally hisher own emotion for reasons of discretion. Referring to the example, afier Cassie’s and Dan’s exams. it was Bob who uttered the ironic comment addressing both of them with a remark in agreement with their performances. Bob tells Cassie, after her success: “My goodness, Cassie! What a ghastly figure! You’re right: you are a real flop”. Bob’s words echo Cassie’s admission of fear. contradicted by the result. In order to express his appreciation of Cassie, Bob chose the strategy of kind irony: in fact, he praised her with words of blame. In spite of his own arrogant declarations, Dan, instead, failed. Bob comments the fact by echoing both his self-confidence and the social expectation of success. “Congratulations! As usual, you’re a great winner!”. Of course, in this case, Bob chose the strategy of sarcastic irony. through which he blamed Dan by words of praise. Resorting to the semantic synchrony process, Bob could praise Cassie ironically by using vocal tenderness combined with linguistic input expressing reproach. Conversely, he could blame Dan ironically by employing a scornful voice associated with a linguistic unit conveying praise. Ironic effect - The last phase of the irony situation, according to the “fencing game“ model, consists in the ironic effect, i.e., the communicative output of the ironic comment. the manner in which the utterance is interpreted by the interlocutor. This outcome depends neither on the ironist nor on the actual ironic value of the utterance. but on the addressee and on the way in which heishe ascribes an ironic meaning to the speaker’s utterance. In this process, as, Anolli, among other scientists, has pointed out (Chapter 1 . this volume), the addressee can choose between different possible interpretations. This plurality of interpretations is a useful device in the hands of the addressee to recognize and ascribe an ironic intention to the speaker’s utterance. In this way he/she has to take hisiher communicative responsibility. since he/she can calibrate the communicative interaction with the ironist him-/herself. Ascription of an ironic intention is an open field. and the addressee is allowed a broad space to justify hisher interpretative choice. In any case. according to the “fencing game“ model. the subsequent possible ~ ’ ~ ~ z / ~ ?are t ~likely ~ ~ ?toi happen ~ ~ ~ ’ in ~ ’ironic ~ interaction.

L. Anolli et al. /Irony as a Miscommunication Design

151

Misunderstanding - Misunderstanding consists in the failure to give the speaker’s utterance an intention that is different from its linguistic decoding, so that the ironic meaning is not grasped by the addressee. Substantially, misunderstanding takes place when the utterance meaning is limited to the linguistic input, without taking into consideration semantic aspects conveyed by the paralinguistic pattern. Such a failure might depend on the speaker’s lack of communicative competence, when he/she, for instance, resorts to a weak variation of supra-segmental profile. It might also depend on the inability of the addressee to grasp the ironic intent of the speaker, due to a physical or cognitive interference (for instance, a noise or a lack of attention) or for some another reason. Denying - Denying consists in the fact that the interlocutor, though clearly grasping the ironic sense of the speaker’s comment, pretends not to understand it out of convenience and interpersonal opportunity. In this situation he/she prefers to confine hidher understanding to the linguistic input as far as its conventional and standard meaning, canceling information linked to the paralinguistic pattern. As a consequence, this countermove represents a “pretence”, in which the addressee tries to neutralize the ironic sense of speech and to become “unreachable”. In fact, in certain situations it is better and more suitable to be considered “ignorant” (or even “dull”) than to be offended, because in this last case one has to defend one’s own dignity. One does not always feel like having an argument or, even worse, a conflict with other people. Touch6 - When the ironic meaning of a comment hits the target, the addressee can recognize it and admit he/she has been struck. Either the witticism amused himher or the shrewdness of the indirect attack wounded h i d e r . He/she may answer with a smile in the case of kind irony, or he/she may counterattack by retorting in his/her turn in the case of sarcastic irony. Modeling irony as a “fencing game”, giving it a time sequence, can have a different sense according to the degree of the speakers’ empathic involvement, as well as constraints and opportunities of the context of use (contextual multiplicity). Within a context of conflict or of cooperation, the speaker, in accordance with the degree of empathic involvement with the interlocutor, can calibrate what can be taken for granted. According to the situation, he/she can choose to emphasize the ironic intention or to mitigate it, by means of different vocal strategies. A scornful voice passing sharp blame and a tender voice passing gentle praise will emphasize the intention; alternatively, both criticism and approval will converge into a bantering voice, with the intention of playing down the effect. In particular, as Anolli, Ciceri and Infantino [66] have verified, a bantering voice can be defined as either kind or sarcastic, depending on how much the intention to mitigate leads to the softening of praise or blame. Referring to the example, Bob’s ironic comment toward Cassie’s and Dan’s exams can generate different effects. Assuming that both Cassie and Dan realized that Bob was present during their exam so as to know the results, they understood the meaning of his wisecrack. Anyway, they can choose between two options: admitting they understand (touche), or, instead, pretending not to understand (denying). Therefore, although in other cases misunderstanding can be a possible outcome of an ironic comment even for the addressee (e.g., when the addressee is not aware that the speaker knows the facts constituting the focal event, or when the interlocutors do not share a stable background of knowledge), in the above-mentioned example the effect of misunderstanding can take place outside the ironist-ironized’s “inner circle”. In fact, other students, passing by them

L Anolli

152

el

al

1

Irony as a Misc.ommunic.crriortDesign

without knowing the facts, can not attribute an ironic meaning to Bob’s utterances, so they are limited to a superficial level of communication. Concerning the addressees’ reaction, there can be different outcomes. First of all. one should underline the fact that a denying strategy may be a crafty countermove by means of which one can avoid the “face-threatening” effect of the ironic remark; therefore, it is more consistent with sarcastic than with kind irony. Dan can reply to Bob’s ironic blame by sayng: “You’re right. Bob. As usual, I’m the number one. Simply, this time the professor was envious because I was more brilliant than him”. Less effectively, Cassie can reply to Bob’s ironic blame by sayng: “Yes, I’m a failure. This time I was lucky; but I remain a loser”. On the other hand. analyzing the possible reactions as “touchk“, Cassie can reply to Bob’s ironic praise by sayng: “Bob, you’re my wittiest friend! I’m so happy! At last I have passed this difficult exam!”. In this way, she admits Bob’s irony has made her happy. Instead, Dan can say: “Stop being sarcastic, Dan: this time it went wrong; but don’t delude yourself. Next time 1’11 be brilliant once more”.

6.6

Conclusions

The “fencing game” model we outlined herein tackles irony as a miscommunication phenomenon according to the MaCHT fi-amework sketched out by Anolli (Chapter 1, this volume). Irony miscommunication is to be considered as a chance, since it consists in sayng without saying. This “say not to say’’ is in the hands of both the communicators. so that the ironic meaning arises at the end of the communicative exchange. when both of them have shared and participated in building the ironic interaction. Within this perspective irony is taken into account as a powerful device for playing communicative meaning and intention in a round-about way and tangentially. It is a kind of metacommunication because of the plurality of semantic interpretations and it depends on the communicative responsibility of interlocutors to choose a certain route of sense instead of another one. In any case, irony offers the opportunity to switch quickly and effectively from a given route of meaning to another, so that one can always save face. By resorting to an ironic utterance you can always fall on your feet. Moreover, the “fencing game” model of irony allows us to tackle some problematic aspects that we have seen in the previous section about theoretical perspectives on irony. First of all, we have to face the subject of semantic inversion in the ironic utterance which carries with it the problem of distinction between the literal and the figurative meaning as it is shown by Booth [I], Grice [lo], and Knox [ 8 5 ] . The “fencing game” model goes beyond this perspective, because it describes irony as a complex pragmatic phenomenon. which cannot be reduced to a mere rhetorical point of view. Irony is neither a ”strange” nor an exceptional phenomenon deviating from the communicative standards in a given culture. On the contrary, it is an integral part of everyday communication. and it is processed as fast as non ironic statements. Gibbs 1861 has correctly pointed out that irony comprehension does not require a supplementary cognitive effort, as if it were the outcome of a twofold processing: first understanding the literal meaning and then the figurative meaning. When irony comprehension requires more time in its elaboration. it might depend on a lack of contextual references, and not on a previous rejection of the literal meaning, as Colston and Gibbs [87], and Gibbs and Colston (Chapter 8, this volume) have underlined. According to this framework, irony is a merely pragmatic phenomenon. as it is derived via standard inferences ( Attardo, Chapter 7. this volume). Secondly, the “fencing game“ model of irony is related to the theories referring to both the significance of irony in the context. and the echo to social belief$. proposed by Attardo

L. Anolli el al. /Irony as a Miscommunication Design

153

[88], Creusere [34], Giora and Fein [89], Kerbrat-Orecchioni [27], Kreuz [28], and Sperber and Wilson [23, 241. In fact, in ironic meaning-making, cultural and contextual assumptions take priority over textual input, which is absorbed and included by the former. However, textual input can function as linguistic hints, especially when they echo a previous utterance or interactive exchange in an explicit way, and, in such a situation, they may favour the appearance of ironic sense. Given its nature, irony is a metacommunicative comment on what is happening in a certain communicative interaction. In this way, communicators can stand at a distance in a communicative exchange, and they can play a role of pretence, as if they were in a theatrical setting, in line with the perspective sustained by Clark and Gerrig [3 11, KumonNakamura, Glucksberg and Brown [ 181. Also for the “fencing game” model irony contains pretence aspects, but carried out in an everyday communication. The ironist can play the part of the critic, or jester or else humorist and so on. Therefore, irony is not necessarily a ruse, a Machiavellian weapon. On the contrary, it is a natural way of communicating that pervades language, by allowing people to express themselves with effectiveness and vivacity. Of course, most often, although not always and not necessarily, the irony voice takes a specific prosodic profile which signals the emergence of the irony intention by the speaker. In particular, “theatricalism” concerns the supra-segmental profile of irony, and not its pragmatic usage, which is however fairly frequent, as recent researches carried out by Gibbs [67] and Infantino [84], have empirically documented. Such research has verified a considerable presence of ironic comments (around 8%) within everyday conversation, especially among friends. Concerning the management of the semantic field and communication emotional intensity [26, 54, 63, 58, 59, 601, the “fencing game” model points out the manifold nature of the ironic remark as a communicative act. We can underline its communicative convenience as miscommunication. This opportunity is given by the possibility at times to soften the intensity of the communicative intention, or to strengthen it, thanks to the semantic synchrony process, which combines verbal and nonverbal elements together in a unitary route of meaning. In such a way, the “fencing game” model is consistent with those recent theories, proposed by Gibbs [13, 671, Lee and Katz [68], and Leggitt and Gibbs [69], which attribute manifold functions to ironic communication (sometimes as a defence, then as a particularly stinging attack, occasionally as a mere witticism). Summing up, the “fencing game” model may offer some advantages in the study of irony. First of all, we can mention its economic value because it rejects what is useless. In fact, according to this model linguistic input and nonverbal signals, social and cultural conditions are put together and synthesized in a unitary semantic pattern. In ironic communication speakers show their communicative competence by using successfully linguistic and extra-linguistic expressions which effectively express their intentions in an indirect way. The “fencing game” model also involves a combination and amalgamation between implicit and explicit levels, by producing a liaison between referential and inferential competence. Thanks to such a combination, the ironist has to interpret a kind of “roleplaying” in the interaction with hisher interlocutor. If one had to reduce communication to an explicit and transparent exchange of messages, the result would be a naive communication, typical of Dostoevsky’s [90] “idiot”, who is neither stupid nor foolish, but naive. He/she lives detached from the world, oblivious of the moves and pretences played out in the “shadow world” of interpersonal relationships, incapable even of lying. From this point of view, an “idiot” would be outside the constraints of civilized society that others are subjected to, free fiom the conditioning of social rules. Hisher language would be characterized by sincerity, simplicity, essentiality and transparency; maybe a

L .4noiii et n l ’Iron\. as a MiscornmunrcatronDesign

I54

paradigmatic instance of Grice’s maxims [ lo]. Irony is substantially a paradigmatic kind of miscommunication as a c0rnple.r communicative act. It involves a plurality of the semantic and intentional levels of an utterance pronounced by the speaker, without compromising hidher own face and leaving the addressee with the responsibility of choosing a certain route of sense among different options. In this manner, the ironist holds an inviolable position, because he/she can always switch from the communicative meaning to the linguistic input, and vice versa. In some way, the ironist is impervious in hisher position. Irony as miscommunication, in line with the MaCHT fiamework (see Anolli, Chapter 1 , this volume), assumes the shape of a complex and kaleidoscopic communicative resource. According to the occasion, it can be employed in order to stress or to weaken derived meanings. Within social relationships, the ironist, director of hisher own image, can adjust communicative patterns for (un)masking himherself to others, as well as trying to reduce the risk of being “wrapped around the other’s finger”.

6.7

Acknowledgments

We would like to thank Fondazione “Piero, Pietro e Giovanni Ferrero” (Alba. Italy) for the support given to the present study.

References W. C. Booth, A rhetoric of’irony. Chicago: The University of Chicago Press. 1974. H. Lausberg, Elemente der literarischen Rhetorik [Elements of Rhetoric]. Miinchen. Germany: Max Hueber Verlag, 1949. L. Olbrechts-Tyteca, Le Cornique du Dircoi4r.s [The comic in discourse]. Bruxelles: Editions de I’Universite de Bruxelles, 1974. C. Perelman, and L. Olbrechts-Tyteca, Traite de I ’argumentation.La Xoui.elle Rhetorique [Treatise on argumentation. The new rhetoric]. Paris: PUF. 1958. R. J. Kreu, and R. M. Roberts, Two cues for verbal irony: Hyperbole and the Ironic Tone of voice. .Metaphor a nd $vmbolic Actirih- 10 ( 1 995) 2 1 -3 I . L. Wittgenstein, Philosophical investigations. Oxford: Blackwell. 1958. D. C. Muecke, The compass ofiron?,.London: Methuen. 1969. D. C. Muecke, irony. Critical Idiom Series. London: Methuen. 1970. M. Mizzau, I1 falso e il finto [The false and the feigned], in: Menzognu e simuluzione [Deception and pretence], M.A. Bonfantini, C . Castelfranchi. A. Martone. I. Poggi. and J . Vincent. EQ. Napli: Edizioni Scientifiche Italiane. 1997. pp. 12 1 - 129. H. P. Grice. Logic and conversation. in: Sintux and Senrnntics. Spetdl Am. P. Cole and J. L. Morgan. Eds. New York: Academic Press. 1975, pp. 41-58, J. R. Searle, Expression and meaning. Cambridge: Cambridge University Press. 1979. S. Fish, Is there a text in this class? Helikon 29 ( 1983) 246-247. R. W. Gibbs, Speakers’ intuitions and pragmatic theory. Cognition 69 (1999) 355-359. S. Glucksberg, Understanding fiCg~irativt7 Iunguagr: From n1etaphor.s t o i d i o m s . New York: Oxford University Press. 2001 R. W. Gibbs, D. Buchalter, J. Moise. and W. Farrar. Literal meaning and figurati\-e language. Discoiirse Processes 16 ( 1 993) 387-403. G. Lakoff, and M. Johnson. Metaphors M V live h\-.Chicago: University of Chicago Press. 1980. M. Mizzau, L *ironia, La c~onfradrli~iorw consentita [Irony. The permitted contradiction]. Milan: Feltrinelli, 1984. S. Kumon-Nakamura, S. Glucksberg, and M. Brown. How about another piece of pie: The allusional pretense theory of discourse irony. .Journal of E.t-perimentnl P.sjdiolop.: General 124 ( 1995) 3-2 I . L. Deckers. and R. T. Buttram. Humor as a response to incongmitieh u.ithin or between schemata. Hiitnor: intrrnatioriirl Jniiniul o l Hrrmor R ~ w u r ~ c3~ (l 14901 i 5-3-64

L. Anolli et al. /Irony as a Miscommunication Design

[54]

155

S. J. Messman, and D. J. Canary, Patterns of conflict in personal relationships, in: The dark side of close relationships, B.H. Spitzberg and W.R. Cupach, Eds. Manwah: Lawrence Erlbaum, 1998, pp. 121-152. M. Mizzau, Eco e Narciso. Parole e silenzi nel conjlitto uomo-donna [Echo and Narcissus. Words and silences in the gender conflict]. Torino, Italy: Boringhieri, 1979. D. Sperber, and D. Wilson, Les ironies comme mentions [Ironies as mentions]. Pobtique 36 (1978) 399-4 12. D. Sperber, and D. Wilson, Relevance. Cambridge: Harvard University Press, 1986. D. Sperber, and D. Wilson, On verbal irony. Lingua 87 (1992) 53-76. G. Almansi, L’affaire mysterieuse de I’abominable ‘tongue-in-cheek’ [The mysterious case of the abominable ‘tongue-in-cheek’].Poetique 36 (1978) 413-426. G. Almansi, Amica ironia [Friendly irony]. Milan: Garzanti, 1984. C. Kerbrat-Orecchioni, L’ironie c o m e trope [Irony as a trope]. Poetique 41 (1980) 108-127. R. J. Kreuz, The production and processing of verbal irony. Metaphor and Symbol 15 (2000) 99-107. R. J. Kreuz, and S. Glucksberg, How to be sarcastic: The echoic reminder theory of verbal irony. Journal of Experimental Psychology: General 118 (1 989) 374-386. H. Haverkate, A speech act analysis of irony. Journal of Pragmatics 14 (1990) 77-109. H. Clark, and R. Gerrig, On the pretense theory of irony. Journal of Experimental Psychology: General 113 (1 984) 121- 126. L. Hutcheon, Ironie, satire, parodie. Une approche pragmatique de l’ironie [Irony, satire, parody. A pragmatic approach to irony]. Poetique 46 (1981) 140-155. L. Hutcheon, Irony ’sedge: The theoly and politics of irony. London: Routledge, 1994. M. A. Creusere, A developmental test of theoretical perspectives on the understanding of verbal irony: Children’s recognition of allusion and pragmatic insincerity. Metaphor and Symbol 15 (2000) 29-45. M. Bakhtin, Voprosy literatury i estetiki [Topics of literature and esthetics]. Moskow: Chudozestvennaja literatura, 1975. S. C. Levinson, Presumptive meanings: The theory oj’ generalized conversational implicature. Cambridge, MA: MIT Press, 2000. L. Anolli, and M. G. Infantino, La pelle della volpe. L’ironia come strategia di comunicazione conveniente [The fox’s skin. Irony as a strategy to communicate in a seemly way]. Studi Italiani di Linguistica Teorica e Applicata 26 (1997) 57-85. M. G. Infantino, L ’ironia. L ’arte di comunicare con astuzia [Irony: The art of communicating craftily]. Milan: Xenia, 2000. C. Lutz, Ethnographic perspectives on the emotion lexicon, in: Cognitive perspectives on emotion and motivation, V. Hamilton, G. H. Bower, and N. H. Frijda, Eds. Nonvell: Kluwer Academic Publishers, 1988, pp. 399-4 19. A. Wierzbicka, Emotions across languages and cultures: Diversity and universals. New York: Cambridge University Press, 1999. R. Rutelli, Messa in scena e struttura bifasica del Witz [Production and two-phasis structure of the Witz], in: La comunicazione spiritosa [Witty communication], F. Fornari, Ed. Firenze: Sansoni, 1982, pp. 187-215. M. Dolitsky, M. Humor and the unsaid. Journal of Pragmatics 7 (1983) 39-48. M. Dolitsky, Aspects of unsaid in humor. Humor: International Journal of Humor Research 5 (1992) 33-43. 0. Ducrot, Dire et ne pas dire [To say and not to say]. Paris: Hermann, 1972. E. Goffman, The presentation of sewin everyday l$e. Garden City: Doubleday, 1959. E. Goffman, Interaction ritual: Essay on the face-to-face hehaviour. Garden City, NY: Doubleday, 1967. B. Allemann,(1956), Ironie und Dichtung [Irony and poetry]. Pfullingen, : Gunther Neske, 1956. R. Barthes, Fragments d ’un discours amoreux [Fragments of a lover’s discourse]. Paris: Editions du Seuil, 1977. H. Bergson, Le rire. Essai sur la signzjication du comique [The laugh. An essay on the meaning of the comic]. Paris: Alcan, 1900. V. Jankelevitch, L ’ironie ou la honne conscience [Irony or the good conscience]. Paris: Flammarion, 1964. M. Jouhandeau, Cartes de I ’Pcrivain [The writer’s paper]. Paris: Gallimard, 1957. S. Kierkegaard, Om Begrebet Ironi med stadigt Hensyn ti1 Socrates [About the concept of irony, referring steadily to Socrates]. Copenhagen: “Magister Artium” Thesis, 1841. L. Pirandello, L ’umorismo [Humor]. Lanciano: Carabba, 1908. P. Brown, and S. C . Levinson, Politeness. Cambridge: Cambridge University Press, 1987.

,

I56

LA. Atiolli et a1 Iron\- as a Miscommunication Design

M. Merleau-Ponty. Phhnomhnologie de la perception [Phenomenology of perception]. Paris: Gallimard, 1945. [561 J. Jorgensen, The functions of sarcastic irony in speech. Journal 0fPragmaric.s 26 ( 1996)613-634. 1571 L. Anolli, and R. Ciceri, Elementi di psicologia della comunicazione: Processi cogniriri e mpetti srrategici [Elements of communication psychology: Cognitive processes and strategical aspects] Milano: LED, 1995. [581 S. Dews, J. Kaplan, and E. Winner. Why not say it directly? The social hnctions of irony. Discourse Processes 19 ( 1 995) 347-367. ~ 5 9 1 S. Dews. and E. Winner. Muting the meaning: A social function of irony. Metaphor and S\*mholic .1ctivi& 10 ( 1995) 3- 19. S. Dews, and E. Winner. Obligatory processing of literal and nonliteral meanings in verbal irony. foirrnal of Pragmatics 3 1 ( 1999) 1579- 1599. R. Giora, On irony and negation. Discourse Processes 19 ( 1995) 239-264. S. Glucksberg. Commentary on nonliteral language: Processing and use. Metaphor and Simholii. .4ctivi@ 10 (1995) 47-57. H. L. Colston, Salting a wound and sugaring a pill: The pragmatic functions of ironic criticism. Discourse Processes 23 (1997) 25-45. 1631 E. Oring, Humor and the suppression of sentiment. Humor: International Journal qf Humor Researc.It 7 ( 1994) 7-26. L. Anolli. R. Ciceri. and M. G. Infantino, Come non dirlo esplicitamente: Le funzioni comunicative dell’ironia [How not to say it explicitly: The communicative roles of irony]. ..lrchhYo di Psicolopiri. Weurologia e Psichiatria 59 ( 1998) 1 - 19. L. Anolli, R. Ciceri. and M. G. Infantino. Irony as a game of implicitness: Acoustic profiles of ironic communication. Journal of Psycholinguistic Research 29 (2000) 275-3 1 1 . R. W. Gibbs. Irony in talk among friends. Metaphor and Svmhol15 (2000) 5-27. C. J. Lee, and A. N. Katz. The differential role of ridicule in sarcasm and irony. Metaphor and Si.mhol 13 (1998) 1-15. J. S. Leggitt. and R. W. Gibbs, Emotional reactions to verbal irony. Di.sc*our.seProcesses 29 (2000) 1 24. L. Anolli, R. Ciceri, and M. G. Infantino, Behind “dark glasses”. Irony as a strate= for indirect communication. Genetic, Social. and General P.yvcho/ogt*Monogruphs (in press a ) J. €3. Bavelas. A. Black. N. Chovil. and J. Mullett. Equir*oealcommunication. Newbury Park. C.4: Sage. 1990. 1721 K. Dindia. The intrapersonal-interpersonal dialectical process of selfdisclosure. in: Qvnamics (?f reintionships. Understanding relationship proc*esses.id. 4. S. Duck. Ed. Thousand Oaks: Sage Publications, I 994, pp.27-57. 1731 K. Dindia, Sex differences in self-disclosure. reciprocity of selfdisclosure. and selfdisclosure and liking: Three meta-analyses reviewed. in: Balancing the .secret.$ of pri\*ate disclosures. LE.l ‘s communication series. S. Petronio. Ed. Mahwah: Lawrence Erlbaum Associates. 2OOO. pp. 2 1-35. C. R. Berger. Planning strategic interaction. .4 ttaining goals throirglt communicati\x~ ac~ion. Mahwah: Lawrence Erlbaum. 1997. M. E. Bratman, What is intention? In: lntentions in communication. P. R. Cohen. J. Morgan. and M . R . Pollack, Eds. Cambridge: MIT Press. 1990, pp. 15-32. P. R. Cohen. and H. J. Levesque. Persistence. intention, and commitment. in: Intentions in c.ommitnic,atiori.P. R. Cohen. .I.Morgan. and M . R . Pollack. Eds. Cambridge: MIT Press. 1990. pp. [551

33-70,

J. 0. Greene, Tactical social action: Towards some strategies for theory, in: The p.yirholop~ql~tac*rical communication. Monogruph.7 in social p.T?rhologi* ot lungurige. id. 2. M. J . Cody and 31. L. McLaughlin, Eds. Clevedon: Multilingual Matters. 1990, pp. 3 1-47. J. 0.Greene. A second generation action assembly theory. in: Message production: .4d\.ance.s iti communication theoq.. J. 0. Greene. Ed. Mahwah: Lawrence Erlbaum Associates. 1997. pp. 15 1 - 170. P. R. Monge, and M . E. Kalman, Sequentiality. simultaneity. and synchronicity in human communication. in: qlwanric. patterns in c.omniirtiic,utiori pro~’t’,m~v. J , H. Watt and C‘. A . \’anLear. Eds. Thousand Hoaks: Sage. 1996. pp. 7 1-92, B. J . Grosz. and C. L. Sidner. Plans for discourse. in: ltitetitions in cort~m~tr~ic~cltion P. R . C‘ohen. J. Morgan. and M. E. Pollack, Eds. Cambridge: MIT Press. 1990. pp. 417-444. i’. R. Waldron. Toward a theory of interactive conversational planning. in: .We.ssupe production: ..Idi.ance.c i n c,otiiniirtiic~Ntiniirhuon.. J.O. (ireene. Ed. Zlah\vah. N J : I..au+ence Erlbaum .Associates. 1997. pp. 195-220. ,-I, (‘utler. O n \a);ing \\hat yoti mean \+ ithout meaning \\hat you
L. Anolli et al. /Irony as a Miscommunication Design

157

meeting of the Chicago Linguistic Society. Chicago: Department of Linguistics, University of Chicago, 1974, pp. 117-127. I. Fonagy, The Functions of Vocal Style. Literary Style. A Symposium. London: Oxford University Press, 1971, pp. 159-174. M. G. Infantino, The king, the queen and the chessboard? Communication strategies and intimacy degree in cross-gender relationships. PhD Dissertation, Faculty of Psychology, Catholic University of Milan, 2001. N. Knox, The word irony and its context, 1500 - 1755. Durham: Duke University Press, 1961 . R. W. Gibbs, The poetics of mind. New York: Cambridge University Press, 1994. H. L. Colston, and R. W. Gibbs, Are irony and metaphor understood differently? Metaphor & Symbol (in press). S . Attardo, Irony as a relevant inappropriateness. Journal of Pragmatics 32 (2000) 793-826. R. Giora, and 0. Fein, Irony: Context and salience. Metaphor and Symbol 14 (1999) 241-257. F. Dostoevsky, Idiot. New York: Penguin, 1956 (Original work published in 1868).

This page intentionally left blank

159

Saj1 Not to Say

L. Anolli et al. (Eds.) IOS Press, 2002

7

Humor and Irony in Interaction: From Mode Adoption to Failure of Detection Salvatore ATTARDO

Abstract. A fundamental definitional problem is examined for humor and irony: in neither case can the subclasses of these phenomena be kept distinct. This indeterminacy is reduced to the indeterminacy of indirect speech and implicature, on which irony entirely and humor, at least largely, rely. The “performance” of humor and irony is investigated by examining the motivations for Ss to use irony and the responses that Hs produce to it, which range from mode adoption to ignoring it (deliberately or not).

Contents 7.1 7.2 7.3

Definition of the object of study................................. Competence.. ....................................................... Performance.. .......................................................

160 162 166

7.4 7.5

References.. ........................................................ Notes ................................................................

176 179

I60

7.1

S .4ttarcio Hirmor- and Iron\. in Interaction

Definition of the object of study

The study of-humor, irony, and other playful forms is plagued by definitional problems (see for example [ 1, p. 162-1631). Often, authors will expend significant energy explaining and justifylng complex terminological distinctions that are bound to crumble at the first close examination. The impossibility of defining the subcategories of a broad class of humorous phenomena has been established; see Attardo [2, ch. I ] for discussion and review of the literature. This state of affairs is probably motivated by the fact that terms such as humor, irony, sarcasm, junny, laughable, ridiculous, etc., are folk-concepts, with fuzzy boundaries, if any.’ Lexicographic studies have shown that the semantic field of what has been broadly defined as “humor” is very rich in closely related. barely distinguishable terms; e.g., [3]; for a review, see [2, p. 2-10], and [4]. Those involved in the academic study of humor have decided to adopt the generic term humor as an umbrella term encompassing programmatically all the semantic field of humor and humorous forms. Irony is generally seen as distinct from humor, but the same definitional problems exist with its close neighbor, sarcasm, cf. [S, p. 7951 for discussion: see also [6]. Irony would however fall under the technical sense of “humor.” While there clearly exists humor that is not ironical and there are ironies that are not perceived as hnny, the issue is not as simple as the intersection of two distinct sets of facts. While the former claim is fairly obvious, the latter may be in need of argument. Consider a case of very aggressive sarcasm; it seems unlikely that it would be perceived as funny. by the target, as it lacks playfulness. Note that I am merely claiming that it would not be perceived as hnny, not that it would not necessarily he hnny.” A prime example is the following: 58-j?ear-old English prqfessor teaching Writing about Biograph?.. .4 student orall\* answ,ered a qtmtion incorrect!i*. Professor: Now that was brilliant. Student: At least I tried.. . (the student immediately became defensive, while the rest of the class snickered to themselves.)

in which the student obviously attends seriously to the ironical meaning, while the rest of the cIass seems to be at least mildly amused by the incident. Needless to say. we are here merely exemplifjmg the old adage that comedy is tragedy that happens to someone other than ourselves. Yet, for all its trite aspect, this example emphasizes the significant point that the perception of humor is not the same as the decoding of it. Let us move to humor, where Hay [8] has introduced a sophisticated four-level model of humor appreciation. each stage of which presupposes the one to its left: recognition -+ understanding

-+

appreciation -+ agreement

In other words, for H to understand a joke, or any other humorous instance. H must have recognized it as such. Consider the following example of humorous insult [9. p. 461: Meena: I’m the only person in this room who freestyled at nationals and came second in the women’s division Dan: let’s face it Meena always comes second Meena: yeah i know that’s cause Sue Willis always beats me + except at distance

S. Attardo /Humor and Irony in Interaction

161

Dan: we were actually making sexual innuendos well I was (Here and elsewhere, I have respected the typographical conventions of the original. The + indicates a short pause.) It is clear that Meena is completely oblivious to the sexual innuendo (“come second”) and therefore cannot understand the humor, let alone appreciate it or agree with it. The step between understanding and appreciation is particularly interesting in this context. Consider again an example [9, p. 391 in which the players are discussing how to call the player with the least desirable position, in this case, Lisa. To be noted is the fact that Meena is a political science PhD student and that Chris’ remark is directed at her: Chris: why don’t we call it something REALLY bad like + pols student Meena: hey you mother fucker Dan: [to Lisa]: yeah you’re a POLS STUDENT Lisa: I AM though Chris: [draws breath in mock horror] Barney: well that’s done it Dan: accidental faux pas + oh no Note how after Lisa brings up the fact that she is potentially offended by the playful suggestion that being a political science student is worse than being an “asshole” or a “scumbag,” (two terms discussed in the previous segment not reproduced here) both C h s and Dan, who have made the potentially offensive joke, and also Barney, who had no part in the incident, comment in various ways on the inappropriateness of the joke. This corresponds effectively to the “that’s not funny” paradox in which someone claims that something is not funny, thus displaying evidence that something was recognized as funny [lo, p. 58-59]. Note, however, that this example refutes the idea that the claim of unfunniness can only come from a person in a position of power. We will therefore have to distinguish between a humor competence [ l l ] , which corresponds with the capacity to recognize and understand humor, and a humor performance, which is the capacity/desire to appreciate it (and possibly to agree with it). Humor performance corresponds to “humor competence” which is opposed to “joke competence” [12]. This move is controversial (cf. [13, p. 391; [2, p. 1971) in humor research, but commonplace in cognitive science (e.g., [ 141). Humor competence is the capacity of a speaker to process semantically a given text and to locate a set of relationships among its components, such that he/she would identify the text (or part of it) as humorous in an ideal situation. This humor competence is analogous and in fact part of the semantic competence of speakers: being able to recognize a sentence as funny is a skill equivalent (but not identical, of course), for example, to being able to recognize a sentence as synonymous with another sentence. Humor performance is, on the contrary, the actual encounter of two speakers (not necessarily physically copresent), in a given actual place and time, i.e., in a given context. In its simplest prototypical form, speaker A says something and speaker B processes the text (what A said) and, having recognized the humor, reacts by laughing. This picture will have to be complicated later, but for the time being let us note that in order for speaker B to perceive the humor, he/she must, at the very least, have understood A (the cases in which B laughs at A for not being meaningful, etc. are irrelevant). Moreover, B must have recognized A’s intention to be funny, and must have evaluated the situation as appropriate for such a behavior.”’ The issues around the recognition of S’s intention(s) are bound to

S. .4 ttrrrdo /i Humor and Iron\. irt lnferactiorr

162

become one of the central points of contention of the theory of irony and humor. Already. studies have started investigating the issues of metarepresentations necessary in irony [ 151 and even of the “theory of mind” necessary to process irony and humor. especially in brain damaged patients [ 14, 16, 1 7, 1 81. This paper attempts to shed some new light on the issues revolving around humor. irony, and other “humorous” modes by briefly reviewing what is known about these modes, drawing from research in linguistics, sociolinguistics, and psycholinguistics. and moreover drawing on a number of studies that have not usually been tied to these modes (such as teasing and jocular abuse) to try to outline a general account of the domain of humor (in its broad technical sense) that is no longer limited exclusively to data coming from jokes. We will deal in some detail with the vagaries of humor performance in the second part of this paper, on the reception of humor and irony. For the time being, we will instead turn to the semantics (and pragmatics) of humor and irony. The following considerations should be seen as a very short summary of some of the central issues in the field but not as a full treatment of said issues. which can be found in [ 5 . 191. for irony. and [2J. for humor.

7.2

Competence

7.2. I Humor Humor consists of two facets, a semantic and a pragmatic one. These two aspects are intermingled in reality and are presented separately merely for ease of exposition.

7.2.I . I

Semantics of’Humor

Linguistic research of humor has coalesced around a variant of the incongruity theories of humor, known as the Semantic-Script Theory of Humor (SSTH) [ 1 11, later extended in the General Theory of Verbal Humor (GTVH) [20]. This is not the place to go into detail in these theories (see [2, 2 1J for discussion and references) but we can limit ourselves to the fact that humor is seen semantically as an antonymic opposition between two scripts/fiames that are compatible entirely or in part with the text. The details of the implementation of the theory would take us too far afield, and we can only refer the reader to the published discussions of the theory. Let us stress however the formal (or at least. semi-formal) character of the theory, which operates starting with a straightforward semantic analysis and builds the analysis of the humor from this foundation in a nonintuitive manner. 7.2.1.2

Pragmntics of‘Humor

Pragmatically, humor is seen as a violation of Grice’s Cooperative Principle (CP). There is a reasonably large literature on the subject, summarized in [22, 21. It should be noted that humor is a real violation, not a flout or a mentioned violation (cf. [22] for discussion) since the CP is violated without the intention to let H arrive at an implicature. Humor differs from other modes of communication that involve violations of the CP, such as lying. in that its purpose (amusement) is largely approved of socially and that significant amounts of humor are incorporated in everyday conversations, exchanges, etc. Therefore, humor is not seen as an antagonistic mode of communication (such as lyng) but rather as part and parcel of communication. This should not obfuscate the fact that humor as a mode is non-

S. Attardo /Humor and Irony in Interaction

163

cooperative (in the sense that S must violate the CP; this is not to say that people cannot cooperate to tell jokes or be funny).

7.2.2 Irony Irony is a purely pragmatic phenomenon, without semantic counterpart. What this means is that the semantics of an ironical sentence and of a non-ironical sentence are indistinguishable. Naturally, in context, an ironical sentence will acquire (inferentially) a meaning that is minimally different and in fact in most cases opposed to that which the sentence would have in a “neutral” context. Contrast this with humor, as exemplified by jokes, in which very frequently the two opposed senses are present in the text (albeit not necessarily so). Thus, the tradition of focusing on the semantics of humor. Humor (jokes) and irony do not differ radically under this aspect, but merely as a matter of degree. Jokes tend to have a richer semantics than irony, which relies almost exclusively on inferential activation of scripts/frames.

