This page intentionally left blank
Meaning, Discourse and Society
Meaning, Discourse and Society investigates the construction of reality within discourse. When people talk about things such as language, the mind, globalisation or weeds, they are less discussing the outside world than objects they have created collaboratively by talking about them. Wolfgang Teubert shows that meaning cannot be found in mental concepts or neural activity, as implied by the cognitive sciences. He argues instead that meaning is negotiated and knowledge is created by symbolic interaction, thus taking language as a social, rather than a mental, phenomenon. Discourses, Teubert contends, can be viewed as collective minds, enabling the members of discourse communities to make sense of themselves and of the world around them. By taking an active stance in constructing the reality they share, people thus can take part in moulding the world in accordance with their perceived needs. w o l f g a n g t eu b e r t is Professor of Corpus Linguistics at the University of Birmingham. His previous book publications include Corpus Linguistics: A Short Introduction (2007, with Anna Cermakova), Text, Discourse and Corpora (2007, with Michael Hoey, Michael Stubbs and Michaela Mahlberg) and he was co-editor of Corpus Linguistics: Critical Concepts in Linguistics: A Reader (2007) and Text Corpora and Multilingual Lexicography (2007).
Meaning, Discourse and Society Wolfgang Teubert
CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo, Delhi, Dubai, Tokyo Cambridge University Press The Edinburgh Building, Cambridge CB2 8RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521885652 © Wolfgang Teubert 2010 This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2010 ISBN-13
978-0-511-77032-6
eBook (NetLibrary)
ISBN-13
978-0-521-88565-2
Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.
Contents
Acknowledgements Introduction
page vii
No meaning without other people The futile quest for a language system Should the mind be an object of scientific enquiry? The discourse community constructs reality The sixteen topics of this book
1 1 5 9 11 12
Meaning, the mind and the brain
31
Part I 1
The cognitive turn
33
2
The long history of mind linguistics
47
3
What do we know about mental concepts?
53
4
Morphing theoretical sémes into ‘real’ concepts
63
5
From mental representations to conceptual ontologies
67
6
What is meaning?
73
7
Where should we look for meaning? Option a: The individual intentional mind Option b: The individual computational mind/brain Option c: The brain where language becomes reality Option d: The discourse as the collective mind
Part II
Discourse and society
88 88 90 95 107 111
8
Language as discourse
113
9
Society presupposes language, and language presupposes society
124 124 127
Language as an essential human feature Language and society
v
vi
Contents Blumer’s symbolic interactionism From the individual mind to the discourse as the collective mind
132 135
10
A closer look at oral societies
140
11
Differences between oral and literate societies
150 152 156 163
The profane and the arcane origins of writing Some effects of literacy Literacy, meaning and reflection
12
Empirical linguistics deals only with recorded language
166
13
Meaning, knowledge and the construction of reality
171 171 176 180
Truth and meaning Meaning and knowledge The collaborative construction of reality
14
The language of the scientific experimental report
190
15
Diachronicity, intertextuality and hermeneutics
199
16
Meaning and the interpretation of a haiku
215
Conclusion
241
Authorship, intentionality and mental states: can the quest for meaning dispense with the investigation of the solitary mind? How real are the realities we experience? The social dimension of experience and intentionality Why the media discourse advertises individual agency We can collaborate to change our reality
Bibliography Index
241 249 258 262 269 273 283
Acknowledgements
The impulse to write this book arose from the quandary of the nature of meaning which had bewildered me for many years. Only when I had moved away from the more navel-gazing disposition of my former research institution in Germany, where I used to look for meaning in the depths of solitary minds, and had been enveloped in a British university’s spirit of sociability, did it begin to dawn on me that just as it is meaning that creates society, it is society that gives rise to meaning. Meaning emerges whenever people interact symbolically, negotiating the signs they use to communicate. As long as such interaction continues, meaning keeps evolving. Therefore there can never be a finite answer to the meaning of meaning. Meaning is only in the discourse, not in people’s minds, and not in a reality out there. If we want to make sense of this discourse and the realities it has in stock for us, we do it not as monadic entities, but by working together, adding layer after layer of interpretation to all the previous layers of which the discourse consists. While such an approach is contrary to that of nativist linguistics, of the cognitive sciences and of the philosophy of mind, it is certainly not new or original. It draws on ideas developed in corpus linguistics and integrationist linguistics and on pragmatism, hermeneutics, social constructivism/constructionism and various brands of discourse studies, to name some key inspirations. I would like to thank the University of Birmingham, its School of Humanities, and most of all my colleagues at the Department of English for the warm welcome they extended to me when I arrived here nine years ago. Without this stimulating atmosphere, the constant encouragement and the generous intellectual support I would never have ventured on this book project. As long as I continue to breathe this academic air, I will, I hope, keep changing and developing my outlook. In this sense the ideas presented here are more work in progress than a conclusive perspective. Over this time I was given ample opportunity to present my ideas to my colleagues at Birmingham and many other places, and not only did they receive them with endless forbearance but they also opened my eyes to many aspects I had overlooked. I am equally grateful to my undergraduate students who engaged in often heated discussions with me, trying to instil a modicum of vii
viii
Acknowledgements
common sense into my ruminations. Many of the topics touched upon in this book have been introduced to me by my graduate students, among them Tony Bastow, Lisa Cheung, Angel Garralda, Nelya Koteyko, Fumiko Kondo, Maria Oset Garcia, Dominic Smith, Laura Strakova, Wang Fang and Wang Weiqun. Friends at a number of Chinese universities, Feng Zhiwei, Li Wenzhong, Pu Jianzhong and Wei Naixing, in particular, have repeatedly given me the opportunity to present my ideas and discuss them in remarkable depth with their students. The University of Birmingham and the Arts and Humanities Research Council granted me a year of study leave which gave me time to organise my ideas. Bill Dodd and Michael Toolan have generously exercised their heartwarming friendship by reading and commenting on earlier versions of many chapters; without their calls to order I would have frequently lost my thread. I owe it to them that the ideas presented here have received some sort of Gestalt. Deeply felt thanks also go to Ruth Mell, who took care of the index, and to Jacquie Mullender, who not only corrected my English but also remedied numerous inconsistencies. Caroline, my wife, had to suffer a lot from my occasional irascible outbursts, and without her unfaltering support I might well have given up this project. Finally, I would like to thank Helen Barton, my editor at Cambridge University Press, for her trust and her imperturbable patience.
Introduction
No meaning without other people This book is about meaning. Probably no society has ever been more concerned with meaning than the one in which we live. Never before have so many people felt such an urge to make sense of the world they live in and of the lives they are leading. They find this sense not so much in themselves as in the discourse, which is the entirety of everything that has been said and written by the members of the discourse community to which they owe their identity. It is communication, this verbal interaction with others, which reassures them about their notions and ideas, and in which they find interpretations they can accept, rework or reject, and in which they can recognise themselves. In principle, everyone has a voice in the discourse. But in reality we find that our modern society is neatly divided into those who are commissioned to produce texts for the media and the rest of us who consume them. While each of us may say whatever we want, it seems to carry less weight than what we are told by the discourse we find on the shelves of our content merchants: newspapers, magazines, television, much of the web, NHS brochures and similar pamphlets issued by our authorities, instruction manuals and even those old-fashioned things called books. Secondary experiences supplied by the media have taken over the role that a person’s own experiences and those of their friends and neighbours had for former generations. Even when we want to find out what our own experiences mean, we trust the texts offered by the content merchants more than our own judgement or that of our friends and neighbours. The media, not the common sense we exercise in conversation with family and friends, will tell me whether feeling fed up with my workload means that I have depression and should take pills. Such a dependence is hardly surprising. Even in the good old days when the media had little power over our thoughts, we always needed a discourse community to make sense of our experiences. Interpretation is inevitably a collaborative act. We do not interpret our experiences for ourselves; we do it for an audience, imagined or otherwise. We want to learn from the interpretations offered by others, and we want others to share our interpretations. There is no meaning without society. My view is that the world, our lives, the things we do or don’t do, and what happens to us, have no meaning at all, in as much as we do not 1
2
Meaning, Discourse and Society
appropriate them through interpretation. Our experiences only make sense when we reflect on them, or when we share them with others by talking about them, or when we weigh them against other people’s experiences as we find them in the media we consume. This is indeed what we do most of the time when we talk with each other: we assign meaning to what we do and what others do. The only reality that counts is the reality we find constructed in the discourse, in this entirety of texts that have been exchanged and shared between the people who make up society. We never cease contributing to this limitless, all-encompassing blog uniting humankind that I call the discourse. The discourse tells us how we experience, how we ‘feel’, what happens to us and what we do. We have learned how to experience things through the stories other people have told about their experiences. For us, too, the only way to communicate our personal experiences is by contributing them to the discourse. We cannot do that without interpreting them, without assigning meaning to them. It is the discourse that makes our lives meaningful. The discourse tells us how we can view the world, our private lives, the things we do or don’t do, and the things that happen to us. Without the discourse, these things, and even life itself, remain devoid of meaning. This is why chimpanzees, lacking language, are not concerned with the meaning of their lives. The word life means what life is for us. The meaning of life is therefore not really different from the meaning of the word life. It is all that has been said about it. Google lists c. 229,000 occurrences of the phrase ‘the meaning of life is’. Here are a few citations, taken from the first fifty entries: The meaning of life is that there is no meaning at all. The meaning of life is to live. the meaning of life is what you make of it The meaning of life is to make life meaningful. the meaning of life is to reach Nirvana The Meaning of Life is the title of a 1983 Monty Python film. The results for ‘life means’ are not so different. Among the first entries (of 1,130,000) listed by Google, we find: Where Life Means Getting a Little Sand in Your Shoes When Life Means Life. Life means so much. Life means suffering. eternal life means serenity Life Means Nothing. Life Means Nothing Behind the Green Wall It is difficult to imagine that by pondering we would find an answer to the meaning of life/life that is not already expressed in the discourse. The same is
Introduction
3
true for all the other things our world consists of. We are aware of no other things, concrete or abstract, beyond those which have already been discussed. By reflecting on them we may hope to find something new. But once we take a closer look we find that what we have taken to be new is no more than a recombination, a permutation, a reformulation of what has been said before. This is how new ideas come about. As long as we keep talking to each other, as long as the discourse goes on, there will always be innovation. That content needs a discourse is not such a new idea The idea that there is nothing ‘really’ new is not new. It has been with us, it seems, since the beginning of time. In Ecclesiastes 1, 9–10, we find it already spelt out: The thing that has been, it is that which shall be; and that which has been done is that which shall be done: and there is no new thing under the sun. Is there any thing whereof it may be said, See, this is new? it has been already of old time, which was before us.
My book is yet another text, in a world flooded with texts on a scale our ancestors would not have dreamt of. Like most other books, it builds on previous texts. The ideas it presents have been around for a long time. What I try to do in this text is to recombine, permute and rephrase them in such a way that perhaps something slightly new takes shape. We can look at the discourse as the memory of all the hands of a deck of cards that have ever been played, and we can see every new text as a new hand, more or less similar to previous hands but not quite identical to them, made up mostly of the very same cards. Sometimes an author may try to smuggle in a new card, and, if she or he is very lucky, the other players will accept it as long as they do not notice. Rearranging existing units of meaning is more than playing with words. It is presenting the world to us in a new light. We always have the power to change this discourse-internal world by adding yet another, our own, interpretation onto previous texts. If other people take notice, if it leaves traces in future texts, it will have had an impact. It is the discourse that turns the stuff of reality out there into objects. As objects of the discourse, they are at our disposition. All we have to do when we want to change things as we find them is to reconstruct them, to assign a new meaning to them. Karl Marx was wrong when he said: ‘The philosophers have only interpreted the world differently, what matters is to change it’ (Marx 1888: 7). The only way to change the world as we encounter it, mediated through the discourse, is to reinterpret it. When we talk, we never start at point zero. We react to things that have been said before. We praise, or accept, or criticise, or reject what has been said before. Perhaps we proffer a counter-example to the example we were given. Even if we
4
Meaning, Discourse and Society
invent a new story, it will be modelled on existing stories. Indeed, unless we had been talked to we would never say anything. For we learn to speak by reacting to those who speak to us first. Whatever we may say, it is made up from the building blocks provided by the existing discourse. All we do in a new text is to rearrange these elements. The text that I contribute to this discourse is a reinterpretation of previous texts, which, of course, were also nothing but interpretations of interpretations of interpretations. For us, the interpretive community, symbolic content never refers to anything tangible. All we can see is how it refers to something said before. The first arbitrary sign used by someone to signal something to someone else referred not to something in the world outside but to an object constructed through symbolic interaction. Any new text is always in some way a comment on previous texts, a re-assignment of meaning. Even those who disclaim the plausibility of this assumption, advancing instead the equally convincing idea that we only speak because we have something new to say, show unwillingly that they are, too, treading no uncharted territory but paths already well-trodden. Noam Chomsky provides an excellent example: The first [creative aspect of language use] is that the normal use of language is innovative, in the sense that much of what we say in the course of normal language use is entirely new, not a repetition of anything that we have heard before and not even similar in pattern – in any useful sense of the terms ‘similar’ and ‘pattern’– to sentences or discourse that we have heard in the past. This is a truism, but an important one, often overlooked and not infrequently denied in the behaviorist period of linguistics, to which I referred earlier, when it was almost universally claimed that a person’s knowledge of language is representable as a stored set of patterns, overlearned through constant repetition and detailed training, with innovation being at most a matter of ‘analogy’. (Chomsky 1972: 11f., my emphasis)
How innovative is this text segment really? According to the Cobuild dictionary, a truism is ‘a statement that is generally accepted as obviously true and is repeated so often that it has become boring’. What Chomsky had to tell us was, it seems, nothing new; it was no more than a reformulation of what we had been told all along. But as such, it was enormously successful. It was the foundational idea of the paradigm of generative transformational grammar which has dominated much of theoretical linguistics for half a century. My aim in this book is to look again at what has been said about meaning in various fields of linguistics, social studies and the philosophy of language. I will focus on two perspectives. The goal of Chomskyan linguistics and of cognitive linguistics (two prominent schools of linguistic thinking) has been to build a model of the language system, seen as the mechanism for turning thought into utterances and utterances into thought. The mind, both schools agree, is the seat of this mechanism. This is the perspective that I will show to be defective. Instead, I will propound the view that meaning is only in the discourse. Our
Introduction
5
world, to the extent that we can make sense of it, is a world we have constructed for ourselves, or, to be more realistic, that others have constructed for us. What we take to be reality is always mediated by what has been said. The futile quest for a language system Is there a language system? This is not a question I am really much concerned with. This book is about meaning, and not about the rules and regularities we find in the field of grammar. The rules we are supposed to observe in a language like English have evolved over centuries. That they are relatively stable and largely accepted wherever English is spoken is on the one hand due to inertia – there is no need to change things as long as they work well. On the other hand, there have been, time and again, efforts to standardise language in cases where we find variants. The rule systems for English and for many other European languages, as we encounter them today, owe much to the creation of the modern nation state in the nineteenth century. A unified school system presupposes not only a unified curriculum, but even more a common language. Every pupil had to be taught an inventory of grammatical rules. To a certain extent, this unification also concerns word usage. While in many English dialects borrow and lend can still be used interchangeably, we are required to use them as converse correspondences in contemporary standard English. What I lend to you is what you borrow from me. In this process of standardisation, dictionaries have played a major role, not only as repositories of the linguistic heritage, but also as voices of authority for the meaning of words. We should distinguish between rules and regularities. Rules are what we find written down in grammar books, and what is taught in school. While we may not always be aware of our own rule-following, we can look rules up. Regularities concern practices we normally follow without being explicitly told. In English, we would normally place the modifying adjective in front of the noun. In French, it is often the other way around. This is not something native speakers have to be taught; they pick it up quite ‘naturally’ when they acquire their first language. What is a regularity for a native speaker often has to be learned as an explicit rule by the foreign learner. This also holds for the vocabulary. We tend to take for granted that words are the ‘natural’ elements of any language. Language acquisition, whether learning one’s mother tongue or a second language, seems to be first of all learning the meaning of words. Dictionaries tell us how they are to be used. Even more than the more loquacious monolingual dictionary, its bilingual cousin fosters the illusion that there is a system behind word meanings. But in spite of all attempts to pin down the accurate meaning(s) of a word, word meanings have a tendency to remain fuzzy. Most frequent words can mean many a thing, but dictionaries rarely agree on the number of word senses and their definitions for any
6
Meaning, Discourse and Society
given word. The reason it is quite impossible to standardise the meaning of words is that many single words in isolation have no fixed meanings. It very much depends on the contexts in which they are embedded, on the words we find to the left and right of the word in question, as to how they contribute to the meaning of a sentence. Indeed, from a semantic perspective, the word as the basic unit of language has been shown to be a rather poor choice. Single words are notoriously ambiguous. Yet we language users normally have no problems with them, as the contexts in which they are embedded tend to tell us how we should read them. The unambiguous units of meaning that we intuitively make out, when we listen to someone speaking or read a text, are often larger than this chain of letters between spaces, often consisting of two or more words, which do not even have to be adjacent. Even though many of these units have never found their way into dictionaries, we use them intuitively as elements of the discourse that we have encountered before, in the same or a similar form. These units of meaning create their own regularities, but not a system. A few lines above I have used the phrase quite impossible. Originally I had written sheer impossible. There is no rule that tells me that one does not use this phrase. After a friend pointed out my ‘mistake’, I checked it on Google. There are about 2,000 hits for it (as compared to c. 600,000 for quite impossible), and most of them are translated from other languages. There is no rule, and certainly no law, that tells me sheer impossible is wrong. How systematic is language? While the founder of the modern discipline of linguistics, the Swiss linguist Ferdinand de Saussure, did not distinguish explicitly between rules and laws, the discovery of language laws has been the main objective of much of theoretical linguistics since the second half of the twentieth century. These laws would have to be universal and valid for all languages; we would have to follow them, largely unaware of them, just as the apple knows nothing about the law of gravity that makes it fall to the ground. Like the apple, we language users would be in no position to violate the rules. I am somewhat unconvinced concerning such laws. Apart from the (obvious) truism that language, at least spoken language, is organised in a linear fashion, there seem to be few candidates for universal language laws that cannot be violated. The jury is still out on this fascinating question. I will revisit it in greater detail in the next chapter. There are few rules concerning the meaning of words or phrases, or what I call units of meaning. Of course, as all academic teachers keep complaining, students often misuse words, particularly rarer words that are part of a specifically academic register. It seems that students are not aware of how these words are normally, perhaps even regularly, used. They may have previously encountered them only a few times, and hardly ever in a situation where it would have been feasible to ask the speaker for a paraphrase. Just encountering a word once or twice, or even more often, is rarely enough to understand its meaning. The user has to be told how to use a word and what it means (which is not always the same).
Introduction
7
There are a number of language theoreticians who would disagree with me concerning these ideas. This is why a discussion is needed. This book is intended as a contribution to a dialogue on meaning. The different points of view allow us, in an exchange of ideas, to pick and choose among the notions that have been used by either side. By recombining and rearranging the elements of which they consist, we will find formulations that will differ more or less from what was there before. The result will be innovation, a new way to look at meaning. It is only possible if the disputants focus on the differences between existing views more than on what may link them together. Only if the dialogue on meaning speaks in many voices and clearly expresses differences, only if it is truly plurivocal, can we hope to achieve some progress. Language in the mind? In the first part of this book I will try to point out what is wrong with a theory which claims that the meaning of a word (or a phrase) is the mental concept to which it corresponds. This is, as I see it, the foundational stance of all cognitive linguistics, in spite of the fact that some varieties of this theoretical framework are more interested in the construction of ‘cognitive’ models and less in the factuality these models claim for themselves. Other ‘cognitive’ schools increasingly delegate the responsibility for the ‘true’ nature of these concepts to the neural sciences. But mental, or cognitive, concepts, or representations, are a staple fare of many twentieth-century language theories that proclaim that there is a mind endowed with a mechanism that processes linguistic input and generates linguistic output. Against this view, I raise two objections. Firstly I insist that meaning is symbolic. What a word, a phrase or a text (segment) means is something that has to be negotiated between the members of a discourse community. Unless I am told, a word means nothing; it is not a sign. Meaning is not what happens in our individual, monadic minds; it is something that is constructed within the discourse. Of course, each of us has individually learnt what words mean. But unless we actually use them in our contributions to the discourse, this passive knowledge will leave no traces. Second, we know nothing about the mind, and there is no way to access what may be in it. Nobody has ever seen a mind. A mind is something we have successfully constructed as an object of the discourse, and as such it serves many good purposes. But we have no way to find out whether minds occur as objects of the reality outside of the discourse, and they are not even objects of a discourse-internal reality shared by everyone. We use the construct ‘mind’ to give a name to a virtual interface between our body and our symbolic, meaningful behaviour. Mental concepts, even if they existed, would not be accessible to any empirical investigation of meaning. Of course we can build one model after another of mental concepts. But they will never be more than models.