7.2.2.1

Pragmatics of irony

The traditional view of irony as a trope, involving the processing of the text, the rejection of its literal meaning in favor of a different, often opposite, meaning (the two-stages theory) has been challenged [23]. This view, supported by some readings of relevancetheoretic work on irony as mentionhepresentation [24, 25, 261, suggested that the ironical meaning was arrived at directly, without first accessing the literal meaning (direct access, one stage theory). Crucial experimental evidence revolves around processing times: if irony involves an extra step (two stages vs. one), this extra processing should be reflected in processing times. The data reported in Gibbs [23] argue forcefully that irony is processed just as fast as non-ironical statements. However, more recent evidence has contradicted this view [27, 28, 29, 30, 31, 321. Similar conclusions in the processing of familiar and unfamiliar proverbs (which have more or less set meanings) had already been found [33]. Despite the differences about the number of stages involved in the processing of irony, most theories converge around a broad notion of contrast or incongruity between the actual situation and the expectations and/or utterance of S. Colston and O’Brien [34, p. 15631 identify as the central component of irony contrast between the “literal” and the figurative meaning, or “between assertion and reality.” Significantly, they use the general term “incongruity” to cover all the various formulations which they gather under the “contrast” heading “incongruity between a remark‘s assertion and reality” [34]. On the role of contrast in irony, see also [35, 36, 71. Gibbs [37, p. 131 quotes “contrast between expectation and reality” ; (cf. also [23, p. 3971 “incongruity”). My own proposal of “inappropriateness” [5] can probably also be reduced to this broad concept, but has the advantage of being formulated in much more formal(izab1e) terms (i.e., in terms of mismatch of presuppositions). The issue of whether incongruity and inappropriateness are interchangeable is in need of discussion. As we have said, irony is pragmatic (i.e., is derived mostly via implicatures and inferences). This raises the question of how to distinguish it from other non-ironical types of indirect speech. We turn now to this complex issue.

I64

7.2.2.2

S. ..ittardo /Humor nnd Iron). in Interaction

On the epistemological statits of iron?..humor, and indirect speech at large

I have come to believe that the same definitional impossibility within the broad category of humorous phenomena is to be found in the broad category of indirect speech, of which irony is a type. Essentially, all implicature is governed by the same inappropriateness relevance dynamics that lie at the bottom of the traditionaVGricean view of implicature (and hence, irony) as a flout of the CP (i.e., a violation that is redeemed by ulterior upholding). The obvious question then is: If irony is to be described as the mechanism of noticing contextual inappropriateness and calculating via relevance the intended meaning, and so are implicatures in general, what distinguishes irony from other forms of implicature? The answer is nothing, textually speaking." In order to explain this mildly counterintuitive claim, we need to back up a little and describe the mechanism whereby implicatures, which are at the core of indirect speech, are calculated. We start with a situation that is widely assumed, but is in fact problematic. I deal with these issues in [38]. For the time being, we can accept the traditional formulation. S utters u in context C meaning p (the situation is a sextuple S, H, C, u, p . p'). Now, u is inappropriate in context C (i.e., there is a mismatch between the presuppositions embodied in the representation of C that S and H share in common ground and the presuppositions of u; see [ 5 ] for discussion of inappropriateness) and H detects this inappropriateness. Furthermore, he/she does not attribute it to random noise or to the incapacity of S (e-g., S is a child, or impaired in hisher speech behavior, etc.). In order to explain and make sense of S's inappropriateness in u, H enters a largely adductive process whereby he/she "guesses" what S is implicating (p')by taking into account u/p. C, and the CP. Abduction is a kind of inference 1401 that follows the form: D is a collection of data (facts, observations. givens) H explains D (would, if true, explain D) No other hypothesis can explain D as well as H does Therefore, H is probably true [4 1. p. 51 Let us provide an example of the adductive process, which has been unduly neglected in the literature on implicature. Consider the following example:

S and Hare sitting in a room. The phone rings. S: "I am not here." S's utterance is clearly inappropriate, as it violates the maxim of quality (it is obviously false). Since S is assumed to be competent, H reasons as follows: if I make the hypothesis that S is violating the CP to implicate that S wants me to behave as if S were not in the room. and more specifically. since the ringing phone is now extremely salient in our common ground. that S does not want to talk to people over the phone at this time, then S's behavior is explained, and since I have no better hypothesis to explain this behavior. I conclude that my hypothesis must be correct and that is what S meant. The significant fact. for our present discussion of adductive inferencing, is that it is nonmonotonic,' unlike deductive logic. Furthermore, since no restriction is put on the nature of the explanatory hypothesis postulated in the second step of the adducti\,e process. there

S. Attardo /Humor and Irony in Interaction

165

is no way to constrain abduction. Each adductive process is essentially open to an infinite number of possible explanations, from which one (or a few) are selected on the basis of contextual relevance and appropriateness (so, the hypothesis that S does not wish to speak on the phone fits the available facts, whereas the hypothesis that S is undergoing an out of body experience doesn’t or does so to a lesser degree). This is significant, because it means that the adductive processes of implicature generate an infinite number of more or less plausible, interesting, significant, etc., implicatures in a given situation (S, H, C, u, p ) . Much has been made of the indeterminate character of ironical meaning (p’) [42] and the intuition is both correct and deep, especially in its connection to what Sperber and Wilson call “loose talk.” We are now in the position to integrate the indeterminacy of irony, loose talk, indirect speech at large, and such noncooperative (deceptive) modes as humor and lying into a broad explanation: all indirect, non-literal discourse is based largely or in part, on adductive, hence, non-monotonic and open-ended, inferential processes. Therefore, all these modes derive their indeterminacy from the indeterminacy of abduction (Le., we can never be sure that a given meaning is exactly what was implicated, ours is a best guess). Furthermore, we can make another significant step, and note that since the meaning of pl is undetermined, significant leeway is given to its interpretation and eventual classification by H. Consider the following example: A is the 80-year-old grandmother of B, a 30-year-oldfemale. B had just run up over a curb driving out of a store. A : They just built that while you were in the store. B: I know.

It seems clear that A is being ironical, as p (“They just built that while you were in the store.”) is patently false. But is A being funny? Is A teasing B? Is A exaggerating? Is A being metaphorical? The answer is probably a little of all of the above, as A may be humorously teasing B about her lack of attention (or of driving skills) by wildly exaggerating the speed with which a curb could have been built and by metonimically having running over a curb stand for being distracteddriving poorly. Any attempt at disentangling this cluster of implicatures would be quixotic, as the completely noncommittal answer of B seems to imply (although there is no way to know what B actually meant; we only can infer that B is playing along with A, i.e., she is mode-adopting, see below). Thus, returning to the question of the distinction between irony and indirect discourse, we conclude that there isn’t one. Irony is just one of the many expressive goals that can be achieved through indirectness in context, by the speakers. In this sense I am in complete agreement with Kreuz [44, p. 1041 “the job of the listener is to recover the discourse goals of the speaker and not to identi9 some rhetorical label like irony or understatement.” It is probably both the existence of folk-taxonomies that distinguish between, say, irony and exaggeration, and the fragmented way in which the domain of indirect speech has been examined that has led to the loss of perspective whereby phenomena such as irony, understatement, teasing, etc., have been seen as existing outside of the intentions of the speakers and the contexts in which their utterances take place. Significantly, recent studies have started investigating less clear-cut modes than plain irony. Gibbs [23] dealt with hyperbole and understatement in his chapter on irony. Gibbs’ [37] study of naturally occurring data, examines, under the heading of irony, jocularity (i.e.,

I66

S. Attardo / H u m o r and Iron\. in Interaction

teasing), sarcasm (negative irony), rhetorical questions, hyperbole, and understatement. Colston and Keller [45] show that exaggeration and understatement can be explained as degrees of contrast and that exaggeration expresses surprise more readily because one can exaggerate more or less open-endedly. Colston and O’Brien [34] have shown that understatement (meiosis) and irony can be reduced to the same pragmatic factors (contrast, see above). Katz and Pexman [46] and Pexman et a/. [47] have shown that metaphorical statements are interpreted as ironical if certain contextual factors obtain, such as knowledge of S’s occupation. (Certain professions are apparently seen as irony-prone. which is expected for comedians. but perhaps less so for cab-drivers).

7.2.2.3

On the relationship between humor and iron!.

We finally turn to a vexing question, for which we are far fiom being able to provide a definitive answer: how do humor and irony correlate? Dews et a/. [48, p. 3481 speculate that the element of surprise “yielded by the disparity between what is said and what is meant” may trigger humor. Colston and associates (see above) have argued that contrastlincongruity is at the root of irony and other forms of figurative language. Giora [49. p. 256-2571 argues that humor and irony share some basic mechanisms. Namely. they both violate the “graded informativeness requirement.’’ but they do so differently: a joke goes from an unmarked meaning to a marked one, while irony does the opposite. It is unnecessary to review the extensive literature on the role of incongruity in humor in this context (see [2, 501 for reviews). It remains to be seen if the incongruity of humor can be reduced to the contrast of irony. and vice versa. Moreover, it is not clear if such notions as inappropriateness, insincerity, etc.. reduce “cleanly” to contrast. Also. if irony is a form of indirect negation [49] and humor is based (in part) on local antonymy [ 1 1 , 501, it follows that both humor and irony include negation as a significant constituent of the phenomenon. However, Colston [35. p. 15631 argues that “contradiction” is not necessarily involved. The connection between irony and humor at the perlocutionary level is fairly commonsensical and is borne out, for example, by empirical results obtained by Kreuz et al. [51, p. 153-1541 who report that, among the goals listed by speakers in ironical utterances, being hnny or witty and to play or be silly were listed much more fiequently than in the case of non-ironical utterances. Along the same lines, Dews et a/. [48. p. 3631 and Kumon-Nakamura et a/. [52. p. 101 show that ironical statements are rated as funnier than literal ones. Toplak and Katz [53] also find that sarcastic speakers are seen as h n n y . On the connections between humor and irony, see also [44,p. 104; 37. p. 24; 54; 55. p. 40-411. Unfortunately, despite the literature mentioned above, little research has spanned both humor and irony (see for example my own culpable exclusion of irony [23). The linguistics of humor has much to contribute to the discussion, since it seems to have a much more formalizedalgorithmic definition of oppositeness than current theories of irony (cf. [50]. for discussion).

7.3

Performance

We have seen that the situation of indirect speech can be described as a sextuple S. H, C. u. p. p’ and the type of indirect speech (metaphor. irony. exaggeration, understatement. etc.) is determined by the relationship between p and p’. as well as by other salient aspects of C. We now turn to the performance (as opposed to competence. not as in theatrical

S. Attardo /Humor and Irony in Interaction

167

performance) of irony and humor (although we focus on irony in what follows, for simplicity). We have established that the relevant situation is that S says something with a certain intention (or that H reads said intention in what S said).” We can ask two distinct questions: why is S acting thus? and what does H do? The next section will address these issues in turn. 7.3.1

Why use irony and humor?

In [2, p. 322-3301, I classified various functions of humor. The list bears repeating, to compare it with the list of functions of irony [ 191. -

-

Social management: this involves all the in- and out-group functions of humor, including but not limited to mediation, social control, establishing solidarity, play, etc. Decommitment, i.e., the possibility of “taking back” something by claiming that one was “just kidding.” Defunctionalization; i.e., the loss of “meaningfulness” that is observed, for example in puns, but in general in ludic uses of language.

Among the secondary functions are the transmission of information and the testing function that have been claimed to exists in humor (see [2, p. 3301 for references). In the following section, I review the uses of irony, following [ 191, which are perhaps less well known than those of humor. - Group affiliation

Irony may serve two opposed purposes: an inclusive and an exclusive one. On the one hand, irony builds in-group solidarity through shared play; on the other hand, it can be used to express a negative judgment about someone [56, p. 4141. Lakoff [57, p. 1731 notes that “irony makes use of presumptive homogeneity and reinforces it: understanding irony communicates ‘You and I are the same.”’ In other words, shared irony serves to create an in-group feeling. Irony can also be used to exclude [55, p. 95-97; 531.

- Sophistication One of irony’s purposes seems to be that of showing off S’s detachment and hence superiority to/from the situation and S’s ability to “play” with language (saying one thing, while meaning another). Dews et al. [48, p. 3471 show that speakers use irony to “show themselves to be in control of their emotions.” An ironical utterance connotes its being ironical (and indirect), and hence its being sophisticated and requiring some mental dexterity to process it. Being associated with humor adds yet another prized connotation to irony, at least in Western society: being able to make other people laugh is a positive trait (obviously, within certain limits). - Evaluation

Grice [59, p. 531 notes that irony is “intimately connected with the expression of a feeling, attitude, or evaluation.” Sperber & Wilson [42, p. 2391 and many others have claimed that the attitude expressed by irony is always negative. Others maintain that a positive irony is also possible [5, 521. Dews et al. [48, p. 3491 note that irony does mute both the negative effect of ironical criticism and the positive effect of ironical praise (see also [60, p. 151. This muting function would then be the point of using irony. The muting function of irony has been called into question [61, 531.

S .4ttardo 1Hirmor and Irony in Intclruc‘tion

168

- Politeness

There has been much discussion about the use of irony as a tool for politeness [62]. It seems that actually irony itself is aggressive (hence, a FTA), but admittedly less damaging of face than overt aggression. Furthermore. irony offers the option of retractability (see below). which also contributes to its use toward politeness. - Persuasive aspect

Carston [63, p. 301 notes that irony is a powefil rhetorical tool because it presupposes the truth of the presupposed proposition to be self-evident. Kreuz et al. [51. p. 1611 note that irony is memorable. Giora [49] sees irony as a highly informative utterance. All of these aspects of ironical utterances can be used persuasively. - Retractabilih?

Berrendonner [64, p. 2381 claims that irony, because it allows one to state something and its opposite at the same time, allows S to avoid any sanctions that may follow from stating directly what he/she thinks. From this perspective. irony allows S to take a noncommittal attitude towards what he/she is saying. 7.3.I . I

On the uses of humor and iroq.

The functions of humor and irony seem to be largely overlapping. Let us note the striking similarities between the two lists: retractability is listed in both cases, and so is social management (in- and out- group hnctions). Humor is certainly a tool for politeness. cf. the “repair” function of humor [2, p. 3241. Obviously the association with humor of irony. noted above, is a common factor between the two lists, as is the mental dexterity aspect. As far as the differences go, humor does not seem to have the persuasive aspect hypothesized for irony, nor the evaluative one. although, of course, evaluative humor is possible. Nonetheless, listing increasingly disparate and elaborate hnctions of humor (and irony. see below) is of course very interesting empirically, as documentation of a given use of humor, but essentially a theoretically vacuous task. since humor (and irony) can be used to convey any meaning and hence for any social hnction which could be accomplished linguistically otherwise. Consider for example that humor and irony can be used both to exclude from and to include in a group, or that both humor and irony can be used to be polite and to offend. Having reviewed some, but by no means all, of the functions of irony and humor. we are facing the general question (originally intended for irony. but that can be expanded to humor as well): “ r c ~ uh ~speaker ~ M*hocould, Ly. hypothesis, have e.ypressed [hid] her intended message [42. p. directlr. should decide instead to sa?. the opposite of’ what [he/] shr meant [.?]’* 2401.

The answer to this question is that people are willing to take on extra processing costs and the risk of being misunderstood because of a number of social and rhetorical functions that being ironical or funny affords. Moreover, to hedge their bets, so to speak. speakers will signal with a number of markers their ironical intent to make sure that no misundentanding takes place.

S. Attardo /Humor and Iron
169

These markers have been reviewed in [ 19, 65, 66, 671. Some of the central motivations to use irony are discussed briefly in what follows; a more detailed discussion can be found in [19]. Markers of humor have been generally neglected in the literature, with the exception of laughter. 7.3.2 Responses to irony

We turn now to responses to irony. It is clear from the characterization of the ironical situation as S, H, C, u, p , p’ that H may or may not catch on to S’s saying p while meaning p’. Now, if H realizes the existence of p’, then he/she may attend to the meaning of p’ (however indeterminate), while obviously if he/she does not realize that there exists a second proposition implied by S, he/she obviously cannot attend to it. The options of H are as follows: 0

react t o p (the said)

0

react to p’ (the implied)

0

laugh

0

not react (e.g., change the topic, be silent, etc.)

Note how laughter and lack of reaction are ambiguous, in that H does not have to have understood p’ (or even p , for that matter) in order to react thusly. So is the reaction to p , since H may have understood p’, but decided to attend to p anyway (i.e., to take the irony literally). If H has not understood p’, then obviously that option is not available to hidher. Thus, we can come to a first conclusion: observation of the reactions of H (or of the audience) tells us very little, since the only non-ambiguous observable behavior of H is hisher reaction to p’. All other behaviors do not tells us whether he/she is even aware of the existence of p’. In general, this is going to be one of the main problems of a theory of the audience’s reaction to S’s linguistic behavior, and namely the fact that vastly different psychological states may result in exactly the same observable behavior. 7.3.3 Ironical Mode Adoption We will begin by discussing the differences between uptake and mode adoption and as a result present a first definition of “mode adoption.” We will develop a more detailed mental space approach to mode adoption in the next section. To start out we need to discuss briefly the Austinian notion of uptake [68, p. 1171 and its differences from taking up and ultimately of mode adoption . Uptake is the recognition of the intention of S and therefore of the illocutionary force of S’s u. The recognition of an utterance as, say, a promise, is obviously based on its (literal) meaning and the context in which it occurs. Recognition of the illocutionary force of an utterance implies no action beyond the mental effort involved in the recognition itself. Taking up involves some other action on H’s part; minimally a display of assent, more generally a more or less explicit agreement to perform some mutually recognized activity. Mode adoption, on the other hand, is essentially an acceptance on H’s part of a possible world, as defined by S, which differs from W,, i.e., the world that S and H mutually know as “reality” in some respect. By acceptance, I mean the agreement to operate, at least momentarily, within that possible world (I am using here “possible world” and “mental

S. .4trar-doi Humor- mnd Iron!. in lnteracriori

170

space” as interchangeable terms). Taking up involves perlocution and is the more general category. Mode adoption is a specific kind of taking up that involves recognition of illocutionary force and then, rather than the acceptance of the illocutionary force (commitment to perform an activity). the performance of another speech act having the same type of inferential path. but not necessarily the same illocutionary force; responding with a promise to a promise is not mode adoption:

.4:Tomorrow, I will wash your car. B: I will buy you a pizza. Let us define differently what mode adoption is: if H is presented with an utterance il by S, which means (“literally’”) p and implies y and H has available the mode adoption strategy, then H has mode adopted u (or S) if H utters u1 which implies q’ and q and y’ belong to the same mode (irony, metaphor, etc.). In other words. H mode adopts S’s i i if he!she answers with an utterance that is in the same mode (uses the same type of implicature as) the utterance by S. I would like to introduce another term, alongside “mode adoption;” namely. “mode factivity” to indicate the propriety of an utterance to allow mode adoption to H. since by uttering an ironical, metaphorical. etc.. utterance. S opens the possibility for H to reply in the same mode. Consider the following example from Kotthoff which illustrates ironical mode adoption [69, p. 121. The extract, of which I reproduce only the translation, comes fiom a dinner conversation at David’s home. He is known for “prefer[ring] a quiet life” and for letting his wife take care of his social life. The “opulent social life” that Maria (one of the guests) mentions is in reality the fact that David attended two dinners at his own home. Maria: You are having such an opulent social life these days. Dalid: a lot. a lot has been going on lately. because I have taken the initiative now. Thus, we see that David, being teased ironically about his lack of social life, reacts by being ironically self-deprecating [69, p. 12- 131. The following example displays mode adoption in playful insulting. The data comes from Hay [9. p. 5 I ] : Meena: i even made notes about what we were doing last time Dan: good you can read them out for the people who weren’t here Barnel!:i was here last time and I still don’t remember Dan: OK + for the people who weren’t here and the congenitally STUPID B n r n e i ~yeah + never want to be excluded from something

7.3.3.I

Not all implicatirres trigger mode adoption

An important point that needs to be highlighted is that mode factivity is not a general property of implicatures. Mode adoption is not the same thing as responding with the same maxim-flouting triggered implicature [2 I]. The difference is that mode adoption does not require the establishment of the flouting at each step: studies of metaphors have found within metaphorical discourse emergent properties (i.e.. aspects of the metaphor that do not follow from either of the domains

S. Attardo /Humor and Irony in Interaction

171

involved in the metaphor). It follows that once one has entered a metaphorical mode, one is free to pursue it within the internal coherence of that mental space (MS),’” to the point where one can develop novel aspects of meaning [71]. Also, with mode adoption, floutings carry over across discourse boundaries (e.g., turns). So, we can legitimately ask the question: what causes the mode factive properties of irony, metaphor, make believe, fiction, etc.? I believe that the answer lies in the peculiar pragmatic presuppositions triggered to accommodate the incompatibility between the presuppositions of u and those shared by the audience in C. On these issues see [21]. 7.3.3.2

A cognitive/mental spatial account

Simply put, H, when faced with a non-mode factive implicature, derives the necessary implicatures that allow himher to preserve the belief that S is CP-compliant within a MS. When faced with a mode factive utterance, H is forced to open a new mental space that allows H to avoid having to reject u as ill-formed and then proceeds as in the non-mode factive case. In other words, when H encounters, say, a scalar implicature in u he checks the MS labeled “reality”(C), finds an interpretation of u that is compatible with it, and moves on. When H encounters an ironical (metaphorical, etc.) utterance, he/she finds that reality cannot accommodate u, and therefore is forced to construct an ironical MS (I) that allows himher to process u. Note that given my analysis of irony [ 5 ] , combined with the presupposition float principle, we know why the presuppositions in I do not float up to M (the base mental space shared by S and H): since under the model of irony [5], an ironical utterance is pragmatically inappropriate to its context, and inappropriateness is defined as having one or more presuppositions which are incompatible with the presuppositions held in the context of the utterance by its participants, it follows that by the second clause of the presupposition float principle the ironical presuppositions are blocked from floating up into M because they are opposite to some of those in M. This phenomenon is of some significance, since it is at the root of irony recognition on H’s part. Faced with the realization that one or several of the presuppositions in u fail to float upwards in M, H has to conclude that the space I is ironical (or belongs to some other indirect type of speech). This opens up the option to either access the space (mode adoption) or reject it. There have been a few promising studies of H’s reactions to irony, to which we turn presently. 7.3.4 What do Hs react to?

Drew’s [72, p. 2191 influential study of teasing analyzes a corpus of 50 naturally-occurring teasing incidents . He outlines a continuum of reactions to teases (“mocking but playful jibes”) ranging from the totally serious reaction to acceptance, laughter and playing along. The continuum can be summarized as follows: 0

Totally serious reaction

0

Ignoring the tease

0

Serious reaction, followed by laughter

0

Laughter followed by serious reaction

172

S. Attardo t Humor and Ironis in Interaction

Laughter. acceptance of the tease. followed by serious reaction Laughter, acceptance of the tease. playing along In Drew’s [72] data, only three instances of the last case, which corresponds to mode adoption in my terminology, were recorded. Extrapolating from Drew’s data, this would correspond to 6% of the data, a figure strikingly close to that in the Nelms [39] corpus (see below). Drew [72, p. 2301 notes that in general the targets of the teases, even if they laughed or treated the tease playfully. they went “on seriously to re-iect and correct the suggestion made in the tease.” As far as the semantics of the teasing. Drew notes that a vast majority of his examples are characterized by exaggerations and contrast. Hay’s [9, p. 32-34] work on “jocular abuse” (i.e., playful insults) is based on 15 hours of conversations among the players of role-playing game. All players knew one another. obviously, but shared different degrees of intimacy. In the transcriptions of her data. Hay finds 133 instances of jocular abuse (multiple occurrences of the same abuse are counted as one instance). An interesting conclusion is that “the overwhelming majority of abuse involved someone who was hlly integrated in the group” 19,p. 381. We will return to this fact. As far as the reactions to jocular abuse are concerned, Hay’s [9, p. 501 results can be summed up in the chart below. I aggregate her data. which are divided by gender (see Table 7.1 ). Note that the total number of reactions (109) is lower than the 133 recorded instances. as Hay notes that in some cases the victim was not in a position to respond. Note also that. in this case, retaliation corresponds to mode adoption. whereas objecting is clearly a refusal to play along. Conversely, agreeing with the abuser is also mode adoption. So. if we add up the cases of retaliation and those of agreement. we get 22 instances of mode adoption, i.e., 20.2% of the data, with 77.8% (i.e.. more than three quarters of the cases) of serious response andor refusal to play along. Table 7.1 Reactions to ]ocular abuse in Hay‘s study

Kotthoff‘s [73] study is based on a corpus of 30 hours of dinner conversations among fiiends and 20 hours of TV discussions. Kotthoff categorizes the responses as addressing the said (i.e., playing along with the irony), addressing the implied meaning (i.e.. responding seriously), mixed (i.e., addressing both the said and the implied meanings). ambiguous responses (which are unclear in that respect), and simple laughter. From her data, it is clear that Hs by no means react exclusively to the second sense (the ironical import), nor do they “play along.” In fact. a mere 50% of turns play along with the irony (mode adopt) in the more friendly and informal context of the conversations. When we turn to the more formal TV debates, only one instance of mode adoption is recorded. showing that Hs lean toward a serious consideration of the implicated sense in the formal environment. To be noted also the relatively high incidence of laughter (a neutral response. nhich acknowledges the ironv. but does not commit to any reaction).

I73

S. Attardo /Humor and Irony in Interaction

Table 7.2 Results of Kotthoff s study

I

Said

It”

Dinner

TV

import

mixed

umbig.

laugh

Tot.

4

10

5

6

51

14

2

5

2

24

It should also be noted that the significant variable seems to be the friendly nature of the dinner conversations: in a friendly environment the sum of the mode adopting (reacting to the said) and mixed turns makes up about 60% of the reactions to irony. However, in the formal setting, the same combination (said + mixed) amounts to only one response in eight. Averaging between the two sets of data is probably meaningless, but gives us roughly one reaction in two as mode adoption. Gibbs [37] reports a study which looks at irony (construed fairly broadly) over a corpus of 289 ironical utterances collected in transcriptions of taped conversations. Interestingly, the overall incidence of irony (8%) is strikingly similar to the 10% figure found in Tannen [74]. The percentages of responses to the ironical utterances are indicated in Table 7.3. With the exception of understatement ironies (which however make up a very small part of the corpus, and one that seems twice or three times as likely to be missed by Hs) as many Hs react by ignoring the irony than those who react by mode adoption. Table 7.3 Reactions to irony in Gibbs’ study

topic

~

Ironic Literal Laughter Missed Change

1

JoGlarity .32 .22 .I2 .05 .29

Sarcasm .33 .23 .13 .04 .29

Hyperbole .2 1 .38 .2 1

.oo .2 1

Rhet. Quest .2 1 .2 1 .25 .09 .33

Understat. .33 .33 .17 .I7

.oo

In Nelms et al. [39] we find a significantly different pattern of responses: 92.3% (336 out of 364 instances of irony/sarcasm) are single turn, i.e., H does not respond ironically, while there are only 25 instances of sarcastic responses (about 6.8 %), and only three are three turn sarcastic exchanges (i.e., S is sarcastic again after H has responded sarcastically). Moreover, 14 out of the 25 sarcastic responses consist of holophrastic “Yeah” thus believing the dispreferred status of sarcastic responses. It is notable that no example of countersarcasm occurred in the corpus. Countersarcasm is a sarcastic response that unlike ironical assent (“Yeah”) “counterattacks” so to speak, by being sarcastic against S. A cartoon example is reported by Yus [75, p. 1331: a man is sitting on a bridge with a large rock tied to his leg. A passer-by arrives, sees him and says “Christ! You’re not going to throw yourself into the Thames, are you?” The would-be suicide replies: “No! I always sit perched on a bridge with a 20 pound rock tied around my leg, now piss off! I don’t need any good samaritan garbage. Thanks!” In the last frame, the passer-by replies: “No! I was

174

S. .4itardo / Humor und Iron\, in Interaction

going to tell you, your rope is to [sic] long and the tides [sic] gone out.” The pattern of responses to irony/sarcasm found in the Nelms et al. [39] study is reproduced below. (Some utterances had multiple reactions). Kinesic responses include rolling the eyes, shaking the head in disbelief. squinting. blushing, giving a dirty look. etc. Table 7.4 Reactions to sarcasm in Nelms’ study

.Won-sarcastic 336

Sarcastic 28

Laughter142

Xo response

52

Smile 24

Kinesic 29

Most work on conversation analysis has stressed the cooperative. playlid. construction of ironies in group settings [76, 691. Gibbs [37, p. 251 comes down on the side of the conversation analysts: “addressees responded to a speaker’s irony by saying something ironic in return [to a large degree],” although he seems to believe that this is a novel finding. In contrast, Nelms et al. [39] sees evidence of the Least Disruption Principle. postulated in Attardo [77] independently of irony, which claims that violations of the CP should be kept to a strict minimum and therefore that ironical utterances should not be responded to with irony. Supporting this view is Hay’s [S] claim that irony does not require humor support. Humor support is defined as signs of appreciation, acknowledgement. sympathy. etc.. that follow instances of humor. One form of humor support is to offer other humor. It is clearly too early to determine which position is more correct, especially when quantitative studies on naturalistic data are a mere handful. A long-term study in this direction is currently underway [78]. However, we may speculate that the differences in the results may be due to the differences in the corpora. Kotthoff s data show clearly that more formal situations result in less irony and in less mode adoption. Gibbs’ data were recorded in conversations between friends and roommates [37. p. 1 I ] whereas Nelms’ data were collected in a broader range of situations and participants. ranging from family members to strangers. It seems probable the irony is a practice one is more likely to engage in n.ith people one is intimate with than with strangers. Consider these independent, but converging sources: in Kotthoffs data. both the number of ironical turns and especially the mode adoptions drop significantly when the data switch from a friendly environment to a formal setting. In Hay’s data, jocular abuse is largely limited to those persons that are well integrated in the group. Jorgensen 1791 showed that sarcasm is more common with intimates. Furthermore, another dimension which needs to be accounted for is the degree of aggression of the ironical utterance. Gibbs [37] does not report degree of aggression. but he reports which utterances were seen as “critical”: half of the sarcastic utterances were seen as critical, whereas about 25O4 of the other categories were classified as such (once more with the exception of understatements, which were never seen as critical). Complete data on aggression in the Nelms corpus are not available yet. but a sample of 60 randomly selected cases was ranked for aggression by 20 judges. The rankings show that on average the samples were seldom considered very aggressive or aggressive. but that a majority of cases were considered slightly aggressive. In general. the sample shows a tendency toward not being very aggressive. However, our data also clearly show some aggression. It seems that aggression in sarcasmirony has a “bulge” distribution [ S O . chart 7 . I ] . If these tentati\.e data can be trusted. \ve can speculate that both factors are significant

S. Attardo /Humor and Irony in Interaction

175

for the production and reception of humor and that therefore there exists a threshold of familiarity, below which humorous exchanges tend to be avoided.

Chart 7.1 Bulge Distribution of Aggression

Perhaps also the familiarity dimension exhibits a bulge, in which case the “humorous space” would be structured as a blob, the outcome of the intersection of the two bulges, orthogonally placed. Attardo [78] will investigate this hypothesis. For the time being, the hypothesis was tested by having two independent judges rate the 60 samples for familiarity, using the following scale: Strangers 0

Acquaintances (e.g., classroom interactions, between teacher and students)

0

Close acquaintances, familiar (e.g., coworkers)

0

Friends Relatives, married couples

The resulting scatter plot having aggression on the x axis and familiarity on the y axis, according to the judges, is reproduced in chart 7.2. The results seem to support the threshold hypothesis. Needless to say, a more sophisticated research tool is necessary, for example, eliciting familiarity rankings directly from the participants, thus avoiding the mediated rankings of the judges.

S. Attardo /Humor and /ran\: in heraction

176

a

1 -’

I) 4

Chart 7.2 Scatter Plot of Familiarity ( Y axis) and Aggression ( X Axis)

7.3.5 Reactions to humor

After reviewing the recent research on the reactions to irony, we now turn in conclusion to the reactions to humor, where we find that there have been few studies of H’s reactions to non-ironical humor. In general, laughter has been the object of sustained attention as well as silence (i.e., ignoring the humor); see Attardo [2, p. 307-31 I ] for references. The discourse analysis of humor has focused mostly on successful exchanges in which the hearers laugh or respond with more humor [76, 8, 121, i.e., Hs mode-adopt. Some studies [8l] have been concerned with performance (in the theatrical sense) of humor and touch upon audience reaction. Very little attention has been given to failed attempts at humor. Theoretically, there should be little difference between irony and humor in the different possibilities of reaction, since H can ignore the humor, take it literally (this is called “suppression of the violation” [22, p. 5531, laugh, or mode adopt, just like with irony. This concludes our discussion of the performance aspect of humor and irony. Needless to say, it is clear that in many ways these first early conclusions are merely hinting at a much richer and rewarding field that is in dire need of further investigation.

7.4 [1]

131 [ 71

References J Holmes, Politeness. power and provocation how humor functions In the workplace. Ditcorrrw Sfudrey 2 (2) (2000) 159-1 8 5 S Attardo. Linguistic Theone7 of Humor Berlin Mouton de Gruyter. 1994 i\‘3chmidt-Hidding. Wit and Humor. in Hrrrnot i t n d H*itz. %‘ Schmidt-Hidding. Ed Ilunich

S. Attardo /Humor and Irony in Interaction

[41

PI C6l [71

[91

u71

177

Hueber, 1963, pp. 37-160. W. Ruch, 1998. Paper presented at the International Society for Humor Studies conference. S. Attardo, Irony as relevant inappropriateness, Journal of Pragmatics 32 (6) (2000a) 793-826. C. Lee and A. Katz, The differential role of ridicule in sarcasm and irony, Metaphor and Symbol 13 (1) (1998) 1-15. A. Utsumi, Verbal irony as implicit display of ironic environment: distinguishing ironic utterances from non-irony, Journal of Pragmatics 32 (12) (2000) 1777-1806. J. Hay, The pragmatics of humor support, HUMOR: International Journal of Humor Research. Forthcoming. J. Hay, Jocular abuse patterns in mixed-group interaction, Wellington Working Papers in Linguistics 6 (1994) 26-55. R. Barreca, They used to call me snow white ...but I drifted. Women’s strategic use of humor. Harmondsworth: Penguin, 1991. V. Raskin, Semantic Mechanisms of Humor. Dordrecht: D. Reidel, 1985. A. Carrell, Joke competence and humor competence, HUMOR: International Journal of Humor Research 10 (2) (1997) 173-185. L. Olbrechts-Tyteca, Le comique du discours. Bruxelles: Editions de 1’Universitede Bruxelles, 1974. B. Bara, M. Tirassa, and M. Zettin, Neuropragmatics: Neuropsychological constraints on formal theories of dialogue, Brain and Language 59 (1 997) 7-49. C. Curcb, Irony, negation, echo and metarepresentation, Lingua. 100 (4) (2000) 257-280. E. Winner, H. Brownell, F. Happe, A. Blum, and D. Pincus, Distinguishing lies from jokes: Theory of mind deficits and discourse interpretation in right-hemisphere brain-damaged patients, Brain and Language 62 (1 998) 89- 106. R. Giora, E. Zaidel, N. Soroker, G. Batori and A. Kasher, Differential effects of right- and lefthemisphere damage on understanding sarcasm and metaphor, Metaphor and Symbol 15 (2000) 63-83. S. McDonald, Neuropsychological studies of sarcasm, Metaphor and Symbol 15 (2000) 85-98. S. Attardo, Irony Markers and Functions: Towards a Goal-Oriented Theory of Irony and its Processing, Rusk. Internationalt tidsskriji for sprog og kommunikation 12 (2000b) 3-20. S. Attardo, and V. Raskin, Script theory revis(it)ed: joke similarity and joke representation model, HUMOR: International Journal of Humor Research 4 (1991) 293-347. S. Attardo, On the pragmatic nature of irony and its rhetorical aspects, in: Pragmatics in 2000: Selected papers from the 7th International Pragmatics conference. Vol. 2, E. Nkmeth, Ed. Antwerp: IPrA, 2001, pp. 52-66. S. Attardo, Violation of Conversational Maxims and Cooperation: The Case of Jokes, Journal of Pragmatics 19 (1) (1993) 537-558. R. Gibbs, The poetics of mind. Cambridge: Cambridge University Press, 1994. D. Sperber and D. Wilson, Irony and the use-mention distinction, in: Radical Pragmatics, P. Cole, Ed. New York: Academic Press, 1981, PP. 295-3 18. D. Wilson and D. Sperber, On verbal irony, Lingua 87 (1992) 53-76. D. Sperber and D. Wilson, Irony and relevance: A reply to Seto, Hamamoto, and Yamanashi, in: Relevance Theory: Applications and Implications, R. Carston and S. Uchida, Eds. Amsterdam: Benjamins, 1998, PP. 283-293. R. Giora, Understanding figurative and literal language: The graded salience hypothesis, Cognitive Linguistics 7 (1 997) 1 83-206. R. Giora and 0. Fein. Irony: Context and salience,Metaphor and Symbol 14 (1999) 241-257. S. Dews and E. Winner, Obligatory processing of literal and non-literal meanings in verbal irony, Journal of Pragmatics 31 (12) (1999) 1579-1599. J. Schwoebel, S. Dews, E. Winner and K. Srinivas, Obligatory processing of the literal meaning of ironic utterances: further evidence, Metaphor and Symbol 15 (2000) 47-61. S. McDonald, Differential pragmatic language loss after closed head injury: Ability to comprehend conversational implicature. Applied Ps-ycholinguistics 13 (3) (I 992) 295-3 12. S. McDonald and S . Pearce, Clinical insights into pragmatic theory: Frontal lobe deficits and sarcasm, Brain and Language 53 ( I 996) 8 1- 104. N. Turner and A. Katz, The availability of conventional and of literal meaning during the comprehension of proverbs, Pragmatics and Cognition 5 (2) (1997) 199-233. H. Colston and J. OBrien, Contrast and pragmatics in figurative language: Anything understatement can do, irony can do better, Journal of Pragmatics 32 (11) (2000) 1557-1583. H. Colston, “Dewey defeats Truman”: Interpreting ironic restatement, Journal of Language and Social Psychology 19 (1) (2000)46-65. H. Colston, Contrast and assimilation in verbal irony, Journal ofpragmatics (Forthcoming).

178

S. Attardo /Humor and Ironis in Interac*tioti

[371 R. Gibbs, Irony in talk among friends. Metaphor and Svrnhol 15 (20oO) 5-27. ~381 S. Attardo, Meaning ,,. impliciture/explicature. implicature, and script-based semantics: on the (lack of) boundary between semantics and pragmatics, in: Empirical, Cognitive-based Studies I n the Semantics-Pragmatics Interface, J. Andor and P. Pelyvis, Eds. Amsterdam: Elsevier. (Forthcoming) ~391 J. Nelms (Brinkley). D. Boxer, and S. Attardo. The least disruption principle: Sarcasm revisited. Paper presented at the Fourteenth Annual International Conference on Pragmatics and Language Learning. April 13-15 . 2000, Urbana-Champaign. IL. U. Eco and T. Sebeok. Eds. The Sign of' Three: Diipin. Holmes. Peirce. Bloomington: Indiana University Press, 1983. J. Josephson and S. Josephson. Abductive h1erunce: Computatiott, Philosophr.. Twhnolop. Cambridge: Cambridge University Press, 1984. D. Sperber and D. Wilson. Relei'ance: C'omniunicatiori and Cognirion. Cambridge: Hanard University Press. 1986. H. Levesque. Knowledge Representation. Cambridge: MIT Press, 1992. R. Kreuz, The production and processing of irony, Metaphor and Svmhol I S (2OOO)99-107. H. Colston and S. Keller. You'll never believe this. Irony and hyperbole in expressing surprise. Journal of Psvcholinguistic Research 27 ( 1998) 499-5 1 3. A. Katz and P. Pexman. Interpreting Figurative Statements: Speaker Occupation can Change Metaphor to Irony, Metaphor and Svmhol12 (1 ) ( 1 997) 1 9-4 1 . E471 P. Pexman, T. Ferretti and A. Katz, Discourse factors that influence online reading of metaphor and irony. Discourse Processes 29 (3) (2000) 201 -323. S. Dews. J. Kaplan. and E. Winner. Why not say it directly'.) The social functions of irony. Disc~oirrst, P ~ O C P ~ 19 S . (1995) T ~ S 347-367. R. Giora. On irony and negation. Discourse Processes 19 ( 1995) 239-264. S. Attardo, The semantic foundations of cognitive theories of humor. HL'.bfOR: In~erna~ionai Joctrrtul qf'HumorResearch 10 (4) ( 1997) 395-420. R. Kreuz, D. Long, and M. Church. On Being Ironic: Pragmatic and Mnemonic Implications. .Metaphor and Svmholic ..Ictivin, 6 ( 3 )( I99 1 ) 149- 162. ~531 S. Kumon-Nakamura, S. Glucksberg and M. Brown. How about another piece of pie: The allusional pretense theory of discourse irony. Journal I ) / .E.rprrimental P . ~ ~ r h o l o fGeneral l.: 124 ( 1 ) ( 1995) 321. 32 (10)( 2 0 0 ) 1467[531 M. Toplak. and A. Katz. On the uses of sarcastic irony. Journal ot'Pra,qmntic.~ 1488.

~ 5 4 1 M. Creusere. A developmental test of theoretical perspectives on the understanding of verbal irony: Children's recognition of allusion and pragmatic insincerity, Metaphor and $i.mhollS (2000) 29-45. % Mizzau. I. L 'ironia: L i r cotitraddizionc consentita [Irony: The allowed contradiction]. Milan: Feltrinelli, 1984. A. Myers-Roy. The function of irony in discourse. Text 1 ( 4 ) ( 1981 1 407-423. R. Tolmach Lakoff. Talking power: The politics yf'language. Basic Books: Harper Collins. 1990. C. Shelley, Forthcoming. The bicoherence theory of situational irony. C'ognitiir Science. H.P. Grice, Studies in the ir*aj.yf'kiwds. Cambridge: Harvard University Press, 1989. S. Dews and E. Winner. Muting the meaning: A social function of irony. .\feraphor and s\mholic. ricti\igy 10 (I) (1995) 3-19. H . Colston. Salting a wound o r sugaring a pill: The pragmatic functions of ironic criticism. Di.vcmr.sr Processes 23 (1997) 25-45. K. Barbe. lronJ*in Corire.vt. Amsterdam: Benjamins. 1995. R. Carston. Irony and Parody and the Use-Mention Distinction. The .2httin,ghoni Litipiristic. C'ircmlrrr~ 10 (1) (1981) 24-35. A. Berrendonner, Elements de pragmutique linguistiyuc [Elements of pragmatic linguistics]. Paris: Les editions de minuit. 198 I . L. .4nolli and M. Infantino. La pelle della volpe. L'ironia come strategia di comunicazione conveniente [The skin of the fox: Irony as a favorable communicative stratep]. Studi lraliani di Linguistica Teoricu e Applicata 16 (3) ( 1997) SO 1-528. L. Anolli, R. Ciceri and M. Infantino, Irony as a game of implicitness: Acoustic profiles of ironic communication. Journal o/'P.sJ,c-holinguisticResearch 29 ( 3 )(2000) 775-3 1 1 . ~ i ~ i . v t i29 c ~ 6 7 1 P. Rockwell.. Lower. slower. louder: Vocal clues of sarcasm. Joirrnul o f ' P . ~ ~ ~ c ~ l t o l i t i ~ResearcA ( 5 ) (2000) 483-495. J . 'Austin. Hoit. To Do Thir1,q.s Il'ith IZiirds. Oxford: Oxford University Press. 1962. t f , Kotthoff. Irony. quotation and other forms of staged intertextuality: Double o r contrasti\ c perspectixation in con\.ersation. / t i f c J t m . f i r ) t icind l-;qpi.$lk,. k i c t i r r c ~5 f 149%)1-2-

S. Attardo /Humor and Iron-v in Interaction

179

G. Fauconnier, Les espaces mentaux. Paris: Editions de Minuit, 1984. Engl. trans. Mental Spaces. Cambridge: MIT Press, 1985. G. Fauconnier, Mappings in Thought and Language. Cambridge: Cambridge University Press, 1997. P. Drew, Po-faced receipts of teases, Linguistics 25 (1987) 219-253. H. Kotthoff, Responding to Irony in Different Contexts, Journal of Pragmatics (Forthcoming). D. Tannen, Conversational Style: Analyzing Talk Among Friends. Nonvood: Ablex, 1984. F. Yus, La interpretacion y la imagen de masas [Mass interpretations and image]. Generalitat Valenciana. Instituto de cultura "Juan Gil-Albert" (Diputacion provincial de Alicante), 1997. N. Norrick, Conversational Joking: Humor in Everyday Talk. Bloomington: Indiana University Press, 1993. S. Attardo, The place of cooperation in cognition, in: ECCS '99: European Conference on Cognitive Science 1999, Sebastiano Bagnara, Ed. Siena: Universita di Siena. Istituto di Psicologia, pp. 459-464. S. Attardo, Aggression and familiarity in irony. In preparation. J. Jorgensen, The functions of sarcastic irony in speech, Journal of Pragmatics 26 (5) (1996) 613-634. N. Wolfson, The bulge: a theory of speech behavior and social distance, in: Second Language Discourse: A Textbook of Current Research, J. Fine, Ed. Nonvood: Ablex, 1988, pp. 21-38. J. Rutter, Stand-up as Interaction: Performance and Audience in Comedy Venues. Unpublished Ph.D. dissertation. Salford: University of Salford, 1997.

7.5

Notes

I See for example the prototype theory of irony in Utsumi [7], in which irony is a function of three variables. The problematic aspects of Utsumi's proposal cannot be discussed in this context, unfortunately. I' I am suggesting essentially that all irony is potentially, but not necessarily actually, humorous. This is because irony shares both the script oppositeness and overlapping requirement of humor; see below. 'I' There are numerous issues which we cannot address in this context: the intentionality of the humor may be distributed across the participants, i.e., the hearer may be the only person intending the exchange to be funny, as in so-called "unintentional humor" (and this analysis itself is not without its controversial aspects), the judgements as to the meaning of the text, the appropriateness of joking, the speakers' intentions, etc. need not occur sequentially and/or consciously, or even at all in routine situations in which expectations are set. I could go on.. . I" Pragmatically, of course, there is a residual violation of the maxim of manner, even after the ironical implicature has been derived. I deal briefly with this issue in Attardo [21] and in Nelms et al. [39]. " i.e., essentially probabilistic, based on the best available evidence and subject to revision. Cf. Levesque [431. "' It should be noted that we are restricting ourselves for simplicity to verbalized humorhrony. There exist situational humor and irony which can probably be reduced to the present formulations. We will not deal with these in this context, but see Shelley [58] on situational irony, and references therein. "I1 The notion of MS was introduced in Fauconnier [70]. This is not the place to discuss this idea, but see [70], [71]. Very briefly we can characterize a mental space as an abstract domain to which information is "relativized" ([70: 381). Thus S may believe he/she is a human being and a lion (metaphorically, or in a dream) without contradiction, by relativizing the contradictory information to two MSs. Fauconnier proposes also a principle of presuppositional projection, called presupposition float which may be stated as "a presupposition floats up until it meets itself or its opposite" ([70: 611).

This page intentionally left blank

181

Say Not to Say L. Anolli et al. (Eds.) IOS Press, 2002

8

The Risks and Rewards of Ironic Communication Raymond W. GIBBS, Herbert L. COLSTON

Abstract: This paper explores the risks and reward of ironic communication. We argue that irony can not be characterized simply as having positive or negative social impact, but can serve multiple communicative purposes, depending on the social context and aims of the conversational participants. Irony may either distance, or bond, speakers and listeners. Contrary to the standard view, understanding irony does not require that people analyze what speakers literally say before deriving their intended meanings. But interpreting irony may require listeners to recognize the second-order, metarepresentational beliefs of speakers. Finally, irony may be a special form of figurative language in the pleasures it affords speakers and listeners.

Contents 8.1 8.2 8.3 8.4 8.5 8.6 8.7

The risks and rewards of ironic communication.. .............. Irony may distance speakers from listeners.. .................... Irony may bond speakers and listeners.. ......................... How is irony understood?. ......................................... The pleasures of irony.. ............................................ Inferring irony where it may not exist.. .......................... Conclusion.. ..........................................................

182 182 183 185 187 190 192

8.8

References. ...........................................................

193

182

8.1

R. U'. Gihhs and H. L Colston ' Risks and Relspard.7 of'lroriic.Communication

The risks and rewards of ironic communication

Irony is a frequently employed form of figurative language that offers speakers many risks and potentially great rewards. When people utter ironic remarks, they assume that others have some ability to infer speakers' communicative intentions. For example. Mary is upset with her husband John for failing to assist in the housecleaning and says to him while he sits idly on the couch "Well, you're a big help around here.'' Mary's remark indirectly conveys her belief that John should be helping with the housecleaning, as well as her annoyance with John for his failure. Speaking ironically in this situation enables Mary to convey complex propositional and interpersonal meanings in a compact manner. Of course, Mary risks John misunderstanding her. because of her nonliteral utterance (i.e.. John may take Mary's utterance on face value as a compliment). Moreover, even if John recognizes Mary's nonliteral intentions, Mary risks offending John by adopting a caustic. hostile attitude in expressing her beliefs sarcastically. The above example illustrates some of the risks and rewards associated with ironic communication. Our goal in this chapter is to explore the unique pragmatic characteristics of irony. We describe why speakers use irony, what listeners typically understand. and misunderstand, when encountering ironic language, how irony is understood. and why irony is particularly useful in conveying both humorous and hostile attitudes in everyday communication. In addition, we claim that people sometimes interpret speakers' messages ironically even when this meaning was never intended. 8.2

Irony may distance speakers from listeners

Consider again the case where Mary says to her husband "Well, you're a big help around here." Many scholars argue that sarcastic irony, such as seen in this example, is especially negative in distancing speakers from listeners (ref). Under this view, Mary's utterance is seen as more negative and hostile than if she stated her complaint more directly. as in "I wish you would help me." or "You're not helping me." Qualitative observation suggests that people do, in fact, sometimes use irony for hostile purposes that distances them from their listeners. Various studies indicate that verbal aggression through sarcasm is directly linked to people's feelings of anger and loss of selfesteem [ l , 21. Some individuals may be more prone than others to angry feelings when hearing sarcasm [3]. Thus, people possessing temperaments viewed as "guardian" (i.e.. melancholic. depressive, industrious, traditional)) have the greatest propensity to react angrily to sarcasm, followed by "idealists" ( i.e, inspired. religious. receptive. friendly). "artisans" (i.e., sanguine, innovative, aesthetic. changeable), and "rationals" (i.e., skeptical. curious, theoretic, tough-minded). respectively. There is some data. from a study examining sarcasm in one workplace setting (staff members at a hospital), that provides partial support for this hypothesis [ 3 ] . Thus, sarcastic irony appears to have negative effects on some social relationships, and is especially hurtful to some individuals. A tragic instance of the negative impact sarcasm can have was the shooting deaths of 13 highschool students by two alienated male teenagers in Colombine. Colorado in 1999. These two adolescents had for years been the object of their classmates' frequent teasing and sarcastic comments. Not all scholars view irony as distancing speakers from listeners. These researchers claim that irony works to mute the negative impact of speakers' meanings [4. 51. For instance, Mary's sarcastic utterance may be less offensive to her husband than if she stated her complaint directly. Irony. in this view. softens the potentially negati\.e impact of \vhat a

R. W. Gibbs and H.L. Colston /Risks and Rewards oflronic Communication

183

speaker intends to communicate. Allcom formulated a list of typical aggressive behaviors in the workplace and observed that sarcasm is an acceptable method for expressing anger, at least in some situations (and compared to lesser condoned behaviors such as bullying, slander, and physical violence) [ 6 ] . Sarcasm is favored in the modem workplace because the serious risks associated with such verbal behavior appear minimal insofar as speakers' comments may be subtle, and even humorous. Experimental studies presents a mixed picture of irony's effects on interpersonal relationships. Some studies suggest that irony does minimize the potentially negative impact of speakers' messages [4]. Yet other studies seem to demonstrate the opposite in that irony appears to enhance the negativity of a comment [7, 81. For instance, some work indicates that the positive direct semantic meaning of a speaker's ironic utterance (e.g., "How pleasing") enhances the degree of criticism perceived in a speaker's message, relative to more direct literal commentary (e.g., "How disgusting"). This enhancement seems to be brought about by a contrast effect, in which the referent situation looks worse when juxtaposed with the ironic remark, again relative to the literal comment. This causing of a shift toward the negative then carries the speaker greater criticism. It also appears that this seeming discrepancy in the literature can be accounted for by subtle contrast and assimilation effects [9]. Some ironic remarks make a referent situation appear positive relative to the situation being seen in isolation or when a literal remark is made about it. Other ironic remarks shift the perceived quality of the situation to the negative as described above. These different shifis are caused by assimilation and contrast effects respectively, and affect the degree of speaker criticism accordingly.

8.3

Irony may bond speakers and listeners

Speaking ironically may have special rewards, because it allows people to communicate "off-record" [ 10, 111. Off-record speaking strategies enable speakers to deny their covert communicative intentions if questioned by someone else. Thus, in the above brief scene, Mary could possibly deny aspects of what John understood from her sarcastic comment if John complained or raised questions about what she meant. But speaking "off-record" using irony has other rewards for both speakers and listeners, beyond allowing speakers to deny the implications of what they say. Consider the following exchange between two college students [ 121. This conversation occurred in their apartment, and focused on some visitors who were staying with them at the invitation of another roommate: Anne: "By the way, were our wonderful guests still here when you came out and ate lunch? Dana: "I had a sandwich and ... 'I Anne: "Isn't it so nice to have guests here?" Dana: "Totally!" Anne: "I just love it, you know, our housemates. They bring in the most wonderful guests in the world and they can totally relate to us." Dana: "Yes, they do." Anne: (laughs) "Like I would just love to have them here more often" (laughs) "I so I can cook for them, I can prepare" (laughs) Dana: "to make them feel welcome." Anne: "Yeah, isn't this great, Dana? Like today I was feeling all depressed and I came out and I saw the guests and they totally lightened up my mood. I was like the happiest person on earth. Dana: "unhuh"

I84

R W Gihhs and H L Colsron Risks and Rewards oflronic Communication 1

Anne: "I just welcome them so much, you kno~y,ask them if thev want anything to drink or eat" (laughs)

Anne and Danak conversation reveals the intense joy that speakers of irony often share. Each person employs different forms of ironic language (e.g., sarcasm, jocularity. rhetorical questions, hyperbole) to indirectly convey their mutual displeasure about the people staying as guests in their apartment. Much of the irony here is humorous, despite its implied criticism of the visitors (and the roommate who invited them). Both speakers are tightly linked in their beliefs and attitudes (i.e., their kommon ground"), and this allows them to pretend that they enjoy their roommate's visitors when this is obviously untrue. In this case, then, irony serves as a mark of intimacy between speakers and listeners. and brings them even closer together. How often do speakers use irony for hostile or humorous purposes'? There are surprisingly few studies on the amount and kind of irony that people use in ordinary conversation. But one detailed analysis of over 60 ten-minute conversations between college students in public and private settings showed the diversity of ironic language [ 121. This corpus revealed 789 different ironic statements as students made ironic remarks 8% of the time. This fact alone shows that irony is not a specific rhetorical device only to be used in unusual circumstances. Yet this analysis of students' ironic utterances demonstrated that irony is not a single category of figurative language, but includes a variety of types. each of which is motivated by slightly different cognitive, linguistic, and social factors, and conveys somewhat different pragmatic meaning. People comfortably use various forms of irony to convey a wide range of both blatant and subtle interpersonal meanings. For instance, students used jocularity (i.e., "I would just love to have them here more often"), where speakers addressed one another in humorous ways 50% of the time; they used sarcasm (e.g., "I would just love to have them here more often"). where speakers spoke positively to convey a more negative intent, 28% of the time; they used hyperbole (e.g.. "They bring in the most wonderful guests in the world and they can totally relate to us"), where speakers express their nonliteral meaning by exaggerating the reality of the situation. 12% of the time; rhetorical questions (e.g., "Isn't it so nice to have guests here?"). where speakers literally ask a question that implied either a humorous or critical assertion. 8% of the time; and understatements (e.g., "James was just a bit late with his rent"), where speakers conveyed their ironic meanings by stating far less than was obviously the case. 2% of the time. Although the conversations analyzed in this study may not be representative of all situations, irony is clearly, often employed for jocular, humorous purposes that bond speakers and listeners closer together. Speakers engage in pretend acts where they adopt different persona to communicate complex interpersonal meanings. Listeners do not always pick up on speakers' ironic messages (about 4% addressees' subsequent utterances suggest that they may have failed to comprehend speaker's ironic meanings). This reflects one of the downsides of ironic communication. Yet addressees frequently responded to speakers' ironic statements by saying something ironic in return (as shown nicely in the above conversation between roommates). This averaged from 21 percent to 33 percent across the five types of irony. Speakers of irony share ironic views of people and events that are jointly extended, and celebrated, upon as the conversation unfolds. This data is consistent with the claim that people use different forms of irony in various discourse situation because they conceptualize situations in ironic terms [ 13, 141. Under this perspective. the reasons for speaking ironically are not solely located in trying to convey specific nuances of meaning. but because people \,lei\ situations ironically and their ironic talkc reflect5 this figurative mode of thought. To the

R. W. Gibbs and H.L. Colston /Risks and Rewards of'lronic Communication

185

extent that speakers and listeners share a similar ironic construal of events, their understanding of what speakers ironically imply by what they say will be greatly facilitated.

8.4

How is irony understood?

In recent years, psychologists, linguists, and philosophers have proposed various theories to explain how people use and understand irony. These theories focus on widely different cognitive, linguistic, and social aspects of ironic language use, even though each theory claims to provide a single umbrella for capturing the essence of irony. The classic pragmatic theory of Grice and Searle maintains that listeners first analyze the literal meanings of speakers' utterances, see these meanings as contextually inappropriate, and then derive the correct nonliteral interpretation given the cooperative principle or the rules of speech acts [ 15, 161. A speaker's tone of voice supposedly provides an important cue to listeners in inferring ironic meaning. A different theory holds that irony is a type of echoic mention, in which speakers echo, or repeat, a previously stated utterance or belief, which in context is recognized as conveying ironic meaning [ 17, 181. Readers, in fact, find it easier to process and judge the ironic meanings of utterances when they echo or paraphrase some earlier statement or commonly-held belief [ 19,201. Another proposal suggests that pretense is the key to irony [21, 221. Under this view, speakers of irony pretend to be some other person or persona and pretend also to be speaking, in some cases, to some person other than the listener. Pretense is not unique to irony. Many forms of indirect, figurative language, including hyperbole, understatement, and certain indirect requests, may communicate their meanings effectively because listeners recognize the nature of a speaker's pretense [ 131. A fourth possibility holds that ironic utterances mostly accomplish their communicative intent by reminding listeners of some antecedent event, even if not all such reminders are echoic or refer to actual or implied utterances [23]. Many ironic remarks merely remind listeners of the attitudes and expectations that they might share with speakers. The allusional pretense theory combines features of both the echoic mention and pretense view by proposing that ironic utterances convey pragmatic meaning by alluding to failed expectations, which is usually achieved by violating the maxim that speakers should be sincere in what they say [24]. One problem with empirical studies of irony understanding is that they mostly address a single factor in support of the authors' primary hypothesis. Few studies systematically explore the above theories within a single set of experiments. This difficulty is certainly understandable because the different theory characterize dissimilar aspects of ironic language. But one study analyzed how the 60 student conversations to see whether there were differences in the degree to which the speakers' ironic utterances involved echo, pretense, or were spoken in a special tone of voice [ 121. This analysis revealed that people using sarcasm and hyperbole adopted pretense much more so than they echoed a previous statement, while speakers of jocularity employed pretense and echo mention with near equaI frequency. An important difference across the various types of irony concerned asymmetry. For jocular utterances, speakers more frequently presented a negative statement to convey a positive message than they spoke positively to express negative meaning. On the other hand, a far greater number of sarcastic utterances were stated positively to convey negative messages than the reverse. Speakers often used various

186

R W Gihhs and H L Colston 1 R i . h and Reward.y of Ironic Communication

special tones of voices with each type ironic utterance, especially with sarcasm, jocularity, and rhetorical questions. It was difficult to determine whether there was a specific intonation pattern associated with each type of irony, and in some cases. speakers did not appear to state their ironic messages using any special tone of voice. Other experimental research suggests that some intonation patterns may be useful in people's communicating ironic messages 125, 261. One implication of the Gibbs findings is that no single theory of irony is currently capable of accounting for the diversity of ways in which ironic language is understood [ 121. This conclusion should not be terribly surprising given the pragmatic and linguistic complexity of irony. There is even an additional pragmatic factor in irony understanding that scholars are now beginning to recognize and study. As described above, pretense is an important, but not necessary, element in people's successful use of irony. For example, in the conversation between the two college students, Anne at one point says "Isn't it so nice to have guests here?" Anne's rhetorical question only pretends to convey her agreement with the implied assertion. Dana immediately extends the pretense by responding "Totally," which prompts Anne to continue "I just love it, you know. our housemates. They bring in the most wondehl guests in the world and they can totally relate to us." Each of these statements are instances of "staged communicative acts" [ 2 1, 271. The college students' scenario is staged in the sense that the Anne, creates for her listener, Dana. a brief improvised scene in which an implied Anne (one actor in the scene) asks a question of an implied Dana (the other actor in the scene). When Dana continues the pretense. she assume the role of co-author of the hypothetical scenario by making an assertion in which an implied Dana performs a sincere utterance within the play. as it were. for an implied Anne. As co-authors of this hypothetical scenario. both Anne and Dana wish for each other to imagine the scene and to appreciate their pretense in staging it. It appears, then, that pretense it fundamental to the teasing, and jocular irony that Anne and Dana communicate. By engaging in pretense. Anne and Dana enable themselves to conceptualize of the visiting guests in a nonserious manner. despite their displeasure with these guests, which should, even if momentarily, help defuse the potentially emotional issue of how annoyed each of them are with the guests. When Anne alludes to. or echoes. some attributed utterance or thought of another person, she creates a representation of a representation (i.e., a second-order belief). By alluding to these implied beliefs. Anne and Dana must recognize the second-order nature of the beliefs if they are to understand what each other intends to communicate. Metarepresentational reasoning of this sort is a special characteristic of ironic communication. Second-order inferences about speakers' intentions do not simply refer to the traditional idea that irony conveys messages that are opposite of what is literally said (see [I31 for a critique of this traditional view). Instead, inferring second-order inferences involves recognition of a thought about an attributed thought. Do listeners infer complex metarepresentations when understanding irony? Colston and Gibbs examined the cognitive processing and the resulting meaning products associated with understanding statements that conveyed ironic. metaphoric, or metaphoric irony in slightly different discourse situations [28]. Participants read stories ending with simple statements, such as "This one's really sharp'' that, depending on the context. conveyed either ironic (e.g., talking about a dull pair of scissors) or metaphoric (e.g., talking about a smart student) meaning. For the ironic context, the teacher's comment "This one's really sharp" reflects her critical attitude toward the idea that the scissors in question might have once been sharp, and considered so. Understanding that the teacher is alluding to some attributed thought or utterance of another individual (a second-order belief). demands that listeners draw a complex metarepresentational inference. Several scholars now contend that iron); and metaphor differ precisely because irony comprehension require5 wch

R. W. Gibbs and H.L. Colston /Risks and Rewards of Ironic Communication

187

metarepresentational reasoning in the way that metaphor does not [29, 301. We hypothesized that ironic meaning would be more difficult to interpret than metaphoric meaning because of the second-order inferences needed when irony is used. A first study showed that readers take longer to comprehend these statements when used as ironies than when conveyed metaphor. A second study probed people's conscious understandings of the different inferences that constitute part of listeners metarepresentational understandings of irony. The results showed that people recognized irony as involving more pretense, more allusion to someone's prior beliefs, more reference to a speaker's multiple beliefs, and more mockery of someone's prior beliefs, than was the case when people read metaphor. These data are overall consistent with previous suggestions that irony understanding requires more complex metarepresentational inferences than does interpreting metaphor [29, 30, 13, 14, 311. Irony understanding specifically demands that listeners draw a second-order inference about a speaker's beliefs in a way that it is not necessary when interpreting metaphor. The idea that ironic communication depends on metarepresentational makes sense given that speakers often echo or allude to others' beliefs when making ironic utterances. Misunderstanding a speaker's ironic remark often is due to listeners' failure to recognize the second-order nature of a person's belief given what that individual has just said. This problem is one of the great risk in ironic communication. On the other hand, correctly recognizing the second-order belief inherent in any ironic remark gives rise to rich, nuanced meaning that often has great rewards for speakers and listeners.

8.5

The pleasures of irony

One finding from the Gibbs study, reported above, was that speakers of sarcasm were significantly more critical and mocking of others than were speakers of jocularity, hyperbole, and rhetorical questions [ 121. Almost all the ironic utterances, with the exception of understatements, were viewed as humorous. Most notably, there appears to be a strong association between an ironic utterance mocking someone or something and it being viewed as humorous. One of the advantages of irony is that it allows speakers to mock others, especially third parties that may not be present. The conversation between the college students, Anne and Dana, who mocked the visiting guests, is a prime example of how people use irony to mock others and enjoy doing so. Another example of how irony mocks absent third parties is seen in the following advertisement. This ad, sponsored by the California Department of Health Services, played in May 1998 on California radio stations. The radio spot is spoken in the voice of a 60year old man in a very sincere tone: "We the Tobacco Industry, would like to take this opportunity to thank you, the young people of America, who continue to smoke our cigarettes despite Surgeon General warnings that smoking causes lung cancer, emphysema, and heart disease. Your ignorance is astounding, and should be applauded. Our tobacco products kill 420,000 of your parents and grandparents e v e v year. And yet, you've stuck by us. That kind of blind allegiance is hard to find. In fact, 3,000 ofyou start smoking everyday because we tobacco .folks tell you it's cool." (Starts to get carried away). "Remember, you%e rebels! Individuals! And besides, you impressionable little kids are makin us tobacco guys rich!! Heck, we're billionaires!!" (Clears throat/Composes himsellj).

I88

R. W. Gihhs and H.L. Colston /Risky and Rewards of'ironic Communication

"In conclusion, we the tobacco conglomerates of America, owe a debt of gratitude to all teens .for their continued support of our tobacco products despite the unfortunate disease and death thev cause. Thank you .foryour understanding. Thank -vou .for smoking. Yours trulv, The Tobacco Industn?." The goal of this advertisement is clearly to make teenagers aware of the persuasive techniques employed by the Tobacco industry to seduce youth to start smoking cigarettes. Ji-ony is particularly useful here for its rhetorical powers to gain listeners' attention. to mock the Tobacco industry for its underhanded methods, and thereby bond listeners with the implicit narrator (Le, the California Health Department). Of course, listeners have varying reactions to this radio message. Many people, especially non-smokers who decry the way the tobacco industry seduces children to smoke, see great humor in the speaker's irony (e.g., "That kind of blind allegiance is hard to find"). Undoubtedly, smokers, and perhaps most teenagers, may feel differently about this message. But no matter how one responds emotionally, the irony here seems more effective in getting listeners' attention. and making them react in a different manner than would be the case if the speaker pleaded his case using literal speech. This relationship between irony and humor arises from people's shared recognition of a contrast and/or contradiction between diverse sets of beliefs. Although there is much debate concerning the viability of the various theories put forth to explain humor. there is nevertheless agreement that propositional, expectational, or other forms of discontinuity, which often manifest themselves in the form of contrasts or contradictions, provide the basis for many instances of humor. Many theories of situational irony agree that contrast and/or contradiction is a necessary condition for irony (see [32] and [ 141 for reviews on verbal and situational irony respectively; see also (331 for a discussion of the necessity of an "ironic environment"). Empirical research has provided evidence for the importance of contrast in the relationship between verbal irony and humor. For instance. one study presented people with written scenarios and statements of verbal irony, that were created to portray a range of levels of contrast between what was described in the scenarios and what was stated by the speakers [34]. Thus, a situation may first be described as being moderately negative: "Rene and her housemate Jeff went grocery shopping for a big party they were having. They wheeled a cart packed full of food up to the checkout to find out that Jeff, who was supposed to have paid, hadn't bought any money." A speaker in the scenario would then comment about the situation with an utterance that was moderately positive, such as Rene saying to Jeff, "Well, this presents us with a great situation." These utterances were referred to as "verbal irony". Other scenarios presented the moderately negative situations with comments made by speakers that were only slightly negative, such as "Well, this presents us with a slight dilemma." These kinds of comments were called "understatement". Still other scenarios, again with moderately negative situations, were presented with comments that were consistent with this degree of negativity, such as "Well. this presents us with a big dilemma." These comments were considered "literal". Participants read these scenarios and gave ratings on the degree of contrast between the comments and their referent situations, and assessments of the humor in the comments. Both measures revealed a predictive relationship between the degree of contrast set up in the commentdsituations, and the particular measure. As the manipulated contrast increased, so did the rated degree of contrast and humor. As the manipulated contrast increased. so did the rated degree of contrast and humor. Thus. there appears to be a direct.

189

R. W. Gihbs and H.L. Colston /Risks and Rewards of Ironic Communication

and linear, relationship between the amount of contrast perceived as present between a remark and it's referent situation, and the amount of humor taken from that instance of commentary. Another study replicated this general finding [351. Participants read moderately negative situations, such as "Fred is really looking forward to a relaxing day at the beach with his girlfriend. He picks her up, but then his car runs out of gas on an empty road on the way to the beach. He then remembers that his brother used the car the day before and probably didn't buy gas. Fred tells this to his girlfriend and says At this point, participants saw either a literal remark (e.g., "This is just terrible"), a hyperbolic comment (e.g., "This is the worst luck ever"), or an ironic comment (e.g., "This is just wonderful"), and again rated the degree of contrast between the scenario and statement and the amount of humor in the final comment. Ironic comments were rated as having the most contrast and the most humor, followed by hyperbolic comments and then literal comments. Another study extended this general finding even further with a comparison between weak and strong forms of irony [35]. This study also more clearly established that the degree of contrast was the causal factor in rated differences in the humorousness of the final comments. Participants were presented with a moderately negative situation, such as "Sheila was looking forward to her boyfriend Walter's visit. When Walter arrived, he was in a terrible mood and was snapping and yelling at Sheila and her housemates. Sheila turned to him and says...: At this point, participants read and gave contrast and humor ratings for either a literal comment (e.g., "Aren't you in a bad mood?"), a weakly ironic remark (e.g., "Aren't you in an agreeable mood?"), or a strong ironic comment (e.g., "Aren't you in a magnificent mood?"). These "weak" and "strong" ironic comments respectively contained moderately positive terms (e.g., agreeable, good, pleased, etc.), and more extremely positive terms (e.g., magnificent, absolutely brilliant, more happy than anyone, etc.). For the analysis, the comments were divided into two groups based upon a different set of ratings given by a separate group of participants. These participants read each weak irony and strong irony term for each item and rated how sirniladdifferent their meanings were. Based upon these ratings, we created two groups of items. One group contained the half of the items for which the weak irony and strong irony versions were least different. The other group contained the items for which the weak and strong irony versions were most different. We then conducted analyses comparing literal, weak irony, and strong irony separately for these two halves of the items. For the items in which the weak and strong ironic version were rated as being least different from one another, contrast and humor ratings by the main In all of the analyses, both kinds of irony were rated as having greater contrast and humor than literal comments. Thus, again, as contrast between statement and scenario increases, so does readers' perceptions of irony as humorous. This pattern holds for remarks that present no contrast (literal), slight contrast (understatement), moderate contrast (hyperbole), somewhat stronger moderate contrasts (weak irony), even stronger contrasts (irony) and relatively extreme contrasts (strong irony). Other studies investigated the role of contrast in provoking humorous reactions to auditorially produced irony [36]. In this study, people listened to a series of short, taperecorded stories that described negative situations, such as: "Your sister Jenny and you are watching a television program where a man who was just about to propose to his girlfriend drops the engagement ring down the toilet. Jenny leans over to you and says One of the characters in the story then made a comment, such as "How romantic." Contrast was operationally defined in this study as the degree of positivity present in the comments, which was manipulated in two ways. First, half of the comments contained ...I'

...'I

190

R R' G i h h and H L Colsron /Risks and Rewards of Ironic-Communication

words with positive meanings (e.g., "How romantic"), where the other half used words with negative meanings (e.g., "How disgusting"). Secondly, half of the comments were spoken with positive intonation (e.g., the actors who created the items on the tape were instructed to speak as if they felt very positively, but to then read the words used for the particular utterance), where the other half were spoken with negative intonation (e.g., the actors were instructed to speak as if they felt very negatively). These two different ways of manipulating the degree of contrast in the utterances each independently worked to enhance the amount of humor perceived in the comments. The results showed that semantically positive comments were rated as funnier than semantically negative utterances. Phonetically positive comments were also judged to be funnier than phonetically negative ones. Similar to written irony, people find more humorous pleasure in spoken irony when these statements expressed greater contrast. The pleasure people take from irony is one of the great rewards found in verbal interaction. But ironic pleasures are also found when people encounter non-linguistic stimuli. In one study investigating peoples' interpretations of imagistic irony, experimental participants were shown a series of images that depicted ironic meanings 1371. The ironic images were derived from a variety of sources, primarily newspapers and magazines, and were selected if they depicted irony without the use of dialogue (e.g., a drawing of a tire service & repair truck stranded with a flat tire). Text was allowed, but only if it was just a contribution to the description of a scene (e.g.. a label on the truck. "Al's Tire Company"). If text was spoken or otherwise involved, then the image was not used. The participants in the study viewed the images in a booklet. and then gave ratings on a variety of factors concerning the images. One group of participants first read a textual description of each image prior to viewing that image (e.g.. "A tire-delivery truck"). These participants were instructed to first read the description, then to imagine what the image would look like based upon the description, then to turn the page and look at the image. and finally to rate how different the actual image was from what they had imagined it to be. The attempt here was to measure the degree to which the images contrasted with people's prior expectations. A separate group of participants viewed the images alone, without textual descriptions, and rated the humorousness of each image. Statistical analysis found that these two measures were significantly correlated. As the difference between expectations and the actual images increased, or put differently. as a form of contrast increased, so did the degree of humor interpreted from the images. These studies demonstrate the importance of contrast in people's experience of pleasure with ironic messages. Such a relationship may arise again because contrast appears to be a common underpinning of humor and irony. Discontinuity is often noted as a common denominator of humor, and contrast is once prevalent form of discontinuity. Thus. irony may be experienced as pleasurable because both the irony and the humor present situations that are discontinuous with expectations.