8
Meaning, Discourse and Society
Language in the discourse community In the second part of this book I want to develop my view that meaning can be found nowhere else but in the discourse. I do not want to distinguish between meaning and knowledge. The word globalisation means all that has ever been said about the discourse object ‘globalisation’. Meaning and knowledge are public. Public knowledge, as I see it, does not have to be true. ‘Truth’ refers to a reality out there, outside of the discourse. But globalisation is not something that we can see, hear or otherwise feel. What I experience is that I can now access my email in an internet café in almost any remote valley in Papua New Guinea, and equally that the gap between the poor and the rich is constantly growing. We can agree to calling it globalisation. But would such an agreement make it a ‘true’ statement? Whatever people say about globalisation, it is neither true nor false. Other people may like it or find it stupid. If it is repeated by others, it will become part of the meaning of the lexical item globalisation. Thus meaning, unlike ‘truth’, is never final; it is always provisional. Whenever we are unhappy with the way someone uses a word or a phrase, or with a longer piece of text, we will open a discussion about its meaning or, more often, about the discourse object for which it stands. We may not be able to convince our interlocutor of our view. But by talking about the word, or the object (which is the same for me), we will jointly come up with a new interpretation of it that will be added to its meaning or our knowledge about the object, and thus modify it. The discourse is the place where new texts react to existing texts, by discussing, questioning or averring what has been said. The discourse has a diachronic dimension and it goes on forever. New interpretations reinterpret earlier interpretations, and new knowledge is constructed in addition to existing knowledge. We are not at the mercy of the reality the discourse presents to us. Together, we have the power to change it. This is why I want to look at the discourse as the collective mind of the discourse community. Unlike the monadic minds of individual people, it is open to our investigation. Linguists are in no way privileged. The discourse is at the disposal of all of us. We all can at least check what Google kindly lists as the meaning of ‘life’ or of life. Linguists are not experts in meaning or knowledge. They do not know more about the meaning of a lexical item than any other member of a discourse community. Since antiquity, it was the trivium of the artes liberales (grammar, rhetoric and logic) that was seen as dealing with language. In the course of the Continental university reforms at the turn of the nineteenth century, the home of these language studies was seen as part of the sciences humaines, or the Geisteswissenschaften. David Hume called them the ‘moral sciences’. But this categorisation was to change. Ever since the end of the nineteenth century, linguistics (like psychology) has been keen to be counted among the hard sciences and to lose its stigma of
Introduction
9
belonging to the human sciences. As the positivists saw it, only the hard sciences, such as chemistry or physics, could lay claim to ‘true’ knowledge, based on brute ‘facts’. Only the hard sciences were seen to be dealing with the reality out there. They were considered important because what they discovered could make a valuable contribution to society; they embodied ‘progress’. A new interpretation of Shakespeare, a new dictionary of ancient Greek, a new look at a painting by Piero della Francesca might enchant a few connoisseurs but could not contribute to the technical revolution. While it is true that some philologists in the nineteenth century professed to have discovered laws of the language system which could match the laws their colleagues in the natural sciences kept discovering, the majority of them set out to interpret the textual remnants of bygone ages, in all their diversity, inconsistency and unruliness. They were discourse linguists avant la lettre. In their thesauri of Sanskrit, Hebrew, Greek and Latin they showed how much the meaning of a word depended on the context in which it was situated. They were more interested in interpreting texts than in language laws as stringent as the second law of thermodynamics. It was this interpretive agenda that Ferdinand de Saussure opposed so fervently. To demonstrate the relevance of linguistics, he set out to rebrand it. Should the mind be an object of scientific enquiry? To establish linguistics as a hard science, however, required one important move. It necessitated the exclusion from the research agenda of people’s intentionalities, of their experience of reality and of their interpretation of what their reality was about. In order to be accepted, linguists had to present their object of study as a system that was independent of unpredictable human intervention. ‘In their [Saussure’s immediate predecessors’] eagerness to achieve scientific status for their linguistic studies by assimilating the discovery of linguistic patterns to the discovery of laws of Nature, they were more than content to sacrifice any distinction between rules and regularities’ (Harris 1987: 109). But there is a categorical difference between language as a system and the kind of systems we find in the natural sciences. As far as the latter are concerned, we observe and describe them from the outside. The entities and their interrelationships that make up the Newtonian system of gravity, the apple and the force that makes it fall to the ground, are not affected by our observation, our deliberations, our scientific discourse, our interpretations. Mainstream linguists in the twentieth century, in their quest for the language system and in their yearning for recognition, insisted on studying language as a zoologist would study the communication system of ants. They postulated an unbridgeable abyss between the object of observation and the observer. For Chomskyan and cognitive linguistics, the language system is something outside of the
10
Meaning, Discourse and Society
reach of language users. But we as linguists can never escape the fact that we ourselves are language users. All language users engage in the generation and interpretation of utterances. Meaning is an integral part of language. Once we take meaning away, language ceases to be symbolic, to be language. Without intentionality, without the property of an utterance to be about some content that can be discussed, language is no longer language. Apples do not have intentionality. They do not mean. Gravity is a force independent of what we say about it. But language is different. We can never escape the discourse in which we are imprisoned whenever we negotiate the meaning of what is said. All we say becomes meaningful only through the discourse. Thus all we say has an impact on the discourse. The discourse is self-referential. This makes language something that belongs to a category different from that of gravity. There is no Archimedean point, no discourseindependent vantage point, from which we can describe language, as we can describe a falling apple. It is true that we will never gain access to another person’s individual experiences. Yet we do not fare better with our own experiences. I experience myself and the world in which I am, the things I do or don’t do, directly and immediately, unmediated. But this immediacy is lost once the moment has passed. Not only am I unable to recall a past experience at will in some future situation, I also cannot let someone else share it. All I can do is to reflect on it, or to give testimony of it, to myself or to other people. This involves, however, translating, or rather re-creating, such an immediate experience into symbolic content, into a representation of an experience, into something that I can recall on a future occasion, and into something whose meaning can be collectively negotiated, something that has inexorably become an indirect and mediated account of my ‘raw’ experience. Intentionality, as I see it, leads us away from ‘raw’ experience; it can be described as being conscious of having experiences. Intentionality is the conscious creation of symbolic content, and the reflection on such content through the act of interpretation. The way in which I re-create my experience, turn it into symbolic content, and interpret it, is unpredictable. There is no mechanism for doing it. Language is not a system for turning my perception of myself or of the world outside into a representation. There are no rules that I follow without being aware of them, as the apple follows the law of gravity. Intentionality is outside of the remit of the hard sciences. We will never be able to say anything about anyone’s immediate, unreflected experiences, not even if they are our own. But there is another way to look at intentionality. The way in which a person transforms their experience into symbolic content will always remain hidden. But how we talk about experiences, assign meaning to them and interpret them, is something that happens inside a discourse, whether an imaginary one in our heads or a real one within a
Introduction
11
discourse community. It is the intentionality of this community that we can study, by analysing this discourse, namely the entirety of texts that have been exchanged and shared between the members of the community. The discourse is real, and we are not prevented from viewing it as a system. It is not a mechanism, though, delivering predictable results, telling us what can be said and what not. Rather it is a system that keeps creating itself, an autopoietic system, a system like Darwinian evolution. This is not what de Saussure or Chomsky had in mind. For them, the mind is an engine that can generate correct, grammatical, sentences whose meaning is equivalent to their semantic representations. While we have to give up the hope of ever gaining access to the working of the individual mind, we can without too much difficulty examine the collective mind in the form of discourse. We can interpret what others have said. Linguistics belongs to the human sciences as much as to the hard sciences. The discourse community constructs reality The reality we experience is not unmediated reality itself but the reality constructed in the discourse. The colour red I see is not the colour red as it may or may not exist in reality out there. It is the colour ‘red’ as it is constructed in the discourse of the English-language community, and this red is in some ways different from the ‘erythros’ of the ancient Greeks or the ‘chì’ of Mandarin. The discourse object ‘globalisation’, which can make life so hard for us, is not something that really exists out there in some discourse-external reality. It is a discourse construct that works well as an argument when we are told why we have to tighten our belts. Only by reading newspapers, not by looking at the world outside, will we find out who belongs to the discourse object called the ‘civilised world’ and who does not. When we look up to the elevation over there it does not indicate whether it is a hill or a mountain. Nature does not come with categories. We have to check our travel guide. The only reality we have at our disposal is a reality which is thoroughly mediated by what has been said in the discourse. For die-hard realists this amounts to unreconstructed relativism. For them, such a constructionist view implies a desire to force others to see the world though our tainted glasses. For John Searle, the American philosopher who seeks to integrate the realism of analytic philosophy with the realism of the philosophy of mind, the case is clear: I have to confess, however, that I think there is a much deeper reason for the persistent appeal of all forms of anti-realism, and this has become obvious in the 20th century: it satisfies a basic urge to power. It just seems too disgusting, somehow, that we should have to be at the mercy of the ‘real world.’ … This is why people who hold contemporary versions of antirealism and reject the correspondence theory of truth typically sneer at the opposing view … [T]he motivation for denying realism is a kind of will to power, and it
12
Meaning, Discourse and Society
manifests itself in a number of ways … [F]orms of postmodernism, deconstruction and so on, are easily developed, having been completely turned loose from the tiresome moorings and constraints of having to confront the world. If the real world is just an invention – a social construct designed to oppress the marginalised elements of society – then let’s get rid of the real world and construct the world we want. (Searle 1998: 17ff.)
Indeed this is the conclusion that I will arrive at. Like John Searle’s real world, the worlds of all the other realists are, as I see it, hardly more than a figment of imagination. We experience it, but even our raw experience of it is already formed by the discourse, by the view it provides of reality, by categories, such as ‘hill’ and ‘mountain’. Thus we can only communicate our experiences in the representations offered to us by the discourse. The only objects available to us, the only objects on which we can exchange our views, are the objects we find constructed in the discourse. But we are not voiceless. This discourse is, at least in principle, at our disposal. Though the media want to make us think otherwise, there is no ‘right’ or ‘wrong’ way to talk about discourse objects. We do not have to obey an authoritative, impenetrable and unchangeable discourse system telling us what we are allowed to say and what we are not. We are, at least ideally, the masters of our discourse and the reality it presents to us. Nothing can prevent us from negotiating and renegotiating this reality. As soon as we refuse to accept the reality we are fed by the content merchants and their text producers, we are free to collectively design a reality that serves our interests and not the interests of those who want to exercise power over us. Totalitarian societies may not want us to publicly question the orthodoxy of the system. But censorship can always be circumvented. The only healthy discourse is a plurivocal and anarchic one. There will, I hope, always be a clash of views. As John Stuart Mill argued in On Liberty, progress and innovation need an unrestrained discourse. Whatever their intentions may be, language theorists committed to the cognitive agenda present us with a blueprint that excludes our collective intentionality from a role in the construction of meaning, reduces meaning to the workings of a mental language mechanism over which we have no control, and replaces our unfettered creativity by forcing language into a straitjacket of (their) rules. Strange as it may seem, it is the ‘antirealist’ view of social constructionism, a view that our reality is what we have constructed (or find constructed) in the discourse, that empowers us to construct a world we may find worth living in. The sixteen topics of this book This book has sixteen chapters and a conclusion. Part I consists of the first seven chapters. Chapters 8 to 16 make up Part II. Here is a short preview of the content.
Introduction
13
Chapter 1 reminds us of the cognitive turn, as it took place in the middle of the twentieth century, following the demise of the previously prevalent paradigm of behaviourism with its impenetrable black box. A key reason for behaviouism’s sudden if untimely demise, and the unstoppable success of the cognitive paradigm, was the appearance of the computer, an electronic brain that would, we were told, soon learn to think and, perhaps not much later, surpass human beings with its superior problem-solving faculties. As I will show, the cognitive sciences used the computer as the blueprint for their model of the mind. They envisaged the mind as comprising both a hard-wired (innate) system of laws, and (acquired) programs containing commands or rules. Minds and computers were seen as functionally equivalent mechanisms that could carry out similar complicated and highly sophisticated tasks. How closely the cognitivists were following the developments in computer technology became evident when the first neural networks were implemented. Suddenly all cognitive scientists became convinced that the human mind likewise had a connectionist structure. Interestingly, the relationship between the computer scientists and the cognitivists also worked the other way around, too. The cognitive models of the mind with their mental concepts were soon sold back to the computer scientists as recipes for the expert systems of artificial intelligence and for machine translation. The early success of the cognitive approach is frequently attributed to Noam Chomsky, the charismatic prodigy of the Massachusetts Institute of Technology. His agenda was the reification of a language faculty as a key component of the mind. He owes his breakthrough to his 1959 philippic against B. F. Skinner’s behaviourism, in which he convinced the larger part of his audience that learning a language cannot be explained by the workings of stimulus and response. In the normal course of language acquisition, there would never be enough input to account for the speed with which young children master a complex syntax plus a huge vocabulary. Instead we have to assume some more or less hard-wired mechanism in our minds, he claimed, a mechanism everyone is born with, that is turned on by verbal interaction in early childhood. This mechanism triggers the release of hard-wired, innate conceptual knowledge. Chomsky’s agenda promised to give us, for the first time, access to the impenetrable black box which the mind was for the behaviourists. The mind quickly became the central arena not only for Chomskyan linguistics but also for other cognitive approaches to the study of language. The models that these newly founded schools were keen to develop were to reflect the workings of the human mind, or even to be more or less isomorphic replicas of them. Common to these models was the idea that the words of natural languages, different as they may be, such as English apple, French pomme or Spanish manzana, all correspond to a mental concept that is the same for all human beings, that is, in a word, universal. It is a view that disregards entirely
14
Meaning, Discourse and Society
the role the discourse plays. In the cognitive paradigm, apple, pomme or manzana are natural language expressions of an innate mental concept. Negotiations of the meanings of these words have no impact on the mental concept, which is predetermined by the innate organisation of the mind. What the cognitive turn brought about was the hypostasising of a mechanism that turns thought (represented by mental concepts) into language and language into thought. In Chapter 2, I will dig into the history of the idea that the mind is the seat of meaning and the arbiter of knowledge. For this axiom is not as new as cognitive linguists tend to believe. It has been around for a long time, even before René Descartes firmly institutionalised it as the one item that makes people uniquely human. As always, the tradition started with the ancient Greeks. For Aristotle, the words of spoken language already referred to mental concepts. This was reiterated time and again through the Middle Ages, after the Arabs in Spain had given their Christian brothers access to the ancient texts. I will refer to Anselm of Canterbury as a medieval representative of this tradition. More interesting was William of Ockham, who actually developed this tradition further. He tried to find solutions to some of the problems that the theory of a lingua mentis inherently poses. A contentious point was whether we would find synonyms among the mental concepts. It separated Platonists, who believed that concepts correspond to ‘natural’ ideas, from the nominalists, who held that the way in which we cut up the world into things is arbitrary. These were questions that again, after the cognitive turn, began to be discussed in the suddenly very popular philosophy of mind. In Chapter 3, I will explore in more detail some conceptualisations of mental concepts as we find them in the philosophy of mind. The starting point is, of course, Jerry Fodor’s very influential book The Language of Thought that, when it appeared in 1975, inspired a huge amount of subsequent work. As he sees it, there has to be a corresponding concept for practically every word, with the sole exception of compounds. Fodor had been a student of Noam Chomsky, and indeed their views on mental concepts were then largely indistinguishable. However, Chomsky was more focused on the question of the innateness of these concepts, and he discussed at great length with the philosopher Hilary Putnam whether we are really to assume that even the concepts for words such as bureaucracy and carburettor were somehow from the very beginning present in the make-up of the human mind. For Putnam such an idea was outlandish, while Chomsky saw it as a serious possibility. While Fodor did not contribute to this discussion directly, his claim that we could not acquire our first language unless we were born with a hard-wired language of thought puts him safely in Chomsky’s camp. His idea of a language of thought has more recently been popularised by Steven Pinker’s 1994 book The Language Instinct, where it is called mentalese and otherwise shows a remarkable similarity to basic English.
Introduction
15
For Anna Wierzbicka, on the other hand, different languages have their own mental concepts. What is innate for her is merely a small set of some fifty basic building blocks, so-called semantic primes, that can be used to configure all the mental concepts we find in any language. Implausible as such a claim seems, there are also the questions that concern not only her semantic primes but all the concepts of any language of thought: how can we find out about them, if they exist only in a form that is independent of any of the natural languages on which we have to rely if we want to communicate? If they embody content, is that content symbolic? If yes, it must have been negotiated; if not, then it begs the question whether it makes any sense to talk about non-symbolic content. European structuralists had also been developing models for turning the meanings of words into concepts represented by neat formulae. As I will show in Chapter 4, structuralist theories developed, for instance, by Algirdas Julien Greimas, and somewhat later by Bernard Pottier, had word senses (sémemes) broken down into distinctive traits called sémes. But these concepts were no more than abstractions, ways to represent the semantic differences between word senses, without any ontological claims. In the work of Manfred Bierwisch we can see how these semantic features gradually morphed in to something real: he began to see in them ‘rather deep seated, innate properties of the human organism’. Without this claim, without the reification of mental concepts as something actually residing in the mind, cognitive linguistics would indeed cease to be cognitive and just be uncontroversially structuralist. The interplay between the study of the mind and the computational work on information processing is also visible in the role concepts play in both environments. One of the early pioneers in machine translation, Alan Melby, eventually came to see the futility of this approach. In Chapter 5, I will discuss my contention that ‘concepts’ can never fully match the meaning of what is said, whether they are constructs in the realm of mind or in the realm of language processing. One problem is that up to this day, concepts are still seen as entities that retain their function (meaning) in isolation, only connected to each other by the relationships between all concepts, as they obtain within a conceptual ontology. But these entities do not map the meaning of the words of which an utterance is made up. For what these words mean is to a large extent determined by the context in which they are embedded. Semantic representations based on single concepts in isolation are never adequate. A more serious problem is that these concepts, whether mental or terminological, are supposed to be language independent, but translatable into all natural languages. But can this idea work? The French equivalent of globalisation is either mondialisation or globalisation, depending on the context. It is hard to see how a language-neutral conceptual representation would take this into account. For our modern language engineers, who are increasingly replacing the traditional linguists in what is sometimes called human language
16
Meaning, Discourse and Society
engineering, concepts are the pure essence of word senses, without contextual contaminations, fuzziness or ambiguity. They are seen to represent ‘the abstract meanings of words, whatever those may be’. It would be stupid not to admit that this approach actually works in environments in which natural language is replaced by a controlled language in which each ‘word’ can be used only as a term, that is in one unequivocally defined way, and in which sentences can contain only one conjunction and sentence length is strictly limited. Aircraft maintenance manuals are successfully translated by machine translation systems. Unfortunately these systems do not fare quite as well when translating Shakespeare. In Chapter 6, I present in detail what has been said about mental concepts in different camps of cognitive linguistics. In particular, I explore the kind of two-level semantics underlying Dan Sperber and Deirdre Wilson’s relevance theory. For them, hearers have in their minds a mechanism that allows them to infer, from the highly ambiguous and fuzzy meaning that they ascribe to verbal utterances, what the speaker really wanted to say. Many mental concepts, they claim, are ineffable and therefore have to be filled in by this mechanism. More recently, Sperber goes as far as implying that understanding an utterance correctly involves some sort of mind-reading. Stephen Levinson, too, is interested in the difference between our thoughts and how we express them. He wants to find the line that separates what he considers a universal mental representation of a thought from what is determined by contingent languagespecific rules for encoding it, rules that have to be learned in the process of language acquisition. For him it could well be an accidental idiosyncrasy that the Mayan language Tzeltal has only one word for the two concepts ‘blue’ and ‘green’, and that French has introduced the complication of tu and vous for the straightforward concept of ‘you’, while English is in these respects more a mirror of universal mental representations. This shows, I believe, a basic dilemma common to all work on mental concepts. Linguists tend to take the way in which their native language cuts up the world out there as basic and universal, while other languages show certain distortions. The third example I present will be the school of Ronald Langacker. For him, the task of linguists is to provide structural analyses and explicit descriptions of thoughts and concepts. Semantics has to ‘elucidate the structure of the complex conceptualizations evoked by linguistic expressions’. In the end, however, these representations do not have to be visited in the mind, but have to be understood as ‘neurological adjustments’ of which the language user is unconscious. Like all the other cognitive linguists and philosophers of mind, Langacker, too, fails to come up with a single example of a mental concept. In Chapter 7, I will briefly recapitulate two concepts of the mind as we find them in the philosophy of mind. One is the mind that we owe to the spirit of the Enlightenment: the mind that turns us into autonomous, moral and rational
Introduction
17
human beings who can recognise truth and act morally. This is the intentional mind, the mind that experiences itself and the mind that reflects. It is the mind that bridges the gap between a monadic, non-symbolic ‘feel’ and its symbolic interactions with other minds. In cognitive linguistics (and in the cognitive sciences generally) this model plays only a very minor role, if any. The standard model of the mind is still very much the mind that is analogous to a computer, in which there are certain hard-wired laws that correspond, for instance, to Chomsky’s universal grammar and semantic universals, things that we find in all languages and over which language users exercise no control. There are also things like computer programs that are responsible for translating universal grammar into the respective language-specific surface structures of natural languages. Together they are the language system, realised in a mechanism that processes thought into utterances, and utterances into thought. But this model of the mind as such a computational device has two flaws. It cannot deal with meaning, because intentionality, aboutness, the awareness of the content of what is said, is excluded from the model as a merely ‘supervenient’ feature. Instead, it reduces semantics to algorithmic operations, disregarding the symbolic content of language signs, i.e. the need to interpret them. The other, even graver, problem is that the mechanism cannot successfully be made an object of ‘scientific’ investigation. We have no direct access to the mind, and indirect conclusions (based on psychological tests) are notoriously arbitrary. In this situation, more and more philosophers of mind and cognitive scientists turn to the neural sciences, linguists being no exception. But there are, I think, strong arguments showing why neurolinguistics will not lead us to a better understanding of meaning. I will analyse Vittorio Gallese’s and George Lakoff’s paper on a neurolinguistic take on the concept of ‘grasp’, and show that the meaning of grasp, its symbolic content, is irreducible to non-symbolic neuron clusters. I will then take a close look at some of the arguments between the neuroscientist Maxwell Bennett and the Wittgenstein expert Peter Hacker on the one side, and the American philosophers of mind Daniel Dennett and John Searle on the other. The conclusion that I draw from their exchange is that the neurosciences cannot provide the answer to the conundrum of meaning. No matter how much we zoom in on the neuronal tissue, or how exactly we can determine which synaptic connections are charged when clusters of neurons are firing, none of these investigations will enlighten us about the meaning of grasp. The brain is not symbolic. It is related to brain-external stuff by physical or chemical, that is causal and material, features. It can store symbolic content, but it does not interpret it. Chapter 8 introduces the second part of the book, in which I attempt to sketch a solution to the two main problems: how we can know what a word, a phrase, a brief text segment or a full text means, and how we can provide the evidence for what we take to be this meaning, in such a way that it can be accepted by others.
18
Meaning, Discourse and Society
This will also necessitate a discussion of the relationship between the discourseexternal reality and what is said in the discourse. The main part of this chapter will be devoted to a discussion of the notion of discourse. This has been no doubt one of the most overused words in the intellectual discourse of the last fifty years or so. For Michel Foucault, discursive formations had little to do with their linguistic surface. For him, it was all content held together by the laws, rules and regularities he saw at work in the discursive practice. It is this inherent order that maintains the unity of the discourse. A similar concept of discourse is found in the relatively new field of critical discourse analysis. Here, the discourse is viewed as a social practice that transmits ideology, as a distorted view of reality, because it reflects the power relationships obtaining in a given society. For me, the discourse, and not the world out there, is the only reality to which we have direct, unmediated access. It is the entirety of spoken, signed or written utterances which have at least one addressee. What I say when I am alone has no effect on others. It is not something other people can react to. The discourse at large is the entirety of all the contributions made ever since people started communicating with each other using language, indeed any language. This all-encompassing discourse is not only infinite in size but also largely unavailable (because lost and forgotten), and therefore, of course, not a suitable object of inquiry. We can only have access to texts that have not been lost. Practically all spoken language disappears as soon as it is uttered. Only the few samples that have been recorded can be revisited. Most of the texts that are available for our investigation are written texts. Some of the texts may be handwritten or typed, while many of the texts we will end up scrutinising will have been printed and/or have been entered in electronic form. The web is a tiny section of this written discourse, and most of what we find there is no older than a few years. Increasingly it gives us access to recorded speech, as well. It also covers many languages. But, in spite of its shortcomings, the texts we find there (or in traditional archives and libraries, or in a box in the attic) are real. They are original documents, or copied versions of original documents. These texts are real language data. They are available for our investigation. They can be exchanged and shared. For most practical purposes of both linguists and lay people, this discourse at large will be cut up into smaller, specific discourses pretty much ad libitum and in line with our research question. We can define a discourse that consists only of English texts, written or otherwise recorded, in the year 1992. We can exclude non-British texts. All kinds of decisions have to be made for a full definition of such a special discourse, a definition that tells us unequivocally for each text if it belongs there or not. What about texts written in Britain by non-British authors? What about texts written abroad by British authors? What about English translations of foreign texts? Even such discourses will be too large to make them a reasonable object of research. Samples of them need to be compiled, in so-called text corpora.
Introduction
19
More interesting are special discourses of texts that have a common topic, although these also tend to be infinitely large. We might, for instance, want to find out about global warming. Thus we can decide that we only want to look at newspaper texts, and we may choose, say, three British, three American and three Indian English-language quality national newspapers. We can agree that we select texts from between 1997 and 2007 in which there is at least one occurrence of global warming and/or climate change. The question is now how many discourses this will give us. That is up to the investigators, but not only to them. The selected texts themselves will also have their say. We have to take into consideration to which other texts they refer, and if they accept or reject what is said in them. In any case, discourses can overlap. A given text can be relevant for any number of them. In Chapter 9, I will try to argue the case that it is language that turns a group of people into a society. Of course, primates and other animals also behave in ways that we can compare to human interaction. But what distinguishes, for instance, grooming within human society from grooming among chimpanzees is that people who groom and who are groomed can make themselves aware that what they are doing is grooming. Labelling a kind of behaviour as grooming requires the availability of categories only language can supply. Humans alone among primates can discuss if what they are doing is grooming or perhaps cuddling. This is not something chimpanzees are known to do. Discourse can assign a meaning to interpersonal behaviour, can make it symbolic. This means that if people behave in a certain way, they can normally tell us what it is they are doing. It is true that primates as well as humans can distinguish ‘us’ from ‘them’. But for primates such a differentiation is part of their immediate experience and not something up for negotiation. Members of the British middle class, on the other hand, can discuss who they regard as ‘them’: the hedge-fund managers who deplete their pension funds, the French who indulge in a 35-hour week, or the asylum seekers who live on the taxes they pay. The ways in which we interact with each other become symbolic by our assigning a meaning to them. It is this ‘sense’ that is the essential characteristic of society. Society is what happens between people and has a meaning. The people themselves, with their individual minds and their individual experiences, however, are not part of society. They are outside, loosely connected to society through the symbolic input they provide, through the texts they contribute to the discourse, and through the effects these contributions have on them. Whatever may be happening inside these individual, monadic minds, is not what explains the ‘sense’ that is continually created inside and through the discourse. As long as ‘sense’ is not expressed, it has no impact. There is no reason to be interested in it in our individual quest for meaning. Indeed, when we try to find out what a particular contribution means, we will not find it in the minds of the speaker or their audience. We will only find it in the intertextual links that connect any
20
Meaning, Discourse and Society
contributions to the other contributions of the discourse. This is how I interpret the systems theory of the German sociologist Niklas Luhmann. For him the discourse is an autopoietic machine that generates ‘sense’ through its self-referentiality, through the ways the texts refer to each other. Luhmann’s ‘sense’ is what I call the intentionality of the discourse as the collective mind of a discourse community. A person’s individual intentionality only concerns that person, and has an effect on society only if turned into a symbolic act. But while we have no direct access to our intentionality, to the way we experience ourselves, society has its own, even if derived, intentionality, namely the symbolic content of its discourse. This intentionality matters because it is tangible and it can be discussed. Once we replace the individual mind of the monadic person with the collective mind of society, the meaning of the contributions to the discourse is at our disposal, whether we are linguists or other members of society. When it comes to meaning, linguists are not in a privileged position. In Chapter 10, I will discuss the relationship between the reality out there, the shared reality, and the primordial speech situation. The way we use the discourse to come to terms with the world outside is the primordial speech situation. There is a group of people, and they engage in verbal interaction. They are in a specific situation, for instance enjoying the calmness of the evening, sitting around the fireplace in front of their cave. Suddenly one of the older men cries: ‘Look, a mammoth!’ None of the young lads has ever seen one. As a result of extensive hunting in previous generations, they are pretty much extinct. ‘Where?’ they ask, and the old man points to a kind of dark spot out there at what they have all learned to interpret as the edge of the forest, half hidden by leaves. ‘Over there, at the edge of the forest’, he says, and points to it. One of the forward young men asks the white-haired gentleman: ‘How can you be so sure it’s not an elephant?’ ‘Because it’s woolly’, says our old friend. What has happened? Using their sense of vision, they registered the stuff out there, the discourse-external reality. But this stuff had already been structured in previous primordial speech situations. All members of the group have already learnt to distinguish edges, including edges of the forest. They have even learnt how to distinguish woolly things from things that are not woolly, though I have no idea how they could have learnt it in the absence of sheep. They have learnt that in the reality presented to them in the discourse it is assumed that stuff normally does exist even if hidden from their view. They have also had elephants pointed out to them on several occasions. And they know that ‘over there’ can mean relatively far away in the direction in which a hand is pointing. What they now add to their knowledge in this specific primordial speech situation is a visual image which they can attach to the word mammoth, which they have heard old people talk about but which they never have seen. There were no paintings in the caves they knew. What they have learned in this
Introduction
21
situation is to distinguish elephant-stuff from mammoth-stuff. Mammoth-stuff is a lot like elephant-stuff, only more woolly. Certainly the mammoth has been an object of their discourse before, just as angels (‘real’ angels, not angel images) are objects of our discourse. They have heard stories about how good a young mammoth used to taste and how much more tender it was than the meat of a young elephant. But so far mammoths have only been objects of their discourse, objects that came up in what old people said, as virtual as dragons. Before their first sighting of such a specimen, it had not yet been pointed out to them as a part of their shared reality. It is this shared reality that allows us to move and behave inside the discourse-external reality without constantly getting into trouble. It teaches us that it hurts to kick a rock. Of course, primates and feral children also do not kick rocks. But they do not know that there is a discourse object ‘rock’, a discourse state ‘hurt’, and a discourse act ‘kick’ (what Bishop Berkeley calls ideas), and that it is entirely conventional to say that kicking a rock hurts. Of course, Dr Johnson is right, too: there is a reality out there. But the only reality we can negotiate in our verbal interactions is not the reality out there but the shared reality, constructed within primordial speech situations. Chapter 11 explores the differences between oral and literate societies. It was when people first started to communicate by signs that the discourse began to evolve. They may have communicated previously by non-symbolic behaviour, by involuntary facial expressions, by grooming and by involuntary gestures. Signs are different. Signs presuppose someone who signs and an addressee. Signer and addressee believe they have to come to agree on the meaning of the sign. Time and again people have been told the meanings of words and phrases. But we do not have to accept what we are told. At any moment, we are free to renegotiate meaning. As long as we find others to side with us, a new twist has been added. Meaning is always provisional. Writing had a profound effect on the discourse, as the collective mind of society. Only writing makes us aware of the fact that there is a discourse, and thus it brings about a fundamental change in the discourse. Writing makes it possible to refer to something that has not been said in our presence but written by someone we do not know at some undisclosed location and potentially a long time ago. Only what is written can be preserved. Writing makes it possible and advisable to note down clever ideas and to let other people know about them. Complicated ways of thinking could evolve because suddenly it made cultural or economic sense to have these thoughts. Writing enabled us to revisit thoughts we had jotted down. We could even start thinking about what we had read. Reflection ceased to be seen as a waste of time and began to be regarded as the foundation of all progress. It was the invention of writing that made the production of ideas marketable. Once we were surrounded by written texts it became more obvious that we were not just bringing to light things hidden in
22
Meaning, Discourse and Society
our own memories, but bits and pieces that we had encountered in previous texts, and that, with little effort, we now could turn them into a new text. Only in a written discourse does a linguistics become possible that compares texts and studies the intertextual links obtaining between them. The misconception attributed to Plato, and frequently repeated to this very day, namely that writing is no more than a way of recording what has been said orally (and what we are too lazy to memorise) makes us blind to the huge gulf separating oral from literate societies. Our discourse, the discourse to which this book is a contribution, is therefore largely the discourse of a literate society. While we continue to converse orally with each other, everything that counts has to be written down. I may tell you that I own this piece of property, but if you contest my claim I will have to produce written evidence. Writing is durable, though not permanent. A written text conveys the image of authenticity (though it can be a forgery like the infamous Zinoviev letter that made the Labour party lose the election in 1924) and, as such, it seems infinitely superior to the human memory, with its deliberate or unintentional propensity to add a spin to the few things we remember at all. Writing, we are brought up to believe, can be relied on. Without writing, we could not conceive the concept of history. Even in the Middle Ages, when writing was restricted to a few professions, people had not yet learned to distinguish between historical events and legends. The chronicles, written centuries after the events, relied on people’s memories and consisted of stories that were transmitted orally. Their truthfulness was not an issue in those days. It takes a literate society to insist that a book on historical events must be based on reliable evidence in the form of documents. There are, as far we know, no surviving eyewitnesses of the Battle of Trafalgar. But there is an abundance of records. The origin of literacy, however, was the invention of writing as an occult and arcane art. The few people who mastered it were in most cases members of the priest class, and they guarded their secret closely. Writing might well have been presented originally to the astounded public as the direct intervention of the deity, as we still find it much later in certain oracles. It was propounded as intricately linked with reality and therefore necessarily true, which is how the holy scriptures in many religions are still seen today. The invention of the printing press made writing ubiquitous. It also established the author as the creator of a written text, and with it came the necessity to distinguish carefully what was written by that person, and what was quoted from other texts. Today, the concept of intellectual property rights (normally not owned by the authors but by anonymous corporations) pervades our market-driven society. Plagiarism is a failing anyone who writes can hardly avoid. It is due to the limitations of our memory, particularly as we grow older, that we often cannot remember the source of an idea that comes to our mind. For
Introduction
23
the most part, we plagiarise unknowingly. Before the ubiquity of the printed book and its easy accessibility in a library not too far away, authors had to rely to a large extent on their memory. They may have written down a few excerpts or summaries; they may even have worked in a monastic library which had parchment copies of key theological and philosophical texts; but the lectures, disputations and examinations to which they had been party had become amalgamated and condensed to the extent that it had become impossible to reconstitute the original sources. Misattributions and misquotations were not only pervasive but readily accepted. Writing before the reification of the autonomous mind, before Descartes, was not seen as an exercise of creativity and ingenuity but of diligence. Authorship made no claim to intellectual property. What was written down were ideas that had been floating in the mostly oral discourse, and this was shared knowledge. The discourse, written and oral, the collective mind as the aggregate of all the members of the discourse community, was still common land, waiting to become enclosed. It was the personality cult of the eighteenth century that turned plagiarising into a wrong. Today we take it for granted that the new meaning, the novel viewpoint that emerges from the rearrangement and recombination of pre-existing units of meaning, is the intellectual achievement of an author’s intentionality. This conviction reflects our belief that the individual mind should be seen as autonomous and fully in charge of its actions. But why should we hold authorship in such high esteem? Bertolt Brecht famously declared that for him authorship was something irrelevant and that he was quite lax in the issue of intellectual property. Shakespeare’s plays are not diminished in their quality by our knowledge that they are sometimes the result of his collaboration with other dramatists. In Chapter 12, I will briefly discuss the contention of many linguists that the true object of their inquiry can only be speech, the spoken language as the primary form of language. This claim is famously made by Ferdinand de Saussure, but also by American structuralists like Leonard Bloomfield and Zelig Harris. But spoken language, speech, leaves no traces that can be studied. Speech is always part of a wider situation, involving, for instance, gestures, ostentation and consequently even mammoth-stuff. Speech extends beyond the narrow field on which language scientists till the soil. They have to work with data records that can be exchanged and shared, even though it is likely that one can acquire linguistic meta-knowledge in oral societies, too, if only up to a point. There must have been multilingual interpreters before writing was invented. It is not necessary to be able to read and write to learn one or more foreign languages, just as we all acquire our first language before we have learned to write. But having a discussion on the meaning of a text presupposes that this text is available for all to join in. Linguists, cognitive linguists, too, work on recorded, i.e. written or transcribed spoken language, if they work at all empirically.