8.6

Inferring irony where it may not exist

The prominence of irony in both thought and language gives rise to some interesting case of miscommunication. People often interpret what others say as irony even though this meaning was never intended. Consider an excerpt from the online publication "Slate" on February, 4 2001, titled "The Perils of Irony. No, Really." This article addressed the controversy over a popular television show and how some American youths interpreted part of it as ironic. "'Jackass' is the name o f a show on MTV in \vhich a guy named Johnny Knoxville and

R. W. Gihbs and H.L. Colston /Risks and Rewards of'lronic Communication

191

his cohorts perform a variety of gross-out stunts and pranks, For example, in an episode that aired last month, Knoxville put on a fire-resistant outfit. This outfit was also covered with steaks. Knoxville was tossed onto a barbecue, sprayed with lighter fluid, and burned. Of course, he had the fire-resistant suit on, so he wasn't injured. A 13-year-old boy who subsequently attempted the same thing, without the special suit, was injured and is reportedly in the hospital with severe burns on his legs. "Now Sen. Joseph Lieberman has written a public letter to Viacom (MTV's parent) attacking the "exploitative and degrading" show. Among other things, he complains that while there is a don't-try-this-at-home disclaimer at the beginning of Jackass , the warning is "self-mocking and trivializes the seriousness of the stunts' potential consequences." "According to a 'Variety' story the disclaimer MTV runs is this: "The following show features stunts performed by professionals and/or total idiots under very strict control and supervision. MTV and the producers insist that neither you nor anyone else attempt to recreate or perform anything you have seen on this show." Apparently it's that phrase "and/or total idiots" that sends up, and thus nullifies, the rest of the language around it. It's a classic example of ironic distancing, a sort of disclaimer within the disclaimer." The author of this article went on to say "It's hard to read such a disclaimer, however straight the wording, and not smirk or even snicker. But that's not the fault of irony." This example of TV watchers, especially adolescents, misunderstanding a warning as being ironic is not farfetched. Listeners and readers often debate with themselves whether some linguistic statement was really intended ironically. Moreover, in a time when ironic modes of thought pervade public consciousness, and influence may aspects of popular culture, it is almost surprising to see people speak literally! Such a phenomenon is illustrated in the following excerpt taken from a popular cartoon, "The Simpsons": Homer is appearing as a cannon-ball-in-the- stomach freak in the sideshow part of the Hullapalooza Tour (obviously taken from the once popular Lollapalooza Concert Tour). The announcer introduces him on the stage as a marvel of nature, one of the greatest acts ever. One teen says something on the order oJ "Yeah right, greatest act ever." A second teen a s h something like, "Did you really mean that, man, or are you just being sarcastic?", to which the first teen replies, "I dunno, man. I can't tell anymore." Thus, the prevalence of irony in contemporary thought pushes people to view events/speech/writing/art as ironic, even when these may not always be originally intended to convey ironic meaning. There are times when listeners recognize some statement as ironic, even though they also understand that a speaker did not intend for what was said to be taken in this way. Gibbs, O'Brien, and Doolittle investigated people's understanding of irony that was unintended [38 1. Consider the following two situations: John and Bill were taking a statistics class together. Before the final exam, they decided to cooperate during the test. So they worked out a system so they could secretly share answers. After the exam John and Bill were really pleased with themselves. They thought they were pretty clever for beating the system. Later that night, a friend happened to ask them if they ever tried to cheat. John and Bill looked at each other and laughed, then John said, "I would never be involved in any cheating." John and Bill were taking a statistics class together. They studied hard together, but John was clearly better prepared than Bill. During the exam, Bill panicked and started to copy answers from John. John didn't see Bill do this and so didn't know he was actually helping Bill. John took the school's honor code very seriously. Later than night, a friend happened to ask them if they ever tried to cheat. John and Bill looked at each other, then

192

R. W Gihbs and H.L. Colston / Risks and Rewards of ironic Communicatiott

John said, "I would never be involved in any cheating." Both of these situations end with the identical statement that in each case is understood as verbal irony. The speaker in the first story specifically intends for his audience to understand what is said as ironic, but the speaker in the second situation does not intend for his utterance to be understood ironically. In the second story, only the addressees and overhearers see the irony in what the speaker actually says. It is quite possible for people to understand a speaker's utterance as irony even though the speaker did not intend the utterance to be understood as irony. Several experimental studies show that people understand utterances in stories like the second one above as having ironic meaning even if the speaker did not intend for the utterance to be understood in this way [38]. In fact, readers rated the final statements in the unintentional stories as being more ironic than were the final statements in intentionally ironic stories. Thus, although irony often reflects speakers' communicative goals to identify aspects of ironic situations, speakers may unintentionally create irony by what they say. These studies on understanding unintentional irony demonstrates an important point about cognitive processes in ordinary language interpretation. People not only can misunderstand a speaker's communicative intentions, they can also consciously recognize. in some cases, that the meaning they have inferred diverges from what speakers intended to convey. The fact that people can both understand what a speaker meant and some other unintentional meaning suggests that some aspects of understanding unintended meanings are connected to people's interpretations of what speaker's meant to communicate. In the irony studies just described, people may have specifically drawn "unauthorized" inferences by virtue of first recognizing the speaker's communicative meaning in context.

8.7

Conclusion

Irony can be a risky way to communicate one's meanings. During the last ten years, there have been tremendous debates among public intellectuals over the "age of irony'' that we all appear to live in. Speakers, writers, artists, and others have taken irony to new heights in highlighting some of the many contrasts, and the failure of reality to meet our expectations. we see in everyday life. Countless controversies have risen over whether or not someone's communicative, and artistic, intentions should best be understood seriously or ironically. We all risk that our ironic intentions may sometimes be misunderstood. To give one example, in February 2001, there is an argument in contemporary music circles about the rap artist Eminem who often expresses violent, homophobic. and misogynist beliefs in his lyrics. Eminem has been nominated for a prestigious "Grammy" award and his recent album has sold many millions of copies. Critics claim that his lyrics are deeply offensive, but Eminem responds that his lyrics are "just a joke'' and that his real fans "don't take them seriously." Yet many people argue in turn that whatever irony the rapper intends in his songs are not clearly understood as such by most of his fans. Others maintain that Eminem's cry that his lyrics are ironic is only an after the fact defense of his originally hateful intentions. Determining whether anyone's speech is ironic can be tricky business. Debates about irony in public discourse will likely always be part of the struggle for certainty in meaning. But despite the many risks that ironic messages entail, it is equally clear that irony has tremendous rewards and benefits interpersonal communication. People use irony to evoke complex social meanings and to bond themselves with others. often celebrating shared beliefs in a humorous way.

R. W. Gibbs and H.L. Colston /Risks and Rewards ofIronic Communication

8.8

193

References

A. Anastasi, N. Cohen, and D. Spatz, Study of fear and anger in college students through the controlled diary method. Journal of Genetic Psychology 73 (1948) 243-249. G. Gates, An observational study of anger. Journal of Experimental Psychology 9 (1926) 325-33 1. K. Calabrese, Interpersonal conflict and sarcasm in the workplace. Genetic, Social, and General Psychology Monographs 126 (2000)459-494. ~ 4 1 S. Dews, J. Kaplan, and E. Winner, Why not say it directly? The social functions of irony. Discourse Processes 19 (1995) 347-367. [51 S. Dews, and E. Winner, Obligatory processing of literal and nonliteral meanings in verbal irony. Journal of Pragmatics 31 (1999) 1579-15. 161 S. Allcorn, Anger in the workplace. Westport: Quorum Books, 1994. ~ 7 1 H. L. Colston, Salting a wound or sugaring a pill: The pragmatic functions of ironic criticism. Discourse Processes 23 (1997) 25-45. H. L. Colston, Comprehending speaker intent in rebuttal analogy use: The role of irony mapping, absurdity comparison and argumentative convention. Language and Speech (in press-a). ~ 9 1 H. L. Colston, Contrast and assimilation effects in verbal irony comprehension. Journal qf Pragmatics (in press-b). R. Gibbs, Intentions in the experience of meaning. New York: Cambridge University Press, 1999. P. Brown, and S. Levinson, Politeness. Cambridge: Cambridge University Press, 1987. R. Gibbs, Irony in talk among friends. Metaphor & Symbol 15 (2000a) 5-27. R. Gibbs, The poetics of mind: Figurative thought, language, and understanding. New York: Cambridge University Press, 1994. ~ 1 4 1 J. Lucairello, Situational irony: A concept of events gone awry. Journal of Experimental Psychology, General 123 (1994) 129-145. H. Grice, Studies in the way of words. Cambridge: Harvard University Press, 1989. J. Searle, Metaphor. In: Metaphor and thought, Andrew Ortony, Ed. New York: Cambridge University Press, 1979, pp. 92-123. ~ 1 7 1 D. Sperber, and D. Wilson, Relevance: Communication and cognition (2nd edition). Cambridge: Blackwell, 1995. D. Wilson, and D. Sperber, On verbal irony. Lingua 87 (1992) 53-76. R. Gibbs, On the psycholinguistics of sarcasm. Journal of Experimental Psychology: General 115 (1986) 3-15. J. Jorgensen, G. Miller, and D. Sperber, Test of the mention theory of irony. Journal of Experimental Psychology: General 113 (1984) 112-120. H. Clark, Using language. New York: Cambridge University Press, 1996. H. Clark, and R. Gerrig, On the pretense theory of irony. Journal of Experimental Psychology: General 113 (1984) 121-126. R. Kreuz, and S. Glucksberg, How to be sarcastic: The echoic reminder theory of verbal irony. Journal of Experimental Psychology: General 118 (1989) 374-386. S. Kumon-Nakamura, S. Glucksberg, and M. Brown, How about another piece of pie: The allusional pretense theory of discourse irony. Journal of Experimental Psychology: General 124 (1995) 3-121. L. Anolli, R. Ciceri, and M. G. Infantino, Irony as a game of implcitness: Acoustic profiles of ironic communication. Journal of Psycholinguistic Research 29 (2000) 275-3 1 1. P. Rockwell, Lower, slower, louder: Vocal cues of sarcasm. Journal of Psycholinguistic Research 29 (2000) 483-489. ~271 R. Gibbs, Metarepresentations in staged communicative acts, in: Metarepresentations, D. Sperber, Ed. New York: Oxford University Press, 2000b, pp. 233-261. H. Colston, and R. Gibbs, Are irony and metaphor understood differently? Metaphor and Symbol (in press). D. Sperber, Understanding verbal understanding, in: What i s intelligence? J. Khalfa, Ed. New York: Cambridge University Press, 1994a, pp. 179- 198. D. Sperber, The modularity of thought and the epidemiology of representations, in: Mapping the mind: Domain specifcity in cognition and culture, L. Hirschfeld and S. Gelman, Eds. New York: Cambridge University Press, 1994b, pp. 39-67. E. Winner, and H. Gardner, Metaphor and irony: Two levels of understanding, in: Metaphor and thought, A. Ortony, Ed. New York: Cambridge University Press, 1993, pp. 425-443. H. L. Colston, On necessary conditions for verbal irony comprehension. Pragmatics & Cognition 8 Ell

194

R W Gihh nnd H L Colvton / Rivks and ReM a r d ~of'lronrc Covimunrcamm

(2000) 277-324. A. Utsumi. Verbal irony as implicit display of ironic environment: Distinguishing ironic utterances fiom nonirony. Journal qfPragmatics 32 (2000) 1777-1 806. [34] H. L. Colston, and J. O'Brien, Contrast and pragmatics in figurative language: Anything understatement can do, irony can do better. Journal qffragmatics 32 (2000a) 1557-1 583. 1351 H. L. Colston, and J. OBrien, Contrast of kind \is. contrast of magnitude: The pragmatic accomplishments of irony and hyperbole. Discourse Processes 30 (2000b) I 79- 199. [36] H. L. Colston, K. Summers, and S. Strickland, S. Pretense and echoic accounts of verbal it-oii!. comprehension: .4n evaluation with auditon presentation. Poster session presented at the meeting of the Psychonomic Society. Los Angeles, 1999, November. 1371 H. L. Colston, Interpreting ironic images. Manuscript in preparation. 2001. [ 3 8 ] R. Gibbs, J. OBrien. and S. Doolittle, Inferring meanings that are not intended: Speakers' intentions and ironv comprehension. D ~ S C ~ UProcesses JW 20 ( 1995) 187-203. [33]

SECTION IV COMMUNICATE IN PRETENDING: COMPUTER MEDIATED COMMUNICATION When a computer network connects people or organizations, it is a social network. Just as a computer network is a set of machines connected by a set of cables, a social network is a set of people (or organizations or other social entities) connected by a set of social relationships, such as friendship, co-working or information exchange. Much research into how people use computermediated communication has concentrated on how individual users interface with their computers, how two persons interact online, or how small groups fbnction online. As widespread communication via computer networks develops, analysts need to go beyond studying single users, two-person ties, and small groups to examining the computer-supported social networks that flourish in areas as diverse as the workplace and virtual communities. Garton, Haythornthwaite and Wellman, I997

This page intentionally left blank

Say Not to Say L. Anolli et al. (Eds.) IOS Press, 2002

9

197

Communicating in CMC: Making Order Out of Miscommunication Giuseppe R N A

Abstract: Everyday, thousands of subjects engage in computer mediated communication (CMC) using a novel linguistic genre, which combines characteristics of both written and oral language. This “virtual” communication may be described as a particular form of miscommunication: a necessarily “pared-down” form of conversation, which lacks the many features and some rules on which traditional forms of interactions depend. According to the “cues-filtered-out” approach, CMC lacks the relational features (social cues), which enable interactors to identify correctly the kind of interpersonal situations they find themselves in. The conclusions of this approach are that CMC occurs in a social vacuum where the personal identities of subjects tend to fade and vanish. This chapter tries to counter this vision by integrating in its theoretical framework five different approaches: the Miscommunication as a Chance Theory (MaCHT), the Positioning Theory (PT), the Situated Action Theory (SAT), the Social Identity Model of Deindividuation Effect (SIDE) and the Social Information Processing (SIP) perspective. In particular, starting from the analysis of the miscommunication processes typical of CMC, the chapter outlines how CMC users are able to make order and create relationships out of the miscommunication processes typical of this medium. Moreover, it presents the emerging forms of CMC - instant messaging, shared hypermedia, weblogs and graphical chats - and their possible social and communicative effects.

Contents

9.3 9.4 9.5 9.6

Introduction.. ........................................................ Flaming, lurking, spamming and identity deception: the dark side of CMC.. ....................................................... Making order out of CMC miscommunication.. ................ The future: emerging forms of CMC.. ........................... Conclusions.. ........................................................ Acknowledgements .................................................

199 204 2 13 222 224

9.7

References.. ..........................................................

224

9.1 9.2

198

G Rita Commimicuting in C.MC'

I98

9.1

Introduction

The growth in the use of global networks of computers, commonly labeled the Internet, has increasingly led to questions about how the Internet can be understood and analyzed as an emergent communication tool. In fact, computer-mediated communication (CMC) has created new opportunities for synchronous and asynchronous discussions among geographically dispersed individuals sitting at their computer keyboards. Everyday, thousands of subjects engage in "written" conversation using a novel linguistic genre, which combines characteristics of both written and oral language. They type their messages, and the text of these messages appears on the screens of their interlocutors, preceded by the "speaker's" nickname or e-mail address. The commonest form of asynchronous CMC is e-mail, in which a sender leaves a message in a receiver's electronic letterbox, which the receiver must open before he can read the message [I]. Another more sophisticated type of asynchronous CMC is Usenet Newsgroup, an electronic notice board on which users can post messages referring to a specific topic or area of interest. Users can read the messages by opening the notice board. and send their own messages in turn. As with e-mail, there is no real-time link between the computers of the interacting subjects. Unlike asynchronous CMC, the most important feature of synchronous CMC is that it does provide a real-time link between users' computers. Although the most frequently cited example is the videoconference, the most widespread system is the Internet Relay Chat. or IRC. IRC is a form of synchronous CMC allowing a group of users to "chat'' by exchanging written messages. They can interact in two different ways: by sending a message either to a specified user, or to all members of the group. Despite the predominance of the textual mode, it has been shown that CMC differs in psychosocial terms from non-electronic written communication, as well as from other existing means of communication. Experimental studies designed to compare CMC and non-electronic written communication [2, 31 have revealed significant differences in their respective degrees of social presence and media richness. Social presence is the user's perception of the ability of the means of communication to marshal and focus the presence of communicating subjects [4]. while media richness is the ability of the means of communication to interlink a variety of topics, render them less ambiguous, and enable users to learn about them within a given time-span [ 5 ] . There is a technical reason for these differences. Communication with a keyboard and computer screen takes longer than normal face-to-face communication, and the absence of metacommunicative features like facial expression, posture and tone of voice encourages users to find other ways of making communication as complete as possible. In this sense, CMC is usually described as an efficient form of miscommunication. 1.e.. a necessarily "pared-down" or, perhaps, more accurately, rarefied form of conversation [6, 71, which lacks the rules on which effective interaction depends. Computer mediation creates an asymmetrical imbalance in the sender-receiver relationship: the sender can transmit information and get cooperation under way, but has no guarantee that the receiver receives the transmission. while the receiver has no guarantee that the sender's declared identity is the real one [8]. How do these differences affect the interaction process'? The chapter aims to answer this question by integrating five different theoretical approaches: the Miscommunication as LI Chance Theoi??,the Positioning Theoyi. the Situated Action Theor?.. the Social identit?. Model of Deindividuation Effect,and the Social information Processing perspective. In particular, starting from the analysis of the miscommunication processes typical of CMC f7aming lurking, spamming and identitlt deception the chapter discusses the miscommunication characteristics of CMC tvwing to identify hou to exploit the full ~

G. Riva / Communicating in CMC

199

potential of the electronic media. Moreover, it presents the emerging forms of CMC instant messaging, shared hypermedia, weblogs and graphical chats - and their possible social and communicative effects. 9.2

Flaming, lurking, spamming and identity deception: the dark side of CMC

The increased diffusion of Internet made CMC very popular. However, what are the communication characteristics of CMC? If face-to-face conversation occurs in a cooperative environment constantly regulated by mutual adjustment and correction [9, lo], CMC occurs in a much less cooperative environment because of the special conditions imposed by the medium itself [ 111. This has a strong impact on the regulation of CMC interaction. In face-to-face communication the regulation of interaction is obtained using a complex system of turn taking and yielding behaviors [ 12, 131. For example, as stated by Patterson [ 13, p. 1081, when a listener is about to attempt to take a turn as speaker, he/she may exhibit some or all of the following behaviors: - a shift of the head away from the speaker; - an audible inhalation; - the initiation of gesture; - over loudness in the first segments of speech. In CMC, almost all of the above behaviors are completely impertinent. This difference in the regulation of the interaction can be explained by the following CMC characteristics [14, 151: - The size of an utterance is determined entirely by speaker. In general, however, in synchronous CMC utterances are rather short: 5 to 13 words per utterances in conversations on MUDS. This increases the feeling of interactivity for participants, and a lets listeners know that the speaker is not idle and not finished speaking. - It is impossible the overlap of utterances. In synchronous CMC two users may be typing at the same time, but it is only upon pressing “return” that their utterance is processed by the MUDARC and displayed to other users. - In synchronous CMC the order of utterances need not be sequentially relevant for meaningful conversation to take place. - Due to the persistent nature of text-based CMC, a communicator need not be present at the time of the utterance(s), but rather has the option of returning to one’s computer later to catch up on what has been transmitted. Moreover, in most CMC environments, and in asynchronous CMC environments especially, two typical features of face-to-face conversation are missing [ 161: - the collaborative commitment of participants and the co-formulation of the message; - the feedback, which allows the social meaning of the message to be processed immediately. As noted by Viegas and Donath [ 171 in a text-only CMC environment, the text element is overloaded as a multiple signifier: In these environments, the participants type messages which are then displayed sequentially on each person’s screen. These messages convey two types of information: one is the content of the message; the other is the presence of the participant. Consequently, if the participant is not actively messaging, he or she is not present on the screen (p. 10).

The possibility of disappearing from screen has a strong impact on the style of

G. Ri1.a ' Communicating in C M C

2 00

discourse. Users, in fact, often send messages only so that the others will not forget them. Contextual information, too, is difficult to perceive in a chat system [ 171. For instance. it is difficult to distinguish among the participants and form a coherent sense of their individual identities: all the users appear as a name in text against the background. Also the interactions among the users are not always clear: in a chat the conversation appears on the screen as a linear progression of lines of text, regardless of the conversation's dynamics. Finally, the temporal information found in oral conversation. such as turn-taking and the negotiation of conversational synchrony by the participants, is not captured by these sequential lines of text. As noted by Mabry [ 181, this leads to a prevalence of argumentative exchanges that are in many cases characterized by a strong and destructive style. In addition, CMC in no way guarantees that a user's declared identity is the real one. The use of false identities, often of a different sex, is widespread in electronic communities and in IRC especially [ 19-22]. In another chapter of this book Marby underlined how the anonymity of CMC, facilitating and facilitated by concealment in selfpresentation, allows distortions to be made through omission of information as well as through selective presentation. In general four particular form of miscommunication are common in CMC [23]: offensive conduct (flaming), listening without making hisher presence known (lurking). delivering a message to someone who would not otherwise choose to receive it (spamming and bombing) and identity deception (spoofing). In the next paragraphs we will try to deepen the analysis of these behaviors.

9.A?. I

Flaming

Even if a meta-analysis on existing studies [24] found that the incidence of offensive conduct is overrated, different areas of CMC are characterized by intense language, swearing, negative or hostile communication. As experienced by many users of Usenet Newsgroup or Internet Relay Chat, the intensity of many communicative exchanges is usually heat. To reduce the number of offending messages, net groups have established a netiquetfe norms of network usage - that specifically addresses how the user can write and post messages. These norms stress obligations for group and self-monitoring to insure that members maintain a correct language, respect for the interlocutor and communicative relevance. The netiquette, however, is not the same everywhere. Some offenses are likely seen as more disturbing than others, and it is equally possible that what one group condemns, another condones. For example, while directing a particularly hostile message at another user is perfectly acceptable on some newsgroups (e.g., ult,fun.warlor&. alt.flame, and a k i r c ) , this approach is usually censured by many socially oriented new sgroups. The typical breach of netiquette involves the use of~lames.With this word Siege1 and colleagues defined "messages that are precipitate, often personally derogatory, ad hominem attacks directed toward someone due to a position taken in a message distributed (posted) to the group" [25, p. 1591. However, the definition of flaming in CMC literature varies. For instance Rice [26] describes flaming as "the tendency to react more critically or with greater hostility, leading to an escalation of conflict" (p. 634). Following this line Walther 1271 defines it as "insults. swearing. and hostile instances of behavior". A more effective definition is the one provided by Thompsen and Ahn: flaming is composed by CMC behaviors that are interpreted to be inappropriately hostile [28]. This definition focuses on an important point: for a flame to take place two separate actions

G. Riva / Communicating in CMC

20 1

must occur. First the behavior has to be created. Then someone else has to interpret the behavior as being offensive. Table 9.1 Offending messages in CMC (adapted from McLaughlin et al., 1995)

Offending messages Ethical Violations

Inappropriate Language Violation of Newsgroup Conventions

Violation of Usenet Conventions

Characteristicsof offending messages Posting private email or personal information individual posters Personal attacks, hostile or coarse language, linguistic affectations which distract or detract from message content Failing to use spoiler warnings, lack of familiarity with and failure to use appropriate subject headers or abbreviations, failing to conform to group spirit or style and group traditions regarding appropriate topics Incorrect or missing subject headers, failing to encrypt offensive material, posting to an inappropriate newsgroup or otherwise

Following this approach McLaughlin and colleagues [29] analyzed characteristics of messages in five popular newsgroups and elaborated taxonomy of reproachable conduct on Usenet. In particular offending messages - those which presented behavior sufficiently in violation of normative expectations to prompt comment and spark remedial discussion and debate - were analyzed for gravity of offense. The result of their analysis is summarized in the Table 9.1. As clearly shown by the table, the use of inappropriate language is only one of the possible offending behaviors in CMC. 9.2.2 Lurking The word “Lurking” is used to define the behavior of subscribers to electronic forums who rarely or never send contributions to the discussions, content to read what others are writing [23]. Herz [23, pp. 4-51 describes lurking as follows: Lurking is a lawalphase in the nethead life cycle. It’s that spooky, voyeuristic time when you haven’t got your bearings yet, but you’re fascinated enough to browse with bovine contentment on the grassy pastures of online discourse. . . . Lurking is like one of those

702

G Rivtr Communicating in

CMC

Sunda-v-night movies on network TV where a p y is struck b?. lightening or to-xic waste and becomes Captain Undetectable, suddenlit able to overhear boardroom conversations and sneak into the lingerie dressing room at Maci*!c.at a single hound.

As clearly stated by this text, in CMC a “lurker” is equivalent to a spy: someone who listens to discussions within a chat room but doesn’t make his or her presence known. The motivations for this behavior are varied: having nothing to say, feeling “outclassed” by scholars who post frequently, or simply enjoying the exchange as a passive reader. Moreover, the use of lurking is a good strategy for getting a sense of what is acceptable in a new environment. However, the drawback to lurking is that, in an entirely text-based environment, if a user writes nothing he/she effectively ceases to exist. As one IRC user noted. ”I post. therefore I am.” By corollary, a lurker does not exist in a way that can be perceived and responded to by others online: he or she is not an active part of the virtual community.

9.2.3

Spamming and bombing

As stated by Marvin, [23] the expression “spam” is used throughout the Internet, on both synchronous and asynchronous forms, for any “excess of words”. In particular the expression “spamming” is currently used for any attempt to deliver a message. over the Internet, to someone who would not otherwise choose to receive it. The term was originally used in Usenet newsgroups to describe identical commercial or off-topic posts made to multiple newsgroups. It has since been expanded to include ordinary email messages. Most of the spam is composed by commercial advertising (unsolicited commercial e-mail or unsolicited bulk email). Potential target lists are created with automated searches by scanning Usenet postings. stealing Internet mailing lists. or searching the Web for addresses. A variant of email spamming is “bombing”. Email “bombing” is characterized by repeatedly sending an identical message to a specific e-mail address. There are two different forms of bombing: - Massmail bombing: a subject sends to a specific e-mail address hundreds, or perhaps even thousands of pieces of email, usually by means of a script and fakemail. It is relatively easy to defend against, since the messages will be coming from just a few email addresses. - Mailing List bombing: a subject subscribes a specific e-mail address to as many mailing lists as possible. This is much worse than a massmail because it requires to unsubscribe from each mailing list. a process that requires information and time. 9.2.4

Identi?. deception

In communication knowing the identity of those with whom you communicate is essential for understanding and evaluating an interaction. However, the common use of false identities in CMC is very intriguing for communication researchers. In fact, the cost of identity deception for the typical CMC user is potentially high. As noted by Donath [30] it reduces both the motivation level and the possibility of information exchange, affiliation and support provision (see Table 9.2). In the chapter on deception Anolli and colleagues clearly analyzed the characteristics of this behavior. However, it is interesting to analyze the possibility of identity deception offered by CMC.

G. Riva /Communicating in CMC

203

Table 9.2 The cost of identity deception (adapted from Donath, 1999)

Areas Information Exchange

costs

The communicator’s identity - in particular, claims of realworld expertise or history of accurate online contributions plays an important role in judging the veracity of a message. Similarly, knowing the writer’s motivation - e.g. political beliefs, professional affiliations, personal relationships - can greatly affect how we interpret his or her statements. Providing affiliation and The creation of a shared community requires that the participants are sympathetic to the ideas around which the support group is based; even if they disagree, there needs to be some hndamental common ground. Trust in the shared motivations and beliefs of the other participants - in other words, their social identity - is essential to the sense of community I In most newsgroups, reputation is enhanced by posting Enhancing reputation intelligent and interesting comments, while in some others posting rude flames or snide and cutting observations enhances it. To the writer seeking to be better known, a clearly recognizable display of identity is especially important. No matter how brilliant the posting, there is no gain in reputation if the readers are oblivious to whom the author is.

I

If we follow Donath [30] we can identify the following forms of deception: - Identity concealment: The most common form of CMC deception is identity concealment: hiding one’s identity. Usually, Usenet and IRC users hide their identity for privacy issues: Usenet and IRC are public forums in which the participants have no control over who reads their posts. However, sometimes identity is hidden to circumvent message filters (killfiles) that allow the skipping of unwanted postings: if you put someone in your filter, you will see no more of hisker postings. In this domain is important to distinguish between two forms of identity concealment: pseudonymity and pure anonymity. A pseudonym, though it may have not a direct link to a real-world person, can build a well-established reputation in CMC and transmit a wealth of contextual information about the sender. A purely anonymous message, on the other hand, is a form of unattributed communication (spoof). As an unattributed line, spoofing attracts attention, is often used for humorous effect, but can be threatening, confusing, or frustrating in normal conversational patterns. - Category deception: Gender deception is the most common form of category deception, especially in chatrooms. In newsgroups, gender deception appears to be much less common, except in forums where sex and gender are the most relevant topics. Also very used are age deception and status enhancement. Typical expression of status enhancement can be easily found in many newsgroups: the “well-built body” in misc.Jitness.weights or the “expert hacker” in alt.2600. - Impersonation: A very dangerous form of identity deception is impersonation. If a subject can pass as you, he can destroy your reputation, both on-line and off. Usually the effect of an impersonation depends upon how defamatory the faked postings are and whether readers believe the false attribution.

304

G. Riva 1 Communicating in CMC

In general, if deceiving someone about your identity in the real world is indeed difficult, it seems to be far easier online, since the deceiver doesn’t have to worry about the high

number of relevant identity clues available offline. As noted by the literature in this area many impersonated postings are made simply by signing the target’s name, without copying the writing style or forging the header information. Very interesting for communication researchers is how successful such simple imitations can be. 9.3

Making order out of CMC miscommunication

The analysis of these behaviors is intriguing for communication researchers. In fact, as underlined by many communication researchers (e.g., Brown and Levinson, [3 11). speakers typically display a “preference for agreement” in social interaction that seems lacking to flaming, lurking, spamming and spoofing. As noted by Mabry [32]: “The question arises as to how mediated groups manage this more adversarial communication context while retaining the discursive coherence and cohesiveness necessary for enacting socially appropriate rational discourse.” As we have seen before, CMC can be described as a necessarily “pared-down” or. perhaps, more accurately, rarefied form of conversation [6], which lacks the rules on which effective interaction depends. This is due to the peculiar characteristics of CMC: communication with a keyboard and computer screen takes longer than normal, and doesn’t support metacommunicative features like facial expression. posture and tone of voice that have an important role in face-to-face communication. Starting from these limits, different researchers considered CMC as unable to support satisfactory socio-emotional and relational communication. if compared with face-to face. In particular, according to the American sociologists Sproull and Kiesler, CMC lacks the specifically relational features (social cues) which enable interactors to identitjr correctly the kind of interpersonal situations they find themselves in [33]. Sproull and Kiesler concluded from this that CMC occurs in a social vacuum where the personal identities of subjects tend to fade and vanish [34]. The most important consequences of this are that: - In CMC, subjects tend to express themselves more openly and freely: “People who interact via computer are isolated from social rules and feel less subject to criticism and control. This sense of privacy makes them feel less inhibited in their relations with others” [34, p. 481. - At the same time, however, loss of personal identity may encourage subjects to break social rules. Siegel, Dubrovsky, Kiesler and McGuire 1251 cite flaming as evidence of this. This vision, usually defined in literature as the “cuessTfiltered-out”approach [35], is based on the concept of social presence that we discussed in the introduction: the feeling of the communicators that other actors are jointly involved in communicative interaction [4]. Social presence theory states that the attention paid by users to the presence of other social participants is directly linked to the channels or codes available within a medium: fewer channels, less attention. However, most criticism has been directed at the alleged absence of social cues in CMC. Mantovani [ 161 asserts that Sproull and Kiesler’s claim that CMC occurs in a social vacuum “[is] unacceptable in general terms because it reduces social reality to some form of physical connection between individuals ... To be excluded from the social context isn‘t simply a matter of being alone in a room” (pp. 1 70- 1 72 ).

G. Riva /Communicating in CMC

205

Mantovani’s criticisms are backed up by numerous studies of social identity and selfcategorization, which show that the relationship between an individual and the social context and reference group is conceptualized in socio-cognitive rather than structural and relational terms [36-401. These studies have shown that subjects are characterized not by one fixed self, but by a variety of selves (self-categories), including the personal and social identities, which emerge from the contexts in which they are rooted [40]. Since they play a crucial role in the creation of context, social categorization and individuation are a major influence on subjects’behavior, irrespective of the co-presence of other people. This point of view is clearly expressed by the Social Identity Model of Deindividuation Eflect - SIDE [41, 421. According to SIDE the actions of a CMC user, while tempered by individual awareness, stem in part from invisible social norms and identity. It is precisely the absence of non-verbal feedback in CMC, which makes these processes even more important than in face-to-face communication. According to Lea and Spears [2], “it could be claimed that the absence of social rules coming from other people, and the uncertain situation which results from this, force people to use social reference norms to regulate their own behavior” (p. 286). Moreover, as underlined by Walther [27, 43, 441 in his Social Information Processing (SIP) perspective, users adapt existing communicative cues, within constraints of language and textual display, to support processes of relational management. This position is strongly supported by research and user experience. For instance, Sherbloom [45] in his study on e-mail, concludes that communicators adapt computer-generated textual signals for specific relational purposes. Chesebro and Bonsall [46] found that a third of all the messages sent in 14 public computer bulletin boards were interpersonal in nature. In general, if sufficient time is available, CMC allows the development of interpersonal relationships, and even intimacy, between the communicators [43]. How is this possible? And in particular, what are the elements required for creating an interpersonal relationship between CMC users? The SIP perspective identified the following factors: - A priori rezational motivators : Possible drives are the affiliation motive, impression management or the need for dominance. - Impression formation. In CMC, this happens by decoding the verbal messages of the communication partner. This is possible by using ‘‘knowledge-generation strategies such as interrogation, self-disclosure, deception detection, environmental structuring, and deviation testing to gather psychological knowledge-level information about other persons” [27, p. 7 11. - Encoding of relational messages : CMC users have to learn how to transmit relational content within the limitations of the available channels. In particular they have to understand how to verbalize relational messages. - Time:to learn how to use CMC effectively requires time. In particular, to get to know each other and to build up trust and friendships via CMC. When this happens, the structure and the contents of CMC interaction change. In particular the following elements are possible indexes of the development of the relation [43, p.561: - Low level of formality: when CMC users feel more comfortable communicating with each other, they will be less focused on the formal aspects of communication. The amount of formality can be evaluated by the attention to general rules, the form of address a communicator chooses, as well as the figures of speech he or she employs. - Trust and receptivity: when CMC users feel more trust in another person, they are more likely to reveal personal details about themselves. The amount of trust is usually expressed through the vulnerability of people’s revelations and their self-disclosing opinions on different issues.

206

G. Rilw jl Communicating in CMC

-

Rate of information exchange: when CMC users create a sufficient level of trust and intimacy the rate of information exchanged increases. This also strengthen the personal relation: when more messages are sent, users grow more comfortable with each other and interesting topics of conversation are brought up. On the other side, a sufficient rate of information exchange is required for supporting any personal relation. The interesting point raised by the two authors, is that CMC communication, in its initial stages, can be more intimate that traditional face-to-face communication [43]. In CMC, the users are less concerned about the impression they are making because of the possibility of identity concealment offered by this medium. These points also support the miscommunication as a chance theor?- (MaCHT) presented by Anolli in the first chapter of this book. According to Anolli, a strategc use of miscommunication may enhance the degrees of fieedom available to the communicators during an interaction. In fact, miscommunication and in particular CMC may offer new communicative tools, that if correctly used, can improve the efficacy of an utterance. If a user handles well the miscommunication processes typical of CMC, he/she may even obtain results difficult to achieve in face-to-face meetings. For instance. CMC may initiate relationships that might never have begun if they required a face to face meeting [47]. In such a situation CMC allows for selective presentation of certain features of a participant’s identity reducing the person’s anxiety over how he or she will be judged. In the next paragraphs we will try to outline how the CMC users are able to make order and create relationships out of the miscommunication processes typical of this medium. 9.3. I

E.vpressing emotions iri CMC

CMC interlocutors are forced to find alternative way for reproducing the metacommunicative features (emotions, illocutionary force, etc.) of face-to-face conversation. According to Utz [48] it is possible identi@ three different forms of emotional expressions in CMC: emoticons, social verbs and emotes (see Table 9.3). Emoticons (also smileys) are the most used textual devices: ASCII glyphs designed to show an emotional state in plain text messages. They consist of various punctuation marks and are viewed by turning the page sideways or tilting one’s head to the left (a list of the most common emoticons are reported in Table 9.3). In particular, Walther and D’Addario [49] describe emoticons as ”graphic representations of facial expressions that many e-mail users embed in their messages. These symbols are widely known and commonly recognized among computer-mediated communication users, and they are described by most observers as substituting for the nonverbal cues that are missing from CMC in comparison to face-to-face communication” (p. 324). As noted by Wolf, [50] many different emoticon collections exist online in several different languages. And as for non verbal emotional expression, there is some confusion in emoticon interpretations: “In some instances the emoticon :-Q means user smokes; others define it as meaning tongue bunging out in nuuseu or sticking out tongue. A more widely used emoticon for user sticking orit tongue is: -P” (p. 829). In general there is a general acceptance in the interpretation of the basic smiley, frowney. and winkey emoticons (see Table 9.4): their respective meaning is humor, sadness, and sarcasm. However, the more elaborate the emoticons become. the greater variation one finds in the interpretations available for them. Even the large amount of information available on the Internet. only a few studies verified the real impact of emoticons in CMC communication. Ut7 [51] examined the impact of emoticons t o the development of friendships in MVDc. The data obtained showed that emoticons are wen as helpful in expressing

207

G. Riva / Communicating in CMC

socioemotional contents. More interesting, the use of emoticons is correlated with development of online fl-iendships. In a different study Walther and D’Addario [49] studied the effects of three common emoticons on message interpretations. The results indicated that emoticons’ contribution to the interpretation was limited and outweighed by verbal content. However they found a negativity effect: any negative message, expressed either verbally or using an emoticon, shifts message interpretation in the direction of the negative element. Table 9.3 Different forms of emotional expression in text-based CMC (adapted from Utz, 2000)

Emotional Expression Emoticons

Social verbs (feelings)

Emotes

Description ASCII glyphs designed to show an emotional state in plain text messages Small pre-programmed scripts to express actions and emotions by simply typing an abbreviation

Narrative descriptions of conversational nonverbal behaviors

Example :-) humor :-( sadness ;-) sarcasm

command: LI laugh appears on screen: You fall down laughing command: smi Aron und appears on screen: You smile understandingly at Aron command: emote/say/pose is so happy that he could embrace the whole world appears on screen : is so happy that he could embrace the whole world.