24
Meaning, Discourse and Society
Whether we deal with written or spoken language, we have no way to investigate the speaker’s intentions. We can never be sure why we are told ‘There are many apples on the tree’. All the evidence we present for the meaning of this sentence has to be shareable. We can no more than surmise the author’s intentions, and even if he or she deigns to tell us what they are, we cannot be sure what intentions loom behind their telling us their intentions. If the immediate context of this sentence does not provide clues for its interpretation, all we can do to figure out the meaning of this utterance is to look up the links it has to other texts in the discourse, and surprisingly we find that this simple and seemingly unambiguous sentence, once queried on Google, can give rise to a number of different readings. Linguists interested in the discourse have to rely on real language data, on the discourse they are investigating, understood as communication, past and present, between members of the discourse community; for there is nothing else for them to investigate. The discourse is the seat of our public knowledge, including our knowledge about language rules and the meaning of what has been said. Indeed, the entire reality in which we find ourselves is the reality that has been constructed in the discourse, which is continually updated by contribution after contribution. This relationship between discourse and reality, between knowledge and meaning, is what I will explore in Chapter 13. From antiquity onwards, this view has competed with the intuitively much more plausible idea that rocks, mammoths and all the other objects are real, distinct entities or categories inherent to the world out there. Of course, there was never an agreement on the meaning of real. For Plato, the first outspoken realist, the ‘true’ signifieds were not the ordinary, contingent, things themselves but the eternal unchangeable ideas that command reality. In the Middle Ages, realists never came to terms with nominalists. At the beginning of the twentieth century, analytic philosophy brought about a host of new realisms. I will look, in some detail, at Hilary Putnam’s unique development from a metaphysical to what he has termed an internal realist. His stance is now probably as close as realism ever gets to the social constructionism that determines my own views on the reality out there. Truth is a somewhat different matter. Ever since Descartes, it is only the individual mind with its power of reason that can determine if a statement is true, i.e. if it conforms with the real world out there. Each man or woman has the autonomy to decide what is true. This was the ground which had been prepared by humanism, and on which the Enlightenment, German Idealism and European Romanticism flourished. It turned the European mind from a collective repository of ideas into a monadic, singular, personal endowment of creativity. To this day, the French and Germans have to struggle to find an equivalent in their languages for what the English call the mind, in the sense described in the OED as:
Introduction
25
[t]he seat of awareness, thought, volition, feeling, and memory; cognitive and emotional phenomena and powers considered as constituting a presiding influence; the mental faculty of a human being (esp. as regarded as being separate from the physical); (occas.) this whole system as constituting a person’s character or individuality
What the French call l’esprit or la raison and the Germans Geist or Vernunft was never, and still is not, the seat of consciousness embodied by an individual’s brain. Neither Geist nor l’esprit are, unless explicitly marked as such, an expression of a person’s individuality. Hegel’s Zeitgeist is an emanation of the Weltgeist and neither of them can be rendered as mind, nor would this be possible for l’esprit de campagne or la raison de loi. Neither in German nor in French folk linguistics would it make sense to assume that lexical meanings are in the head, in the form of mental concepts or cognitive representations. Meanings are the common property of language communities. Members of such a community can learn them from the input they receive. As Robert Solomon sees it, our current obsession with the mind as the engine of our individuality may well be the result of the sublimation of the Enlightenment in German Idealism. The invention of the autonomous person as the ultimate judge of truth and the sole creator of symbolic content can, he thinks, be attributed to Immanuel Kant: The first critique, the Critique of Pure Reason, tackles the problem of knowledge and Hume’s scepticism. At the heart of that formidable book, and the dramatic shift in Western thinking that begins with it, is an enormous expansion in the concept of the self, its scope power and richness. The primary change is a shift from the passive to the active mode, a rejection of the traditional idea of the human mind as a receptacle or receiver, to the insistence that the mind imposes its order on nature, an order that is fixed and immutable to all of us. (Solomon 1988: 27)
Descartes in conjunction with Kant; reason combined with the idea of an autonomous and isolated self: these are the sources of the modern conceptualisation of the mind, as we find it in the West. But is the idea of a collective mind, a mind that connects people, a thing of the past? Not really. Rather, it is a widely accepted theory in child psychology, and can lay a claim to scholars such as Jean Piaget and L. S. Vygotsky. Anyone who analyses how children learn to speak, how they acquire knowledge and how they do things together, cannot but accept the fact that cognition is distributed (Salomon 1993) and collaborative (Bearison and Dorval 2002). Without interaction, without the exchange of knowledge, without being members of a discourse community, human beings remain mute, devoid of consciousness and unable to plan ahead. An alternative to the view that words, or word senses, correspond to mental concepts of which the people using them are unaware can be found in social constructionism or constructivism, both of which have been around for quite some time. Peter Berger and Thomas Luckmann’s groundbreaking book The
26
Meaning, Discourse and Society
Social Construction of Reality was first published in 1966, and the two authors stress their indebtedness to the sociology of knowledge of Karl Mannheim’s Ideologie und Utopie, first published in 1929 (an English edition appeared in 1936 (Mannheim 1936)). Philosophers working in the paradigm of analytic philosophy, like John Searle in his book The Construction of Social Reality (1995), reject social constructivism (without ever referring to Berger and Luckmann’s book) because, for them, it smacks of relativism. Indeed, social constructionism, the younger sibling of social constructivism and closely associated with the work of Kenneth Gergen, for instance in his 1994 book Realities and Relationships (Gergen 1994), purposefully avoids the issue of truth, even more than those in the sociology of knowledge camp (who accept truth in the context of the hard sciences). It is this school in which the discourse object becomes conflated with the lexical item that stands for it. It does away with the distinction between lexical and encyclopaedic knowledge. This means, in effect, that meaning and knowledge are no more than different ways of looking at the same thing. Among the first to renounce the presumption that there is an absolute truth in Descartes’ sense were the American pragmatists. Charles Peirce, William James and John Dewey weakened, in slightly different ways, his main claim that our mind, if isolated from external input and from the body in which it resides, is a mechanism that can distinguish between what is the case and what is not. Pragmatism and constructionism, in particular, have dealt Cartesian epistemology a major blow, with dire consequences for the status of the natural sciences. If there is no methodology to tell us whether statements are factual or not, then our trust in the dependability of them is misguided. Knowledge does not have to be true in order to count as knowledge. Books on unicorns and angels can tell us much more than logarithmic tables. We owe our illusion that we deal directly with the reality out there to the primordial speech situation. Because it provides us with a shared reality, it is easy to take one for the other. We experience that kicking a rock hurts, but this non-symbolic experience becomes symbolic knowledge once we tell other people and hear from other people that kicking a rock hurts. What is negotiated in the discourse, our collective mind, is never our experiences, only what we tell about them. The discourse of the hard sciences has always demanded particular attention. We are all brought up in the belief that this discourse, more than all other ones, links what is said directly with the discourse-external reality. Scientific discourse is about truth, factuality and the world as it really is, not as we see it. Such a perception, however, does not sit well with my contention that the only reality to which we have access is the reality constructed in the discourse. In Chapter 14, I will argue that the language of the scientific report should be viewed as an attempt to re-create a virtual primordial speech situation. What is described in the form of the lab environment and the events that take place in it
Introduction
27
can be seen as an extension of the discourse as it takes place when the experiment is actually carried out, with the readers as the virtual participants in this situation. For me, the data that scientists work with is originally no more than a handful of the (unstructured) stuff that reality consists of. Some kind of agreement has been reached in a scientific discipline as to what is supposed to be detected, counted and measured in respect of the data, and also in respect of the technical gadgets that carry out these procedures. The report has to be detailed enough to allow each reader to repeat the experiment. It will have revealed a scientific truth if the same results are achieved each time. My contention is that a scientific report succeeds if its readers share a reality with the experimenter, a situation made possible by the use of deictic devices which are part of the (extended) discourse, not outside it. What is observed is not the stuff out there but objects of a shared reality, objects of the discourse. The results obtained by carrying out the experiment are an interpretation of the stuff out there, not a mirror of it. My way of looking at the scientific report is at odds not only with the more conservative varieties of philosophy, but also with more progressive perspectives, that come under the label of social epistemology. While there is an acceptance of the fact that we cannot expect a scientific community to recognise truth, in their view it is still truth, and not the acquisition of any kind of knowledge, that is seen as the goal of the scientific enterprise. In Chapter 15, my focus will be on the role of hermeneutics in relation to the social construction of reality. The abyss between realism and constructionism seems difficult to bridge. Despite the mostly American paradigm of pragmatism, Anglo-American philosophical communities today seem to be more reluctant than their Continental cousins to entertain the idea that our reality is only a discourse construct, apparently because they are afraid of the relativism disease. While they readily admit that our grasp of reality must always be biased by the cultural and societal constraints enveloping us, and that the reality in which we find ourselves is, at least to the extent that we rely on writing, determined by the discourse, they still maintain that truth, not meaning, is what knowledge is about. On the Continent, too, there are voices insisting that the quest for truth should not be prematurely abandoned. Jürgen Habermas’s Theorie des kommunikativen Handelns (Theory of Communicative Action) (1981) conjures the vision of an ‘ideal speech situation’, a symmetric, rational dialogue between equal members of a discourse community, leading to incontrovertible, unbiased and therefore true results. The goal, Habermas insists, must be a society in which all psychological shortcomings of its members have been treated successfully (by psychoanalysis), and social power structures have given way to a truly democratic order. But even such a utopian scenario still begs the question of to what extent our concept of rationality, this heirloom of the Western Enlightenment, can be called universal. Is not rationality a discourse construct just as much as the reality that is the object of the discourse?
28
Meaning, Discourse and Society
This is, of course, not only the staple fare of hermeneutics but also the stance of post-structuralism. The discourse creates itself, and it is only within the discourse that we find meaning. For Jacques Derrida there can be no question of the signifiants of which our discourse consists signifying anything in some discourse-external reality. They refer to nothing else but other signifiants within the discourse, so that each signifier is just one single event in an infinite iteration of occurrences. There is no point, he claims, in interpreting the discourse in relationship to some metaphysical reality (cf. Derrida 1988). French post-structuralism is heavily indebted to hermeneutics, the art of interpreting texts. Hans-Georg Gadamer and Jacques Derrida both developed their philosophical positions in their endeavour to distance themselves from Martin Heidegger. While Derrida insists that understanding is impossible, Gadamer is more interested in defining the conditions, utopian as they may be, that would enable it. They both agree that all interpretation takes place inside the discourse as an inescapable prison. Their stance conjures up Hilary Putnam’s mind-in-the-vat conundrum (Putnam 1981: 1–21), his example of the impossibility of knowing whether our (individual, monadic) mind is connected to a mind-external reality, or whether our perceptions are no more than mere illusions. This also holds for the discourse as the collective mind. Whether as individuals or as a discourse community, we will not be able to escape the world that we have created for ourselves, except in primordial speech situations. But even there, all we gain is a shared reality, not access to the reality outside of the discourse. Gadamer and Derrida, I believe, also agree that a text’s authorship is irrelevant; and that the author’s intentions are not only irretrievable but inconsequential. For Derrida, however, interpretation is a random process that has to be viewed as taking place without agency, while Gadamer sees it as an intentional act carried out by people who make conscious decisions. They share the belief that there is no true and final interpretation; for them, what is being interpreted is always an interpretation of foregoing interpretations, so that each interpretation is, by necessity, provisional, as long as the discourse goes on. What for Derrida is a random process producing an unpredictable outcome, is for Gadamer the creativity of the interpreter in their dialogue with the peer community. His magnum opus, the book Truth and Method, originally published in 1960, rejects both the idea that we should ponder the truth of texts and the idea that it is possible to design a methodology (in the sense of a list of instructions) which would lead to the one and only correct interpretation. In each new interpretive situation, the text has to be seen from a new horizon. Thus for Derrida and for Gadamer, though for opposite reasons, any interpretation has to be contingent and will in turn become the possible object of some future interpretation. Finally, in the last chapter, I present a long and rather tiresome example of how we can try to come to terms with the meaning of a text. The text I will
Introduction
29
analyse is very short; it is a conventional haiku consisting of 5 + 7 + 5 syllables, altogether eight words. My aim is to make it plausible that the meaning of this or any piece of text can be extracted from its intertextual links and the paraphrastic content we find in these references. However, interpreting this meaning is not based on a strict methodology that would guide each interpretive act to the same conclusion. All we can do is to detail the discourse from which we extracted the intertextual links, to show how we have selected the citations we found relevant, and how reading these citations contributed to our reading of the text in question. Interpretation is never a solitary exercise. There is always an interpretive community within which an interpretation takes place as a collaborative act. My suggestion for how to read the haiku will be met by other people’s suggestions, and in the end we will have, as a result a multitude of views, a small number of general trends and perhaps even a few points upon which everyone agrees. This end, however, must not be final. At any time, the haiku can be reinterpreted, and this reinterpretation may well include an interpretation of all previous interpretations. Interpretations are always provisional. In my conclusion, I try to show what consequences my constructionist approach entails. The reality we find constructed in the discourse does not have to be the reality that would serve us best. To the extent that we define ourselves as nothing but consumers of the texts sold to us by our content merchants, we have little say in how this reality is constructed. Only if we reclaim our right to make sense of our world, can we take our destiny into our own hands. To make an impact on the discourse, however, we need others to take notice of what we say. Interpretation is a collaborative act. We can only hope to exercise an influence on the reality in which we find ourselves, the reality of the discourse, if we work together making sense of it and reconstructing it wherever and whenever we find it necessary. Single monadic minds achieve nothing.
Part I
Meaning, the mind and the brain
1
The cognitive turn
In this chapter, I want to argue that the cognitive turn that characterises the second half of the twentieth century has not been able to clear up the mystery of meaning. In the 1950s, the cognitive sciences replaced previous paradigms trying to make sense of human interaction such as the American traditions of pragmatism (Charles Sanders Peirce, William James, John Dewey, George Herbert Mead and Herbert Blumer) and of behaviourism (particularly B. F. Skinner). Both behaviourism and pragmatism have a social focus. Cognitivism, on the other hand, is about the working of the individual mind. It has become a prominent scientific paradigm in many disciplines of the human and social sciences, for instance in psychology, linguistics, anthropology, and philosophy, but also, interestingly, in biology and the computer sciences. The pragmatists look at symbolic interaction. This is the kind of behaviour to which people assign meanings, and includes the interpretations we, the people, assign to the interactive behaviour of other people. In his investigation of verbal behaviour, the behaviourist B. F. Skinner attempted to correlate the verbal and other symbolic input to which a person is exposed, with their linguistic output (the utterances that person makes). For behaviourists who wanted to observe social interaction, an individual’s verbal behaviour seemed to offer the only access to an understanding of how meaning is assigned to interaction. Behaviourism is based on the assumption that the behavioural scientist’s only sources of information are what happens to people (the input they are exposed to) and the behaviour they exhibit (their output). What is taking place inside them cannot be observed; it remains a ‘black box’. Both pragmatists and behaviourists refrain from making assumptions about the mind. The date usually identified with the demise of behaviourism is the publication of Noam Chomsky’s 1959 review of Skinner’s book Verbal Behaviour, which had been published in 1957 (Skinner 1957). Human behaviour, and linguistic behaviour in particular, claims Chomsky, is so complex that a theory focusing on stimulus and response can never explain it. For while linguistic behaviour 33
34
Meaning, Discourse and Society
may need some stimulus, it is to a very large extent determined by a mechanism all human beings have in their minds. So he can see no other explanation for the ease with which people generate new, original sentences, than that their minds must be equipped with an innate faculty. They cannot have learnt the rules or the lexicon from their environment, due to the dearth of relevant stimuli. Unless they were endowed with an innate language faculty (sometimes called by him a language organ), it would be impossible for them to produce grammatical sentence after grammatical sentence. It is this view that leads Noam Chomsky to reject behaviourism. Today, even Chomsky would agree that Skinner and he were talking about different issues. In the words of a recent paper, co-authored with Marc D. Hauser and Tecumseh Fitch, Chomsky accepts that ‘it is important to distinguish between questions concerning language as a communicative system and questions concerning the computations underlying this system, such as those underlying recursion’ (Chomsky, Hauser and Fitch 2002: 1569). The refusal to make assumptions on the mind is a core tenet of behaviourism, and certainly of Skinner himself. This is evident when he speaks about understanding. The mind and its mechanisms are not invoked. For him, ‘[t]he listener can be said to understand the speaker if he simply behaves in appropriate fashion’ (Skinner 1992: 277). For him language ‘has come to refer to the practices of a linguistic community’, i.e. what is happening between people, while he states as his aim the description of the observable verbal, i.e. symbolic, behaviour of individual speakers (p. 2). While Skinner tried to analyse a person’s linguistic output in terms of previous stimulating input, for Chomsky it is the mental mechanism that makes up a person’s innate language faculty, not their verbal behaviour. For him, language is not so much the E[xternal]-language, i.e. the natural language that a person is speaking, as in English or Pirahã, but the I[nternal]-language, the language of the mind. As all humans share the same genetic make-up, there must be, Chomsky claims, language universals, linguistic features common to all natural (human) languages but not found in the communication systems of animals. Language thus is not defined as symbolic, negotiated behaviour, but as a biological endowment unique to human nature. He makes a distinction between on the one hand substantive universals, which have to do with the physiological realisation of language, for instance the claim that there is a maximum number of tones (five) for (tonal) languages, and on the other hand formal universals involving the character of rules determining the syntax of the I-language. The latter are universals of the mental language faculty. As such, they also determine the syntactic structures of E-languages. Half a century ago, Chomsky and his disciples postulated the existence of many such universals. One such universal is that all grammatical operation is structure dependent. An example of this is the question rule as we find it in English. It states that when we turn a statement into a question, it is the verb
The cognitive turn
35
belonging to the main clause, and not that of a subordinate clause, that needs to be put into front position. To take the example offered by Geoffrey Sampson, one of Chomsky’s fiercest critics, the sentence The subjects who will act as controls will be paid. can be turned by moving the second will to the front: Will the subjects who will act as controls, be paid? but not the first will, the finite verb of the relative clause: *Will the subjects who act as controls will be paid? This universal rule states that sentences are organised in structures of phrases, and phrases in structures of words, and that this structural organisation must be preserved in a transformation of a statement into a question. It implies that the syntax of the I-language, and, inferentially, of all E-languages, is organised in a structure-dependent way. Yet there is no necessity, says Chomsky, for syntax to be organised this way. It is just a matter of our genetic layout. One problem is that once we try to formulate these rules explicitly, we find that we quickly have to make adjustments for different E-languages. While in English practically all sentences have a syntactic subject, there are languages like German or Chinese in which we find sentences without a subject noun phrase. Such languages obviously do not conform to the deep structure rules that Chomsky postulated in his 1965 book Aspects of a Theory of Syntax (Chomsky 1965). There we learn that all deep structure sentences consist of two constituents, a subject noun phrase and a verb phrase containing all that is not part of the subject noun phrase. Of course it is always possible to say that a sentence without a subject noun phrase (e.g. ‘Mir [dative] ist kalt’ [I am cold]) is a transformation of a deep structure sentence (e.g. ‘Es ist mir kalt’) in which the subject noun phrase (es) has been turned into a zero element. Such an argument, however, would make the concept of language universals appear rather arbitrary. If only B is observable, and we have transformation rules that can turn almost any kind of A into almost any kind of B, then almost anything could be said of A. As Geoffrey Sampson puts it (in Popperian terms), such a theory of language universals cannot be falsified and is therefore useless (Sampson 2005: 57ff.). This leads us to the next problem. As long as we take the syntax of the I-language, that is the mental mechanism responsible for the structure of sentences, as an instance of metaphysical reality, we have to accept that it cannot be accessed directly. All we can do is to build a model that we hope is functionally equivalent to it. For instance, the phrase-structure syntax model proposed by Chomsky claims that in I-language, and therefore in all E-languages, simply words are organised in phrases, and that
36
Meaning, Discourse and Society
structure dependency is a universal feature of language. Yet there are alternative models that do not employ the notion of a phrase structure. Dependency grammar is such a model, in which a sentence is defined by the words it consists of and the relationships obtaining between these words. Every word of a sentence depends on another word, with the exception of the main verb of the main clause of a sentence. The underlying assumption is that (conjunctions excepted) each sentence has only one main clause, and each main clause has only one main verb, while all other words depend, directly or via other words, on this main verb (cf. for instance, Ulrich Engel 1994). This model can describe how statements are turned into questions just as well as Chomsky’s model. The important issue therefore is whether the postulated language universals ‘really’ exist in some actual language faculty as part of a ‘real’ entity called the mind, or whether they are simply constructs of linguistic theories, of models reflecting this mysterious entity, the language system. The evidence accumulated over the past fifty years by linguists working on a vast range of natural languages has consigned all but one of the original Chomskyan language universals to the dustbin. Within the framework of his paradigm, the one language universal that has survived is recursion. Recursion, originally a mathematical concept, is the property of languages that allows the replication of the syntax determining the structure of a sentence within that sentence, thus making it possible to embed a subordinate clause in a main clause, or a modifying noun phrase within a head noun phrase, and another subordinate clause within the subordinate clause, or another modifying noun phrase within the modifying noun phrase, and so on, so that we can obtain a hierarchical structure of theoretically unlimited depth. In this connection, Chomsky frequently quotes Wilhelm von Humboldt’s famous apophthegm that language is a system that makes infinite use of finite means. For Chomsky this means that recursion allows us to always generate a new sentence, a sentence that has never been generated before. An example would be a sentence like ‘The idea is good’. Into this one can embed the sentence ‘He has this idea’, generating ‘His idea is good’, and the sentence ‘That there is a structure is an idea’, leading to ‘His idea of a structure is good’. Into this sentence, we can embed the sentence ‘Language has a structure’, leaving us with ‘His idea of a structure of language is good’, and so on, without end. With Dan Everett’s findings concerning the Pirahã language, the language of a small Amazonian tribe, a new discussion has ensued over the universality of recursion (Everett 2008; Colapinto 2007). As Everett sees it, Pirahã does not allow recursion. In this language (a language, though, that does not have abstract expressions such as idea, structure or language) it is only possible to add sentence to sentence: instead of ‘The mother of the boy bitten by the snake cries’ we would have: ‘The snakes bites the boy. The boy has a mother. The mother cries.’ Assuming that Everett’s representation of the Pirahã language is
The cognitive turn
37
correct, would that be an argument against recursion as a language universal? Is what this principle claims ‘really’ the process that someone’s language faculty undergoes when a person says ‘The mother of the boy bitten by the snake cries’? Is there a way we could find out what happens in someone’s mind in the instants before the utterance is said? Skinner’s and the behaviourists’ key argument was that the mind and its workings cannot be an object of scientific research any more than the soul. Just as Chomsky reproaches Skinner for not explaining verbal behaviour in terms of the workings of the mind, so would Skinner assign Chomsky’s claims about the language faculty to the realm of fable. Both Skinner and Chomsky are generalising language acquisition; they are not interested in individual differences. For Skinner it seems plausible to assume that children acquire language through a favourable environment, and the more time carers, teachers and more advanced peers spend with a child practising the relevant skills, the faster and better they learn to speak it. Chomsky, on the other hand, believes that external stimuli are never sufficient to account for the speed with which a child acquires a new word, or instantiates a new grammar rule (such as to place an adjective in front of the noun, as in English, or behind the noun, as often in French). This is why he insists on an innate mechanism that, as he sees it, is triggered by no more than superficial exposure to a word, a phrase or a sentence. (For more details on the innateness of mental concepts, see the next chapter.) Once the language faculty is put into motion it enables children to come to accumulate a large vocabulary often through no more than a single exposure to a new word, and this enables them to construct their thoughts in I-language sentences syntactically organised on the basis of innate rules. The language faculty of the mind is conceptualised as a computer carrying out complex processes on minimal external prompts. In the article co-authored with Hauser and Fitch mentioned above, eleven pages long, the words compute, computation and computational occur over forty times. Chomsky’s understanding of the mind as an engine of reason is indebted to René Descartes, as we can read in his 1966 book. Descartes, to whom we owe the still vibrant perspective of a body–mind dualism, was among the first philosophers in modern times to extol the individual, monadic mind which alone is in a position to determine truth. Nothing we learn from others can be known as true unless it passes the test set by la raison, our god-given power of reasoning. The celebration of reason gave rise to the Western Enlightenment movement, and also to moral sense philosophy. In both traditions, each person is seen as a monadic, autonomous, individual entity, and it is not so much their interaction with other people as their own singular creative power that allows them to become a person in the modern sense. It was in the seventeenth and eighteenth centuries that the idea of a creator’s intellectual property, of their authorship, became the prevailing paradigm. Whether it was violins,
38
Meaning, Discourse and Society
cathedrals or books, all were perceived as individual creations. As long as we keep believing in the autonomy of the individual person, we also have to posit the individual mind as a sophisticated, highly complex mechanism, as a kind of stand-alone computer, that is able to carry out such achievements on its own. It is this view that prevails even now. Whoever wants to find out the nature of our creative powers has to investigate the mind. Over the past fifty years, the philosophy of mind and the cognitive sciences have been busy drawing up one blueprint after another of this elusive entity: the individual mind. Not many people have dared to express their doubt about the reality of the mind. In 1941, an otherwise unknown author published in the Thinker’s Library a book with the title The Myth of the Mind. There he writes on page 2: ‘There is no more an entity called the mind, from which what we describe as mental phenomena proceed, than there is an entity called digestion presiding over the digestive processes, or an entity called respiration governing the respiratory processes’ (Kenyon 1941). At the end of that decade, the BBC ran a series of broadcast talks on The Physical Basis of Mind, which was subsequently edited and published by Peter Laslett (1950). A. J. Ayer’s discussion of the contributions of seven highly reputed physiologists came to the conclusion that ‘mind and body are not to be conceived as two disparate entities between which we have to make, or find, some sort of amphibious bridge, but that talking about minds and talking about bodies are different ways of classifying and interpreting our experiences’ (Ayer 1956: 74). For Ayer the mind thus is a convenient construct of our discourse, not something we could find in the reality out there. Gilbert Ryle in his famous book The Concept of Mind (1973) is equally convinced that there is no such thing as a mind, and he likens this idea to believing that there must be a ghost in a machine that makes it work. For him this myth, or dogma, as he calls it, is caused by mixing two distinct categories, perhaps comparable to the sentence ‘She came home in a flood of tears and in a sedan chair’ (Ryle 1973: 23). The dogma of the ghost in the machine, he says, similarly ‘maintains that that there exist both bodies and minds; that there occur physical processes and mental processes, that there are mechanical causes of corporeal movements and that there are mental causes of corporeal movements. I shall argue that these and other analogous conjunctions are absurd’ (p. 23). But in spite of Kenyon, Ayer and Ryle, it is the philosophy of mind that has dominated much of the philosophy of the past fifty years, and it is the cognitive sciences that have conquered much of what used to be the territory of the human sciences, by developing model after model of this enigmatic entity, the mind. I will now have a closer look at the concept of the mind that we find in cognitive linguistics. There is a rather disquieting aspect to the foundation of the discipline of cognitive sciences in general. When, in the 1950s, this new field replaced behaviourism, it was claimed that it had now finally become possible to demonstrate and explain directly the working of the mind, while also
The cognitive turn
39
promising that such an inquiry, like that of behaviourism, could be based on scientific principles. But because it is not so easy to look into people’s heads, a model was needed to spell out how the mind works. The computer had just been invented, and its prospects seemed boundless. People believed it was only a matter of time and computing power before it competed with human reasoning. More memory, more operations per second, more complex programs, and computers would emulate if not surpass human thinking. As they were believed to be, in principle, functionally equivalent to the human mind, they were seen as the perfect model of cognition, offering the additional advantage of blocking out the erratic impact of hormones and emotions. The same computer programs could run, in principle, on any computer, delivering identical results. The materiality of the computer was insubstantial. Its abstraction was the Turing machine, Alan Turing’s Platonic idea of a deterministic finite automaton. Computers were, from Descartes’ dualistic perspective, as close as one could get to the pure mind, the mind not contaminated by the body. That was the hour of birth of the computational theory of the mind, a theory which quickly became a doctrine stating that the working of the mind can be understood and described by analogy with the working of the computer. Ray Jackendoff was influential in relating this model to the study of language, for example in his book Consciousness and the Computational Mind (Jackendoff 1987). Many philosophers, psychologists, sociologists and linguists saw in the sudden availability of the computer as a model for human minds the unique opportunity to free their disciplines further from the haut goût of the increasingly unattractive humanities: finally they could look their colleagues in the natural sciences straight in their eyes. The computer seemed to provide the final breakthrough. It enabled the cognitive sciences to place the study of the human mind firmly on a par with chemistry and physics. What had been a mostly philosophical discussion, with all the arbitrariness of interpretation, could now be based on solid fact. This was a welcome opportunity for many linguists to jump on the bandwagon of scientific and technological modernity. Not unimportantly, it was there, and not in the humanities, where the big research funds were on offer. With the newly available computational model the mystery of the mind could now be finally solved. As Paul Smolensky tells us, ‘[t]he hypothesis driving most of modern cognitive science is simple enough to state: the mind is a computer’ (Smolensky 1994). Further rewarding consequences soon began to show. Thanks to the new paradigm, other disciplines began to take the cognitive sciences seriously. Therefore it came as no big surprise that when, some years later, computer scientists set out to invent the new research field of artificial intelligence, they now looked to the cognitive sciences for inspiration. The cognitivists naturally were happy to sell back the blueprints they had copied
40
Meaning, Discourse and Society
a few years earlier from the computer engineers, together with some new annotations, to the emerging artificial intelligence community. The new goal was to transfer the now established view of how humans perform mental operations and solve problems of all kinds, to computer scientists eager to develop ‘intelligent’ machines. This cross-fertilisation continues to the present day, as does the ignorance about the circularity involved. So when the computer scientists developed the concepts of parallel processing and connectionism, these new ideas were immediately taken over by the cognitivists. Thus Pierro Scaruffi tells us: ‘[A] connectionist structure such as our brain works in a nonsequential way: many “nodes” of the network can be triggered at the same time by another node. The result of the computation is a product of the parallel processing of many streams of information’ (Scaruffi 2003). The new developments were greeted by many active in the philosophy of mind. They seemed to vindicate the computational theory of mind. ‘Connectionism’, says Daniel Dennett, ‘is a fairly recent development in A[rtificial] I[ntelligence] that promises to move cognitive modelling closer to neural modelling, since the elements that are its bricks are nodes in parallel networks that are connected up in ways that look rather like neural networks in brains’ (Dennett 1993: 268, Dennett’s emphasis). The belief that parallel processing and neural networks are all that it takes to make our computers truly intelligent was permeating the whole of Western(ised) society. Those were the days when everybody was talking about a new generation of computers believed to be able to finally surpass human intelligence. Experts kept promising that the next generation of computers could independently learn (program themselves) how to do things instead of just carrying out a programmer’s instructions. They could be trained to reason based on common sense, thus enabling them to pass reliable judgements on matters too complex to be thought through by humans. These new ‘virtual machines’ would employ an ‘architecture’ of processors operating in a parallel, interactive way. The relationship between the initial and the resulting states of such a machine would not be determined by pre-programmed commands, but would develop itself on the basis of huge amounts of training data. This is the idea behind connectionism. Once the computer has ‘learned’ which initial stages lead to which resulting states, this knowledge can be applied to any new initial state displaying the same properties, and the computer will deliver the correct final state. Eric Pederson and Jan Nuyts, Conceptualization, would not disagree:
the
editors
of
Language
and
Thus, while ‘classical’ cognitive theories would consider representations to be virtual ‘objects’ of some type, manipulated by a ‘machinery’ of procedures or rules which are somehow implemented in the human brain, connectionist and parallel distributed
The cognitive turn
41
processing theories consider representations to be simply the resultant characteristics of peculiar states of the ‘conceptual system’ distributed across the neural networks of the brain … In the latter view, if the notions of knowledge and representations are to be used at all, any characterizations of them beyond the vague ones given above are no longer acceptable as descriptions of actual cognitive mechanisms creating human behaviour. (Pederson and Nuyts 1997: 1f.)
So can we be sure that the brain/mind is isomorphic with a (connectionist) computer? Perhaps not quite, Ray Jackendoff thinks, but that does not harm this model. He accepts that ‘eventually the neuronal basis of mental functioning is a necessary part of a complete theory’. But neuroscience today is ‘far from being able to tackle the question of how mental grammar is neurally instantiated’. Therefore ‘the formal/computational approach is among the best tools we have for understanding the brain at the level of functioning relevant to language, and over the years it has proven a pragmatically useful perspective’ (Jackendoff 1997: 9). For the John Searle of the late 1980s it was exactly this equation of computers and human minds that he saw as the foundational error of the cognitive sciences: If one looks at the books and articles supporting Cognitivism one finds certain common assumptions, often unstated, but nonetheless pervasive. First, it is often assumed that the only alternative to the view that the brain is a digital computer is some form of dualism. The idea is that unless you believe in the existence of immortal Cartesian souls, you must believe that the brain is a computer. Indeed, it often seems to be assumed that the question whether the brain is a physical mechanism determining our mental states and whether the brain is a digital computer are the same question. Rhetorically speaking, the idea is to bully the reader into thinking that unless he accepts the idea that the brain is some kind of computer, he is committed to some weird antiscientific views. Recently the field has opened up a bit to allow that the brain might not be an old fashioned von Neumann style digital computer, but rather a more sophisticated kind of parallel processing computational equipment. Still, to deny that the brain is computational is to risk losing your membership in the scientific community. (Searle 1990: 21–37)
So how much can we trust this model, used by both the cognitive and the artificial intelligence communities? Some misgivings are allowed. In spite of billions of dollars invested in artificial intelligence and machine translation systems, results are far from satisfactory. For forty years the artificial intelligence community has told us that success is ‘just around the corner’. We are still not there. The epochal endeavour to teach computers ‘common sense’, the megalomaniac CYC project of a comprehensive ontology of knowledge, never came close to the originally envisaged results (www.opencyc.org/), nor did EUROTRA, the European project for automatic translations from and into all the EU languages, which cost the taxpayers hundreds of millions of euros (www-sk.let.uu.nl/stt/eurotra.html). Could it be that the whole approach was faulty? In Chapter 5, I will come back to the question of artificial intelligence.
42
Meaning, Discourse and Society
Electrical impulses change the states of computers, something we tend to call carrying out the commands of a computer program. They are no more conscious than a clock is aware of the time it is showing us. Computers do not act. Neither have they objectives, nor do they plan, and nor are they aware of the computations they carry out. Computers do not have intentionality. Texts, text segments or lexical items mean nothing to a computer. It is hard to see how these machines can be a model of the mind. When the cognitive sciences emerged in the fifties of last century, they quickly spread to many areas. To the extent that they deal with language, we have to distinguish between cognitive linguistics with a lower-case c and l and an upper-case Cognitive Linguistics. The former label applies to all attempts to model the language system as a mental faculty, regardless of whether such a model is an accurate representation of the mind. Any formal mechanism, such as Chomsky’s generative transformational grammar, for instance, would be such a device, as long as it is functionally equivalent to the mind’s processing of language. To the extent that the many versions of the philosophy of mind are concerned with language, they can be seen as part of cognitive linguistics. Upper-case Cognitive Linguistics claims to go further, namely to describe and explain not just the language mechanism but the workings of the mind itself. In the words of the editors of the Cognitive Linguistics Reader, ‘the cognitive commitment asserts that models of language and linguistic organization proposed should reflect what is known about the human mind, rather than the purely aesthetic dictates such as the use of a particular kind of formalisms or economy of representation’ (Evans, Bergen and Zinken 2006: Introduction, 4). We, the observers, do not have direct access to the mind. Indeed, in many respects the mind appears to be as elusive as the soul. All we can do is to ‘scientifically’ compare input to output, i.e. sensory input and behavioural output, and draw our conclusions about what is happening ‘inside’. Ever more complex techniques may enable us to catch a glimpse of the workings of the brain, for instance of neurons firing, and we can attempt to correlate these findings with the (linguistic) behaviour people exhibit. But the brain is not the mind. The mind is no more than a theoretical construct, a model of a possible interface between body and behaviour. No technology can monitor what is happening in the mind. I will argue that any claim to empirical or ‘scientific’ factuality, whether in cognitive linguistics or in Cognitive Linguistics, is quite unfounded. It cannot be otherwise, as long as we agree that the mind cannot be inspected. No introspection lets us see inside our mind. Whatever we think we find there is less mental reality than a reflection of the theories that fall under the label of folk psychology, i.e. of certain beliefs we entertain and communicate in the discourse at large. In the words of Scott M. Christensen and Dale R. Turner (1993: xvi), folk psychology is ‘the tag given to ordinary talk about the mind … [I]t
The cognitive turn
43
talks about beliefs and desires, intentions and fears, wishes and hopes. It is essentially the vocabulary we use to talk about and explain ourselves and others. It is the vocabulary of the mental’. But our discourse about our thoughts and feelings reflects the ways we make sense of our interpersonal relationships. The mind is no more an object of scientific investigation than the soul. As I will show, any attempt to describe what goes on in the mind has to fail because it will run into contradictions that cannot be resolved. Even if we could gain access and if we found a way to describe consistently the workings of the mind, it would only tell us about a particular person’s first-person experience, for instance of something that has been said, but not the meaning of something that has been said. For language is symbolic; it signifies something to someone. But what any symbol, including a text or a text segment, a phrase or a word signifies, what it means, is not something we can figure out by ourselves. It is not like the presence of smoke which can signify fire. This is something we can learn from our own experience. All kinds of animals can make the connection between smoke and fire. The relationship between the two is not a symbolic one. Smoke is a symptom that can indicate that fire is nearby. The word apple, on the other hand, is not a symptom indicating the presence of an apple. No animal will grasp the meaning of the word, though it may be possible to train an ape to take the spoken word as a symptom for the promised fruit. The word apple is not causally or by similarity connected to the fruit ‘apple’. For those who can speak English, the word apple is therefore more than a symptom. For them it is the fruit of the apple tree, a part of their lunch, the key prop in the story of William Tell and more. It is a sign, and we need other people to tell us what it means. It is not something we can figure out by ourselves. What it means we learn not just by a deictic act but also from its frequent occurrences in the discourse to which each of us is exposed. In this sense language presupposes society. Language signifies; it is symbolic. What it signifies is something that the members of the discourse community have agreed upon. Speakers of English are expected to know what apple means. When someone uses the word, I know what they are talking about, and if I am asked to buy a pound of apples, I will not come back with pears. On request I can talk about apples. Almost everyone who speaks English will have had experiences of apples, the fruit, and apples, the word. But what someone’s mental representation of an apple, of or the word apple, is, I will never know. Even if it were possible to gain access to the mental representations that cognitive linguists claim correspond to a text or a text segment, the only way to make them an object of investigation is to render them, not in some ‘language of thought’, but in a natural language such as English. It is the only way in which people can share and exchange symbolic content. Whenever we want to point out the meaning of gestures or other language-free symbolic behaviour, we must use language. There is no alternative. Never do we communicate mental representations.