In the last research, Utz [51] studied gender differences in emoticon use. After analyzing the messages posted in different mixed-gender newsgroups, the author verified that: if women most frequently use emoticons to express humor (35%), men use emoticons most often to express teasing or sarcasm (3 1%). But for our goals the most interesting part of the paper is the Conclusion [51, #1197; p. 8321: What emerges from a closer inspection, however, is that while emoticons are defined as vehicles to express emotion-hence “emotional icon ”--their actual function hinges on the definition of the word emotion. As noted by the author, both males and females users have altered the definition of emoticon to suit their conception of emotion. On one side males have expanded on the conventional definition of emotion to include sarcasm and teasing. On the other side female users have expanded on the male definition of emoticons and their use “adds other dimensions including solidarity, support, assertion of positive feelings, and thanks” (p. 833). These results seem to support two different approaches to CMC we discussed before. First, they are coherent with the social information processing perspective [27, 43, 441: in CMC users adapt existing communicative cues, within constraints of language and textual display, to support processes of relational management.

G. Riva /i Communicating in CMC

’08

Table 9.4 The most common emoticons

Svmbol :-) :-( I

(:& (:-*

Meaning Basic smiley face; used for humor and sometimes sarcasm Basic frowney face; used for sadness or aneer

Symbol AM

Meaning Laughter

:-D

Laughing

Angry Kiss

Y O -(

Confused Death

I

1

8-##

Second, they also agree with the miscommunication as a chance theon, presented by Anolli in this book. According to this perspective, communicators try to manage their communicative *focusin the best possible wav , given the contextual constraints and their respective encyclopaedia of knowledge [521. Communicative focus is concerned with how the speaker lets the addressee know what in particular he/she notices about the prominent aspect of the communicative act: a communicative act arises as the focus moves through the field of structures of beliefs, driven by the communicator’s goals and guided by thoughts and communicative devices. Using the concept of communicative focus we can predict that users, according to the contextual constraints and their respective knowledge of the situation, will try to communicate using any available tool. For instance, MUDs offer to communicators more expressive tools than traditional text-based IRCs. In fact, in them. we can find new emotional tools - emotes and social 1.erh.Y - that are used together with emoticons to improve the communicative focus [53]. Emotes are narrative descriptions of conversational nonverbal behaviors typed-out by the users. For example, if my name were “Joe” in a MUD and I type “emote cries out loud,” the result for others on the MUD would be ”Joe cries out loud” - giving my “character” action and movement - and even emotion. Another possibility offered by MUDs is the use of social verbs (also called.fee1in.p or feature objects):small pre-programmed scripts to express actions and emotions by simply typing an abbreviation. There are several hundreds of verbs and adverbs which can be combined at pleasure. For example, “smi iro” results in “smile ironically”; it is possible to smile sadly, happily, knowingly, innocently, and so on. Other feelings are hug, laugh. cry. poke, kick, kiss, or sigh. In advanced CMC, such as in Instant Messaging or Shared Hypermedia (see paragraph 9.4). emoticons have e\.olved in a list of graphical rich icons that are a standard feature of

1

i

1 j

G. Riva / Communicating in CMC

209

the client used. In these new forms of CMC, emoticons are used both in the textual chat to express a specific emotion, and as mood indicators attached to the user ID/avatar.

Figure 9.1 Advanced emoticons (MSN Messenger XP, left, and Odigo, right)

9.3.2 Definition of a common context in CMC

As already suggested in other chapters of this book, during the last fifty years the most famous communication model - the “parcel-post model” [54, 551, which describes communication as the passage of information from one person to another - became obsolete. First, as noted by Anolli in the first chapter of this book, any communication theory must face and solve the dilemma of opposition between meaning stability and meaning instability of a word, an utterance, or a gesture. Within this perspective, meaning consists of afuzzy set: a class of communicative units with a continuum of grades of membership. Second, as underlined by many authors, the information-transfer model does not take into account the cooperative component of communication, which stimulates reciprocal responsibility for successful interaction and a series of subtle adaptations among interlocutors. As Doheny-Farina [561 notes: “The theory of communication as information transfer separates knowledge from communication; it treats knowledge as an object that exists independently of the participants in the innovation venture. With this independent existence, information becomes an object that can be carried through channels” (p.8). However, it is possible to communicate only to the extent that participants have some common ground for shared beliefs, recognize reciprocal expectations, and accept rules for interaction, which serve as necessary anchors in the development of conversation [57]. Ghiglione’s definition [58] of communication as “the co-construction of a reality using systems of signs and a mutually acceptable set of principles which make exchange possible and provide the rules needed to govern it” (p. 102), applies equally well to CMC as a constructive form of miscommunication. The main difference is that in CMC, as we have extensively seen before, the reality is asymmetrically co-constructed. In fact, the receiver can decide at will to terminate interaction, or continue it by turning himself into a sender. This decision is far from casual: it depends on how the receiver interprets the situation, what his aims are, and the social rules that govern his behavior. Some researchers have even used the term “electronic opportunism” to describe this feature of CMC [59]. In this sense, CMC may be defined as a process by which a group of social

710

G. R h o Communicating 111 C.4K

actors in a given situation negotiates the meaning of the various situations. which arise between them [60]. Stasser's definition of CMC may seem straightforward enough, but it has two important implications that have had a decisive effect on CMC studies [61]. If CMC is a process of negotiation: - the only way to understand it is by analyzing the subjects involved in it, and in the environment in which they operate, meaning that the social context in which CMC occurs plays a crucial role; - new processes and activities will develop. challenging and modifjmg the initial relationship between subject and context. Most researchers would broadly agree that these two statements are true. According to Mantovani [16], the early 1990s saw changes in the study paradigms of person-computer and person-computer-person interaction. The main outcome of this has been the understanding that interaction can only be fully understood through detailed analysis of the social context in which it occurs: *'... at this point we should no longer see people simply as 'users' of given systems, but as social 'actors'. In other words, whether expert computer users or not, people act independently and have their own reasons for what they do. and it is computers and systems that have to adapt to people. not vice versa" [ 16. p.633. Based on traditional cognitive analyses of information processing and symbolization. Sitirated Action Theoi?2 (SAT) introduces a change of perspective very interesting for our analysis: action is not the execution of a ready-conceived plan, but the subject's adaptation to context [62]. As Suchman notes, "instead of separating action from the circumstances in which it occurs as the execution of a carefully thought out plan ... [SAT] tries to study how people use circumstances to develop an intelligent course of action.' (p.167). This necessitates profound changes in how "social cotttext'' has previously been defined. In SAT, social context is not something physical and stable like an organization or the power structure within it. As Mantovani [ 161 stresses, contexts are not given, but made. Thus: - context is conceptual as ~ v l as l phj:yical: actors perceive situations using cultural models, and act accordingly in cultural ways: - context is zrnstahle: cultural models are constantly modified by subjects' actions and choices. In this sense, social context may be considered as the symbolic system of a given culture that is continually being altered by practical human intervention. Applying SAT to CMC, Mantovani [ 16. 631 concludes that CMC participants cannot be regarded simply as technology users. Rather, they are social actors with their own aims and autonomy in situations, and it is technology, which must adapt to them. This idea poses serious problems, however. If social actors actively respond to their environment and end up changing it, how can context ever be analyzed properly? Mantovani meets the dificulty with a three-level model of social context that links situation and social norms to the use of computer technology. The first level is social context in general, the second, ordinary situations of everyday life. and the third. local interaction with the environment via computers. The links between the three levels can be studied in either direction, starting from use of computers or from social context. Thus. the use of computers may be regarded as part of everyday life, which is in turn part of the broader social context. By interacting with each other. the physical environment and the social context. subjects activate a spiral of actorenvironment exchanges. First-level person-computer interaction leads to interaction in everyday qituations. and thence to cultural changes.

G. Riva / Communicating in CMC

21 1

Working in the opposite direction, social context supplies the elements needed to interpret situations correctly, and situations generate the aims, which determine local interaction with the environment via computer. So, it cannot be explained exclusively in terms of the interpersonal relationships, or physical environment, in which information exchanges take place. Social context is a prerequisite of communication, “a shared symbolic order in which action becomes meaningful, and so generates meaning” [ 16, p.1061. Thus, SAT implies a radical redefinition of the meaning of communication. Context may be co-constructed by social actors, but they use communication to exchange meanings, not pieces of information. More precisely, the content of communication is interpretations of the situations which actors are involved in. In this sense, the most effective way of clarifllng the meaning of messages is to relate them to a shared context of meaning. Studies of Positioning Theory (PT) have served to reinforce this view. As recently formulated by Rom Harre [64, 651, PT replaces the traditional concept of role with the concept of positioning. The main difference between the two is that a role is a stable and clearly defined category, while positioning is a dynamic process generated by communication. Developing on Bakhtin’s ideas and Vygotsky’s studies, PT identifies two distinct processes underlyng social activities. The first, naturally enough, is discourse-generated positioning, which Harre defines as “the way in which subjects dynamically generate and explain their own and other people’s behavior” [65, p.4051. Harre defines the second process as the rhetorical redescription by which subjects shape their social context, “the discursive production of stories about institutions and macro-social events undertaken to make them intelligible in the form of social icons” [65, p.3941. As in SAT, context is not given in PT, but is constructed socially in ways, which are endlessly different because of the changes communication brings about in the structuring of the cultural context. The main difference between SAT and PT lies in the role attributed to discourse production. PT sees conversation as the most important human activity of all because it encompasses virtually all known mental phenomena. As Harre & Van Langenhove [65] stated, “many mental phenomena like attitudes and emotions are immanently present in discourse production” (pp. 394-395). 9.3.3 Identity awareness and construction in CMC

As we have seen in the previous paragraphs, CMC allows identity deception and anonymity. But there is a problem here: how can you communicate and activate the positioning process without staking your own identity on the outcome? As we have seen, communication always requires a framework of rules and meanings, and this is especially true of CMC in which many features of face-to-face conversation are “rarefied”. One solution is to represent you by “coding cultural expectations at a symbolic level” [66, p. 1021. In constructing a false identity, the subject has to make wider use of social stereotypes than would be the case in normal conversation if he wishes his identity to be recognized and accepted. This means that CMC, may force subjects to resort to massive use of stereotypical attitudes and behaviors if they are to achieve any shared understanding of actions and situations [22]. What are the possible effects of this process? According to the Social Identity Model of Deindividuation Effect - SIDE theorists [41, 421, a social or a group identity will replace individual identity in CMC. Through experimentation, SIDE researchers found a salient

212

G. Riva / Communicating in CMC

shift from an individual identity to a group identity with individuals adhering to group norms [67]. A recent research investigated the intergroup properties of flaming in CMC. In particular the researchers tried to investigate how some form of identifiability affects flaming language: name, e-mail address or geographical location [68]. Using for their analysis the SIDE model the authors verified that communicators produced more stereotype-consistent (group-normative) descriptions of out-group members’ behaviors when their descriptions were identifiable to an audience. In particular. identifiability to an in-group audience was associated with higher levels of stereotypeconsistent language when communicators described anonymous out-group targets. According to SIDE, this result can be explained by strategic reasons: to gain acceptance from the in-group, to avoid punishment from the out-group, or to assert the identity to the out-group. In a different study, 75 group members were primed with a certain type of social behavior (eficiency vs. prosocial norms). Consistent with the model, anonymous groups displayed prime-consistent behavior in their task solutions. whereas identifiable groups did not [67]. These results suggest that anonymous individuals in CMC are inclined to accept ingroup norms and identity and reject out-group norms and identity. In-group favoritism increases, while stereotyping and bias between groups prevails. The SIDE model can be applied to broader social categories such as gender, race and nationality. For instance. when anonymity maintains and gender is revealed in CMC, individuals tend to behave in terms of gender norms. The prediction is that these social categories remain firm or become fossilized in anonymous CMC. At the same time, however, the difision of CMC may produce changes in how personal identity develops. In fact, external language and interior dialogue are intimately related, and the link plays a crucial role in the formation of the subject’s identity and higher mental processes [69, 701. The way interaction with other subjects mediates meaning is hndamental to this shift from external language to interior dialogue. As Davies and Harre [71] point out, during conversation subjects’ selves “participate in an observable and subjectively coherent way in the joint production of story fines” (p. 48). In this phase, which uses interlocution in the manner described by Jacques [72], subjects see themselves as “contradictors” [7 I , p. 471 and use the positioning process to construct “a variety of selves” (p. 47) closely linked to the outcome of interaction. This is very similar to the “transactional conte-rtualism” developed by anthropologists and sociologists. For example, Rosaldo [73] says that the notion of self develops not from some internal essence relatively unaffected by the social world, but from experience accumulated in the world of meanings, images and social relationships in which each person is unavoidably involved. Hsu [74] defines this unbroken link between self and environment as “psychosocial homeostasis”. the unremitting effort to establish a balance between satisfaction of intrinsic needs and the demands of socio-cultural context. In psychology, these ideas have carried over into the work of Gergen [75] and Bruner [76]. Gergen in particular has looked in detail at the construction of self, in studies of how an individual’s self-esteem and concept of self vary in a set of different situations. These studies show that the concept of self varies both in relation to the kind of people the subject spends time with, and in response to the positive and negative comments they make. On the whole, then, the self may be seen as a product of the situation in which the subject acts. For his part, Bruner, though accepting the subject’s autonomy. speaks of “creatures of history” whose selves are both “a guarantee of stability and a barometer reflecting changes in the cultural climate“ (p.108).

G. Riva / Communicating in CMC

213

Markus and Nurius’ concept of possible selves [77] offers a theoretical explanation of the relationship between identity and context. According to these authors, possible selves “give a specific cognitive form to our desires for control, power and belonging, and our widespread fears of failure and incompetence” [77, p.9601. Although possible selves constitute our repertoire of different selves, their main feature is that they are exempt from direct social control and social negotiation. As Markus and Nurius say, “individuals have ideas about themselves which are not firmly anchored in social reality. As representatives of the self at some hture time, possible selves are visions of the self which have not been tested and validated by social experience” [77, p.9551. Potentially, a subject may be in a position to create an infinite number of possible selves, but in normal circumstances the repertoire of possible selves is a combination of the subject’s personal experience, and the living and communication environments he is familiar with. As well as being a source of more or less appropriate behavioral models, the media also offer a range of images and symbols that people can identify with easily. As Meyrowitz [78] points out, communication technology has changed our social context. Especially in younger people, the influence of social context on the construction of identity is beginning to wane as reference communities like the family, school or church, which in the past anchored social contexts in shared sets of rules, gradually loosen their gnp. Recently, Kraut and colleagues examined the Internet’s impact on emotional well being [79]. The results, discussed in the American Psychologist, showed that greater use of the Internet resulted in small but statistically significant increase in depression and loneliness and decreases in social engagement.

9.4

The future: emerging forms of CMC

The present situation would seem to be that the new media are accelerating the dissolution of traditional rule-based social contexts, and that this dissolution is itself draining the media of content and meaning. Doheny-Farina [80] argues that once we begin to divorce ourselves from geographic space and start investing ourselves in virtual communities, we further the dissolution of our real communities. However, as noted by Wallace [Sl] “the most important mediator of behavior in Internet environments is the purpose of the people who visit or inhabit them” (p. 5). Particularly, their use depends on how they are interpreted, what projects are in them, and what we think about daily reality [ 161. In this sense, as we have seen before, the CMC experience may be defined as a process by which a group of social actors in a given situation negotiate the meaning of the various situations which arise between them [60]. One of the consequences of this approach is that new processes and activities will develop out of this negotiation process, which challenge and change the initial relationship between subject and context [82]. To understand better what are these processes and activities, the following paragraphs will analyze the emerging forms of CMC: instant messaging, shared hypermedia, weblogs and graphical chats. 9.4.1 Instant messaging As we have seen before, IRC is a form of synchronous CMC allowing a group of users to “chat” by exchanging written messages. They can interact in two different ways: by sending a message either to a specified user, or to all members of the group. In e-mail,

2 13

G. Rita / Communicating in C.MC

instead a sender leaves a message in a receiver's electronic letterbox. which the receiver must open before he can read the message The mixing of IRC and e-mail has produced Instant Messaging (IM) . Instant messaging is a particular form of CMC that enables the user to create a private chatroom with one or more users (for a list of available IM clients, see Table 9.5). How does it work? Each user defines a list of people that he/she wishes to interact with: IM users can send messages to any of the people included in this predefined list (buddl?list or contact list) as long as that person is online. Typically, the instant messaging system alerts you whenever somebody on your private list is online. You can then initiate a chat session. Sending a message opens up a small window where the interlocutors can type in messages that all of them can see.

Figure 9.2 Instant Messaging interface (ICQ 200 I 1

Instant messaging appeared in the CMC scene in November 1996, when iCQ (a combination of letters that is shorthand for the phrase ''I seek you"). a tkee instantmessaging utility, was launched. Historically, the difksion of text based CMC is linked to the low bandwidth available to many Internet users. However. the increasing availability of fast Internet connection (ISDN, ADSL, Cable, etc.) is pushing software developers to integrate new communication tools in standard IM clients. The first step in this trend was the inclusion of audio communication: second generation IM clients allow talking with users anywhere in the world using the computer microphone and speakers. Other typical features of second-generation IM clients is real time file sharing and e-mail support. A further step in this trend is the creation of Video instant Messaging: Video IM also allows both video chat live, and the video messages recording/sending to the users who are in the chatrooms. To use the video features of a Video IM the user needs a video capture card and a normal camera, or a webcam. Third generation IM clients have even more features. Below are listed the more interesting from a communicational viewpoint: - Remote Assistance: a user can see and, if permitted. take control of the computer of another users. That's like sitting next to each other and looking at the same screen

G. Riva / Communicating in CMC

215

- Application sharing: Use the same application in real-time. If a user open a program, for example a word processor, the connected users can work on the document together. - Shared sketching: Draw diagrams with other user at the same time. Communicators can

sketch their ideas simultaneously as if they were both drawing on the same whiteboard.

Figure 9.3 Video IM Interface (Eyeball chat)

G. Rira 1 Communicating in CMC

216

Table 9.5

IM clients

Channels Developer Website Tool AOL http://www.aoI.com/aimlhomenew.adp AOL Instant Text i Messaging Text, NovaW iz http:/lwww.eyeballchat.com EyeBall Audio, i Video Text, ICQ h ttp ::/web. icq .c o w i ndex/ 1 ICQ2001 I Audio I I I Text Jabber .com http ://www .jabber.org Jabber Text, Microsoft http://messenger.msn.com I MSN Messenger 4 Audio Text, Microsoft http://messenger.msn.com MSN I Aud i 0, Messenger Video XP Mi crosofi http: //www.microsoft.com/windows/netmeeting! Netmeeting Text, Audio. Video Paltalk.com http:hww.paltalk.com PalTalk

1

.:%1

Yahoo! Messenger

1

~

Text. Audio, Video

Yahoo

http://messenger.yahoo.corn/

t

I

I

I

9.4.2 Shared H\pernrediii

The term “hypermedia” refers to an “on-line setting where networks of multimedia nodes connected by links are used to present information and manage retrieval” [83, p. 6621. In fact this term can be considered an umbrella term. referring to any computer-stored information related and retrieved by links. The World Wide Web is a well-known hypermedia environment that users can access through the Internet by using interactive browsers. such as Microsoft’s Internet Explorer or Netscape’s Navigator. If the information is textual in the first place, we talk of a hypertext, and if there are certain visual, musical. animation elements or the like included. we talk of a hypermedia [84]. Hypertexts and hypermedia are structured around the idea of offering a communication environment that mimics human thinking-that is, an environment that allows the user to make associations between “concepts” rather than move sequentially From one to the next. as in an alphabetical list. Hypermedia concepts are thus linked in a manner that allows the user to jump from subject to related subject in searching for information. For example. a hypermedia presentation on navigation might include links to such topics as astronomy. bird migration, geography, satellites. and radar. For this reason. hypermedia tools are widely used in distance learning. How is it possible to integrate the advantages of hypermedia with traditional synchronous CMC? A possible answer comes from Shared Hvpermzedia (SHY): new Internet tools attaching computer mediated communication to Web browsing [85.86] In SHY. different users. who are simultaneously browsing the same Web site. can communicate with each other and $hare files or web addresses (see Table 4.6 for 3 list of

G. Riva / Communicating in CMC

21 7

available SHY). Using a simple interface, usually resembling a little remote control (see Figure 9.4), SHY users can get a constantly updated list of all the other online users who are visiting the same Web site [ 8 5 ] . All they have to do is click on any person icon, open a message window, and start the communication as happens in traditional instant messaging. The twist is that the chatters automatically have something in common: the Web page they're reading. Table 9.6 Shared hypermedia clients

Tool ICQSurf Odigo

Developer ICQ NovaWiz

Website http://www.icq.codicqsurf http ://www.odigo.org

Usually a SHY lets the user conduct group and private chats, exchange information or files, and share the same web pages (for a more detailed characteristics of a typical SHY, see Table 9.7).

Figure 9.4 Shared Hypermedia interface (Odigo)

G. Rirta Communicating ill CMC

’18

Table 9.7 Features of shared hypermedia (Adapted from Riva. 2001 )

One-on-one or multi-user chat Users can make calls to multiple people up to 100/1000 (text, audio and video) users. In multi-user chat one ore more moderators can control group participation by sharing the microphone. i It is also possible to broadcast a radio-style Internet talk 1 i 1 show where the host maintains control and invites listeners to participate to the event. Users can send text and voice messages to users who are Email (text and \voice) ! not online. I Users can create their own Web Tour and escort other users through a list favorite web sites. tozrr Search engine Users can find other users with a specific sex. age andior similar interests. I Transfer of files Users can upload and download from other users documents and files.

1

1

1

I

I

1 f

~

On any web site, SHY users can see a list of other users and talk with them on group and private levels. SHY further enhances the user experience by consolidating different form of computer mediated communication (e-mail, Instant Messaging, etc.) into one fully integrated interface. Many SHY also have a search engine that can be used to find users with a specific age and/or similar interests. In this way it is really easy to set up a group with a common interest. like Social Psychology, or get online to practice a foreign language with a mother-tongue users. By assembling people with similar interests and surfing habits. this new Internet platform transforms Web browsing into a social activity. In this sense a SHY can be the starting point of community-centered environments based on communities of practice

[W.

What are communities ojpractice? At the simplest level, they are a small group of people who share a common goal over a period of time. They are not a team, a task force. or an identified group but attend the same course, collaborate on a shared task or work together on a product [87, 881. More specifically, they are peers in the execution of “real activities.” What holds them together is a common sense of purpose and a real need to know what each other knows. 9.4.3

Wehlogs

If SHYs are the product of the union of hypemedia with synchronous CMC, the union of hypermedia and asynchronous CMC produced weblogs. A weblog (also blog) is a web page made up of usually short, fiequently updated posts that are arranged chronologically as in newsgroups (see Table 9.8 for a list of Weblog clients). The content and purposes of blogs are mainly expressive: links and commentary about other web sites. news about a company/person/idea,photos. poetry, project updates. even fiction. Most weblogs are personal: public diaries of web pages to recommend to other users (see Figure 9.5). Some others are collaborative efforts based on a specific topic or area of mutual interest. They help small groups communicate in 3 \vay that is simpler and easier to follow than email or discussion forums.

G. Riva / Communicating in CMC

219

Figure 9.5 Weblog interface (Diaryland)

Using a private weblog on an intranet allows team members to post related links, files, quotes, or commentary. As noted by blogger, one of the sites where is possible to set up a weblog for free [89], this new form of CMC “can help keep everyone in the loop, promote cohesiveness and group culture, and provide an informal voice of a project or department to outsiders”. Table 9.8 Weblog clients

Tool Blogger Diary land Livejournal GrokSoup

Developer Pyra Labs. Diaryland Livejournal.com GrokSoup.com

Website http://www.blogger.com http ://www.diaryland.com/ http://www .livejournal.com/ http://www.groksoup.com/

9.4.4 Graphical chats Graphical chats (see Table 9.9) are a different variant of classical text-based IRC. As we have seen extensively before, text chats lack nonverbal cues that facilitate face-to-face conversations, such as gestures or physical distance. To overcome these limitations graphical chats add visual representations for physical bodies and spaces to a text chat window. In particular, graphical chats provide a visual representation of both the rooms as a 2D/3D space, and the different users. In fact, each user can choose a graphical representation - picture, drawing or icon - of himself or herself that is called avatar. Usually it is possible to select a standard avatar provided by the program, an avatar created by another user, or to create a custom avatar. In graphical chat clients, however, text is still used for the actual conversation; users communicate with others via typed text that appears in “speech balloons” that pop up next to the participants’ avatars. All users within the same 2D/3D space can see each other’s messages (with the exception of whispers - private point-to-point messages), irrespective of the distances between avatars. Moreover, most avatars have a preset range of expressions - happy, angry, etc. - and behaviors -jump, fight, kiss, etc. - that can be used to convey non verbal cues.

G. Riva 1 Communicating in CMC

220

Table 9.9: Graphical chats

Graphical 3 0 chats Active Worlds ChatPOP Cyber Net Worlds Galaxy Worlds Outerworlds Talk World ~

Web address http://www .activeworlds.com/ http ://www .he1lopop.com/ENGLISWi ndex .asp http://www .cybernetworlds.com/ http://www.galaxyworlds.com/welcome.htm http ://www.outerworlds.com/ http://www.talkworld-online.com

Figure 9.6: ZD Graphical Chat interface

However, as noted by Viegas and Donath [17] the use of avatars introduces new problems: "Space needs to be allocated for every user's avatar as well as for their speech bubbles. The screen becomes quickly cluttered. which can hinder communication. More subtly, the avatars can distort expression and intent by providing a small range of (often broadly drawn) expressions that overlays all of a user's communications. Even if an avatar has several expressions. and many do. it is still a far cry from the subtlety of \mbal expression. let alone our physical gestures." (p. 10).

i I

1 1

1

1

G. Riva /Communicating in CMC

22 1

Churchill & Snowdon [90] recently identified a series of key issues a 3D chat developer has to face for supporting effectively the communication process (pp. 5-7): - the transition between shared and individual activities. Actors should know what is currently being done and what has been done in the context of the task goals. - flexible and multiple viewpoints and representations. Tasks often need use of multiple representations each tailored to a different point of view and different subtasks. - a shared context. The shared context is composed of symbolic references which allow actors to orient and coordinate themselves. It includes the shared knowledge of each other’s current activities, shared knowledge of each other’s past activities, shared artifacts and shared environment. - the awareness of others. This awareness includes both the knowledge of shared task related activities and the sense of co-presence. - the support to communication activities . Negotiation through face-to-face talks is important for collaboration. In fact, conversation analytic studies of negotiation at work have detailed how subtle verbal and non verbal contribute to such negotiation.

Figure 9.7: 3D Graphical Chat interface

Also in this vision the key content of communication is the interpretation of the situations which actors are involved in. So, the most effective way of clarifying the meaning of messages is to connect them to a shared context of meaning. Generally, development of CVEs calls for conceptual mechanisms with which groups can be built and vehicles through which groups can express themselves [91]. To address this issue, 3D chat designers usually use some tricks. For instance, even if 3D chat fail to convey eye contact, facial expressions and real body images of people, some of the effort in their design is focused to the development of tools for the creation of faces. This choice reflects the considerable societal attention on the face as medium for expression and

information display. Particularly, facial expressions exceed verbal reports to enhance context comprehension. Many developers of graphical chats systems are also aware of the need to “create community’’ in the context of their efforts [91]. Even if many traditional means for creating community are not available, a great effort is given to the creation of virtual town squares or meeting rooms. According to Coate [92] the work of maintaining virtual communities is similar to the one of an innkeeper: facilitating interaction and keeping order among patrons. In fact. if graphical chats have to serve as community for their users, they have to embody, or replace with adequate substitutes, some functions of community life that parallel those commonly provided by “traditional” communities [93].

9.5

Conclusions

The emergence of information technology is changing the way people interact with computers [6 I]. Technological advances have gradually shifted the focus from computers that have become less of an end in themselves, and more of a means in terms of what people actually do with them. The most evident sign of this change has been the diffusion of the CMC. The technological evolution of the media leads us to believe that CMC could become in the very near future, the predominant medium, or rather, it is possible that will become a general communication interface: an interface used for interpersonal relationship and for the creation and management of information. Its success is creating a new psycho-social space that is the fertile ground for social relationships, roles, and a new sense of self [6, 821. As recently noted in a recent interview [94] by Sherry Turkle, a Massachusetts Institute of Technology researcher, the CMC and in particular Internet, “is the identity technology - much of what people do online, is self-explanation and presentation. from searching and e-mailing, to chatting or creating a home page” (p. 17). “The Web is a safe place to try out different roles, voices and identities” confirms John SuIer [94], psychologist and Web researcher for the Rider University. “It’s a sort of like training wheels for the self you want to bring out in real life” (p. 17). The result of these new selves is a new sense of presence that fills the space with a fluid form of networkkommunity that is usually called cyberspace [6, 8, 61, 951. Cyberspace is a universe made up of things that can be seen and heard, but they are neither physical objects nor necessarily a representation of physical objects. They are built of information coming partly from operations of the physical world, but largely fiom the accumulation and exchange of knowledge arising from human initiatives in the fields of culture, science and art. In this sense. a key goal for psychology and communication researcher is doing more thinking and theorizing about how to get people to make better connections between the cyberspace and the rest of their lives [96]. However, this is not an easy task. In fact. CMC is a new form of communication with significant differences from non-electronic written communication, as well as from other existing means of communication. There are two main reasons for these differences: - communication with a keyboard and computer screen takes longer than normal face-toface communication, and; - the absence of metacommunicative features like facial expression. posture and tone of voice. This is why CMC can be described as a form of miscommunicafion. 1.e.. a necessarily “pared-down” or, perhaps. more accurately, rarefied form of conversation [6. 71. which lacks the rules on which effective interaction depends.

G. Riva / Communicating in CMC

223

According to the "cues-filtered-out" approach [3 51, CMC lacks the specifically relational features (social cues), which enable interactors to identify correctly the kind of interpersonal situations they find themselves in [33]. The conclusions proposed by these authors are that CMC occurs in a social vacuum where the personal identities of subjects tend to fade and vanish [34]. The chapter tried to counter this vision using in its theoretical frame five different approaches here listed in alphabetical order: - the Miscommunication as a CHance Theory (MaCHT) presented by Anolli in the first chapter of this book: a strategic use of miscommunication may enhance the degrees of freedom available to the communicators during an interaction. If a user handles well the miscommunication processes typical of CMC, he/she may even achieve results difficult to obtain in face-to-face meetings. - the Positioning Theory (PT): PT replaces the traditional concept of role with the concept of positioning. The main difference between the two is that a role is a stable and clearly defined category, while positioning is a dynamic process generated by communication [64, 651. - the Situated Action Theory (SAT): action is not the execution of a ready-conceived plan, but the subject's adaptation to context [62]. - the Social Identity Model of Deindividuation Effect (SIDE): a social or a group identity replaces individual identity in CMC [4 1,421. - the Social Information Processing (SIP) perspective: users adapt existing communicative cues, within constraints of language and textual display, to support processes of relational management [27,43,44], Starting from the above theories, the chapter outlined how the CMC users are able to make order and create relationships out of the miscommunication processes typical of this medium. Moreover, it presented the emerging forms of CMC - instant messaging, shared hypermedia, wehlogs and graphical chats - and their possible social and communicative effects. This analysis enables us to reach the following conclusions: - Communication is as the outcome of a complex coordinated activity, an event that generates conversational space within the weave of personal and social relationships. Thus, communication is not only - or not so much - a transfer of information, but also the activation of a psychosocial relationship, the process by which interlocutors coconstruct an area of reality. In CMC this happens inside a rather special kind of container - Cyberspace - that tends to rarefy the structural and process features of communication. - the CMC experience may be defined as a process by which a group of social actors in a given situation negotiate the meaning of the various situations, which arise between them. In this sense, the most important mediator of behavior in CMC is the purpose of the people who use them. Particularly, their use depends on how the CMC processes are interpreted and what projects are in them. Obviously, the issues raised in this chapter are just a first step towards a definitive analysis of CMC and its effects on our lives. But it allowed us to demonstrate that communication technologies are no longer seen by researchers as rigid prostheses external tools marking the limits and limitations of users who are slaves rather than masters - but as an opportunity: ways of genuinely enhancing the communication of the interlocutors who use them.

224

9.6

G. Rira 1 Communicating in Ch4C

Acknowledgements

I would like to thank Fondazione "Piero, Pietro e Giovanni Ferrero" (Alba. Italy) for the support given to this study. The present work was also supported by the Commission of the European Communities (CEC), in particular by the IST programme (Project VEPSY UPDATED. IST-2000-25323 - http://www.psicologia.net - http://www.vepsy.com). Moreover, the author has benefited from discussions on topics relevant to this chapter with other colleagues. Their support and contributions are hard to overestimate. These people include Luigi Anolli, Carlo Galimberti and Fabrizia Mantovani. 9.7

References G. Riva and F. Davide. Communications through Virtual Technologies: Identity. Community and Technology in the Communication Age, in Emerging Communication: Studies on ,Vew Technologies und Practices in Communication, G . Riva and F. Davide. Eds. Amsterdam: 10s Press. Online: http:i/www.emergingcommunication.com/volume 1 .html. 200 1. M. Lea, Rationalist assumptions in cross media comparisons of computer mediated communication. Behavior and Information Technolorn9 10 (1991) 153-1 72. R. R. Rice, Media appropriateness - Using social presence theory to compare traditional and new organizationa media, Human Communication Research 19 ( 1993) 45 1-458. J. Short. E. Williams. and B. Christie. The social p . ~ ~ ~ h o /of'relecnmmunicc~tions. ofl* London: Wile?. 1976. R. L. Daft and R. H. Lengel, Organisational information requirements. media richness and structural design. Management Science 32 ( 1986) 554-57 1. G. Riva and C. Galimberti. Interbrain frame: interaction and cognition in computer-mediated communication. CvherPs.vchologv & Behavior I ( 1998) 295-3 10. G. Riva and C. Galimberti, Towards CyberPsychology: Mind. Cognition and Society in the Internet Age, in Emerging Communication: Studies on Ne1.r. Technologies urid Prtrctic~rsin C'i~mrtirrrticariorr. G . Riva and F. Davide, Eds. Amsterdam: 10s Press. Online: http:iiwww.emergingcommunication.com/volume2. html, 200 I . G. Riva and C . Galimberti, Computer-mediated communication: identity and social interaction in an electronic environment, Genetic. Social and General P.sJ*cholog,* Monographs 124 ( 1998) 434363. C. Goodwin and J. Heritage. Conversation analysis. ..lnrttrrrl Rei*iriisn/'.4nthropn/oL~~ 19 I1990) 28.3307. C. Galimberti, La cotiversazione [Conversation]. Milan: Guerini e Associati. 1992. S . E. Brennan, Conversation with and throuFh computers. L'.ser modt~llinp a t i d riser-udrrpred interaction I ( 1991 ) 67-86. M. L. Patterson. Nontrerhal hehai?or: a.fiitic.tionalper.YpeL"ii'e.New Y ork: Springer-Verlag. 1983. M. L. Patterson. Functions of non-verbal behavior in social interaction. in Handbook q/'Languag:e and Social P.~Jrho/ogi*, H. Giles and W . P. Robinson. Eds. Chichester. UK: Wile>-. 1990. pp. 239270. L. Chemy. Coni~ersationand Comntiirtiti.: Chit i n il \irtual \t.orld. Stanford. CA: CSLI Publications. 1999. L. Cherny. The MUD register: Conversational modes of action in a text-based virtual realih. in Linguistics Department. Palo Alto. CA: Stanford University. 1995. G. Mantovani. .Ire\z. romniunicution ~ t i ~ ~ i r ~ ~ t i n i froin c ~ r i t sc\vt?*dii,1. : to iirtuui. London: Taylor & Francis, 1996. F. B. Viegas and J. S. Donath. Chat Circles. presented at CHI 99 conference on Human factors in computing systems, Pittsburgh. PA, USA. 1999. E. A. Mabry, Textual framing as a communication climate factor in online groups. in Perspecrira o ~ i mrdiu and our understanding ?/'the sncitri tc.or/J. S . D. Reese and 0. H. Gandy. Eds. Hillsdale. NJ: Lawrence Erlbaum Associates. 2001. pp. 323-336. P. Curtis. Comunicazione via MUD: i fenomeni sociali delle realta virtuali basate su testa {Communication \ria MUD: Social phenomena of \,irtual reality hased on text. 1. Sirrtwri lt1rtd/isynri 8 (1906)229-25.3,

G. Riva / Communicating in CMC

225

P. Curtis, Mudding: Social phenomena in text-based virtual realities, in Culture of the Internet, K. Sara, Ed.: Lawrence Erlbaum Associates, 1997, pp. 121- 142. M. R. Parks and K. Floyd, Making friends in Cyberspace, Journal of Communication 46 (1 996) 8097. G. Mantovani, Virtual reality as a communication environment: Consensual hallucination, fiction, and possible selves, Human Relations 48 (1 995) 669-683. L. A. Marvin, Spoof, Spam, Lurk and Lag: the Aesthetics of Text-based Virtual Realities, Journal of Computer-Mediated Communication 1 ( 1995) on-line: http://www.ascusc.org/jcmc/vol1/issue2/marvin.html. J. B. Walther, J. F. Anderson, and D. W. Park, Interpersonal effects in computer-mediated interaction: A meta-analysis of social and antisocial communication, Communication Research 21 ( 1 994) 460487. J. Siegel, V. Dubrovsky, S. Kiesler, and T. W. McGuire, Group processes in computer-mediated communication, Organizational Behaviour and Human Decision Processes 37 (1 986) 157-187. R. R. Rice, Computer-mediated communication system network data: Theoretical concerns and empirical examples, international Journal of Man-Machine Studies 32 (1990) 627-647. J. B. Walther, Interpersonal effects in computer mediated interaction: A relational perspective, Communication Research 19 (1992) 52-90. P. A. Thompsen and D. K. Ahn, To be or not to be: An exploration of E-prime, copula deletion and flaming in electronic mail, ETC: A Review of General Semantics 49 (1992) 146-164. M. L. McLaughlin, K. K. Osborne, and C. B. Smith, Standards of conduct on Usenet, in Cybersociety, S. Jones, Ed. Thousand Oaks, CA: Sage, 1995, pp. 90-1 1 1. J. S. Donath, Identity and deception in the virtual community, in Communities in Cyberspace, M. A. Smith and P. Kollock, Eds. London: Routledge, 1999, pp. 29-59. P. Brown and S. Levinson, Universals in language usage: Politeness phenomena, in Questions and Politeness, E. Goody, Ed. Cambridge: Cambridge University Press, 1978, pp. 256-3 10. E. A. Mabry, Framing Flames: The structure of argumentative messages on the net, Journal qf Computer-Mediated Communication 2 ( 1997). L. Sproull and S. Kiesler, Reducing social context cues: Electronic mail in organizational communication, Organizational Behavior and Human Decision Processes 37 (1986) 157-187. L. Sproull and S. Kiesler, Connections: New waq’s of’ working in the networked organizations. Cambridge, MA: MIT Press, 1991. M. J. Culnan and M. L. Markus, Information technologies, in Handbook of organizational communication: An interdisciplinary perspective, F. M. Jablin, L. L. Putnam, K. H. Roberts, and L. W. Porter, Eds. Newbury Park, CA: Sage, 1987, pp. 420-443. D. Abrams and M. A. Hogg, Social identity theory. Heme1 Hempstead: Harvester Wheatsheaf, 1990. M. A. Hogg and D. Abrams, Social identifications: a social psychology of intergroup relations and group processes. London: Routledge, 1988. R. Spears and M. Lea, Social influence and the influence of the “social” in computer-mediated communication, in Contexts of’ computer-mediated communication, L. Lea, Ed. Hemel Hempstead: Harvester Wheatsheaf, 1992, pp. 30-65. H. Tajfel and J. C. Turner, The social identity theory of intergroup behaviour, in Psychology of intergroup relations, S. Worchell and W. G. Austin, Eds. Chicago: Nelson-Hall, 1986, pp. 7-24. J. Turner, M. A. Hogg, P. J. Oakes, S. D. Reicher, and M. S. Wetherell, Rediscovering the social group: A selfcategorization theory. Oxford: Blackwell, 1987. T. Postmes, R. Spears, and M. Lea, Breaching or Building Social Boundaries? SIDE-Effects of Computer-Mediated Communication, Communication Research 25 ( 1998) 689-7 15. S. Reicher, R. M. Levine, and E. Gordijn, More on deindividuation, power relations between groups and the expression of social identity: Three studies on the effects of visibility to the in-group, British Journal of’Social Psychology 37 (1 998) 15-40. J. B. Walther and J. K. Burgoon, Relational communication in computer-mediated interaction, Human Communication Research 19 (1992) 50-88. J. B. Walther, Computer-mediated communication: Impersonal, interpersonal, and hyperpersonal interaction, Communication Research 23 ( 1996) 3-43. J. Sherbloom, Direction, function, and signature in electronic mail, Journal qf Business Communication 25 (1988) 39-54. J. W. Chesebro and D. G. Bonsall, Computer-mediated communication: Human relations in a computerized world. Tuscaloosa, AL: University of Alabama Press, 1989. F. Mantovani, Networked seduction: A test-bed for the study of strategic communication on the Internet, CyherPsychology and Behavior 4 (200 1) 147-154. S. Utz, Social information processing in MUDS: The development of friendships in virtual worlds, Journal qf Online Behavior 1 (2000) online: http://www.behavior.net/JOB/v 1n 1 /utz.html.

236

G R ~ v a Communicating in CAK

J. B. Walther and P. D'Addario. The impacts of emoticons on message interpretation in computermediated communication, Social Science Computer Review 19 (200 I ) 324-347. A. Wolf, Emotional expression online: Gender differences in emoticon use. Ci.her/?~irho/op~ and Behavior 3 (2000) 827-833. S. Utz, Der Aufbau von interpersonalen Beziehungen in MUDs: Die Rolle von Motiven und Kommunikationsstrategien [The formation of interpersonal relationships in MUDs. The role of motives and communication strategies]. Gruppendvnamik 32 (200 1 ) 145- 160. R. E. Turner, C . Edgley. and G. Olmstead. Information control in conversation: Honesty is not always the best policy. Kansas Journal yf'Sociologv 11 ( 1975) 69-89. J. B. Walther and L. C . Tidwell. Nonverbal Cues in Computer-Mediated Communication. and the Effects of Chronemics on Relational Communication. Journal ofOrgani,-ationalComputing 5 ( 1995 ) 355-3 78. [543 C. E. Shannon and W. Weaver. The mathemuticnl theor?. ol'commitmication. Urbana: University of Illinois Press. 1949. [ 5 5 ] D. G. Tatar. G. Foster, and D. G. Bobrow, Design for conversation: Lessons from C'ognoter. international Journal ofMan-MachineStudies 34 ( 199 1 ) 185-209. 1561 S. Doheny-Farina, Rhetoric, innovation, technoloa: i m e studies of tecknicul cwnmuniculron irt rechnokogv transfers. Cambridge, MA: MIT Press. 199I . H. H. Clark and E. F. Schaefer, Contributing to discourse. Cogtiiti\+eScience 13 ( 1989) 259-794. R. Ghiglione, L'homme communiquant.Paris: '4. Colin. 1986. E. Rocco and M. Warglien, La comunicazione mediata da computer e I'emergere dell'opportunismo elettronico [The growth of electronic opportunism in computer-mediated communication]. Sisfettii /nte/ligenri 7 ( 1995) 393-420. G. Stasser. Pooling of unshared information during group discussion, in Group processes und prodiictil-iti..S. Worchell. W.Wood. and J. A. Simpson. Eds. Newbury Park. CA: Sage. 1992. pp. -1867. G. Riva and C . Galimberti. The psychology of cyberspace: a socio-cognitive framework to computer mediated communication. h ' e ~/dea.s . in P . ~ , d w l o15 ~ 'f 1997 14 1 - I 58. L. Suchman, Plans and situufed action. Cambridge, UK: Cambridge University Press. 1987. G. Mantovani, Social context in HCI: A new framework for mental models. cooperation and communication. Cognitive Science 20 ( 1996) 337-296. R. Harre. Language and science of psycholog!.. ./oiirnul /Or thc Theor?. of- Sociul Behai.iour unti Personalih. 4 ( 1989) 165-1 88. R. Harre and L. Van Langenhoit. Varieties of positioning. ./ournu1 for- tlici Theon. of S o c . i t r l Behuiiour 21 ( 1991 ) 393408. A. R. Stone. Will the real body please stand up? Boundary stories about Lima1 cultures. in (:t.her.space:Firsf steps. M. Benedikt. Ed. Cambridge. MA: MIT Press. 1991 . T. Postmes. R. Spears, K. Sakhel. and D. de Groot. Social influence in computer-mediated communication: The effects of anonymity on group behavior. Per-.sonalit~~ & Social P.~?diologi. Bulletin 27 (2001) 1243-1254. I(. M. Douglas and C. McGarthy. Identitiability and self-presentation: Computer-mediated communication and intergroup interaction. British ,/ourrial ol'Sociu1 Ps?rhoIop 40 (200I ) 399-4 16. C. Fernyhough, The dialogic mind: .4 dialogic approach to the hisher mental functions. .~CJH.I r f m c in P.s?.cholo~~~ 14 ( 1996) 47-67. A. Saito. Social origins of cognition: Rartlett. evolutionary perspective and embodied mind approach. Jour-nu1qf'rhe Theog, of'Social Behuikwr 26 ( 1996) 399423. B. Davies and R. Harre. Positioning: the discursive production of selves. ./ourmil tor f l i e Theor?. of Sociul Behaiiour 20 (1990) 43-63. [731 F. Jacques, L'espace logiqiro de 1'intrrloc.iitioii.Paris: PLI F. 1985. ~ 7 3 1 M. Rosaldo, Toward an anthropology of self and feeling, in Cultitre theor?.: E.vsuy.v on the mind. .sel/' cind cmotion. R . A. Shweder and R. A. LeVine. Eds. Cambridge. UK: Cambridge University Press. 1984. pp. 124- 136. ond self: A . Manella. G. DeVos. and F. Hsu. 1741 F. Hsu. The self in cross-cultural perspective. in C.iilfrirf> Eds. New York: Tavistock. 1985. K . J . Gergen. Toiturd tran.~fimnutiottin social knoichd,qe. New York: Springer. 1987. J. Bruner. , 4 r t s of',neaning.Jerusalem:: Harvard University Press. 1993. H. Markus and P. Nurius. Possible selves. .4mericun P.stdiologi.qr 41 ( 1986) 954-969. J . Meyrowitz. .Vo stvise c?f'plucr:Tlie irnpm'r of' t~ltv-tt*onictnetliri o n .voc.iirl hc4iui.ior. Set\. Yorli: Oxford University Press. 1985. R. Kraut. M. Patterson. V. Lundmark. S. Kiesler. T. Mukopadhyay. and W . Scherlis. Internet paradox: A social technology that reduces social in\ ol\ einent and psYcholo_eical \\ell-heing'.'. . I m t ~ r i ( y i ~P7~ \ ~ ~ / i o / o q23 i s r( l W 8 ) 1 0 1 '-1031

G. Riva /Communicating in CMC

227

S. Doheny-Farina, The wired neighborhood. New Haven: Yale University Press, 1996. P. Wallace, The psychology ofthe Internet. New York: Cambridge University Press, 1999. G. Riva, The mind in the Web: Psychology in the Internet age, CyberPsychology and Behavior 4 (2001) 1-6. P.-A. Federico, Hypermedia environments and adaptive instructions, Computers in Human Behavior 15 (1999) 653-692. Encarta, World English Dictionary. Seattle, WA: Microsoft Press, 1999. G. Riva, Shared Hypermedia: Communication and interaction in Web-based learning environments, Journal of Educational Computing Research 25 (2001) 205-226. 1861 G. Riva, From real to virtual communities: Cognition, knowledge and intention in the World Wide Web, in Learning and teaching on the World Wide Web, C . R. Wolfe, Ed. San Diego, CA: Academic Press, 2001, pp, 131-151. 1871 J. Lave and E. Wenger, Situated Learning: Legitimate Peripheral Participation. Cambridge, MA: Cambridge University Press, 199 1. H. McLellan, Situated Learning Perspectives. Englewood Cliffs, NJ: Educational Technology Publications, 1995. Blogger.com, What is a webloghlog?, vol. 2001: Pyra Labs., 2001. E. F. Churchill and D. Snowdon, Collaborative virtual environments: an introductory review of issues and systems, Virtual Reality 3 ( 1998) 3- 15. 1911 J. A. Oravec, Virtual individuals, virtual groups: Human dimension of Groupware and Computer Networking. Cambridge, UK: Cambridge University Press, 1996. J. Coate, Innkeeping in Cyberspace. Proceedings of Directions and Implications of Advanced Computing. Palo Alto, CA: Computer Professional for Social Responsibility, 1992. G. Riva and G. Mantovani, The need for a socio-cultural perspective in the implementation of virtual environments, Virtual Reality (2000) 32-38. B. Murray, A mirror on the self, Monitor on Psychology 31 (2000) 16-19. S. Kiesler, Culture of the Internet. Mahwah, NJ, US: Lawrence Erlbaum Associates, Inc., Publishers, 1997, pp. xvi, 463. S. Turkle, Parallel lives: Working on identity in virtual space, in Constructing the selfin a mediated world. Inquiries in social construction, T. R. L. Debra Grodin, Ed.: Sage Publications, Inc, Thousand Oaks, CA, US, 1996, pp. 156-175.

This page intentionally left blank

229

Say Not to Say L. Anolli et al. (Eds.) IOS Press, 2002

10

Cyber-Attraction: The Emergence of Computer-Mediated Communication in the Development of Interpersonal Relationships Fabrizia MANTOVANI

Abstract: The main goal of this chapter is the analysis of cyber-attraction: the emergence of Computer Mediated Communication (CMC) in the development of interpersonal attraction. According to recent theories and studies, it would seem that not only does CMC support emotional and intensely involving communication between people, but also that it would be characterized by and offer a specific allure, a special element that makes it so fascinating. In particular, different forms of miscommunication - implicit, say not to say, and obliquity - play an important role in this process. Starting from this background, the chapter addresses the effects of CMC on interpersonal communication, and especially on the development of interpersonal attraction, aiming at identifying the specific features that this process has in cyberspace. At the same time, the chapter underlines the fact that online and oMine are not two separate dimensions to be dealt with separately and almost dichotomously. A broader approach to the study of interpersonal communication in CMC is needed: flexible and complex enough to account for the use of communication strategies in interaction contexts characterized by different levels of virtuality and by the presence of interfaces.

Contents

10.5 10.6

Introduction.. ........................................................ 230 Supporting “relationa1”communication in CMC ............... 23 1 From cyber-attraction to cyber-seduction.. ..................... 233 Towards a new perspective in cyber-seduction: communicating 237 and miscommunicating in cyberspace. .......................... Conclusions.. ....................................................... 24 1 Acknowledgements ................................................ 243

10.7

References.. .........................................................

10.1 10.2 10.3 10.4

243

230

10.1 Introduction

The social nature of Computer-Mediated Communication (CMC) is more and more evident. As Rheingold [ 11 asserts, “It is all about people communicating with other people. in any way they can and for many purposes: exchange pleasantries and argue, engage in intellectual discourse, conduct commerce, exchange knowledge, share emotional support. make plans, brainstorm, gossip, feud, fall in love, find friends and lose them, play games. flirt.. .”. (p.3). The activities today carried out over the Internet are many and complex. in a way that could not be imagined at the beginning. Certainly one of the emerging features of the Web is its relational and communicative nature: the initial centrality of the information exchange is moving to the building of on-line relationships, from friendship up to romantic (and even sexual) relationships. As pointed out before, this evolution was literally unthinkable only ten years ago. when the first studies on Computer-Mediated Communication (CMC) in collaborative tasks [241 had found CMC to be an inadequate way for people to share emotional content, let alone develop meaningful, long-lasting relationships. According to the authors. the main reasons of this inadequacy had to be found in the lack of nonverbal cues and in the diminished sense of social presence. Despite that preliminary vision, which conveyed the idea of CMC as cold and poor communication medium, the Internet is emerging as “a community of chronic communicators’’ [ 5 ] , and we see that interpersonal relationships can and do develop online

PI. Parks and Floyd [7] pointed out that personal relationships conducted via CMC are “common,” with just over 60 % of people in their sample reporting that they have formed a “personal relationship” with someone they had initially contacted through a Usenet newsgroup. The result of research, as well as the now many accounts of personal experiences show that personal relationships are being established online, from friendship to romantic and even sexual relationships. As Chenault [8] affirms, “CMC becomes a new way to “find” each other, a way for personal relationships to build”. The development of interpersonal (sentimenWfriendship) relationships on the Internet. is a very important and complex issue, seen also the growing of this phenomenon in these very last years. In order to investigate it effectively, a broad perspective is needed. Many elements are involved in this process: among these, surely interpersonal attraction is recognized as a key issue and a major push to interpersonal relationships. As Levine [9] points out, “Relationships can begin in many different environments - from the grocery store to the workplace to an online chat-room”. But, according to Brehm, “the first big step toward a relationship is always the same: interpersonal attraction, the desire to approach someone” [ 101 In other words, without attraction. there are no relationships. online or off. We know that in usual face-to-face interactions interpersonal attraction strongly relies on physical factors (especially in the first phases). Since online these features are not present, communication features become the crucial element to influence whether we are attracted by someone or not. Similarly, when showing interest to someone. other means than smiles and glances have to be used. Object of this chapter is thus cJther-atfraction: the emergence of CMC in the development of interpersonal attraction. According to more recent theories and studies. it would seem that not only CMC does support emotional and possibly rich and intensely involving communication between people. but also that it would be characterized by and offer opportunity for a specific c-tllirr-P. a special element that makes it so charming. There seems to be a specific room for implicit. w. r i o t to 5 m . . and obliquity. ,411 thow

F. Mantovani / Cyber-Attraction

23 1

communication forms which can be defined as miscommunication do play a central role in cyber-attraction. Since you don’t see the other person, there is much room for imagination and idealization, as well as the possibility of daring more “behind a wall”; there is an interface, which can act as a protecting barrier between the partners. So, the chapter has the main goal to get a better understanding of this phenomenon: on the one hand, it addresses the effects of CMC in interpersonal communication, and especially in the development of interpersonal attraction, aiming at identifylng the specific features that this process undertake in cyberspace. At the same time, the chapter underlines the fact that online and offline are not two separate dimensions, to be dealt with separately and almost dichotomously; instead, a broader approach to interpersonal communication is needed, flexible and complex enough to account for the use of communication strategies in different interaction contexts, characterized by different levels of virtuality and by the presence of interfaces. Literature that can help shedding light on these processes consists of the early work of CMC [ 11- 131, theories on interpersonal attraction [ 10, 14-171, psychosocial theories such as SAT [ 18, 191 and interpersonal communication psychology theories (Anolli, Chapter 1 in this volume) [20-231. On the one hand we will focus on the specific devices and tools CMC offers to interpersonal communication (in terms of possibly richer communication options): the communication options which can be used to convey socio-emotional meaning and to foster the reduction of interpersonal distance will be analysed. On the other hand we will investigate how online and offline dimensions merge in a number of psychological processes which are recognized as important for attraction and further steps; we will try to understand, how they develop in cyberspace and how traditional theories of interpersonal attraction apply online.

10.2 Supporting LLrelational”communication in CMC

A first step in order to understand the possibility of “online relationships” is investigating the consequences of CMC for interpersonal communication, with special focus on social and emotional aspects of communication, and on communication strategies involved in the reduction of interpersonal distance and the achievement of growing intimacy. This has mainly to do with being emotionally intense, open, express feelings and convey emotions, states of mind, etc. Much CMC research [24] is concerned with how the technical features of different communication media might influence what it is possible to convey via each medium. Among these characteristics are the richness of cues that a medium conveys (e.g., whether it conveys texts or whether it includes visual and auditory cues), the visibility or anonymity of the participants and the timing of exchanges (e.g., synchronous or asynchronous communication). Results from early research in computer-supported cooperative work (CSCW), examining the effect of different media on groups working remotely on collaborative tasks [2], on consensus building in business contexts [3], on cooperation in education [4] were not very encouraging and led to an early assumption that textual CMC systems support communication poorly, particularly socio-emotional communication. As pointed out by Preece [25], two closely related theories that help to explain these observations are socialpresence theory [26] and media richness t h e o v [ 121. Social presence theory is concerned with how successfully media convey a sense of the participants being physically present; face-to-face communication is used as the standard

232

F .Manto\,arti C\+er-..lrtractiori 1

for the assessment. The dimension of social presence depends not only on the verbal features of the exchange but also on vocal and nonverbal cues, body language, and context [ 13, 271. Reduced social cues (i.e. gestures, body language, facial expressions, appearance and so on) in CMC are caused by low bandwidth, which affects communication [28]. Media richness theory is similar to social presence but takes a media perspective [ 121. It addresses the media’s capacity for immediate feedback-how well it conveys cues. and how many and in which ways senses are involved. Social presence hndamentally affects how participants feel emotion. intimacy. and immediacy [ 131. Early studies in CMC reported fewer personal messages with lower socioemotional content (e.g. [29]) and lacking cues about social context. The “cuesfiltered-out theo? [30] was used to explain these observations [ 1 I]. Bandwidth was insufficient to carry all the communication signals needed for conveying social, emotional. and contextual content. In text-only systems, for example, both task information and social information are carried in the same single verbal/linguistic channel, which, though adequate for most task information, cannot transmit nonverbal information such as body language, voice tone, and so on. [3 1. p.4761. It is clear that this filtering out social, emotional and contextual information can have important consequences on the interaction, especially where the development of an interpersonal relationship is the main focus. According to Preece [25], there are three main ways that this affects communication: first, signals needed to understand conversation may be missing; second, the management of speaking turns may be modified; third. the impossibility of seeing and hearing the speaker can make it more difficult to infer information regarding the context of the conversation and the speaker’s feelings. In face-to-face communication. the subject can use multiple communication channels. through which he can convey emotional content. On the phone. speakers rely mainly on linguistic and non-verbal vocal cues and make use of a synchronous style of interaction. Differently, as broadly discussed in the previous chapter, communicative environments characterized by higher virtuality levels are more rarefied than the kind of interaction that happens in normal conversation. Furthermore, while face-to-face (and, partially. telephonic) conversation occurs in a cooperative environment constantly regulated by mutual adjustment and correction, where feedback and tuning are constant [32]. CMC occurs in a much less cooperative environment because of the special conditions imposed by the medium itself [33]. What is especially critical in this context is the lack of direct and immediate feedback: it can thus be more difficult to infer the intentions of the partners. Nevertheless the absence of physical and nonverbal cues should not be taken to mean that the computer medium is (completely) impersonal or devoid of social cues. or that the cues it transmits lack the subtlety of those communicated face-to-face [34. p.2 161. In fact. there is a high degree of socio-emotional content observed in CMC (e.g.[ 1 , 35-37] ). even in organizational and task-oriented settings [34, p.2 171. A number of textual and graphical devices were created. for example. to reproduce the meta-communicative features of face-to-face conversation and to convey an emotional dimension with the literal meaning of messages (for an extended analysis see Chapter 9). Emoficons (or ‘smileys’) are graphical signs combining punctuation marks and symbols into miniature sideways faces are extensively used in chat, e-mails and newsgroups to show sender’s mood. Ending a sentence with ;-) for example, “lets other know that you‘re joking or feeling cheerful“ [38]. Adding a frowning face. such as :-( does the opposite effect. As noted by Riva in Chapter 9, in Multi-User Dungeons (MUDS).text-based multi-user environments, people can use narrative descriptions of conversational nonverbal behakiors ( ’emoting’).They consist in typed messages that appear surrounded by quotation marks ”

F. Mantovani / Cyber-Attraction

233

and preceded by dialogue tags, such as “emote” or ‘pose’ [8]. Emoting, in MUDs, is a way to use commands to bring action and emotion to language. For instance, for a user whose nick-name is Mary78, typing “emote laughs loud” would result for the other users as “Mary78 laughs loud”; the character can thus be given action and movement- and even emotion. The use of poses as well as words to convey meaning gives MUDs an “odd but definitely usefbl kind of disembodied language”. More sophisticated CMC environments are represented by multi-user two- and threedimensional worlds: within 2- and 3 - 0 chats such as “The Palace” or “Active Worlds” the users are represented by avatars (graphical representations of the user), and they’re immersed in a virtual environment, where they can move around like in a videogame and act with a predefined set of choices (usually 6 or 7, including actions such as “Wave”, “Happy”, “Angry”,etc.). This can enhance the feeling of immersion and co-presence and re-introduces spatial elements as well as others nonverbal cues usually filtered out by traditional CMC. For example, as Krikorian [39] points out, the management of interpersonal distance in a virtual world can be a powerful element for the development of interpersonal relationships.

10.3 From cyber-attractionto cyber-seduction We can define seduction as the passage from a preliminary state of attraction and curiosity, to the search for contact and intimacy [14, 171. More in detail, it can be considered as an intentional and strategic behaviour induced by the attraction (usually sexual) for a subject. Main goal of this behaviour is to establish “an intriguing bond with the partner” (Ciceri, Chapter 4 of this volume), in order to reach an intimate relationship, through a progressive reduction of interpersonal distance [40]. According to the Steps Theory of Givens [41] and the bio-social model of Kendrick and Trost [42], it is possible to describe seduction as a timed sequence of interactions characterized by different “steps” or phases: after the phase of choice of an attractive partner and the phase of attracting hidher attention and interest, the third phase consists in trying to reduce the interpersonal distance in order to obtain a higher level of intimacy (and possibly establishing a relationship). The first step is thus attraction: almost all theorists agree that interpersonal attraction is a “positive or negative attitude toward another person” [ 151, p.2. “Attitude” can be defined as a “person’s readiness to respond toward an object, or a class of objects, in a favourable or unfavourable manner”. Interpersonal attraction is defined as “an individual’s tendency or predisposition to evaluate another person or symbol of that person in a positive way” ~ 1p.2., This attitude usually leads to the desire of being noticed, “to change status from being anyone to being someone” [43, 441 through a careful management of self-presentation (exhibition) and of the show of interest into the other person. This with the aim of facilitating self-disclosure (which circularly leads to an increase in liking) and reducing interpersonal distance, in order to an intimate relationship. As Ciceri points out in the fourth chapter of this volume, seductive interaction con be considered and analysed “as a flexible plurality of behavioural patterns, corresponding to the variety of communicative intentions: exhibition, approaching the partner, deepening reciprocal knowledge and reaching of a level of intimacy.” The strategies behind seduction processes are usually focused on self-disclosure through a reciprocal filtering of the level of intimacy of the information provided. This communicative approach, described and discussed by the Uncertainty Reduction Theory,

234

F. Mantovani / Crher-Attraction

allows for the reduction of the uncertainty level without the risk of being excessively vulnerable to the decisions of the partner [38]. Given the impossibility of knowing the intentions of the partner, the seductive goal is manifested gradually, through the use of an “oblique language [45]. Main characteristic of this language is ambiguity: the subject discloses himherself without an excessive exposure. In this way it is possible to reduce the risk of being rejected, avoiding a possible reduction in the self-esteem. According to Ciceri (Chapter 4 in this volume), “the act of seducing is a subtle and enticing game, which requires using the right amounts of exhibition or disguise, flaunting or revealing slightly, saying something but not saying too much“. In face-to-face communication, a significant role is played by nonverbal communication: the subject has both to be convincing and attractive; this is obtained by a subtle negotiation process in which nonverbal cues are of central importance [ 14: Anolli. Chapter 1 of this volume; Ciceri, Chapter 4 of this volume]. But what happens to this complex process within cyber-space, where a number of new opportunities ( e g more careful management of self-presentation) and constraints (miss nonverbal channels) are present? It is now time to investigate the implications CMC has on important relational processes deeply involved in the development of interpersonal attraction. such as first impression formation, showing interest into someone, self disclosing etc. We will discuss which form they take online and whether and how traditional theories apply in this context: the general aim is however to highlighten the globalit?.,of the communication e-rperience and to try to explain the continuity of the communication processes. from online to offline contexts of seductive interaction. As a general hypothesis, in fact, the mechanisms involved in offline communication are not different from the ones implied in the online communication. In particular, we assumed that a unitary and coherent theoretical framework and some general principles should be used to analyse seduction and interpersonal attraction in these different communication contexts. ”

10.3.1 First impression formation: choosing an attractille partnerWhen first getting to know a new person, a central role is played by the development of impression formation. In everyday face-to-face interaction the partners rely on different types of cues, such as beauty, voice, way of moving, way of talking, clothing, etc to build the first impression of the other. But how do these impressions develop in different environments and especially on-line? As we have seen in Chapter 9, according to the “cues filtered out” theon. [30], also called “reduced cues” theory, the computer filters out aspects of communication that are of central importance in face-to-face communication (eye-contact. self-contact gestures. posture. voice pitch, intensity, stress, rhythm and volume). Absence of non-verbal cues makes the process of first impression formation take on partially new forms. Using alternative media, such as CMC. the actors rely on different cues, such as language, style. timing and speed of writing, use of punctuation, use of emoticons. etc. According to Lea and Spears [34, p.2171, “even first-time users form impressions of other communicant’s dispositions and personalities based on their “communication style”. Furthermore. although it is true that the cues conveyed in face-to-face interaction are generally more than in CMC, in “real life settings”. they are overwhelmed by the so-called “phJ*ukn/attractiseness stereoype” 1461: physical attractiveness was found to be one of the essential factor influencing first impression formation and development of

F. Mantovani / Cyber-Attraction

235

interpersonal attraction. The influence of physical attractiveness leads to a spiral by which more attractive people are addressed more favourably and thus induced to show the best of them and leading to a hrther increased liking. Online, the lack of visual cues about physical appearance can give the partners an opportunity to interact without the weight of physical attractiveness stereotype. Finally, though we can generally affirm that the richness of cues conveyed in face-toface interaction is hardly present in CMC, the impression of our partner built in mediated communication is not for that less complete and articulated. According to Lea’s and Spears’ “SIDE theory” [34], it is possible to predict that the smaller amount of cues that appear in CMC can take on great value, possibly leading to an “over-attribution” process, building stereotypical impressions of partners based on language content of CMC messages. As far as rhythm and differences are concerned [8], Walther [28] also found evidence, within a “social information processing perspective ”, that computer-mediated groups gradually increased in impression development to a level “approaching that of face-to-face groups (p. 381). Social information processing suggests different rates and patterns of impression development using alternative media, such as CMC. Thus, it takes longer to find enough information to be able to form impressions via CMC, but that it does happen, according to Walther, and that process by which we form impression is not actually “altered” via CMC, only slowed down. 10.3.2 The management of seductive interaction: attracting attention and showing interest After getting in touch with someone for the first time, how can we think these relationships develop online? And how can the game of courtship develop? What are the means to expresshhow interest into someone into such a “rarefied” communication medium? These are some of the questions we’re going to address in this section. Usually, as pointed out by Altman, [47, p.271, as people go on in their interaction and start establishing a relationship, they “gradually move toward deeper areas of their mutual personalities through the use of words, bodily behavior, and environmental behaviors”. In the management of seductive interaction, there are some important measures of attraction that do not come into play in CMC, nor can they be measured, including: eyecontact (duration and type), proxemic indexes such as “inclination” to one another (leaning towards and other body language), the distance between the partners, etc. According to Ciceri (Chapter 4 in this volume), “the form of seductive message is multimodal, depending on the relation between signals belonging to different systems of expression (vocal, verbal, visual, kinetic), which involves the whole person. During the course of seductive interactive game, communication is influenced by verbal and nonverbal components .” In CMC, bodily and environmental aspects are reduced or removed, giving worddtext all the importance”. Nevertheless, there are specific communication tools that have been identified in the literature [8] as able to compensate for the lack of nonverbal cues and to convey emotion. Furthermore, “the seducer should be able to act in an implicit and intriguing manner because enticing the other person needs an attentive and sensitive process of negotiation”. In the following brief description there are some elements that are recognized playing a crucial role in the development of attraction, online as well as offline. Proximity and frequency of contact: as noticed by Wallace [46], one obvious explanation for the proximity effect is that it provides the opportunity to meet and know

236

F. Manrowni / Crher-Attraction

significant others. This makes the other person become familiar, predictable and develops a positive effect of exposure [lo]. Moreover, one reason the real-life proximity promotes attraction is that the person’s nearness makes people expect -and anticipate- fbture interaction. In the online world, proximity is not defined by physical location, but instead by the shared use of a particular tool (chat room, newsgroup, etc.). This particular proximity is also called intersection frequency. In the design of new CMC environments this feature is supported through the insertion of “buddy lists”, allowing people to know whether hisher friend are online, making it thus much easier to stay in touch. Similariv: according to Brehm 1101, people are more attracted to someone they believe has attitudes similar to their own than to someone whose attitudes differ: the greater the proportion of shared attitudes, the greater the attraction to that person. On the Internet. where it is more difficult and time consuming to learn how people think about multiple issues, the law of attraction may cause many “false starts” in friendship and love. In a virtual medium, in fact, as Schnarch points out 1481, there is no way to really know whether someone’s attitudes are similar to one another’s, because it is not possible to compare a partner’s self-presentation with what one can see for himself. This limitation. on the one hand produces a large amount of idealisations and expectations, and on the other hand allows a strategic management of the similarity perception. Express liking: as Wallace [46] points out “when someone likes us. we tend to like this person back“. Partly because we’re flattered and the other person’s attraction to us raises our self-esteem, just as physical attractiveness creates a spiral of positive treatment and positive response. In real life, there are many cues that people can rely on to detect that someone likes them: smiles, attention, glances, etc. On the Internet people must rely on other methods to show that they like someone else. The most important is probably attention: flatteries, compliments and signs of interest, which can be expressed in various forms (a nice compliment/positive comment on a message posted by someone in a forum) and multimedia devices (e.g. virtual flowers).

Humour: the role humour plays in making people more attractive is well recognized [49]. Especially as far as self- or object directed humour are concerned, also on the Internet, humour can be a powefil force in interpersonal attraction, particularly because it

isn’t overshadowed by the physical appearance factor. Humour can be expressed very easily on the Net in just typed text [50]. and those who excel at it are probably increasing their score on the interpersonal attraction scale. 10.3.3 The paradox of self-disclosingin cvher-space As noted above. self-disclosure is a central process in the development of attraction and crucial to the establishment of relationships [ 161. In fact. developing a close relationship with another person calls for a certain level of intimacy, or self-disclosure. Over time, the exchange deepens and the two people disclose more and more information to one another. The management of self-disclosure. with the filtering of personal information is extremely delicate. As Ciceri (Chapter 1 in this volume) underlines, “it is necessary to be able to get the right balance between allowing oneself yet denying oneself. putting oneself forward yet restraining oneself’; it is all a matter of “saying a little but not saying too much” [45]. People can manage the timing. the breadth and depth of self-disclosure. which is rewarding to the listener because it makes the other person feel worthy of receiving

F. Mantovani / Cyber-Attraction

231

intimate information, thus raising self-esteem. Jourard and Lasakov [5 11 hypothesized that “liking” another person is mainly a result of having “disclosed” to the person, almost “independent of that person’s reaction to the disclosure” (cited in Altman and Taylor, 50). Altman and Taylor posit an interactional theory: “Revealing leads to liking and liking leads to revealing, “as a cyclical and continuous set of events (50). As Chenault [8] points out, “because often in CMC people are interacting with relative “strangers” (people they have never met “in real life”), the dynamics of when to disclose, and what to disclose come into play. Within cyber-space, self-disclosure seems to be influenced in a quite complex and partially controversial way: on the one hand, selfdisclosure is strictly tied to the development of trust and according to Mehrabian [52], “people are discouraged from expressing personal feelings to strangers and so it becomes necessary to rely on implicit behavior to infer how another person feels and how to pursue a relationship further” (p.156). On the other hand, it would seem that anonymity and the reduced sense of presence given by the interface between communicators allows a more dishinibited behavior, fostering the possibility to open oneself (there would be less or no fear of “losing face” [53]).