44
Meaning, Discourse and Society
Whether we speak or write or use sign language, it is always natural language. It does not help to use formal expressions, such as mathematical formulae, to represent mental representations. Formal languages, as they have been used by numerous linguistic schools in the second half of the twentieth century, offer no alternative, because all formal notations have to be translatable into natural language in order to mean something to someone. Whenever the cognitive sciences deal with language, they tend to neglect or even ignore the social aspect of meaning. This is even more the case today than ten or twenty years ago. For now, with a growing recognition that neither cognitive linguistics nor Cognitive Linguistics can give us the answers they were supposed to provide, the cognitive sciences turn increasingly to the neural sciences. If it has been futile to try to put a tap on the mind because the mind does not render ready data, the brain is now becoming the object of linguistic research. But the brain does not know what the word apple means. Scientific evidence that certain neurons are firing, or that there is measurable synaptic activity, is not the meaning of apple. The brain may store the experiences a person has. But no brain scan will ever reveal these experiences. They certainly will not tell us what the word apple means. The meaning of apple is a convention that people of a given discourse community normally abide by, a convention, however, which can always be questioned and renegotiated by this discourse community. The meaning of apple is socially constructed. To learn how the word apple is stored, be it in the brain or in a computer or on a magnetic tape, does not help us to understand its meaning. As I will argue in this part, neither the brain nor the mind itself will help us to unlock the mystery of meaning. Why then have linguists been so enchanted with the mind? One of the reasons is the fascination with consciousness. There may be moments when we talk without actually making ourselves aware of what we are saying. But the possibility of consciousness is always there. Bees, ants or apes may communicate as well, but we do not expect them to reflect on what they are communicating. They are not even aware that they are communicating. They lack, for all we know, intentionality, this elusive faculty of being conscious of themselves, of their thoughts, their beliefs, their feelings or of symbolic content. Only human beings seem to possess intentionality. They set themselves a goal, knowing that they do so; they form a plan, knowing that they are forming a plan, and do what is necessary to achieve it. This is what it means to act. Computers, bees, ants, and, I would say, even primates, do not know that they act. Acting, including speaking, writing, listening and reading, are activities that require agency. Acting requires intention. Acting, in this sense, is opposed to processing, being involved in a process without being aware of it. This is how bees communicate and computers operate. Humans can make themselves conscious of what they do. We can speak, but we do not have to; we can read, but we can also let it be. Like every
The cognitive turn
45
other activity we choose to carry out, the way we use language is contingent. It is up to us what we say, and it is up to us how we understand what other people are telling us. It is true that we sometimes, particularly in recurrent standard situations, take refuge in automatic reflexes. Normally, I will greet the people I come across in predictable ways. I will not make a conscious decision as to which of the possible options I will use. But there is no way to predict what I am going to write as my next sentence when I am working on a new paper. It is this property of intentionality, as it is commonly called, that makes the individual mind so fascinating. But it is precisely this property that neither the cognitive sciences nor the philosophy of mind has ever come to terms with. Some philosophers like Colin McGinn believe that we will never come to terms with this concept (McGinn 1991). This is why cognitive linguistics has a problem. Consciousness and intentionality, this strange phenomenon of being aware of oneself, escapes all scientific scrutiny. Therefore the goal cognitive language research sets itself is not what it means to be aware of the meaning of an utterance, but to model the mechanism that allows us to correctly generate and process utterances. This mechanism is the language system. It consists of various components, including phonetics, syntax and semantics. Each of these components contains a list of elements and rules determining how these elements can be configured. If these rules are properly applied, the result will be the generation of grammatical sentences, which, by reversing these rules, can then be understood by a hearer. As a mechanism, this language system does not depend on the intentionality of its users. It is rather like a randomised computer program that ensures that certain procedures are carried out successfully. There are, however, two ways to envisage the language system. For Ferdinand de Saussure, the language system to be explored was that of a given natural language, such as French or English. Saussure’s perspective on the language system was synchronic. He did not look at the historical, the diachronic dimension of language. Therefore he was not interested in the nature of the rules determining the language system. They are like the laws of nature, for instance the laws underlying mechanics. It could be said that the linearity of spoken language is just such a universal, unchangeable principle. Each spoken utterance must have a beginning preceding its end. It must, of necessity, have a temporal extension. But there are also rules that can undergo change. The French language as it was spoken in Saussure’s time was a long way from Old French. Old words had disappeared, new words had been introduced; vowels and consonants had changed their qualities; the tense system had changed, and even the syntax had been affected. These changes had come about largely without people being aware of them. If there was consent, it was largely tacit. But there were also changes introduced by rules set up on purpose. The impenetrability of French orthography was designed to cut out competition posed to established printers by uninitiated rogue printers. Other rules were
46
Meaning, Discourse and Society
designed to protect the privileges of those in power, by favouring a certain dialect and a certain accent, by prescribing un bon ton to separate the elites from the masses, and by introducing syntax rules that would separate the educated classes from those who could not afford a good education. Thus the system of a natural language features different kinds of rules, inalterable ones as in the natural sciences, that people are normally not aware of, and ones based on explicit conventions (and therefore, in principle, open to renegotiation). From a synchronic perspective, these differences do not count. What counts is that people abide by them. They even abide by rules they are not conscious of. Therefore the language mechanism that makes us speak and understand must be in people’s heads. How it gets to be there was not one of Saussure’s great concerns. This is why Ferdinand de Saussure can perhaps be seen as a precursor to cognitive linguistics, but not as a cognitive linguist. Noam Chomsky, on the other hand, is not interested in the system of English, or any other language as such. His concern is the language faculty shared by all people, regardless of the language they speak. The language faculty, as he sees it, is defined by the mechanism encapsulating the laws underlying all natural languages. This is why languages, apart from their superficial differences, all employ the same deep structure. As he says in his book New Horizons in the Study of Language and Mind (actually a collection of papers published earlier in various places): [This] initial state is a common human possession. It must be then, that in all their essential properties and even down to fine detail, languages are cast to the same mold. The Martian scientist might reasonably conclude that there is a single human language, with differences only at the margins. (Chomsky 2000a: 7)
It is this universal, and therefore innate, language faculty that is the focus of cognitive linguistic research. It is a mechanism that transforms a thought into language, into an utterance, and then, on the part of the hearer, converts this utterance back into a thought. This mechanism is there, regardless of whether we are conscious of our thoughts, and the thoughts we are fed with, or not. For cognitive linguists of all feathers, meanings must be in people’s heads. The meaning of an utterance can be nothing else than its mental representation, whether we are aware of it or not. This idea is not as new as it appears. Aristotle was one of its early protagonists. It suggests itself. Where else should meaning, should the whole mechanism that lets us correlate a sentence with that which it is about, be situated, if not in the mind? If people use language, then the language system is to be found, and can only be investigated, in people’s minds. This has been, apart from very few exceptions, the way Western philosophers from Aristotle to the Middle Ages have looked at language.
2
The long history of mind linguistics
In his text Peri Hermeneias (‘On Interpretation’) Aristotle tells us: ‘Spoken words are the symbols of mental experience.’ For him, as for Chomsky, the language system has a material location: the human mind. Without human minds, there would be no language system. The belief that the human mind is, or contains, a mechanism that makes us speak and understand has been with us ever since Aristotle. Innateness and universality are by no mean novel ideas. The philosophy of the Western world has, with few exceptions, always adhered to the belief that language as such, in its pure and perennial form, is found in the mind, and that all the language decay in its spoken and written reality is due to the objectionable influences of social deterioration. Aristotle’s Peri Hermeneias is worth quoting in more detail, as an early example of the belief that the words we use correspond to innate and universal mental concepts. These are the first sentences of Aristotle’s tract: Now spoken words are the symbols of mental experience, and written words are the symbols of spoken words. And just as written words are not the same for all men, neither are spoken words. But what these are in the first place – signs of mental experience – is the same for all; and what these experiences are likenesses of – actual things – are also the same.
This is the first text in our Western history claiming that we all, regardless of the language we are speaking when communicating with each other, have in common the same language of thought, that the mental representations of natural language utterances are all composed of the same universal building blocks. In his commentary on Aristotle’s text, Ammonius, an Alexandrian philosopher of the third century AD, also accepts the notion of innate concepts, autophues ennoiai (literally: ‘emerging by themselves, naturally’) (Ammonius 1998: 109 [147, 29]). Averroes paraphrases in his commentary Aristotle’s sentences quoted above in these words: Thus, we say that spoken utterances signify primarily the ideas that are in the soul, while written letters signify primarily these utterances. Just as written letters – I mean, script – are not one and the same for all nations, so, too, the utterances by which ideas are expressed are not one and the same for all nations. Therefore, the meaning of both script 47
48
Meaning, Discourse and Society
and spoken utterances comes from convention rather than nature. And the ideas which are in the soul are one and the same for all people, just as the beings [objects] which the ideas in the soul are examples of and signify are one and exist by nature for all people. (Averroes 1998: 125 [16a, 4–9])
While Anselm of Canterbury also subscribes to Aristotle’s views, he expands them, too. Words for him are signs (Monologion 10): For in ordinary usage we recognize that we can speak of a single object in three ways. For we speak of objects either (1) by perceptibly employing perceptible signs (i.e., [signs] which can be perceived by the bodily senses) or (2) by imperceptibly thinking to ourselves these same signs, which are perceptible outside us, or (3) neither by perceptibly nor by imperceptibly employing these signs, but by inwardly and mentally speaking of the objects themselves – in accordance with their variety – either through the imagination of material things or through rational discernment. For example, in one way I speak of a man when I signify him by the name ‘man.’ In another way [I speak of him] when I think this name silently. In a third way [I speak of a man] when my mind beholds him either by means of an image of a material thing or by means of reason – by means of an image of a material thing, for instance, when [my mind] imagines his perceptible shape … Each of these three kinds of speaking has its corresponding kind of words. Yet, words of that [kind of] speaking which I mentioned third, and last, are natural and are the same for all races, if they are not words for unknown things … Moreover, [these natural words] can without absurdity be called truer the more they resemble, and the more expressly they signify, the objects for which they are words. (Anselm of Canterbury 1980: 21)
For Anselm, and this is new, a word no longer needs be read aloud. It is enough to perceive the written sign and think of its spoken form. It is also possible to use these words, i.e. the words of a natural language like Latin or English, in order to think. But for him, the truly natural ‘words’ are the mental images, the representations we have of things, and ‘these are the same for all races’. What makes these concepts so natural is their resemblance to the objects for which they stand. They are, in modern terminology, not so much symbols as icons. Thus we see that for medieval language philosophy the naturalness of words had nothing to do with their form. The meaning of spoken or written words is the mental image of the object they designate or for which they stand. It is, in modern technical jargon, the conceptual representation. Here we have, in a nutshell, the foundational agenda of cognitive linguistics. As I will show below, these ‘natural words’ are equivalent to the words of Jerry Fodor’s language of thought (Fodor 1975) and of Stephen Pinker’s mentalese (Pinker 1994). They are the same for all humankind. But already here we are confronted with the paradoxical nature of this proposal. Are these natural words signs themselves, so that spoken and written words have, as their referent, again a sign, namely the corresponding ‘natural word’, or are these natural words, these mental images, content without form? In the first case, we are confronted with Derrida’s theory of the infinite iterability of signification (Derrida 1988), namely his contention
The long history of mind linguistics
49
that signs only ever refer to other signs. In the second case, in the case of content without form, we have to deal with an image that does not represent anything but itself, and thus would have to be identical with the object for which it stands. Anselm’s answer to this conundrum is found in Monologion 31: ‘[A]ll such words by which we mentally speak of objects (i.e. by which we think of them) are likenesses and images of those objects for which they are words. And every likeness and image is true in proportion to the exactness with which it imitates the thing whose likeness it is’ (Anselm of Canterbury 1980: 47). These images are signs with form and content, of the things they represent. But if they are signs, they are in need of interpretation. They are like spoken and written words in that they have a meaning which we have to be told. William of Ockham’s logic marks, according to Stephen Meyer-Oeser’s entry ‘Medieval Semiotics’ in the Stanford Encyclopedia of Philosophy (http://plato. stanford.edu) are: an important, though not the only important, step in the process that might be described as a progressive ‘mentalization’ of sign. The idea behind this process is the contention that without some sort of ‘intentionality’ the phenomena of sign, signification and semiosis in general must remain inconceivable. This tendency of relocating the notions of sign and signification from the realm of spoken words to the sphere of the mind is characteristic of the mentalist logic arising in the early 14th century, and remaining dominant throughout the later Middle Ages. (Meyer-Oeser 2006)
A question that has been frequently discussed in this connection is whether for Ockham there is just one concept for each real-world object or whether his lingua mentis allows for synonymy. As Paul Vincent Spade in his entry ‘William of Ockham’ in the same encyclopaedia sees it: For Ockham, the terms of mental language are concepts; its propositions are mental judgments. Whereas the signification of terms in spoken and written language is purely conventional and can be changed (hence in English we say ‘dog’ whereas in Latin it is ‘canis’), the signification of terms (concepts) in mental language is established by nature once and for all. Concepts ‘naturally signify’ what they are concepts of; this ‘natural signification’ is thought of as a kind of representation relation, based on the fact that concepts are in some way ‘naturally similar’ to their objects. This arrangement provides an account of synonymy (both interlinguistic and intralinguistic) and equivocation in spoken and written language. Two terms (whether from the same or different spoken/written languages) are synonymous if and only if they are subordinated to the same concept; a single given term of spoken/written language is equivocal if and only if it is subordinated to more than one concept simultaneously. (Spade 2006)
This raises an interesting question: is there synonymy or equivocation in the mental language itself? (If there is, it will obviously have to be accounted for in some other way than for spoken/written language.) This is obviously a question of some interest to today’s philosophers of mind. David J. Chalmers, who has
50
Meaning, Discourse and Society
developed his own controversial theory of consciousness, believes that Ockham’s approach requires the possibility of at least some synonyms in his lingua mentis, and consequently in the language of thought/mentalese that some cognitive linguists espouse (Chalmers 1999: 76–99). But he admits that his arguments are more a discussion within the philosophy of mind than an analysis of Ockham’s writings: ‘My argument for synonymy is largely theoretical, although it has a textual element’ (p. 77). On close reading, Claude Panaccio comes to a qualified conclusion: ‘The only form of redundancy explicitly excluded is synonymy between simple terms’, while ‘there may even be certain sorts of ambiguities in Mentalese’ (Panaccio 1999: 73). The question of synonymy raises important questions about the nature of mental concepts. One way to see them is as reflections of the things that have been in the world and/or are still there (while one of Aristotle’s much-discussed arguments has always been that future events are not part of our reality). Such a realist (Platonist) stance implies that the ways we divide the world up into things (the categories we use for putting them into pigeonholes) are internal to the world, and not assigned from the outside (for instance, by the negotiations of the discourse community). This stance would not allow for synonymic concepts, as there can be only one authentic categorisation. In this version of a language of mind, Peter’s terrorist would have to correspond to the same concept as Paul’s freedom fighter. But Ockham was a nominalist, not a realist. For him, concepts embody our perceptions of things, not the things themselves. They are universal, because the same things cause the same perceptions. Ockham’s concepts thus are signs which do not point to things in the real world but to internal representations. Like many cognitivists, Ockham does not address the issue of the loop into which his approach leads him: if concepts are signs, then they have to have a meaning; that is for him, an internal representation. But is not such an internal representation again a sign standing for something else, and so on, in an infinite regress? This is a question to which I will return below. The reason why, after all these centuries, there is still much disagreement on the finer niceties of Ockham’s theory of a lingua mentis is a result, as I see it, of the unavoidable contradictions one becomes engulfed in if one sets out to come up with a consistent model of the workings of a mind. While for Spade, Ockham’s rejection of the possibility of synonymous concepts leads to an ‘interesting failure’, Claude Panaccio, as we have already seen above, thinks he has found ‘very strong textual evidence that, according to Ockham, there can be synonymy in mental language’ (Keele 2007: 659). It is a pity that few philosophers of mind and even fewer cognitive linguists engage in a discussion of Ockham’s views and their representation in the history of philosophy, because it would throw light on unavoidable inconsistencies that cannot fail to have an impact on their theories as well. Peter King points out the convergence between his approach and that of Jerry Fodor:
The long history of mind linguistics
51
Ockham is a committed nominalist who is also attracted by what Fodor calls ‘the language-of-thought hypothesis,’ and it is his philosophy of psychology underlying Mental Language, not the well-formed formulae of Mental Language itself, that explains how unwelcome ontological commitments can be avoided. On this reading Ockham has more in common with contemporary cognitive science than with [Russell’s] atomism of a century ago, and shares many of its substantive theses. (King 2005)
When the cognitive sciences replaced the behaviourist model, some fifty years ago, they obviously found the ground well prepared. There was, in the medieval dimension of the philosophical realm, little opposition to the idea that we must look for the language system in the mind. However, apart from very few exceptions, cognitivists neglected to explore this ground in any depth. To some extent, their efforts today look like a reinvention of the wheel. Of course, what we find foregrounded in the more philosophical approaches to language ever since the Enlightenment are the ways in which language can be seen to refer to a reality outside the discourse. This interest became a cornerstone in the philosophy of the Vienna Circle, and the core element of all subsequent analytic philosophy. Particularly in the American philosophy of much of the twentieth century, the many varieties of realism demanded more attention than the question of the role of the mind as an interface between language and reality. The early Hilary Putnam of the twin-earth thought experiment, who used to refer to himself as a metaphysical realist, never tired of telling everyone that ‘meanings just aren’t in the head’ (Putnam 1981). In Europe, on the Continent, however, there was a strong interest among certain scholars involved in various disciplines of the humanities, the Geisteswissenschaften, in establishing themselves as hard sciences. This was the case both for psychology, including Sigmund Freud’s psychoanalysis, and for Ferdinand de Saussure’s brand of synchronic linguistics. The psyche, the individual mind, came to be seen as something as real as the society, this other new object of research, introduced by Auguste Comte in the novel academic field of sociology. Saussure’s contemporary, the leading French sociologist Emile Durkheim, differed from Max Weber by insisting that society is more than the aggregate of the individuals that comprise it. For him, language was a social fact, a reality that cannot be (fully) reduced to the intentionalities of the people who make up a discourse community, i.e. a group of people participating in the same discourse. This was also the way in which Saussure looked at language. In §2 of his Course in General Linguistics we find this very clear statement: The individual’s receptive and co-ordinating faculties build up a stock of imprints which turn out to be for all practical purposes the same as the next person’s. How must we envisage this social product, so that the language itself can be seen to be clearly distinct from the rest? … It is a fund accumulated by the members of the community through the practice of speech, a grammatical system existing potentially in every brain, or more
52
Meaning, Discourse and Society
exactly in the brains of a group of individuals; for the language is never complete in any single individual, but exists perfectly only in the collectivity. (Saussure 1983)
What Saussure comes to call la langue is different from the language system we find in cognitive linguistics, due to this affinity with sociological thinking. It is not a universal system, firmly implanted as an innate faculty into every human being’s mind, but a contingent system, a system that could be different, indeed a system that is different for each natural language and for each synchronic plane of that language. Thus it is the superstructure of the language of a specific discourse community, not of humankind as such. While Saussure holds that it can be studied only from a synchronic perspective, it is, as is society, subject to change. It is neither Ockham’s oratio mentalis nor Chomsky’s language faculty nor the language module of the cognitive sciences. But the idea of a language of the mind looms in Saussure’s head, too. Somewhat later, in Chapter IV, he says: A language, as a collective phenomenon, takes the form of a totality of imprints in everyone’s brain, rather like a dictionary of which each individual has an identical copy. Thus it is something which is in each individual, but is none the less common to all. At the same time it is out of the reach of any deliberate interference by individuals. (Saussure 1983)
A language system over which its users do not have any control, with which they cannot interfere, and which cannot be accessible to consciousness, is indistinguishable both from Ockham’s lingua mentis and from Chomsky’s language faculty. It is now time to turn our attention to this strange construct, the mind. Is it more open to scientific or at least intellectual exploration than the idea of the soul, once ubiquitous, but now only found in the habitat of religious notions? Google lists 1.3 million occurrences of philosophy of mind. Should we take this as proof that there is such a thing as the mind?
3
What do we know about mental concepts?
In 1975, Jerry Fodor, a linguist/philosopher close to Noam Chomsky, published his highly influential book The Language of Thought. Here we find the ancient and medieval idea of a lingua mentis in a new wrapping, incidentally without any reference to earlier sources. Indeed it seems almost to have been a complete reinvention. In this book, Fodor pursues the idea that even very young children, before they acquire the natural language spoken in their environment, already possess a language, namely this language of thought, now sometimes also called mentalese. There must be a language of thought, he argues, because ‘you cannot learn a language whose terms express semantic properties not expressed by the terms of some language you are already able to use’ (Fodor 1975: 61). As this is true and accepted for learning a second language, it must also hold for the first (external) language. People are born with an innate language, a mentalese which is universal, i.e. the same for everyone regardless of their ethnic origin. When we speak, we actually formulate an utterance in the language of thought which we subsequently translate into the language we have acquired, in order to communicate: ‘English has no semantics. Learning English isn’t learning a theory about what its sentences mean. It’s learning to associate its sentences with the corresponding thoughts’ (Fodor 1998: 9, Fodor’s emphasis). For Fodor, natural language words are secondary, and mental concepts are primary. Thus he disagrees with the twolevel semanticists like Sperber and Wilson, whom I will discuss in the next chapter. For his purposes, ‘word meanings are just concepts’ (p. 2). But what do we know about these concepts? Fodor is aware that we have no direct access to them. Therefore he proceeds to investigate our natural language, assuring us that the findings will also hold for mentalese. ‘I have argued that the language of thought cannot be a natural language. Nevertheless, facts about the latter provide us with some of our best data for inferences about the former’ (Fodor 1975: 100). The nature of concepts therefore mirrors to a large extent the evidence of word meaning. Each natural language word, unless it happens to be a compound (Fodor mentions, as an example, airplane) corresponds to a single concept (p. 152). This is a claim that makes little sense. If it were true, then each natural language would feature the same concepts, an assertion everyone who 53
54
Meaning, Discourse and Society
has ever translated knows to be rather questionable. In the end, Fodor unfortunately misses the chance to give his readers a natural language representation of a mental concept. Thus we never find out what they look like. This must not make us grieve: he consoles us with the … suggestion I regard as entirely speculative but very interesting to speculate about: viz. that the language of thought may be very like a natural language … If [this hypothesis] is true, it goes some way towards explaining why natural languages are so easy to learn and why sentences are so easy to understand: The languages we are able to learn are not so very different from the language we innately know, and the sentences we are able to understand are not so very different from the formulae which internally represent them (p. 156).