10.4 Towards a new perspective in cyber-seduction: communicating and miscommunicatingin cyberspace 10.4.1 Seduction and miscommunication

As we have seen, establishing a sentimentahtimate relationship over the Internet is not an easy task. As pointed out before, whether on- or off-line, seduction is a very complex communication act: it can be defined as a strategic and intentional sequence of moves, primarily pushed by attraction (usually sexual) towards another person (usually of the opposite gender), having the main goal to attract himher in order to reduce interpersonal distance and establish an intimate relationship. As we have seen, the strategies behind seduction processes are usually focused on the management of self-presentation and on self-disclosure through a careful reciprocal filtering of the level of intimacy of the information provided. Given the impossibility of knowing the intentions of the partner, the seductive goal is manifested gradually, through the use of an “oblique” language [45]. Main feature of this linguistic style is its ambiguity and enigmaticity: the subject discloses hidherself without an excessive exposure. In this way it is possible to reduce the risk of being rejected, avoiding a possible reduction in the self-esteem. It is a very subtle task, where in order to achieve his goals, the subject has to run along a border area between implicitness and explicitness, ambiguous and direct style, communication and miscommunication. As Anolli points out in Chapter 1 of this volume, communication is a complex game of intentions, including both simple communication acts characterized by the pursue of a single intention (either instrumental or self-revealing) and complex acts (multi-intentional). In this second case, two separate intentions are present and coordinated within a hierarchical structure, which allows the pursue of an intention (self-revealing) through the achievement of the other (the instrumental intention). Seductive communication is made of complex acts (multi-intentional), where usually all first-level level communication acts (making questions, compliments, offering flowers, etc.) are super-ordinated by the higherlevel intention (being noticed, bringing to self-disclosure in order to reduce interpersonal

238

F Mantotuni C:i*her-..l trrcrcrion

distance. etc.). For example, manifesting an (instrumental) intention (for example asking for a glass of water) is used not only for the simple achievement of this intention by the means of its communication (to have a glass of water), but as an instrument to communicate a different intention at another level (I ask for some water to signal my presence, etc.). At the same time, manifesting a self-revealing intention (presenting oneself as rich. or powerful. etc. ) is aimed at appearing more interesting to the partner’s eyes, in order “to be chosen”. In this case. the management of self-presentation carried out during courtship, where people tend to present themselves at their best, enhancing their qualities and minimizing defects, has many point so f contact with deceptive communication. The “courtship dance” is a context where people tend to resort to “say not to say, where the attitudes of the speakers seem to be intentionally ambiguous. confused. misdirected and pretended. As pointed out by Anolli, in Chapetr 1 of this volume, this is a typical miscommunication situation, where discursive interactions have their opaque border areas where communication still succeeds without being openly and explicitly strived for. It seems to be a normal phenomenon in our everyday communications. where communication can be successful without being perfect. We are in a situation where the success and the value of a communication outcome are not in the transfer of a “package of information” (Shannon & Weaver’s perspective [54]). but rely in the play of negotiation of the interactants’ intentions. Understanding the meaning of an utterance does not imply the understanding of a transparent semantic content determined solely by its autonomous truth conditions. Understanding it is more similar to an interactive play in which the players have to negotiate reciprocal intentions. (Anolli. Chapter 1 of this volume). On the one hand, this makes “communication a “risky” task” in which the borderline between communicative and noncommunicative intentionality in speakers” are very feeble. On the other hand this game of intentions, and the continuous process of reciprocal re-interpretation of the partners’ intentions increases the possibilities in human communication. As suggested by the MaCHT (Miscommunication as CHance Theory) by Anolli (Chapter I in this volume). ”miscommunication is not plainly a defect or a mismatch; actually, it is also a chance. because it enhances the degrees of freedom available to the communicators in their interaction”. 10.4.2 Seductive (mis)communication in cYt6erspace and Situated Action Theor?.

From the analysis carried out in the previous paragraphs, we can see that people engaged in a seductive interaction online as well as offline are faced to specific opportunities and constraints, which can enhance both the risks and the opportunities of miscommunication. In CMC, communicating to attract and seduce implies being able to manage the interaction and the relationship in the context of a different and new communication environment. It means adapting to and interacting with other people through an interface. in a complex way and adopting new channels. new codes. in a context which offers new opportunities as well as many limitations and constraints. as compared to the environments Lve’re used to (e.g. face-to-face). etc. There are in fact, in cyberspace, some specific elements that can influence miscommunication, enhancing opportunities and risks for the management of this complex communication act offered by it. For example. we could think that the presence of an interface would lower the fear of ”losing the face” [53]; at the same time. it makes e\.en more difficult than usual to detect the other person‘\ feedback.

F. Mantovani / Cyber-Attraction

239

Baym [ 5 5 ] writes about the “egalitarianism”, in that many see CMC allowing peoplemaking aspects like appearance mute points and giving everyone who can type an equal chance. However, the very equalizing aspects of CMC that are seen as positive can also, as the reduced-cues [30] perspective points out, cause problems and a number of possible misunderstandings. The lack of context leaves a lot of room to misunderstanding, illusions, projections and also deception and lying. It is enough to remark that, for example, in faceto-face interactions we (usually, at least) know for sure the gender of the person we are talking to. It is a fundamental element for the choice. This is not the case in CMC, where gender swapping (presenting oneself as a person of the opposite gender) is not so rare a practice. Although the described effect of CMC features on the definition of opportunities and limits for the interaction are generally accepted and considered true, the situation is somewhat more complicated. Each communication environment (face-to-face, chat, 3D chat, etc.) is different: it offers different signalling channels, possesses different features. But also each interactant is different: he/she has different personality characteristics, different technological competence, culture, etc.; furthermore, each interaction is different, according to context and interactants. The specific characteristics of the interactants and the unique features of their interaction come into play in determining the meaning of these features (in terms of opportunities and constraints) and in exploiting them in order to reach their goal. Do different people univocally interpret these features as opportunities and limits? And how do people interact with these opportunities and limits in order to reach their goals? A theory that helps understanding and investigating interactions typical of CMC is Situated Action Theory [19], as implemented in the model by Mantovani [56], which analyses this process (of interaction with artefacts) in a complex, multi-level way. According to the SAT, the action is an adaptation of the subject to the context he’s placed in, a context which offer him a number of specific affordances/opportunities and limits/constraints. The focus of the SAT approach is to “investigate how people use the affordances to develop an intelligent course of action”. According to the author, the way people use and interact with the artefact (computer/characteristics of CMC), the so-called “local interaction with the artefact ”, is always situated in a certain context. Interaction depends on the interpretation of the situation by the actor. According to this interpretation, the subject gives a personal meaning to the characteristics of the environments and identifiedperceives affordances and limits. This happens according to a number of elements, such as: personality characteristics, socio-cultural contexts, objectives and goals of the actors. 10.4.2.I Perceiving opportunities and limits The same CMC feature (e.g. presence of the interfacebarrier) can be interpreted differently and represent something different, either an opportunity or a limit according to a number of subject’s characteristics (the most salient are: extroversionhtroversion, level of competence, physical appearance, previous experience with the media involved). For example, someone who is very pretty (or handsome) can feel more comfortable in face-to-face interaction, because of the importance physical appearance has. He/she can see the fact of not being seen in chat as a limit, where as an ugly or not very attractive girlboy could find it a resource. This, also because some of the elements that are overshadowed in face-to-face, such as the weight of words for words, do have a great importance and permanence in CMC. As another example, this perception can also change also according to the personality

characteristics of the people. For instance, we could expect an extroverted person could have no problems either in face-to-face or in other media, or probably miss the "human touch" when chatting; on the other hand, an introverted people could find himself more at ease in the context of a lower-emotional involvement situation, such as the one in chat. where you don't either see or hear the person you are communicating with. Obviously, a very important element is also the level qf competence (typing and writing) and chat experience. A person familiar with emoticons can find them as an opportunity. a useful device to convey emotions; otherwise they could just be meaningless signs. 10.4.2.2 Exploiting affordances and managing constraints Furthermore, the next step is to see whether the actors are able to exploit these affordances. depending on a complex inter-relationship of factors. Once perceived and identified opportunities and constraints. they decide how to exploit them and how to structure the interaction. We could assume that the subjects are not always able to communicate most effectively in different communication environments, either because of a general lack of knowledge and technical competence or because of contextual elements that make the situation uncomfortable to them (their perception/interpretation of the situation), or also because they can't tune to the other person. We can hypothesize that the different interpretation of the features of varied communication environments can influence and modify (broadenhestraidmodulate) the repertoire of communication strategies that one person has. It is possible that this perception of opportunities and limits can bring people to communicate in different ways. Changing the communication environment and thus changing the context can be a powerful experience of discovering of new communication opportunities that are usually made difficult by the emotional involvement or other factors, such as personality characteristics. For example, being ironic to seduce (in face-to-face conversation) is an incredibly complex act: irony in itself is a multi-level, meta-intentional act [57]. In the case of "being ironic to seduce". this act is even more complex and demanding: the ironic intention by itself becomes a sub-ordinated intention as compared to the higher-level seductive intention. This requires a very high level of attention to manage facial expression, gestures and voice in the interaction, attentional resources that are not always present in face-to-face seductive interaction, where usually the emotional involvement is high. In CMC, being ironic could mean carefblly underlining certain words with punctuation. an attentive use of capital letters or making use of emoticons, in much more controlled situation. The same person who. for many reasons, can be embarrassed in face-to-face interaction can experience in CMC a reduction in the relational stress and find a way to be ironic. The same person who cannot make an effective use of her voice and facial expressions, but is very effective with writing can become perfectly patron of the situation and of the managing of the codes in chat, and be so a very ironic and brilliant person. He/she can thus use this powerful weapon to seduce, a weapon she could never use in faceto-face where she would possibly prefer a different style. As pointed out by Anolli in the first chapter of this volume, it is very important to consider the range of signalling systems. and the plurality of expressive means available to the communicators in order to express their communicative intention. The focus is "on the semantic synchronv process, intended as the speaker's competence to organise. coordinate, and make the different communicative systems converge on each other in order to make explicit the meaning of one's own communicative intention in a unitary and coherent 14 ay.

F. Mantovani / Cyber-Attraction

24 I

It would thus emerge, as Mantovani [58] suggests, that seductive communication in cyberspace usually requires two tasks: the analysis of the characteristics of the communicative environment in which the play of interpersonal attraction develops, the exploitation of the affordances offered by communicative environment according to opportunities and constraints. Furthermore, the communication style used by the subject, in order to be effective, must be tuned and flexibly adapted according both to the partner’s personality and competence. The key effort of the subjects involved in a computer-mediated interaction is always the negotiation of the meaning of the situation they are involved in. Within a communicative interaction, people are constantly involved into a process of negotiation: negotiation of the meaning and negotiation of the intention Situations are always conjointly defined by both partners, and the meaning for the interaction always emerges as a result of the process of negotiation [ 18,591 in order to reach a shared meaning This is coherent with the dialogic principle of human communication. As Anolli and Ciceri [20] have pointed out, communication is not simply an act but it is an inter-act, where both partners constantly co-determine the construction of the communicative reality and operate on it. Miscommunication as a founding feature of indeterminacy in human communication is the key element: it offers the basis for both misunderstanding and enrichment of the human communication at the same time. In the very moment that communication leaves space for indeterminacy, on the one hand this implies the risk of misunderstanding; on the other hand it allows a creative interpretation and construction of the interaction, through the process of negotiation. This allows the partners to constantly re-define the context they’re people are placed in. The relationship between context and communication is a bi-directional one: on the one hand, the context defines and influence certain elements of the interaction, on the other hand, communicators during their interaction constantly re-define and re-build the context.

10.5 Conclusions

As we have seen in this chapter, cyber-attraction is a very complex issue, in which a rich network of elements is at stake. In order to analyse this process effectively, we assumed that it was not possible to analyse it separately from “offline attraction”: there should be a unitary framework to investigate this phenomenon online and off. Along this path, we thus tried to investigate this phenomenon within a general and comprehensive perspective, in order to forward a unitary vision for on-line and off-line communication and behavior. This meant considering the existing differences but investigating them in a global and coherent framework, where insights and theories developed or concerning one level can enhance reflection and understanding of the other levels. CMC is becoming more and more part of our life: thus “on-line” and “off-line” dimensions will mergehtegrate more and more in our everyday experience and use; we thus need to look for comprehensive models to investigate and understand them. In order to understand the specific features of online seduction and especially to explain the paradoxical co-existence of charm and risks, it was necessary to look up at a concept/perspective that could unify the analysis of seductive communication in different communication environments. From the analysis carried out, it emerged that miscommunication (understood as the characteristic of indeterminacy and rarefaction of communication) can represent a powerful perspective for the comprehension of this process.

On the one hand, it explains how in cyberspace it is possible to overcome many of the constraints of face-to-face communication: first of all, the importance that physical appearance has in the choice/selection of the partner and in being noticed (which is a disadvantage for the non-beautifid); secondly, the time which is usually needed to reach intimacy. Because of the reduced cues, it is possible to be attracted byiattract more many people than in face-to-face, independently from physical appearance and it is possible to dare more with less fear of “losing face”. At the same time, it is possible, as literature on online support groups [60, 613 shows very clearly, to reach very high levels of empathy and intimacy in shorter time as compared to face-to-face. On the other hand, the concept of miscommunication accounts for the existence of specific risks/limits in online attraction: the first one is represented by the risk of making illusions grow to far; although the cues which are given are only few. and anyway less that in face-to-face, this leads to make great use of fantasy and idealization to draw a complete picture of the virtual partner [34]. Theorists generally agree that “romantic love is inexorably tied up with fantasy” [ 15, p.153.J. As Waller and Hill [62] put it “Zdealizafionis an essential element in romantic love” (p.120). How better to idealize than through CMC. where one is left to paint his or her own mental picture of someone? Obviously, this can be risky when deciding to be faced to reality. This remark explains also some results from literature [6], reporting that many people do prefer to “stay online” and never “go back to earth”. Meeting offline means reducing the space for idealization and illusion. and can sometimes be disappointing, when expectancies don’t meet reality. This leads us to a second limit of cyber-seduction. which consists of the impossibility. online, to compare what the partner says with what we see or with information got from other people/elsewhere: the only source is what the partner says. There is thus much room * self-presentation; this brings to for deception. or at least distortion and sa! not to s u ~ in specific problems for the development of trust. It would seem that. coherently with the vision forwarded in this book. miscommunication, in its dark and bright side. is part of our everyday communication experience, online or off. It is simply a matter of managing opportunities and risks of this intrinsic feature of human communication. In trylng to understand the interpretation and perception of the affordances and limits for such a particular communication as CMC, we also looked up at Situated Action Theory. According to this reflection, these opportunities and limits are not strictly and deterministically given, but are the result of a process of perception and interpretation by the people engaged in the interaction (who must possess new and specific skills). They finally are also the result of the process of negotiation of a shared meaning for the context and the interaction, and of reciprocal, continuous interpretation of the partners’ intentions. Also. it emerges how there is an interdependent. circular relationship between interaction and context (and thus the interpretation of the situation: there is always space for a redefinition of the situation and of the framework of opportunities and limits of the media). For example. this should adapt to different people and different steps in the relationship. Probably, the factors of competence and previous experience are likely to impact heavily on our CMC interactions and relationships. Since the use of CMC (to meet people. discuss topics of interest, etc.) will become more and more familiar in the future. a medium among the others, it is likely that this will allow people to balance more objectively its opportunities and limits. Through growing use and experience, they also could acquire new specific skills and become able to perceive and manage more effectively the communication resources available within that communication environment. Laurel N. Hellerstein [63 J found that

F. Mantovani / Cyber-Attraction

243

heavy users of e-mail and electronic conferencing, in a university setting, were more likely to use the computer to “initiate new friendships, make new friends, and communicate with others,” whereas light users tended to do build relationships in other ways. She also saw that many of these relationships formed and started online then grew off online (friends or romance). This seems to be an interesting finding indeed, because it could support the idea (still to be tested) that, apart from pathological use, increased use in CMC leads to acquisition of specific skills in the management of online relationships and in their integration into their everyday life. Although current research focused on the use of CMC to meet new people (and possible development offline), a very interesting field of study is represented by the uses of CMC in different contexts than getting in touch with “strangers”. In fact, e-mail, chats, etc., are more and more used also to support the development of relationships born offline (occasional acquaintances), or to stay in touch with people with whom there already exists an intimate relationship. As we can see from these remarks, there are no strict and fixed boundaries between online and offline communication. Future research should be aimed at trying to understand and detect how interaction develops according to the repertoires and possibilities allowed by different media and communication environments, as well as at identifylng critical dimensions of the interactants which can account for a different perception and exploitation of these affordances. Also, we should define measures of communicative effectiveness in relationship with the achievement of relational goals in online relationships. Further studies should be carried out, that investigate these processes in different environments, from face-to-face to different forms of CMC, in order to explore more into depth the role and the complex interaction of the variables pointed out in this chapter. This approach would allow for the continuity of study of communication in different interfaces, in order on the one hand to see how insights and categories from traditional literature can be applied to online behavior, and on the other hand to use the findings from studies of cybercommunication to refine our theoretical models.

10.6 Acknowledgements

I would like to thank Fondazione “Piero, Pietro e Giovanni Ferrero” (Alba, Italy) for the support given to the present study. It was also supported by the Commission of the European Communities (CEC), in particular by the IST programme (Project VEPSY UPDATED, IST-2000-25323 http://www.psicologia.net; http://www.vepsy.com ). I thank Professor Luigi Anolli for his contribution and support, as well as for his useful comments on earlier versions of this chapter. I also thank Giuseppe Riva for discussions on issues related to this chapter and Valentino Zurloni for his help in formatting and editing it.

10.7 References [ 11

[2]

[3]

[4]

H. Rheingold, The virtual community: Homesteading on the electronic .frontier. Reading, Mass: Addison-Wesley, 1993. G. M. Olson and J. S. Olson, Research on computer-supported cooperative work, in Handbook of Human-Computer Interaction, M. Helander, T. K. Landauer, and P. Prabhu, Eds., 2nd ed. Amsterdam: Elsevier, 1997. L. Sproull and S. Kiesler, Connections: New Ways of Working in the Networked Organization, vol. 32. Cambridge, MA: MIT Press, 1991. S. R. Hiltz, Online communities: A case study of the Qflfficeof the Future. Norwood, NJ: Ablex Publishing Corp., 1985.

244

F. Mantovani / Cvher-Attraction

M. Strangelove, The Internet, electric gaia and the rise of the uncensored self, Computer-Mediared Communication Mugurine 1 ( 1994) online: http://sunsite.unc.edu/cmc/mag/1994/sep/self.html. A. Baker, Two by two in cyberspace: Getting together and connecing online. Cvherpsychology. & Behavior 3 (2000) 237-242. [71 M. R. Parks and K. Floyd. Making friends in cyberspace., Journal of’ Compurer-Mediated Communication 1 ( 1996) online: http://www.ascusc.org/jcmc/voI 1/issue4/voll no4.html. B. G. Chenault, Developing personal and emotional relationships via computer-mediated communication, CMC Magazine 5 ( 1998) on-line: http://www.december.com/cmc/mag/ 1998/may/chenault.htmI. & Behavior 3 (2000) 565[91 D. Levine, Virtual Attraction: What Rocks Your Boat. C\*herps?rholo~~ 573. S. S. Brehm, Intimate relationships. New york: McGraw-Hill Press. 1992. M. J. Culnan and M. L. Markus. Information technologies. in Handbook qf’ organizational communication: A n interdisciplinary perspective. F. M. Jablin. L. L. Pumam. K. H. Roberts. and L. W. Porter, Eds. Newbury Park. CA: Sage, 1987. R. L. Daft and R. H. Lengel, Organizational information requirements. media richness and structural design, Management Science 32 ( 1986) 554-57 1 . R. E. Rice, Media appropriateness. Using social presence theory to compare traditional and new organizational media. Human Communicarion Research 19 ( 1993) 45 1484. 1141 K. Grammer, The Courtship Dance: Patterns of Nonverbal Synchronization in Opposite-sex Encounters, Journal of Nonverbal Behavior 14 (1999) 253-267. E. Berscheid and E. Hatfield, Interpersonal attraction. 2nd ed. New York: Random House. 1978. K. Dindia. The Intrapersonal-InterpersonalDialectical Process of Self-Disclosure. in Lhwamic. of Relationships, S. Duck, Ed. Thousand Oaks, CA: Sage. 1994. S. Duck. Steady as (s)he goes: Relational maintenance as a shares meaning system. in Communication and relational maintenance, D. J. Canary and L. Stafford, Eds. New York: Academic Press. 1994. G. Mantovani, Yew* communication environments: from e\.er?da?s to \ir[uul. London: Taylor & Francis, 1996. ~ 9 1L. Suchman, Plans and situated action. Cambridge: Cambridge University Press. 1987. [20] L. Anolli and R. Ciceri, I1 gioco delle intenzioni nell’atto molare. in Elemertti di psicologia della comunicazione. L. Anolli and R. Ciceri, Eds. Milan: LED. 1995. J. L. Austin, How to do things with word.s.Oxford: Clarendon Press, 1962. H. P. Grice. Meaning, in Semantics: A n interdi.wiplinar3. wuder. D. D. Steinberg and L. A . Jakobovits, Eds. Oxford: Oxford University Press. 197 1 . pp. 53-59, D. Sperber and D. Wilson. Mutual knowledge and relevance in theories of comprehension. in Miitira/ knowledge. N. Smith. Ed. London: Academic Press. 1982. pp. 6 1-87. L. Garton. C . Haythornthwaite. and B. Wellman. Studying On-Line Social Networks. in Doing Internet Research: Critical I.ssiies and .Wethod.y for E.ramining the .Vet. S . Jones. Ed. Thousand Oaks. CA: Sage, 1999. J. Preece, Online Communities. Designing Lhnhiliti.. Supportitig Sociahilit\*. Chirchester: John IViIey & Sons. Ltd., 2000. J. Short, E. Williams. and B. Christie. The social ps?.cholofl. o/‘ telecommioticatioiis. London: John Wiley & Sons. 1976. R. E. Rice, Computer-mediated communication and organizational innovations. ./ournu/ of Communica~ion37 ( 1987) 85- 108. J. B. Walther, Impression development in computer-mediated interaction. ISe.rtern ./ournu/ of Commimieations 57 i1993) 38 1-398. S. R. Hiltz, K. Johnson, and M. Turoff. Experiments in group decision making: Communication processes and otucome in face-to-face versus computerized conferencing. Hriman C~~nrmunictrriott Research 13 (1986) 2 2 5 - 2 5 2 . L. Sproull and S. Kiesler, Reducing social context cues: Electronic mail in organizational communication, Management Science 32 i1986) 1492-1 5 12. J. B. Walther, Anticipated ongoing interaction versus channel effects on relational communication in computer-mediated interaction. Human Communication Research 20 ( 1994) 473-501 . C . Goodwin and J. Heritage. Conversation analvsis. .4nniral Rei.ie\c. o/’..liit/irr.,pol~~~~, 19 ( 1990) 283307. S. E. Brennan. Conversation with and through computers, C ‘ser .Wodelling crnd C .rer-.Jdapted Interaction 1 ( 199 1 ) 67-86, M. Lea and R. Spears, Love at first byte? Building personal relationships oser computer neworks. in I -nthv.ytirrficw’ rt~hriorvhips.Off tho hcarrvr t m c k. 1. T \I’ood and S Duck. Eds. Xcwbur! Park. (‘:I:

F. Mantovani / Cyber-Attraction

245

Sage, 1995, pp. 197-233. J. Ord, Who's joking? The information system at play, Interacting with computers 1 (1989) 118-128. N. B. McCormick and J. W. McCormick, Computer friends and foes: Content of undergraduates' electronic mail, Computers and Human Behavior 8 (1992) 379-405. R. E. Rice and G. Love, Electronic emotion: socioemotional content in a computer-mediated communication network, Communication Research 14 (1987) 85-108. Cybershrink, Psychology Today 28 ( 1995) 20-2 1. D. H. Krikorian, J.-S. Lee, and T. Makana Chok, Isn't that spatial?: distance and communication in a 2-D virtual environment, Journal of Computer-Mediated Communication 5 (2000) online: http://209.130.1.169/jcmc/vol5/issue4/krikorian.html. D. Morris, I1 comportamento intimo. Milano: Arnold0 Mondadori Editore, 1986. D. B. Givens, The Nonverbal Basis of Attraction: Flirtation, Courtship, and Seduction, Psychiatry 41 (1978) 346-359. D. T. Kendrick and M. R. Trost, A Reproductive Exchange Model of Heterosexual Relationships: Putting Proximate Economics in Ultimate Perspective, in Close Relatioships, C. Hendrick, Ed. Newbury Park: Sage Publications Ltd., 1989. E. Battisti, J. Baudrillard, A. Finkielkraut, M. Perniola, and G. Pettena, Ipotesi di Seduzione. Bologna: Cappelli Editore, 1981. J. Baudrillard, in I1 Sogno e la Merce: antologia di scritti sulla pubblicita, V. Codeluppi, Ed. Milano: Lupetti & Co., 1987. J. Baudrillard, De la seduction. Paris: Editions Galilee, 1979. P. Wallace, Thepsychology ofthe Internet. Cambridge: Cambridge University Press, 1999. I. Altman and D. A. Taylor, Social penetration: The development of interpersonal relationships. New YorK: Holt, Rinehart and Winston, 1973. D. Schnarch, Sex, intimacy and the Internet, Journal of Sex Education and Therapy 22 (1997) 15-20. M. B. Wanzer and S. Booth-Butterfield, Are funny people more attractive? An examination of humour orientation, loneliness and social attraction, Communication Quarterly 44 ( 1996) 42-52. N. K. Baym, The performance of humour in computer-mediated communication, Journal of Computer-Mediated Communication 1 ( 1995) online: http ://jcmc.mscc.huji.ac.il/vol 1/issue2/baym.htm 1. S. M. Jourard and P. Lasakov, Some factors in self-disclosure, Journal of Abnormal and Social Psychology 56 (1958) 91-98. A. Mehrabian, Silent messages: Implicit communication of'emotion and attitudes, 2nd ed. Belmont, CA: Wadsworth, 1981. E. Goffman, The presentation of self in everyday life. New York: Doubleday, 1995. C. E. Shannon and W. Weaver, The mathematical theory of communication. Urbana: Illinois University Press, 1949. N. Baym, The emergence of community in computer-mediated communication, in CyberSociety: Computer-Mediated Communication and Community, S. G . Jones, Ed. Thousand Oaks, CA: Sage, 1995, pp. 138-163. G. Mantovani, L'interazione uomo-computer. Bologna: I1 Mulino, 1995. L. Anolli, R. Ciceri, and M. G. Infantino, Behind the 'dark glasses'. Irony as a strategy for implicit communication, Genetic, Social and General Psychology Monographs (in press). F. Mantovani, Networked Seduction: A Test-Bed for the Study of Strategic Communication on the Internet, Cyberpsychology &Behavior 4 (1) (2001) 147-154. G. Riva and C. Galimberti, Computer-mediated communication: identity and social interaction in an electronic environment, Genetic, Social and General Psychology Monographs 124 ( 1998) 434-464. J. Preece, Empathic Communities: Reaching out across the Web, Interactions Magazine 2 (1998) 3243. J. Preece, Empathic Communities: Balancing emotional and factual communication, Interacting with computers 12 (1999) 63-77. W. Waller and R. Hill, The jumily: A dvnamic interpretution, 2nd ed. New York: Holt, Rinehart and Winston, 1951. L. N. Hellerstein, The social use of electronic communication at a major university, Computer and the SocialSciences l(1985) 191-197.

This page intentionally left blank

247

Say Not to Say L. Anolli et al. (Eds.) IOS Press, 2002

11

Ambiguous Self-Identification and Sincere Communication in CMC Edward A. MABRY

Abstract. While sincerity is a necessary taken-for-granted element of face-toface communication, its status in CMC is more likely to be a ponderable (or arguable). In fact, the anonymity of CMC, facilitating and facilitated by concealment in self-presentation, allows distortions to be made through omission of information as well as through selective presentation. To understand the characteristics of this process, we analyzed the message contents of men and women participating in asynchronous online discussion. Recent research indicated men and women differed in their fidelity of message concealment. Level of concealment did not significantly predict expected relationships between identity and attributes of online message style. Findings regarding concealment and message style led to a consideration of communicator status and gender identity needs. A new message style variable, openness, was also examined. Research findings indicated that gender identity and communicator status were strongly related to message openness.

Contents 11.1

11.5 11.6

Self-presentation, gender identity and concealment in CMC ................................................................. Gender identity and concealment: new research. .............. Gender, status, and concealment in CMC.. ..................... Gender identity and status implications for communicator openness ............................................................. Conclusions.. ....................................................... Acknowledgments. .................................................

11.7

References.. .........................................................

11.2 11.3 11.4

248 252 259 262 262 263 263

11.1

Self-presentation, gender identity and concealment in CMC

Sincerity is an inextricable part of social interaction. According to GotTman [ I , 23. sincerity frames the assumption of trust underlying self-presentation. The assumption that we are as we appear provides a rationale for not questioning the motives of another person's social behavior. Only when someone's behavior appears to violate social norms or is erratic do we question its sincerity. There are times when individuals feel compelled to be less than forthcoming in their self-presentations. These instances comprise situationally induced episodes of miscommunication. Breaches of sincerity are less problematic in faceto-face (F2F) communication. Interactants have the advantage of observing a wide range of kinesic and vocalic cues in order to assess the congruency and consistency of communication. Sincerity can be seen as having two potentially contradictory ends: sincerity of selfexpression and sincerity of self-presentation. When a person strictly attends to their own identity imperatives, they ostensibly function honestly in their constructive representation of self-attention. The extent to which a person can be sincere depends up the extent to which the person is capable of self-expression without constraint. A person is far freer to perform impulsive acts when there are no consequences linked to the performance of such acts. Such freedom to act is reduced by the revelation of identity. For example, once the identity of a philanthropist is revealed the person may receive unwanted requests for other charitable donations. On the other hand, when individuals present themselves to others, they increase the likelihood that their actions will have consequences in the form of shaping others' opinions of them. Furthermore, the more honest a person is in their self-presentation, the more they are likely to believe the consequences of an action are not merely a possibility but a prohahilip. Thus, revealing one's identity compels the individual to frame their actions in such a way as to minimize the likelihood of creating negatively valenced perceptions of that identity (or its prior concealment). If, for instance, the hypothetical philanthropist referred to above does not conceal their identity, but makes contributions to, say, only one. or one type, of charity, they risk being publicly criticized for having a "pet cause" and not a truly charitable nature. 11. f .I Inferring sincerih*in F2F cwnmunication

In F2F communication, sincerity (in general) is inferred from observations of possessions and social interactions. According to Belk [3]. we are the sum of our possessions; our identity is inferred from the presentation of our possessions. We make determinations about a person's sincerity through information derived from our observations of the match between revealed identity and behavior [I]. We question poorly concealed or explained inconsistencies between identities and behaviors. A person's sincerity is inferred from two dimensions of self-presentation: congruency and consistency. Congruency refers to the parallelism between how a person verbally symbolizes a relationship and how they actually conduct the relationship [ 11. Sincerity also depends on consistency, or the extent to which a person's self-presentation is consistent with their representative behavior. Sincerity can thus be seen as honesty in selfpresentation and relational intent that endures relational longevity and the accumulated knowledge of others. Honesty in self-presentation is most easily inferred from assessing the match between what people say about themselves and what can be visually observed about them. In a FZF meeting. L4.e can determine much about a person from a myriad of unobtrusive measures

E.A. Mabry /Ambiguous SelJlldentification

249

like artifacts, gestures, and mannerisms. These observations give us personal information that makes inconsistencies difficult to conceal or explain [3]. Information gleaned from visible behavior assists us in making perceptual attributions that either confirm or disconfirm the apparent truth-value of a person's self-presentation. Summarizing to this point, sincerity in F2F interaction is inferred from consistent honesty and openness about one's self. Whereas, duplicity uses sincerity as a front for egodefensive or nefarious ends advanced by concealment or self-conscious interaction management. Yet, how is real sincerity discerned from contrived sincerity? This question is difficult enough to ask in terms of F2F communication. With the growing popularity of computer-mediated communication (CMC), the question becomes even more important to ask and most difficult to answer. 11.1.2 Sincerity and anonymity in CMC

Although the term CMC refers to a variety of communication channels that are electronically mediated, this assessment focuses on text-based asynchronous electronic mail [email]. The fact that email typically is uncensored and unregulated makes it an ideal channel for managing self-presentation. On the other hand, email also fosters intimate and supportive interpersonal relationships (Bechar-Israeli [4]) in which the disclosure of identity is purely optional. While sincerity is a necessary taken-for-granted element of F2F communication, its status in CMC is more likely to be a ponderable (or arguable). Many computer-mediated contexts are predicated on anonymity. Spears and Lea [5, 61 propose that the identifiability of one's self-presentations directly influences the ways in which the individual engages in online communication. As discussed in Chapter 9, their social identity and deindividuation (SIDE) model argues that anonymity facilitates deindividuation . Postmes and Spears [7] explain the mechanism for deindividuation as the capacity to act with relatively low risk of identity threat. Reduced threats to one's identity also lowers an individual's need to engage in self-monitoring. A decline in self-monitoring increases the likelihood of disinhibition as a group member whenever disinhibited behavior is supported by group norms. Feeling like one is part of the group frees an individual from mindful identity management in deference to one's group identity and the group's collective identity. Postmes, Spears, and Lea [8, 91 have shown the amount of identity protection afforded by communication channels dictates whether an individual's communication is more strongly affected by identity needs or group norms. However, the SIDE model assumes that online communicators are completely in control of personally revealing online information or fully skilled at self-concealment. Not even F2F interactants are capable of maintaining their social masks all of the time. This leads to the question that grounds the present research: What tools exist for assessing sincere selfpresentations in computer-mediated communication? Vicissitudes of CMC technology can sometimes accommodate interactants' desires for anonymity by allowing users to override email system default settings that includes personalizing information in email message headers. Yet, users of this concealment feature put their anonymity at risk if they choose to be too open in their self-presentations. Anonymity comes with a cost. One's self-presentation is constrained by the self being presented [6]. Behavior inconsistent with the presented self is suspicious. Honesty in relational intent is also inferred through self-presentation. Walther's [ lo] summary of the relevant research notes that an actor's anticipation of subsequent interaction increases the amount of personal information they exchange. Anticipated future social contact causes actors to present themselves in ways that produce positive

250

E.A. Mahn. i ..imhiguous Self-Identlficntion

attributions and heightened desires for future affiliation. Regardless of the observability of a person’s forthrightness, some people are duplicitous and intentionally misrepresent themselves. Intentionally successful miscommunication. duplicity, depends on a person being able to make sacrifices for the duplicitous end, keep secrets, and attempt to present a flawless finished product [I]. Success requires that the machinations of duplicity be kept concealed. Presumably, a person that is not honest in his or her self-presentation will not be relaxed or comfortable and will appear nervous from engaging in highly absorbing self-monitoring (Walther and Burgoon [ 1 11). Yet, Goffinan [ 11 warns that the closer an impostor’s performance approximates the real thing, the more intensely others may be threatened. A convincing impostor weakens moral connections between the presumed legitimacy of authorization to play a part and the capacity to play it. Computer-mediated communication is not the only channel in which people use the ability to remain anonymous as a way to conceal themselves. But, standard features in CMC make concealment easier than in other channels. Nonverbal and relational cues are substantially absent in CMC. Cues of social presence, the senses of inclusion that actors are jointly involved in communicative interaction (such as facial expression, direction of looking, posture, dress and non-verbal, vocal cues) are often missing or difficult to emulate in CMC. Social context framing cues in communication settings, and the social status of communication partners, are likewise often difficult to establish (Mabry [ 121). Culnan and Markus [I31 have referred to this removal of information as the ‘*cues filtered out” perspective. As presented in Chapter 9, the Cues Filtered Out perspective provides insight into how interactants distort messages they receive as well as messages they send. These distortions typically involve attributions about the information value of meanings that can or cannot be conveyed on particular channels. Lea and Spears [6] note that participants with too few cues invest the cues filtered-in that are available for interpretation with inflated value leading them to have distorted impressions of their online partners. Similarly, Walther [ 141 asserts that reduced communication cues in asynchronous communication facilitates selective self-presentation. Therefore, miscommunication involving what ultimately results in insincere selfpresentations may not always be mindful because it emanates from misperceptions associated with message cue filtering. I 1.I .3 Anonj~rnih~ and concealment in CMC

The three themes emerging fiom the cues filtered out perspective. anonymity, concealment and distortion of self and other, call into question how honesty in self-presentation can be trusted in CMC. The anonymity of CMC. facilitating and facilitated by concealment in self-presentation, allows distortions to be made through omission of information as well as through selective presentation. One of the main questions surrounding the selectivity of presentation is who distorts, in what way, to what extent, and for what outcome? In order to be truly concealed. one must be invisible to others. Invisibility can only be achieved if one does not participate. As soon as a person interacts, his or her presence is publicly accessible to others even when their identity is shielded through anonymity (O’Sullivan [ 151, Scott [ 161). Thus. interactants cannot be either revealed or concealed because the role they assume is more complex and not fully under their control [ 161. Instead, people reveal themselves to a greater or lessor extent through their participation. Sincerity, genuineness in self-presentation. or concealment and distortion are discernible through the analysis of their message hehatvior (Buller, et. al. [ 171, Reicher and LeLine [ 181). Mam [19] i k w s anonymity. and thereby revelation and concealment. as

E.A. Mabry /Ambiguous SeEf-Identification

25 1

scaleable. A nominal range appears to span from how implicitly or explicitly an individual’s styles of communication function to reveal or conceal their identity in the context of how their self-presentation is framed [ 161. However, there does not appear to be a uniform approach for advancing a precise metric. In fact, Buller, et. al. [17], studying F2F interaction, failed to find a general pattern of deceptive interaction. Instead, they found that the specific type of deception enacted (i.e., falsification, concealment, equivocation) produced relatively different patterns of deceptive interaction. Of central concern to CMC users is how to distinguish between people strategically concealing their identities and those revealing their true identities [ 11. Effective concealment depends on observers not being able to tell the difference. Yet, participants forming intimate relationships with others apparently feel adequately informed about them using no other means than textual messages. Interactants in CMC do, however, have one clue that may give them insights into a person’s self-presentations and relational intent: level of formality. Goffman [2] asserts that social rituals, like greeting rituals, become abbreviated, and informal, when probability and frequency of social contact are high. A person’s self-image can be served by the formality of their self-presentation. An informal presentation appears learned and credible, and focused on the task, rather than the social, aspects of the interaction. Formality and informality seem to parallel the gender descriptors of “agency” and “communality,” thus suggesting that formality may be gendered [21]. The idea that clues allow inferences to be made about a person in terms of selfpresentation and relational intent forms the basis for another theoretical perspective regarding the determination of honesty in self-presentation. Walther [20] proposed that CMC often supplemented the communication channels of people who see each other regularly. CMC merely provides an efficient means of sending messages without the delays of unanswered or returned phone calls or visiting empty offices. Thus, cues filtered out research effects have been limited to initial interactions among strangers. Walther [14] offered an alternative theory stating that relationships in CMC develop despite the lack of social cues, because time and rate of message exchange attenuate the effects of missing information. The alternative he proposed is Social Information Processing theory. In the Social Information Processing perspective, lack of social information changes the pacing of relational development by forcing interactants to adapt their remaining communicative cues, language and textual display, to the processes of relational management. As a result, relational communication changes from impersonal to more personal levels as relational developed forms in CMC. According to Spears and Lea [51, relational development follows a pattern of relational motivation. Exchanging messages leads to impressions derived from decoding text from which psychological knowledge is inferred. In turn, this leads to the management of relational changes and the encoding of relational messages. Social Information Processing theory assumes that interactants rely on disclosed information to determine honesty. Once information is obtained, then, according to Walther [20] strategic probes are used to test the relationship. Over time, however, with many exchanges, relational development should converge with F2F relationship development. The assumption is that, over time, inconsistencies of text and selfpresentation will become evident through self-disclosure and relational knowledge. The cues filtered out perspective questions the viability of inferring believability from a message that is purely text-based. Social information processing claims that, over time and self-disclosure, a high level of relational knowledge can occur. These conflicting views provide a strong argument for investigating the effect of computer-mediated communication on honesty in self-presentation and the communication of relational intent.

252

E.A. M a h n /Ambiguous Self-identification

I 1.I . 4 Assessing gender, anon-vmi?,.and concealment in CMC messages Neither cues filtered out nor social information processing perspectives address the production of congruity or consistency despite the centrality of these ideas to the notion of honesty. For the purpose of this project, congruity is the match between the level of relational involvement, topical involvement, or level of formality observed in a message. Consistency is the extent to which self-presentation is consistent with gender-revealing identity information. A primary assumption advanced in this project is that attempted concealment, specifically as it relates to gender identity, can be successfully extrapolated from an actor's communicative performances in online contexts. The constituents of sincere messages are the actor's congruity and consistency of language use serving to disambiguate gender. This can take place either through self-referential identification or message constructions indicative of communicator style attributes like relational involvement, topical involvement, and formality of expression. Spears and Lea [6] suggest that gender may be one subject on which people engaged in CMC are more likely to be evasive in their self-presentations. As an ego-defensive maneuver, how this evasiveness is manifested may be gendered with women being more likely to conceal their gender identity and men more likely to "gender bend" or present themselves as women. Based on this assumption. women or men who reveal too much risk exposing themselves. A person is free to explore the possibilities of the CMC environment only to the extent that he or she is able to j7awlessfv enact the persona he or she has adopted. A flawless performance disallows openness [ 11. Women, according to Spears and Lea [ 5 , 61, conceal their gender to prevent harassment in a largely male-dominated social environment and men experiment with alternative sexual identities for personal gratification. They contend that women are more selfprotective in situations like online discussions where they perceive greater risks fiom being attacked as gender role-incumbents unless group norms mitigate those risks. This reticence leads women to be less forthcoming in their self-attribution of gender and apply greater effort at managing their self-presentations. Thus, it is reasonable to expect that: men and Ii*omen in online discussion groups will difer significantlv in their manner of selfidentzfiing gender. Consistent with the above reasoning, we also would expect that the style of messages men and women contribute in online interaction will be related to the apparent clarity they wish to portray in their appropriation of gender. If the foregoing assumption is correct: levels of relational involvement, topical involvement. and-formalityshould dlffer according to whether and/or howl men and women provide identih?cues in their online messages.

11.2

Gender identity and concealment: new research

The assumptions regarding gender and identity management discussed above were tested in a study analyzing attributes of messages sent to online groups. Data in this study was obtained as part of an international, computer-mediated collaborative research project. The principle goal of the project was to facilitate the investigation of CMC by examining messages contributed to computer-mediated bulletin boards and discussion groups accessible on the Internet. Over a period of months, research objectives and methodological practices were chosen, then a trained group of researchers analyzed samples of messages obtained from electronic discussion groups. A complete explanation of the project is contained in Rafaeli. Sudweeks. Konstan. and Mabr?,1221.

E.A. Mabry /Ambiguous Self-Identification

253

I I .2.I Research data A complete discussion of sampling is available in Rafaeli, et. al. [22] and will be abbreviated here. Various computer bulletin boards, lists, and news-groups were canvassed for a period of approximately one-month. Messages were randomly sampled across days and times. Only lists in English were retained for the message pool. An N = 3000 messages, from 30 different online discussion groups, comprises the database. A standardized, message content analysis coding protocol was collaboratively developed by a subgroup of the research team [22]. The content analysis measured 46 message variables; 40 variables were hand-coded and six were machine coded. The coding protocol required trained coders to read the literal text of a message and apply all applicable code to each message. A message was evaluated on whether it contained content descriptive of facts, opinions, humor, challenges, meta-communications, presence of graphic art, formality of composition, quoted material, emotional tone (or flames), sender characteristics (e.g., gender, status), and stylistic various factors (e.g., appropriate subject line attribution, presence of personalized signature lines). A subset of the project's message variables were used in this study are listed in Table 11.1. Over 40 researchers participated as coders. Coders were provided online training, proficiency tests, and expert guidance. Training involved coding a set of sample messages chosen to cover the range of code variables. Coders rated the messages and returned their results via electronic mail. Low concordance on the training sample of messages led to additional coaching and more training messages being sent for test coding. High agreement with preferred responses to training messages qualified a person as coder for the purpose of receiving messages to be analyzed in the main study. Coders were electronically sent sets of 100 messages. In addition to the set of messages, coders were supplied with various reporting style formats for submitting their work, and a post-task questionnaire requesting impressions of the list coded and information about coding and reporting practices. Completed sets of formatted codes were returned via electronic file transfer to a host computer system. Work was automatically screened using custom software to debug technical errors (e.g., off line formats, typographical errors); rejected codes were returned to the coder for correction. Given the methodological approach taken in the project, coding reliability proved to be a complex task. Two conventional methods for assessing reliability were used: Brennan and Prediger's [23] modification to Cohen's kappa, qn), coefficient, and Cronbach's [24] Alpha coefficient, &a). The kappa coefficient permits an assessment of inter-rater reliability under conditions where marginal values of an n x n coding matrix are free to vary. Kappa was computed for nominal and ordinal variables. Alpha was computed for scaleable variables.

Table 11.1. Study 1 : Variables

item

Range

Operational Description

Retiabilie

Measurement Variables

Relational Involvement Argumentativeness Promoting Cohesiveness

Scaled: Friendly to hostility Scaled: Strong AgreemenrsStrong Disagreement

R,, u

=

1-6 1-5

.58 67

u

=

75

Topical Involvement Opinion Referencing Messages Connecting Prior Message Lines Quoted

1-3 1-4 1-9 1-4

No: peripheral: central to message

(1

=

54

None; 1: 2-: sequential string No: yes None: 1 - 10: 1 1-2s;26-

K

=

60

K

-

76

K

=

xx

K

7

.x7

K

=.so

Formality Emoticons Emodevices Articons

R,,

53

R,,

1-3 1-3

1-3

.52

Smiley faces. winks. etc.: None: 1 : 1CAPS, !!!!. etc..: None; I : -7Keyboard graphics: None: 1 : -7-

h' = 97

1 = Female; 2 = Male

K

I = Not; 2 = name/signature; 3 = stated: 3 = indirectly implied: 5 = 2. 3. and/or 4

K

Research Design Variables

Gender Gender Identification

1-9 1-5

T

.'6

.66

A sample of 1,000 messages, 100 messages cross-coded from 10 discussion groups/lists, constituted the data for reliability analyses. Because coders were not fixed across lists, reliability calculations were performed on a list-wise basis (for each 100 crosscoded message sample) and averaged. Some difficulties were encountered as both list-wise and variable-specific results yielded substantial variability and, for some variables, moderate to low reliability values. List-wise coefficients could not be computed for some variables due to attenuated variances caused by high percentages (exceeding 95%) of intercoder agreement. Reliability results were within acceptable ranges. Table 1 1.1 contains reliability results for Study 1 message variables contributed from two separate studies that have used the project's database (Mabry [ 2 S ] . Karge, et. al. [26]) plus additional analyses conducted for this study.

I 1.2.2 Memiires of gender. conc*ealrnent.and C.WCv?wssagt~.s

Two types of analysis variables are used for Study 1 analyses: research design variables and measurement variables. These variable types are identified in Table 1 1 . 1 . Variables used in the research design were: gender of the message author. and the mode of gender identification of the message author (if present). Measurement variables were constructed from Relational Involvement. Topical Involvement. and Formality. Constructing summated measures did lead to a loss of data due to uneven patterns of unreconciliable or missing data. Gender. Both research design variables for this study are related to gender: apparent gender and the mode of its expression in a message. The latter is important because it involves how forthcoming communicators were in their disclosure of gender. This variable notes whether and/or by what methods online interacts re\.ealed their gender. The coding

E.A. Mabry /Ambiguous Self-Identification

255

protocol differentiated names and/or formatted signatures, direct references to the individual’s gender in the message text (e.g., “I am a woman.. ..”), indirect references to the individual’s gender in the message text (e.g., “Us men ought to stick together....”), the presence of two or more types of attributions (mixed modes of gender identification), or the absence of any identification of gender in the message body or headerdfooters. Relational Involvement. This variable was designed to measure the relational tone of a message expressed through the presence (or absence) of interpersonally sensitive language or language that promotes group solidarity: argumentativeness and messages promoting cohesiveness. The items were reflected and summed to calculate a profile variable measuring each message’s relational involvement. The summated variable produced a scale range with higher scores indicating the actor’s greater value for pro-social behaviors. Topical Involvement. Both adherence to, and amount of expression about, a message’s central them was used in constructing a measure of topical involvement. Message content variables used in constructing this index included: opinionation, presence and length of matter quoted from prior messages in the group, extent of prior message referencing, and references to continuities observed in preceding message strings. The relative presence of these variables indicates attempts not only to stay on topic but also increase credibility with other members of the group. These items were reflected and summed to form a profile measure of topical involvement. A high score corresponded to a high level of topical involvement. Formality. This variable measures the level of formality (or informality) present in the message and relied on the presence (or absence) of graphical expressions of emotion denoting departures from conventional message composition styles. Content variables used in constructing this item included: presence of emoticons (e.g., “smiley” faces), emodevices (irregular use of punctuation or capitalization for emphasis), and articons (artistic or illustrative use of symbols). Items were reflected and summed to compose a measure of the relative amount of formality or informality present in each message. This measure was directionally scaled so higher scores reflected less formality. 11.2.3 Research findings on gender and concealment

In order to facilitate a precise assessment of the assumptions this research addressed, data analysis was subdivided into tests of specific hypotheses regarding the potential effects of gender and identity management on message style. To confirm the independence of the summated message style items for separate analyses, Cronbach’s Alpha coefficient was calculated for the relational involvement, topical involvement and formality scales. The Alpha coefficient indicated the three scales were significantly independent (R, = -.O 1) of each other and could be analyzed separately. The first hypothesis contends that women would perceive online discussions as holding greater risks for them as interactants, compared to men, and lead them to engage in less disambiguated, sincere, self-presentations through their self-attribution of gender. Preliminary analyses of the N = 3,000 messages indicated that messages produced by men outnumbered women’s message output by a five to one ratio (Men = 2,162, Women = 427). Thus, men were clearly dominant on-line interactants in the discussion groups sampled. The contingency table comparison for men and women according to their manner of self-identification produced a significant Chi-square value ( ~ ~ ( =4 1) 19.77, p < .0001). Approximately the same percentage of messages from men and women contained no embedded gender attributions (Women [W] = 9.8% versus Men [MI = 9.4%). Women

756

E .4 .Mahn, /.4mhrguoiis Self-Identification

were less likely to explicitly provide their names compared to men (W = 74.5% versus M = 87.5%). And, women were four times more likely to enact indirect references to their gender (W = 6.8% versus M = 1.6%). Yet, women were also considerably more likely to use either direct gender references (W = 4.0% versus M = .05%) or produce messages containing more than one form of gender reference (W = 4.9% versus M = 1 .Ooh).These findings are consistent with basic thrust of the hypothesis. Consistent with the forgoing results, we would expect the styles of message enactments by men and women to vary according to the state of apparent clarity in their appropriation of gender. That is the underlyng argument advanced in three remaining hypotheses on the presence of relational involvement, topical involvement, and message formality. However, results for the test of hypothesis one clearly suggest that the statistical interaction of gender and mode of gender identification is the proper model for testing the remaining hypotheses. Therefore, the data for the hypotheses were analyzed in a 2 (Gender) x 5 (Mode of Identification) univariate analysis of variance [ANOVA] with relational involvement, topical involvement, and formality as dependent variables [27]. Hypothesis two proposes that the amount of relational involvement in the messages of men and women would be significantly related to the clarity of their gender appropriation. The results are inconclusive. The ANOVA interaction effect was not significant ( F (4. 1,203) = 1.62, p < .15). This indicates that any differences in the identity management of men and women in online groups did not affect the stylistic content of their messages. Yet, it is possible that the styles of online messages varied between men and women and mode of identity management without interacting. The main effect for gender also was not significant (F ( 1. 1,203) = 2.57, p > .lo). Means for this analysis are contained in Table 11.2. However, the main effect of mode of identification was significant (F (4, 1,203) = 5.65. p < -0001. o’= .15). Table 1 1.3 contains the means for this analysis. Multiple comparisons between means indicated that the relational involvement mean for mixed, or multiple, forms of gender appropriations (M = 4.07) was significantly lower than those for all other identification modes. This indicates that message senders were more likely to be confrontational when gender identity was not concealed or not consistently concealed throughout a message. Table 11.2. Means for Men and Women on Message Variables

Message Variables

Men

Women

Relational Involvement Topical Involvement Formality

4.92 I .76

4.70 1.78 1.36

1.15

The third hypothesis asserts that the amount of topical involvement in the messages of men and women would be significantly related to the clarity of their gender appropriation. The ANOVA interaction effect for topical involvement was not significant (F (4, 1.203) = 1.10, p < .40). Therefore, the underlying hypothesis could not be supported by the data. The main effect for gender also was not significant (F ( 1 , 1,203) = .33, p < .60). The main effect for mode of identification was significant (F (4. 1,203) = 6.23. p < .0001, (0’ = .14). Table 11.3 contains the means for this analysis. Multiple comparisons indicated that the topical involvement means fur direct appropriations (M = 2.03) and mixed modes of identification (M = 1.98) were significantly different from those of other identification

E.A. Mabry /Ambiguous Seljlldentyication

257

modes. However, the results make it impossible to support this hypothesis. Table 11.3. Means for Modality of Gender Identification on Message Variables

Message Variables

None

Relational Involvement Topical Involvement Formality

4.67 1.74 1.25

Gender Identification Name Direct Indirect Mixed 4.95 1.75 1.14

4.52 2.03 1.96

4.50 1.88 1.51

4.07 1.98 1.68

Note. None = no explicit gender reference; Name = a name attributed to a gender included with message; Indirect = reference to one's own gender only by comparison or implication; Direct = explicitly stating one's gender; Mixed = a combination of two or more modes (except None).

The final hypothesis argues that the amount of formality of expression in the messages of men and women will be significantly related to the clarity of their gender appropriation. The ANOVA interaction effect for message formality also was not significant (F (4, 1,203) = 1.75, p < .15). Again, there was no significant difference in the way men and women engaged in identity management that affected the formality of messages they expressed in online groups. The main effect for gender was significant (F (1, 1,203) = 23.68 p < 0001, a2= .23). Means for this analysis are in Table 11.2. A comparison of the means showed that men were significantly more formal in their message enactments than women. The main effect of mode of identification also was significant (F (4, 1,203) = 53.45, p < .001, o2 = .42). Table 11.3 contains the means for this analysis. Multiple comparisons indicated that message formality means for messages with no gender identification (M = 1.25) or identification only through a name (M = 1.14) were significantly lower, indicating greater formality, than other forms of gender identification. And, the mean for indirect appropriations was significantly different from all other means. Therefore, once again, while there were substantial differences on formality scores related to mode of gender identification, gender and mode of identification did not statistically interact and the research hypothesis could not be supported.

I I .2.4 Gender, identity management, and concealment: research implications The results of analyses of mode of gender and mode of gender identification and message variables were inconclusive. Men and women participating in online discussion groups tend to use different forms of self-referential message constructions. Therefore, the overall amount of and potential for miscommunication caused by identity management was not substantially isolated to one gender. Also, it would not be accurate to conclude that only men used more disambiguating strategies. Women's use of gender-revealing self-referential messages showed a more complex trend. First, the majority of messages from women paralleled the content patterns observed for men. Second, women were more likely to create messages including either direct references to their sex or containing mixtures of references with varying types of self-identification. Women also were four times more likely than men to use indirect selfreferences to their gender in the messages they sent to online groups. The three hypotheses specifically asserting relationships between gender, self-

identification, and language variables measuring relational involvement, topical involvement, and message formality were not supported. It would be tempting to attribute these results to factors like the attenuated range of the summated scales (based on items with only two or three categories). However, this explanation would be inconsistent with the significant ANOVA main effect results obtained using these variables. The research design variables of gender and mode of identification of the interactant were robust. They produced significant ANOVA main effects for all of the message style variables. These results provide strong, though not unequivocal, support for the presumptive rationale underlyng the hypotheses. Results for comparisons between men and women's use of relational involvement language were not in the direction expected. If women in on-line groups are disposed to be more guarded with their presence. they would be expected to have lower relational involvement scores than men. The results did not confirm that expectation. Women's scores were significantly higher than men's scores. Results for the topical involvement and formality seemed consistent with the basic assumptions underlying the strategic self-presentation hypothesis suggested by Spears and Lea [ S , 61. Women were significantly less likely to be as involved in the topicality of online discussions. They were also less likely to use formal language in their messages. However, there is evidence from F2F research that deception and informality are positively related [ 171. These outcomes seem consistent with the expectation that, in order to enact a more viable means of self-protection online, women will appear more detached in the presence they project through their messages. Yet, this distance may be tempered by a less formal style of self-presentation. This finding is also consistent with the assumption that as women perceive less risk in a group setting they may become less concerned with identity management thereby reducing their social distance as members by using less formal messages. In general, the results of the data analyses appear to bear out these implications. Looking broadly at the overall pattern of results. we have some reason to be optimistic that, even with failures in the potency of certain statistical tests. the underlying rationale motivating this project has been reinforced by these findings. There is some evidence to suggest that both men and women approach on-line discussions as though sincerity was not overly problematic. This suggests that online messages are influenced by a sort of normative halo-effect regarding taken-for-granted assumptions of genuineness accorded F2F interaction. Second, and just as important, we also have demonstrated that the exportability of assumptions about managing identity that govern F2F contexts are open to manipulation in online contexts (albeit for unarguably good reasons). The self-referential nature of communication can present a Janus-faced dilemma for many people. Genuine selfpresentation can be precipitous in contexts that are relatively inaccessible to conventional social control mechanisms. The results of this study clearly indicate that participants in online groups apparently make choices regarding their persona and presence. These choices are instanciated in participants' communication styles evidenced either mindfully or unmindfully in the ways they construct messages that are shared publicly with. for the most part. strangers or cursory acquaintances. Thus, while not usually intended. the pragmatic realities of online contexts can make miscommunication either difficult to avoid or socially necessary.

E.A. Mabry /Ambiguous Self-Identification

11.3

259

Gender, status, and concealment in CMC

We use information obtained from observing people's F2F social interactions to make inferences about the genuineness of their self-presentations [2]. An important clue to a person's sincerity is the way in which he or she performs facework (Goff'man [28]. Facework is a general term for negotiations between social actors involving their autonomy, solidarity, and approbation (Lim and Bowers [29]). Solidarity and approbation refer to positive face, which addresses desirability needs, including regard for a person and respect for a person's abilities [28]. Autonomy refers to negative face, or a person's need not to be imposed upon by others. Attention to face can be observed through the level of openness with which he or she engages others [30]. Finding from the preceding study reported that the percentages of men and women sending messages containing no embedded gender attributions was approximately equal. Women were less forthcoming than men in explicitly providing potentially revealing information. Similarly, women were four times more likely than men to indirectly refer to their gender. Yet, women were much more likely than men to use direct gender references as well as author messages containing more than one form of gender identification. Most compelling, however, is the finding that confrontational messages were more likely to contain identity information. This result is not easily reconciled with theoretical stances like the cues filtered out perspective. It is consistent with SIDE model reasoning only if we can assume that online participants felt a strong enough sense of in-group identification to mitigate risks associated with revealing identity information [71 Findings from the previous study also suggest that participants in CMC groups become more willing to engage in sincere self-presentation as they become more ego-involved with their participation. However, using gender as the only proxy for identity management may not yield the most powerful measure of identity. In fact, evidence bearing on the management of anonymity suggests that social power can mitigate anonymous activity [18, 191. This reasoning is consistent with the SIDE model and Social Information Processing [ 141 perspectives. In the latter case, providing information is a way of garnering prestige. Anonymity becomes a hurdle to accumulating status through information sharing because it inhibits the formation of credibility impressions [ 151. Status is also a corollary to facework and, by extension, openness in self-presentation. Higher status people have greater face needs. However, their social power also provides more social capital for negotiating face needs and more presumed control over threats to their sense of face. Yet, women's status resources as communicators are different than men's. And, differential uses of language often leads to miscommunication between men and women [30]. The potential for such miscommunication seems even greater in online group contexts. As the preceding research has demonstrated, both gender and identity management shapes the styles of CMC messages enacted in online groups. Men and women also can choose to reveal or conceal other social status cues besides gender. Gender is also an interactive status marker that is often only manifestly evident during communication [31]. Other status cues, based on conferred or achieved social position, can also have a great bearing on individual and group identity. Therefore, status cues not specifically related to gender could enhance, or even subsume, the status value of gender. There is also evidence that social status and concealment decisions are related [ 191. And, evidence from research on deception in F2F interaction shows that status markers like expertise influence perceptions of messages from both senders and receivers [ 171. Literature regarding concealment and uncertainty indicates that interpersonal openness is more likely to be signaled in the patterns of message content and exchanges that define

E.,4. Mahn /Ambiguous Self2dentjficarion

260

personal relationships [32]. Openness also can define the value of the social capital one brings to a social context. Higher status is more likely to stimulate openness when the source(s) of one’s status are not perceived as targets of ego-threatening reactions from others. The working assumption is that: non-gendered status markers should enhance social power, thereby making gendered identip less tenuous, and lead to greater openness in online social communication. Results from another study investigating the hypothesized effects of gender, status. and identity management on communication are discussed below. However, the message variables previously analyzed were not replicated in this study. The underlying rationale for this hypothesis is not broad enough to be extended to relational or topical involvement. Message formality is easily understood as a stylistic corollary to status. However, the potency of the message formality variable was already established and was not re-tested in this study for the sake of brevity. I I . 3. I Assessing status, identip management, and openness Data for this study was obtained from the same database used in previous study. Testing the hypothesis required the use of gender and gender identification variables, plus the addition of a variable to measure the research design factor of status. Message analysis required the construction of a new summated scale measuring openness. See Table 1 1.1 for information regarding variables carried over the previous study. Three research design variables were needed for testing the hypothesis regarding status, gendered identity management effects on communicative openness: gender of the message sender, mode of gender expression (if present), and communicator status. One analysis variable, communication openness, was tested. The openness variable is explained below. Status was measured using a message content item coded to denote the presence or absence of any form of information about the communicator’s personal status (e.g.. degrees, positions, titles, and so forth). Messages were rated “no” when they did not contain such information and “yes” when they did. The kappa statistic for reliability on this measure was K = .8 1. As was the case in the preceding study, there was some data loss due to aggregating items in scale construction. Table 1 1.4. Openness Variable Item Inclusiveness Personal Addresses Challenges Exhorting Action Openness

Range

1-2 1-2 1-2 1-3

Operational Description

Reliability

Using 1“. person plurals: No; yes Acknowledging others: No; yes Demands authorizatiodverification:No; yes Out-of-group responsiveness: No: peripheral; central to message

K =

.79

K = .71 K = .95 K =

I?,=

.83 40

One summated scale to measure openness was developed by screening relevant message analysis variables included in the database. Four variables that operationalized a sense of interpersonal acceptance and inclusiveness, and willingness to expand thought and proactive behavior, were reliably aggregated to form the openness variable. The variables, brief explanations, and reliability information are presented in Table I 1.4. individually. the items demonstrate strong reliability. The summated scale is weaker but allows for an efficient metric suitable for the statistical design.

26 1

E.A. Mably /Ambiguous Self-dentification

11.3.2 The effects of status and gendered identity management on openness

Previous analyses of the N = 3,000 messages showed that men's message output outnumbered women's by a five to one ratio. Those findings suggested that there is an interaction between participant's gender and mode of gender identification. Therefore, the hypothesis was analyzed using a 2 (Gender) X 5 (mode of identification) univariate analysis of variance (ANOVA) [27]. The hypothesis proposed that the amount of openness in message enactments of men and women would be significantly related to the clarity of their gender appropriation and apparent status. The results are intriguing. The ANOVAs for main effects were significant: gender (F (1, 1,194) = 9.77, p < 002, o2= .15); mode of gender expression (F (4, 1,194) = 21.23,p<01,a2=.29);andstatus(F(1,1,194)=8.74,~<003,02=.13). However, there were significant two- and three-way interaction effects that made it impossible to interpret main effects. The interaction between status and mode of gender expression was significant: F (4, 1,194) = 5.85, p < 0001, o2= .34. And, there was a significant three-way interaction: F (3, 1,194) = 3.86, p < 009, a2= .lo. Only the highest order of interaction effect can be directly interpreted [27]. Table 11.5 contains the means for the interpretable ANOVA analysis. Overall, the effect, while statistically significant, was only moderately strong in yielding interpretable comparisons. One reason for this outcome is the lack of data for men directly mentioning status markers in their messages. This is a somewhat surprising result as it better fits patterns that have been established for women in online groups. Post hoc comparisons of means indicated men indirectly mentioning status (M = 1.83), and women directly mentioning status (M = 1.69), contributed most to the significant interaction effect. Table 11.5. Openness Scale Means for Gender, Mode of Gender Identification, and Status Gender: Male

Modalities

Status

None

Name

Direct

Indirect

Mixed

Not Mentioned Mentioned

1.17 1.42

1.18

1.33

1.19

.OO

1.15 1.83

1S O 1.38

1.22 1.39

1.22 1.25

1.54

1.27 1.33

1.53 1.58

Gender: Female Status Not Mentioned Mentioned

1.69

The only clear pattern to emerge for either gender was that men and women were more open in their communication under very different combinations of status and gendered identity representations. Men were more likely to evidence greater openness in their online group messages when their mode of gender identification was indirect. The patter for women was more complex. Openness in women's messages was highest under both indirect and mixed identification conditions. Therefore, it appears the triggering mechanism stimulating openness may be readily identified for men but could be somewhat less obvious for women. Moreover, the fact that different ways of revealing gender were related to the presence of status markers suggests that status empowers each sex differently

262

E.,4. Mabn. / .4mhiguous Selflldentification

in online groups.

11.4

Gender identity and status implications for communicator openness

The hypothesized assumption that there is a relationship between participant's gender, selfidentification, status cues, and level of openness was supported. Not surprisingly, men and women in the anonymous and perfunctorily identified conditions were the least open. This is consistent with the possibility that either on an individual basis across groups, or due to the social climates in particular groups, there were greater risks to self-image regardless of gender. It is clear that women are more likely to enact openness in their communication because they reveal themselves more than men. Findings that connected gender and status also were revealing in another way. It was surprising to learn that status did not universally empower members of either gender to act more openly as communicators when they had made revealing remarks in their messages. This finding suggests group members that do not participate anonymously probably have specific interaction goals in mind. Such goals may explain why they choose to participate in a particular group. Mabry [ 121 has shown online groups with a clearer apparent sense of purpose have messages with more internal structuring (framing) and higher topical involvement, but lower formality, scores.

1 1.5

Conclusions

Finally, we must ask, where do the findings of these studies lead? First, it is clear that lines of theoretical inquiry embracing egoistically nuanced explanations for online communication have an advantage over theories that more narrowly focus on the presumed communicability of message channels. Second, the fact that message style and openness were influenced by decisions involving communicator choices for concealment or revelation suggests that self-presentation often is actively managed. Status factors that could be meaningful to an online group (whether related to gender or not) enter into a communicator's message constructions. Third, the notion that sincerity as it has been discussed here is a problematic facet of social interaction indicates that miscommunication is not always passive or accidental. Clearly. it would be prudent to reassess both the conceptual and operation definitions of our primary variables. Admittedly, it is hard to imagine that our measures did not potentially over-specify the concepts of relational involvement. topical involvement. and formality. Obvious too is the need to more directly address the ways consistency and congruency operate within a message. Here again, additional study might lead us to different patterns of individual message content measures available in the database. And, of course, the opportunity to revisit the original messages and reanalyze them by applying new measures (and even methodologies) for understanding sincerity and its constituents is exciting. As mediated technologies become increasingly complex, and capable of integrating video. voice, and data transmissions into broadband mediated channels, it becomes important to understand how the physical and symbol producing attributes of these communication technologies influences communication choices [ 3 2 ] . This becomes particularly necessary as new technologies increase the visibility of participants in online groups. The level of accessibility a communicator chooses to permit \ \ i l l become a context

E.A. Mabry /Ambiguous Selfldentification

263

defining decision. Such decisions will shape communication goals and perceptions of utility attributed to how technologies are being used to enact those goals. The notion of miscommunication could well shift from an emphasis on behavior to underlying intentions associated with channel selection. Finally, there is the need to explore other communicative implications of electronic messaging. Research needs to address both the message-centered (and thus interactant-centered) and the contiguous and processual dimensions of the computer-mediated communication. One limitation to the data reported here was that data sampling procedures inferred with the collection of contiguous message streams. Future research needs to examine intact message streams in order to provide more precise analyses of message patterns and any embedded meanings such patterns might reveal.

11.6

Acknowledgments

The author wishes to thank Ms. Martha Karge and Ms. Drue-Marie McKelton for their assistance in early phases of this project.

11.7

[51

References E. Goffman, The presentation ofselfin everyday life. Garden City: Doubleday and Co., 1959. E. Goffman, Relations in public. New York: Basic Books, 1971 R. K. Belk, Possessions and the extended self. Journal ofConsumer Research 15 (1988) 139-168. H. Bechar-Israeli, From to : Nicknames, play and identity on internet relay chat . May, 1996. [On-line] Available E-mail: [email protected] R. Spears and M. Lea, Social influence and the influence of the 'social' in computer-mediated communication, in: contexts of computer-mediated communication, M. Lea, Ed. London: Harvester Wheatsheaf, 1992, pp. 30-65. R. Spears and M. Lea, Panacea or panopticon? The hidden power in computer-mediated communication, Communication Research 21 (1994) 427-459. T. Postmes and R. Spears, Deindividuation and antinormative behavior: A meta-analysis, Psychological Bulletin 123 (1998) 238-259. T. Postmes, R. Spears and M. Lea, Breaching or building social boundaries: SIDE-effects of computermediated communication, Communication Research 25 (1998) 689-7 15. T. Postmes, R. Spears and M. Lea, The formation of group norms in computer-mediated communication, Human Communication Research 26 (2000) 34 1-371. J. B. Walther, Anticipated ongoing interaction versus channel effect on relational communication in computer-mediated interaction, Human Communication Research 20 ( 1994) 473-50 1. J. B. Walther and J. K. Burgoon, Relational communication in computer-mediated interaction, Human Communication Research 19 (1992) 50-88. E. A. Mabry, Textual framing as a communication climate factor in online groups, in: Framing public life: Perspectives on media and our understanding ofthe social world, S. D. Reese, 0 . H. Gandy, Jr. and A. E. Grant, Eds. Hillsdale: Lawrence Erlbaum Associates, 2001, pp. 323-336. M. J. Culnan and M. L. Markus, Information Technologies, in: Handbook of organizational communication: An interdisciplinary perspective, F. M. Jablin, L.L. Putnam, K.H. Robert and L.W. Porter, Eds. Newbury Park: Sage, 1987, pp.420-443. J. B. Walther, Computer-mediated communication: Impersonal, interpersonal and hyperpersonal interaction., Communication Research 23 (1996) 3-43. P. B. O'Sullivan, What you don't know won't hurt ME: Impression management hnctions of communication channels in relationships, Human Communication Research 26 (2000) 403-43 1. C. R. Scott (as Anonymous), To reveal or not to reveal: A theoretical model of anonymous communication, Communication Theory 8 (1998) 38 1 -407.

2 64

E .4 Mabn .4mbigrious Self-Identification

[ 171 D. B. Buller. J. K. Burgoon, C. H. White and A. S. Ebesu. Interpersonal deception VII. Behavioral profiles of falsification. equivocation. and concealment. Journal qf'Lunguage and Social P.qrholo,q. 13 (1994) 366-395. [ 181 S. Reicher and M. Levine. Deindividuation, power relationships between groups and the expression of social identity: The effects of visibility to the out-group. Briti.yh Journal qf Social f.y,vcholo@.,33 (1994) 145-163. [ 191 G. T. M a n , What's in a name'? Some reflections on the sociology of anonymity. Tlic Informa/ion Science, 15 ( I 999) 99- 1 1 2. [20] J. B. Walther, Interpersonal effects in computer-mediated interaction: A relational perspective. Communication Research 19 ( 1992) 52-90. [ 2 1 ] A. H. Eagly, Se.x diflerencus in .socinl hehatior: .-I .soc.iul-role interpretntion. Hillsdale: Lawrence Erlbaum Associates. 1987. [22] M. S. Rafaeli, F. Sudweeks, J. Konstan. and E. Mabry. Appendix: Project H overview: A collaborative quantitative study of computer-mediated communication. in: ,vehrork and netplarb: I 'it-tidalgroups on rhe Internet. F. Sudweeks. S . Rafaeli. and M. Mclaughlin. Eds. Menlo Park: A A A I PresstMIT Press. 1998. pp. 265-28 1 . [33] R. L. Brennan and D. L. Prediger, Coefficient kappa: Some uses. misuses. and alternatives. Educational and Psychological Measurement 41 ( 198 1 ) 687-699. [24] L. J. Cronbach. Coefficient alpha and the internal structure of tests. Psj-chometrika 16 ( 195 1 ) 297331. [ 2 5 ] E. A. Mabry. Frames and flames: The structure of argumentative messages on the Net. in: .Yenwrk and netplq>;Virtual groups on the Ittiernet. F. Sudweeks. S . Rafaeli. and M. Mclaughlin. Eds. Menlo Park: AAAI PressMIT Press, 1998, pp. 13-35. [26] M. Karge, N. Burrell, M. Allen and E. A. Mabry. Gender differences in communication styles in computer-mediated communication. Department of Communication. University of \IrisconsinMilwaukee. unpublished report. October. 200 1. [27] G. Keppel. Design and analvsis: .3 re.warc-her!yhandbook. Englewood Cliffs: Prentice-Hall. 199I . [ 2 8 ] E. Gof€inan, Interaction rittial: E.r.srij*s on ~fuct.-ro-Iuceheha\ior. Garden City. NY: Doubleday and Company. 1967. [29] T. S. Lim and J. W. Bowers, Facework: Solidarity. approbation. and tact. Humari C'ommrrnic*ation Research 17 ( 199 1 ) 4 15-450. [ 301 G. Hiemstra. Teleconferencing. concern for face, and organizational culture, in: Commtinication ).carhook. 6. M. Burgoon. Ed. New Brunswick: International Communication Association. 1982. pp. 874-904. [ 31 ] N.M. Henley and C. Kramarae, Gender, power. and miscommunication. in: ".~~i.~communIc.afiorl" and problematic talk. N. Coupland. H. Giles and J. M. Wiemann. Eds. Newbury Park: Sage Publication. 1991. pp. 18-43. [ 3 2 ] J . R. Brown and L. E. Rogers, Openness, uncertainty. and intimacy: An epistemological reformulation. in: ",l..li.scornmunicatiori *' and problematic raik, N . Coupland. H . Giles and J. M. Wiemann. Eds. Newbury Park: Sage Publication, 199I , pp. 146-165. [33] E. A. Mabry. Group communication and technology: Rethinking the role of communication modality in group work and performance. in: N m . directions in group conimtitiicatiori. L. R. Frey. Ed. Thousand Oaks: Sage Publications. in press.

265

Author Index Anolli, L. Attardo, S. Balconi, M. Carrera, P. Casado, C. Ciceri, R. Colston, H.L. Fernandez-Dols, J.-M. Ferrero, G. Gibbs, R.W. Infantino, M.G. Mabry, E.A. Mantovani, F. Reboul, A. Riva, G. Zaninelli, S.

ix,3,73,135 159 73 117 117 ix,73,101,135 181 117 V

181 135 247 229 43 ix, 197 vii