The underlying notion behind Fodor’s idea about a language of thought is that for him ‘the primordial function of language is not communication but the externalisation of thought … In fact, I think that the thought-first view simply has to be right … One can, in short, make sense of thought without [natural] language (or rather, of thought without communication), but not, I think, of language without thought’ (Fodor 2005). I disagree. All the accounts we have of feral children provide evidence that the singularly human faculty of abstraction rests overwhelmingly on language, on the language we learn from our interaction with other people. Another linguist/philosopher close to Chomsky’s view on language is bestseller author Stephen Pinker. With Fodor he shares some of the views on the language of thought: People do not think in English or Chinese or Apache; they think in a language of thought. This language of thought probably looks a bit like all these languages; presumably it has symbols for concepts, and arrangements of symbols … [C]ompared with any given language, mentalese must be richer in some ways and simpler in others. It must be richer, for example, in that several concepts must correspond to a given English word like stool or stud … On the other hand, mentalese must be simpler than spoken languages; conversation-specific words and constructions (like a and the) are absent, and information about pronouncing words, or even ordering them, is unnecessary. (Pinker 1994: 81–2)
This quote shows nicely how cognitive linguistics multiplies entities without need, and thus violates Ockham’s razor. Natural languages have words (stool or stud), mentalese has symbols representing the various meanings of stool or stud, and these symbols stand for concepts. Does that mean that first we have to translate a natural language sentence into mentalese, and then link this mentalese sentence to its cognitive representation, and then we have understood what the sentence means? This would amount to a triplication of our semantic apparatus. Again, an infinite regress looms (as above, in the case of Ockham, in Chapter 2): the meaning of a natural language expression corresponds to a
What do we know about mental concepts?
55
symbol in mentalese, which in turn points to a concept, i.e. again a sign whose meaning must be interpreted. But who or what is interpreting this sign? Fodor seems to assume that the number of concepts is pretty much the same as the number of words we find in a natural language. In his most recent book, The Stuff of Thought (2007), Stephen Pinker makes it clear that for him only a limited number of concepts are innate. Indeed, he pokes fun at Fodor, calling the chapter in question ironically ‘Fifty thousand innate ideas’, which he takes to be Fodor’s view. Instead, Pinker advocates a view that more complex concepts are composed of a very limited set of innate basic concepts, somehow similar to Kant’s a-priori categories. In his review of the book, the Oxford philosopher Colin McGinn sums Pinker’s view up in these words: Pinker sees that something conceptual has to be innate if language learning is to be possible at all, but he doesn’t believe that it can be anything parochial and specific; so he concludes that only the most general categories of the world are present in the genes – the categories that any human being (or animal) needs to use if he or she is to survive at all. Among such categories, for example are: event, thing, path, place, manner, acting, going, having, animate, rigid, flexible, past, present and future, causality, enabling and preventing, means and end. (McGinn 2007)
McGinn is not unsympathetic to such ideas. He maintains that ‘it is far more plausible to suppose that our language is caused by our thoughts – that we can only introduce words for which we already have concepts. Words express concepts; they don’t create them’ (McGinn 2007). This begs the question whether we were all born with a dormant concept of globalisation before this neologism made it into our newspapers. The question is, of course, what innateness means. Do we have to assume that there are distinct genes for these categories, or basic concepts, that we are born with? In his review of Pinker’s book, Philip Lieberman points to a massive counter-argument: If Pinker’s hard-wired elementary concepts actually existed, genetic variation would ensure that some people lacked a concept or two. We would encounter people who could not acquire the meanings of entire sets of words or think or act upon these concepts. A child missing the gene coding for the primal concept kinship would never be able to understand words such as family, cousin, mother and so on. (Lieberman 2007)
Fodor’s idea that for each word there is a concept is also threatened by comparing and contrasting the words of natural languages. For once we proceed from a monolingual to a multilingual perspective, we are bound to recognise quickly that as long as we look at words in isolation they hardly seem to map onto each other from one language to the next. Whatever there might be, in terms of mental concepts in our heads, there is hardly any evidence to call it genetically inherited, innate or universal. Take the German words Kummer, Trauer and Gram. Bilingual dictionaries tell us that their English equivalents are mainly grief and sorrow. There is no
56
Meaning, Discourse and Society
one-to-one relationship, however. The English sorrow translates into Kummer when a young girl is left by her lover; it translates into Gram, if an old man cannot accept his fate, and it translates into Trauer if someone post puberty suffers the loss of someone dear and seemingly irreplaceable. German–English dictionaries invariably offer sorrow and grief as the equivalents of Trauer, but never provide a distinction between the two equivalents. But there must be a distinction. In most sentences featuring either grief or sorrow, native speakers insist that one cannot replace the other. Native speakers tell me that one normally cannot replace grief by sorrow in the sentence: ‘She heard the news with regret, with sadness, but hardly with grief.’ Similarly, sorrow cannot normally be replaced by grief in: ‘Her fervent soul was full of sorrow for the world and its sinfulness’ (Teubert 1999). Again we are confronted with the fact that there is no way to deal with the meaning of single words in isolation. Words, or other lexical items, do not correspond to mental content items that would have the same fixed content in any (and without any) context. There are contexts that allow grief, and other contexts that allow sorrow. It is the context, the situation and an infinity of peripheral conditions, which have to be matched. Kummer in Thomas Mann’s novels is different from Kummer as we find it in tabloids. Words out of context, in isolation, are, to a large extent, vacuous, waiting to acquire a specific meaning from the wider and the narrow context, in particular from the collocates they co-occur with (Sinclair 1996). How should we imagine the mental concepts corresponding to lexical items in isolation, outside of their contexts? Would they not have to be similarly indeterminate? Or do they have a specified meaning, regardless of the context in which they occur, just as H2O is always the referent of the word water, regardless in which context it occurs? Is there anything to gain from a position that grief or sorrow correspond to one concept or two or more concepts which are universally the same for all humankind, or from a position that each native English speaker may have their own concept, or concepts, corresponding to grief and sorrow? Should we think people are aware of their concept (or their concepts)? Should we believe that people can be mistaken about their mental concepts? Can we trust their reports on their introspective quests? However we phrase such questions, it seems impossible to arrive at a consistent answer. For Noam Chomsky, it was a well-advised decision to exclude, over decades, the semantic component from his specifically grammatical investigations into the workings of the language faculty. That their different vocabularies keep me from understanding people speaking another language is, for the Chomsky of the Aspects of a Theory of Syntax, simply a surface phenomenon, something quite irrelevant from the perspective of a Martian linguist, as Chomsky never gets tired of telling us. Just as all languages share the same syntactic deep structure, Chomsky the mentalist seems to believe that they also share a common pool of mental concepts. Whenever he discusses the question of the universality of
What do we know about mental concepts?
57
mental concepts, he invokes the elusive innate language organ or language faculty, of whose reality he has no doubts. This has far-reaching consequences. In his article ‘Language and interpretation’ (Chomsky 2000b), Chomsky presents himself not only as a linguist but also (and perhaps primarily) as a philosopher of language. There he takes up, in modified form, Jerry Fodor’s claim that concepts are holistic and cannot be decomposed into more basic, primeval, concepts, and he agrees with Fodor that all concepts are somehow already present (even though dormant, like certain DNA sequences in our genome) in the human language faculty. ‘There is, it seems rather clear, a rich conceptual structure determined by the initial state broken down of the language faculty (perhaps drawing from the resources of other genetically determined faculties of mind), waiting to be awakened by experience’ (Chomsky 2000b: 64). In another contribution, ‘Language as a Natural Object’, reprinted in the same volume, he explains why this must be the case: The linkage of concept and sound can be acquired [by children] on minimal evidence … However, the possible sounds are narrowly constrained, and the concepts may be virtually fixed. It is hard to imagine otherwise, given the rate of lexical acquisition, which is about a word an hour from ages two to eight, with lexical items typically acquired on a single exposure, in highly ambiguous circumstances, but understood in delicate and extraordinary complexity that goes vastly beyond what is recorded in the most comprehensive dictionary, which, like the most comprehensive grammar, merely give hints that suffice for people who basically know the answers, largely innately. (Chomsky 2000b: 120)
This is, to say the least, highly speculative. If children around eight years of age really had a fully working vocabulary ‘with lexical items … understood in delicate and extraordinary complexity that goes vastly beyond what is recorded in the most comprehensive dictionary’) of 26,280 words (12 words × 365 days × 6 years) we should wonder why they do not put it to better use. Be this as it may, though, the underlying question is what it means when Chomsky says that children typically acquire a lexical item in a single exposure, which can, he insists, only be explained by the fact that they ‘basically know the answers, largely innate’. What exactly does the sentence ‘the concepts may be virtually fixed’ mean? On the same page, we also find ‘There is reason to believe that the computational system [of the mind] is invariant, virtually’. It is perhaps just one of those typical Chomsky-sentences whose main purpose seems to be to add rhetorical fervour to his argumentation but which should not be taken too literally in terms of its content, such as this sentence, also on page 120: ‘But there is evidence that the languages [English, German, Latin, Greek, Sanskrit and Chinese] have basically the same inflectional systems, differing only in the way formal elements are accessed by the part of the computational procedure that provides instructions to articulatory and perceptual organs.’ While Chomsky may not have been aware that the first five languages he mentions
58
Meaning, Discourse and Society
belong to the same family, he must have known that there are no data confirming his claim in the case of Chinese. However, his clever employment of hedging adverbs such as basically and virtually make every argument invulnerable. ‘Virtually all walls’, we could say, ‘can be painted black or white; therefore black and white are basically the same colour’. Unsurprisingly, Hilary Putnam strongly disagrees when it comes to the question of mental concepts: ‘Contrary to a doctrine that has been with us since the seventeenth century, meanings just aren’t in the head’ (Putnam 1981: 19, Putnam’s emphasis). This is what Hilary Putnam has been saying consistently over many decades. If meanings are not in the head, then the idea of innate mental concepts does not make sense: ‘A Chomskyan theory of the semantic level will say that there are “semantic representations” in the mindbrain; that these are innate and universal; and that all our concepts are decomposable into such semantic representations. This is the theory I hope to destroy’ (Putnam 1998: 5). This is how Putnam describes Chomsky’s (and Fodor’s) theory: Mentalists who follow Fodor’s lead are committed to the idea that there is an innate stock of semantic representations in terms of which all our concepts can be explicitly defined … How could such concepts as carburetor be possibly innate? Primitive peoples who have had no acquaintance with internal combustion engines show no difficulty in acquiring such concepts. On Fodor’s account this means that their ‘language of thought’ contained the concept ‘carburetor’ prior to their acquiring a word for that concept, even though nothing in their evolutionary history could account for how the concept ‘got there’. (Putnam 1998: 15, Putnam’s emphasis)
How could one possibly argue against so much common sense? Can Chomsky underpin his claim to the contrary? Can he repudiate Putnam’s injunction? He would not explicitly refer to it if he thought he could not: ‘Some, for example Hilary Putnam, have argued that it is entirely implausible to suppose that we have “an innate stock of notions” including carburetor and bureaucrat’ (Chomsky 2000b: 65). But Chomsky’s repudiation avoids straightforward argumentation. Instead he reverts to a parable: Notice that the argument is invalid from the start. To suppose that, in the course of evolution, humans come to have an innate stock of notions including carburetor and bureaucrat does not entail that evolution was able to anticipate every future physical and cultural contingency – only these contingencies. That aside, notice that a very similar argument had long been accepted in immunology: namely the number of antigens is so immense, including even artificially synthesized substances that had never existed in the world, that it was considered absurd to suppose that evolution had provided ‘an innate stock of antibodies’; rather, formation of antibodies must be kind of ‘learning process’ in which the antigens played an ‘instructive role’. But this assumption might well be false. Niels Kaj Jerne won the Nobel Prize for his work challenging this idea, and upholding his own conception that an animal ‘cannot be stimulated to make specific antibodies, unless
What do we know about mental concepts?
59
it has already made antibodies before the antigen arrives’ (Jerne 1985: 1059), so that antibody formation is a selective process in which the antigen plays a selective and amplifying role. (Chomsky 2000b: 65)
Thus in the very moment when we experience a particular stimulus or trigger, the corresponding mental concept comes to our rescue. The trigger can be the perception of something ‘real’, as in the case of a carburettor, or it can be an idea, like the notion that we are burdened by too much admin, as in the case of bureaucracy. Certainly not all cognitive scientists will be ready to follow him that far. I myself have no idea how realistic Jerne’s antibody theory is. Yet to believe that the whole infinity of future discourse objects is somehow already present in our DNA seems to overstress the suggestiveness of Chomsky’s charisma. Fodor himself, who shares with Chomsky the belief that concepts such as ‘bureaucracy’ and ‘carburettor’ have to be regarded holistically, and cannot be decomposed into semantic primitives, would, I think, hesitate to underwrite this claim, at least since the publication of his 1998 book Concepts: Where Cognitive Science Went Wrong. While he still strongly stresses the innateness in his earlier book, The Language of Thought, mentioned above, it seems now that he prefers not to become involved in this tricky argument. Concerning the innateness of the concepts ‘carburettor’ and his other favourite example, ‘doorknob’, he only tells us in his more recent book on concepts: ‘A lot of people have Very Strong Feelings about what concepts are allowed to be innate … [T] here is, at present, a very strong consensus against, as it may be, DOORKNOB or CARBURETTOR. I have no desire to join this game of pick and choose since, as far as I can tell, it hasn’t any rules’ (Fodor 1998: 28). In those few instances in which he is more specific, his peculiar metaphoric way of speaking makes it hard to pin down his true position. Concerning the concept ‘doorknob’ he explains: [W]hat has to be innately given to get us locked to doorknobhood is whatever mechanisms are required to come to strike us as such. Put slightly differently: if the locking story about concept possession and the mind-dependence story about the metaphysics of doorknobhood are both true, then the kind of nativism about DOORKNOB that an informational atomist has to put up with is perhaps not one of concepts but of mechanisms. (Fodor 1998: 142)
Thus Fodor implies that the question whether doorknobs correspond to an innate concept is the wrong question. For him, the important issue is the innateness of the mechanism that links concept and object. For him, this mechanism cannot be the result of evolutionary adaptation. In an article recently published in the London Review of Books, ‘Why pigs don’t have wings’, he rejects the adaptationist component of contemporary Darwinism. Evolution, he seems to say, progresses not so much by adaptation as by inherent properties of the genome (‘endogenous variables’), comparable to the dormant nature of mental concepts before they have been ‘triggered’. In his view,
60
Meaning, Discourse and Society
the classical Darwinist account of evolution as primarily driven by natural selection is in trouble on both conceptual and empirical grounds. Darwin was too much an environmentalist. He seems to have been seduced by an analogy to selective breeding, with natural selection operating in place of the breeder. But this analogy is patently flawed; selective breeding is performed only by creatures with minds, and natural selection doesn’t have one of those. The alternative possibility to Darwin’s is that the direction of phenotypic change is very largely determined by endogenous variables. (Fodor 2007)
Both Chomsky and Fodor posit mental concepts which exist independently of a person’s intentionality, of what this person is aware of and their plans to do something. People are reduced to automata, to computers carrying out providence’s programs. Not all cognitive linguists, not all researchers working on mental concepts, would agree with the holistic notions of Fodor and Chomsky. Many linguists, perhaps more on the Continent than in America, prefer a theory that assumes a rather limited number of semantic primes, or basic conceptual bricks, that are seen as the building blocks of all more complex concepts. The big advantage of such a theory is that while these bricks are innate and universal (in line with the tradition of Kantian a-priori ideas) natural languages and individual speakers can develop their own unique ways of looking at things. One of the cognitive linguists taking this stance is Anna Wierzbicka. In her model of the mind (1996), we find about fifty semantic primes, among them variables like sometimes, someone, something, verbs such as think, want, feel, say, happen, move, four adjectives: good, bad, big, small, nouns such as: part, kind, people, two pronouns: I and you, and a medley of connectors like where, above, after, if, because, etc. This, then, would be the translation of the natural language sentence ‘X felt guilty’ into a representation by semantic primes: X felt something sometimes a person thinks something like this: I did something because of this, something bad happened because of this, this person feels something bad (http://rhm.cdepot.net/knowledge/theory/ NaturalSemanticMetalanguage/definition.html) Of course, Anna Wierzbicka is well aware of the problematic nature of such a theoretical construct. If there ‘really’ were language-independent concepts, of a more primitive or even of a complex nature, either holistic or not, how would we know what they ‘mean’ and how they would translate into a natural language? For her readers, Wierzbicka has kindly translated them into English. Or perhaps these expressions only look like English, but are in fact a language-independent code, the ‘real’ mentalese. Perhaps only in Texas it is the common understanding that English is the language of thought (cf. Texas
What do we know about mental concepts?
61
Governor Ma Ferguson, who is famously quoted as saying ‘If English is good enough for Jesus, it is good enough for Texas.’). What semantic primes mean outside of God’s own country we will only know once they have been translated into those natural languages. As long as we think of them as languageindependent, their meaning remains hidden from us. We can only come to terms with them once they are translated into a natural language, and we are confronted with the rather serious problem that we will never be able to control the appropriateness of this translation. Furthermore they are, in translation, of necessity as ambiguous as natural language tends to be. Thus we are left in doubt of the ‘true’ meaning of Wierzbicka’s concepts. Can we really be sure that there is a distinct single mental representation of the English verb think in every human being’s mind? Can we be sure that all other natural languages have equivalents for think, corresponding to the pertinent mental concept? As I see it, there is no method that could possibly tell us if her translation of the complex mental concept given above into X felt guilty is correct. Neither can we be sure that it is ‘really’ true that bad in ‘something bad happened’ is the same bad as in ‘feels something bad’. Is X felt guilty really the same as German X fühlte sich schuldig? Does it mean the same as X had a bad conscience, X had pangs of conscience, X had a sense of guilt, X felt remorse and X repented? Are these just synonyms, or do they correspond to different concepts? Wierzbicka would point out that she is only sketching a model of mentalese, not mentalese itself, and that this model is perhaps not even functionally equivalent to the ‘real’ mental representation. Widely read as she is, she has repeatedly related her approach to Leibniz, for example in this quote: ‘Im wesentlichen geht diese Idee auf Leibniz zurück und auf seine Vorstellung “eines Alphabets menschlichen Denkens”, das heißt, einen Katalog der Begriffe, die aus sich selber verstanden werden können, und aus deren Kombinationen unsere anderen Vorstelllungen entspringen’ [In all relevant aspects this idea is based on Leibniz and on his model of an ‘alphabet of human thought’, i.e. a catalogue of concepts which can be understood out of themselves, and whose combinations engender our other ideas] (www.humboldt-foundation.de/kosmos/titel/2002_003.htm). Even if Leibniz never distanced himself from the youthful folly of his doctoral dissertation Ars Combinatoria (1666), it was as unsuccessful as all other endeavours in the last millennium in constructing a perfect language. This is the sobering conclusion we can draw from Umberto Eco’s The Search for the Perfect Language (1995). While most other linguists exploring models of a mentalese of basic concepts present their concepts as isolated items that can be assembled in a list, Wierzbicka’s original contribution is her suggestion of a mentalese syntax that informs the relationship her semantic primes have with each other when they are composed into a more complex mental concept. Is this how we should imagine mentalese, the language of thought? Do we find there the same
62
Meaning, Discourse and Society
categories that we use when we describe natural languages? Does the context in which it is embedded affect the way in which a basic concept has to be understood? Do we find there finite verbs, as we have them in English, but not in Chinese? Ray Jackendoff seems to know that the system of cognitive representations, mentalese, or, in his terminology, ‘conceptual structure’, does not have parts of speech: ‘Whatever we know about this system, we know it is not built out of nouns and verbs and adjectives’ (Jackendoff 1997: 31). This is not how Steven Pinker sees it (see above). If cognitive linguistics were a science comparable to physics or chemistry, there would have to be an agreed methodology to decide between such irreconcilable assertions. But perhaps our claims about semantic primes and conceptual structures are more a conjectural hypothesis than empirical knowledge, just like our hypotheses concerning life after death.
4
Morphing theoretical sémes into ‘real’ concepts
If we disregard Wierzbicka’s mental syntax, we may well compare her semantic primes to those which were at the core of the mainstream Continental European semantic theories of the 1960s and the 1970s. Usually we trace this semantic feature theory back to Louis Hjelmslev’s Prolegomena (Hjelmslev 1963 [1943]). His phonological analysis and his concept of the phoneme became the model for semantic analysis and the concept of the séme, as it quickly became popular in the linguistics of Romance languages. Bernard Pottier combined Hjelmslev’s approach with the Prague school of structuralism. He was also the first to call the ‘distinctive semantic features of lexemes’ sémes. This is how he describes the meaning of chair: chair: {s1, s2, s3, s4} (‘to sit on, on legs, for one person, with a backrest’). Relative to the set containing easy chair, chair is defined as without the séme s5 (‘with armrests’) and so on. (Pottier 1978: 86, my translation) Thus meaning can be analysed in terms of differences, through the presence or absence of sémes. It is this focus on difference which grounds this theory in de Saussure’s structuralism. Algirdas Julien Greimas also uses the concept of sémes (Greimas 1983 [1966]: 22ff.). He distinguishes between the presence of a séme, the negation of this presence (‘negative séme’) and a state in which a given theme is neither present nor absent (‘neutral séme’). For Pottier and Greimas the séme is thus the smallest feature (trait distinctif) to distinguish the meaning that accounts for the difference between one word such as chair and another, semantically related lexical item, easy chair [French chaise vs. fauteuil; German Stuhl vs. Sessel]. Sémes here are understood as heuristic constructs. In the second edition of Theodor Lewandowski’s Linguistischem Wörterbuch (1976) we find this entry for semantisches Merkmal [‘semantic feature’]: Bedeutungsatom, Bedeutungskomponente, Element des Begriffs bzw. Inhalts, der als in sich (mikro)strukturiert aufgefaßt wird, Basis-Element und Konstrukt einer semantischen Theorie, das sich mit Konstrukten wie Atom, Gen usw. vergleichen lässt. Bei 63
64
Meaning, Discourse and Society
der Konzeption des semantischen Merkmals handelt es sich um eine Übertragung des Prinzips der distinktiven Merkmale auf den Bereich der Semantik … Bei Bierwisch (1967: 3) sind semantische Merkmale ‘certain deep seated, innate properties which determine the way in which the universe is conceived, adapted, and worked on.’ [An atom or a component of a meaning, an element of a concept of content which is understood as having a (micro-)structure, the basic element and construct of a semantic theory comparable to constructs such as atoms or genes. This conception of the semantic feature is a transference of the principle of distinctive features to the field of semantics. For [Manfred] Bierwisch, semantic features are ‘certain deep seated, innate properties which determine the way in which the universe is conceived, adapted, and worked on.’] (Lewandowski 1976: 3, 663)
What is interesting in this entry is the naivety with which semantic features (sémes) are fused with the semantic primes we find in cognitive linguistics. For these features are, unlike the primes, no more than theoretical constructs within a linguistic model to which no ontological reality is ascribed. This is the understanding we find in the entry Merkmal [‘feature’] in Lewandowski’s linguistic glossary: Begriffliches Konstrukt, ein Begriff, der für das richtige Verstehen der Sprachstrukturierung unentbehrlich ist (Martinet); für die Konstruktion und Funktion sprachlicher Einheiten als notwendig betrachtete begrifflich-hypothetische Mikroelemente. [Theoretical construct, a concept indispensable for the proper understanding of structuring language (Martinet); conceptual-hypothetical micro-elements considered as essential for the construction and function of linguistic units.] (Lewandowski 1976: 2, 446)
Would Manfred Bierwisch agree? Developing further the contention put forward by Jerry Fodor and Jerold Katz, namely that it is possible to ‘construe a meta-theory containing a list of semantic features from which we can take the theoretical vocabulary of any special semantic theory’ (Katz and Fodor 1963: 208), Bierwisch explains: This does not mean, of course, that the dictionary of each given language must show exactly the same distinction as that of any other language. It implies only that, if a distinction is made, that property can be characterized in a nontrivial way in terms of a universal set of semantic markers. If we accept this view, then two different questions immediately arise: What is the theoretical status of the universal semantic markers; how must they be interpreted? What are the elements of the universal set and how can they be established? … The question here is: in what way, by what type of phenomena, are they motivated outside of the structure in the narrower sense? In other words: what is the interpretation of semantic markers, how are they connected with thought? There are good reasons to believe that the semantic markers in an adequate description of a natural language do not represent properties of the surrounding world in the broadest sense, but rather certain deep seated, innate properties of the human organism and the perceptual
Morphing theoretical sémes into ‘real’ concepts
65
apparatus, properties which determine the way in which the universe is conceived, adapted, and worked on. (Bierwisch 1967: 3)
(The German version here reads: ‘welche Beziehungen bestehen zwischen ihnen und den kognitiven und perzeptiven Leistungen des Menschen?’ [‘which relationships exist between them and the cognitive and apperceptive acts of a person?’] (Bierwisch 1967: 2, 1970: 270–1).) Bierwisch, it seems, is no longer troubled by the question of whether the semantic features are theoretical constructs of the linguist, which can be derived from the analysis of a natural language. For him they have become ‘real’ objects, located in human cognition. Are they learnt or inherited? He is obviously very sure: ‘Not only is there no reasonable explication of how semantic markers are learned. It is also very difficult to explain in a natural way such well known facts as displaced speech, fictitious objects and in general all gaps between meaning and reality’ (Bierwisch 1967: 3). For Bierwisch, there is no alternative to the innateness option. This was not so unexpected at a time when the attraction of Chomsky’s model had its first peak in Europe. Linguists in East Germany and other Eastern European countries in particular embraced it because it located their field securely within the sciences, outside of the Geisteswissenschaften, with their suspected bourgeois affinities. As real scientists, linguists could thus exempt themselves from the obligation to justify their approach by adopting a specifically Marxist–Leninist perspective. Bierwisch fully subscribed to this fairly pervasive Anglo-Saxon entrenchment in a realism born out of common sense, which still to this day transforms large parts of American philosophy into an exercise in various forms of realism, thus making relativism appear almost a swearword. This abyss between (largely) Anglo-Saxon realism and (not only) Continental constructivism, in the guise of nominalism, hermeneutics or (post)structuralism, is rarely bridged. Bierwisch’s semantic features no longer correspond to the sémes hypothesised by Pottier or Greimas. Lewandowski’s dictionary passes silently over this crucial difference, namely that sémes are viewed as the linguists’ constructs, derived from the analysis of a natural language by applying heuristic procedures, while the cognitivists’ mental concepts are seen as ontologically real entities. Whether it makes sense to posit a given séme in order to account for the difference between two semantically closely related words can be negotiated. But whether there is an innate mental concept meaning ‘with an armrest’ is not a matter for discussion; it has to be proven in a scientific sense. It is this insistence on ontological reality as opposed to hypothetical models that distinguishes the programme of cognitive semantics from that of structural semantics. Behind innate mental concepts we find looming the postulation of universality. No longer are we analysing a particular natural language, but the language of thought, the mother of all languages. Does this unverifiable claim really help us to reveal the mystery of meaning? What do we gain if we set out to
66
Meaning, Discourse and Society
investigate some elusive model of the workings of the mind, rather than sticking to the real language data to which we share access? It does not matter how we have to imagine mental concepts. They may be holistic, as in the case of Chomsky’s example of the carburettor, or we may imagine them as concepts composed out of semantic primitives; there is never any kind of empirical evidence about them that could be objectified. Perhaps this is the most plausible explanation for the lack of consensus among cognitive linguists as to the nature of these mental/cognitive concepts/representations.
5
From mental representations to conceptual ontologies
Natural languages leave a lot to be desired. They are, as we have learned, full of vagueness and ambiguity. They are subject to constant change, and under closer scrutiny seemingly regular features tend to get lost in an almost infinite diversity of regional, situational, social, genre-specific and domain-specific variation. We are always encountering language usages which appear foreign to us. For thousands of years people have been complaining about the decay of language. Often when we question language use (normally the way other people use language) we look into the past. Then we ask ourselves what the word in question ‘really’ means, and for really we could read originally. Apparently we long for a Golden Age when there was still a natural, uncorrupted relationship between the word and what it stood for in the world outside. This explains the popularity of etymological dictionaries. But no matter how far we go back, we never seem to arrive at the ‘true’ meaning. Like everyone else, Plato, in his dialogue Cratylus, could not answer convincingly when asked what makes such a relationship natural. To Socrates’ question ‘For the gods must clearly be supposed to call things by their right and natural names, do you not think so?’ Hermogenes responds: ‘Why, of course they call them rightly, if they call them at all’ (Plato, translated by B. Jowett on Project Gutenberg online). In which language do the Olympians converse among themselves? Do we have to imagine these ‘natural names’ as equivalent, not to words of a natural language, but to the language of thought, the universal mental concepts? In this context it might be worth having a look at machine translation and artificial intelligence. Perhaps there we can learn why concepts seem to be so much more attractive than natural language expressions. For these are fields in which concepts have been key features from an early stage. These are not the mental concepts we have encountered so far. They are akin to the concepts we find in terminology. These concepts are language-independent units of knowledge defined by a unique combination of characteristics. The key characteristic of such concepts is that they are context-free. Their definition will always be the same, regardless of where they are found. It does not matter whether they are denoted by an English, a French or a Turkish term: all terms refer to the same concept. Concepts refer to abstract or concrete things in the ‘real’ world. They 67
68
Meaning, Discourse and Society
can be defined by linguistic expressions as long as they are unambiguous; by formulae, as in chemistry; by images, or by their position in a standardised, i.e. universally accepted, taxonomy, like the Linnaean system. The definition of a concept can only be changed by a rightfully appointed group of experts. Concepts relate to each other in defined ways. The concept ‘mammal’ is the hypernym of the concept ‘cow’; ‘the concept ‘cow’ and the concept ‘donkey’ are co-hyponyms; the concept ‘foot’ is a meronym of the concept ‘body’, etc. All concepts, together with their relationships to each other, are called a conceptual ontology. It is obvious that the computer can process a text consisting of concepts much better than a text consisting of unruly natural language words. Translation becomes a simple algorithmic task, and it is not necessary for the computer to know what it is translating. These concepts are not the mental entities discussed above; they are the staple fare of terminology. Ever since its introduction as a new discipline in the early decades of the twentieth century, the necessity of terminology as such has never been in doubt. In engineering, it is the essential means of ensuring that a finished product is the exact replication of its blueprint. No ambiguity or fuzziness is permitted when the development phase gives way to the production process. The engineers and skilled workers involved have to know the exact meaning of the instructions they follow. This is even more true in the case of distributed production processes. While the blueprint may have been developed somewhere in a head office, various parts will then be produced at different production sites and finally be assembled in a special plant. Only stringent terminology makes it possible that at the end all the parts fit together and form the intended result. But terminology is the exception. If it could be applied generally to natural language texts, if it were possible to cut out all the ambiguities and fuzziness, then all the translations of Shakespeare into German would be identical. They are not. The meanings of the words and phrases and sentences used by him are still, after 400 years, negotiated by thousands of scholars. There are no straightforward, unchangeable and clearly defined concepts into which they could be converted. The crux is that there is, apart from Fodor’s language of thought and Pinker’s mentalese, no language-independent way to define something. As hard as one may try, the words used in any definition always have a tendency to be ambiguous, and definitions often do not exactly map onto each other from one language to the next. Translating natural language is not always simple. On the other hand, terminological concepts can be easily processed by computers, at least in principle. Terms can be translated on a one-to-one basis and they can be used in information retrieval and knowledge extraction. Unlike terms, however, natural language words pose problems. They are fuzzy and ambiguous. A word in one language hardly ever maps onto a word in another language. In each occurrence of a word, its meaning is largely determined by its
From mental representations to conceptual ontologies
69
context. This ambiguity and this context-dependency is something we, the language users, hardly ever notice. This is because we are aware of what is being said, of its meaning. Computers process language without being concerned with meaning. They process without knowing what they process. Like a huge clockwork delivering its tasks, driven by the cogs of countless wheels, computers (or computer programs) work by permuting, rearranging and transforming the concepts the sentence strings consist of. While for us these concepts are symbols in need of interpretation, for the computer they are just meaningless, uninterpreted elements causing certain permutations to happen. Therefore for a long time the standard approach to the machine translation of natural language has been to convert natural language words into the concepts of a consistent conceptual ontology. This way, people thought for a long time, everything disturbing, unclean, fuzzy and ambiguous could be filtered out, so that we were left with nothing but the true, authentic, uncorrupted meaning of a concept. Even though this expectation has rarely led to the envisaged results, it explains the popularity of conceptual ontologies in the artificial intelligence community. The idea is that once fuzzy, ambiguous language is turned into orderly and formalised conceptual representations, the job of language processing can be passed on from traditional linguists with their aspirations for ever-increasing complexity to modern, sober, task-oriented language engineers. Building an expert or a machine translation system then becomes similar to producing a new car out of its many parts. Even robots can be trained to do that. Concepts in language engineering are thus seen as the language-independent, spiritual, angelic natures of natural language words which have become unclean in the cause of being used by a sinful humankind. The names Adam and Eve that were taught in Paradise have gradually changed their original meanings. The Tower of Babel symbolises the decay our languages have since undergone. What is needed is a rescue operation that will undo this process of contamination. This is how the difference between the word and its concept is commonly described: Ontologies describe concepts, not the way these concepts are expressed in words in a natural language. Therefore it is usually assumed that the ontology is languageindependent. (Weigand 1997) Concepts represent the abstract meanings of words, and lexical entries represent the surface realizations of these meanings … Concepts represent word meanings, whereas the lexical knowledge they have represents ways to express these meanings with words. (Agnesund 1997)
Concepts, it seems, are pure meanings, cleansed from the impurities which they contracted through the contingencies of change afflicting natural languages. In these ‘language-independent’ ontologies there is no room for doubt as to what is a concept, how it is defined and how it is related to other concepts. Concepts are neither fuzzy nor ambiguous. Every proposition is either correct, ‘grammatical’,
70
Meaning, Discourse and Society
or not. If machine translation is still unsatisfactory then, according to this claim, it is only because we still have problems in converting natural language sentences into their conceptual representations. I have to admit that this approach can work. Today, many aircraft maintenance manuals are machine translated, and experts claim that this process is less error prone that human translation. One of them, John Smart, has this to say: In 1988, the European aircraft alliance, called Airbus Industrie[s], followed Caterpillar and developed a Controlled English vocabulary to write aircraft maintenance manuals. Today, the ASD-STE100 Simplified Technical English is widely deployed by 800 airlines and mandated by Airbus, Boeing, Embraer, Bombardier and other aerospace companies. To avoid product liability, these companies must mitigate the risk of human error in maintenance tasks. For example, ASD-STE100 Simplified Technical English does not contain the common English auxiliary verbs, might, should or may. These verb forms do not exist in other languages. The ASD-STE100 Simplified Technical English allows the use of valid part names and terms in addition to the basic 985 words. You are not reading a typographical error. There are only 985 basic terms. However, the dictionaries expand by component with terms for hydraulic systems, avionics and engines. An average STE vocabulary can reach 8,500 to 12,500 terms. (Smart 2006)
The language that is translated by computers is a ‘controlled’ language. It is free of synonymy. Only one conjunction of the kind and is allowed per sentence, and sentence length is strictly limited. For controlled language, machine translation works. But machine translation, or expert systems, cannot process the kind of general language we tend to find in texts elsewhere. Even though the language of administration and law is full of terminology, this terminology is much more flexible than the language used for detailing an industrial production process. EUROTRA, the overly ambitious European project for machine translation of texts emerging from the European institutions from and into all the EU languages, which cost taxpayers hundreds of millions of euros between the late 1970s and 1994, never led to a functioning product (www-sk.let.uu.nl/stt/ eurotra.html, http://en.wikipedia.org/wiki/Eurotra). Could it be that the whole approach to the construction of an interlingua consisting of languageindependent concepts was faulty? Alan Melby, one of the founding fathers of machine translation, gives a convincing account of the reasons why conceptbased machine translation cannot work for normal, unrestricted language, in his co-authored book The Possibility of Language (Melby and Warner 1995). In her paper ‘Machine translation: its past, the potential and the problem’, Nancy Brigham sums up his view in these words: The failure of the interlingua model provides insight into the difficulties of machine translation. Alan Melby was part of a team that, beginning in 1970, spent five years reducing language to sememes (language-independent concepts) that could be translated into several target languages. By 1978 he hit a ‘Wall,’ and concluded that ‘the language-independent universal sememes we were looking for do not exist!’ Melby
From mental representations to conceptual ontologies
71
says that the interlingua researchers failed ‘because of a fundamental misunderstanding of … languages.’ ‘I had a degree in mathematics,’ explained Melby. ‘The world was supposed to be a nice, tidy place.’ But that approach ‘leave[s] true creativity behind. All meaning becomes mechanical combinations of atomic word senses.’ He developed a new respect for the Whorf-Sapir hypothesis that languages are not interchangeable and one’s world view is influenced by one’s language, and his continuing work focuses on productivity tools for human translators. (Brigham 2000)
Mutatis mutandis, this analysis also holds for artificial intelligence projects using a conceptual approach. The most lavishly funded of them was CYC, designed to develop common sense skills and thus be able to emulate human reasoning. In spite of billions of dollars poured into it, it never achieved its goals. The key problem is the proposition that we can construct a conceptual ontology, a language-independent system that would be a true mirror of the reality out there. But is that reality really accessible to us? Even in New Scientist, a publication almost invariably inclined to herald technological progress, we find an article on 15 April 2006, expressing some doubt: It also begs a crucial question: can a computer really acquire common sense without experiencing the world directly, given that the latest neuroscience shows we base our judgements on gut feeling and emotion rather than a rational assessment of the facts? (www.newscientist.com/channel/opinion/mg19025471.700-the-word-common-sense. html)
Others, like Justin Mullins, writing in the same journal just about one year earlier, still cling to their indefatigable optimism: For years researchers have been promising to deliver technology that will make computers we can chat to like friends, robots that function as autonomous servants, and one day, for better or worse, even produce conscious machines. Yet we appear to be as far away as ever from any of these goals. But that could soon change. In the next few months, after being patiently nurtured for 22 years, an artificial brain called Cyc (pronounced ‘psych’) will be put online for the world to interact with. (Mullins 2005)
How exactly do we expect such a system to perform? It may well be able to tell us, in a not-too-distant future, the average temperature on 7 July 2008 in Birmingham, Alabama. But will it ever answer reliably the question: ‘Is there a monster in Loch Ness?’? Conceptual ontologies can tell us a lot about the ways in which their compilers view the world. But what is a beast to some of us may well be a beauty to others. What ontology engineers such as Mattias Agnesund and Hans Weigand surmise about the relationship between words and concepts surely also accounts for the attraction of the innateness theory. Both camps believe that language ceases to be unruly once it has been transferred into a language-independent universal representation. As soon as a natural language sentence is translated into mentalese, we
72
Meaning, Discourse and Society
have something similar to a mathematical equation, an expression that can be decided on the basis of its formal properties. In addition, it can claim reality and universality. In formal calculi, we observe the workings of immutable laws. The linguistics of formal languages is a pure science as close to mathematics as one can get. Is understanding really that simple? Leaving aside, for the moment, the question of whether concepts are innate, holistically acquired or composed of inherited semantic primitives, what then is the ‘real’ content of the concept ‘carburettor’? Is it enough to know that a carburettor is an important part of an internal combustion engine which has to be repaired or replaced when it stops working, or does it include a comprehensive representation of its functionality? Is the meaning of the word carburettor identical with the (content of the) concept, or, if not, what is the difference? What could we gain from an analysis of the concept ‘carburettor’ that we could not gain from the analysis of the word carburettor? What is the use of mental representations of concepts, for linguists? One reason, I believe, why linguists do not like to ask these questions is that they long for firm ground, for formulae as simple as the ones they find in the laws of physics or chemistry. There is, however, a fundamental difference. The objects of the natural sciences do not talk back. The scientific community can define what hydrogen or uranium is as they see it fit. They can make sure that everyone using these terms in internationally refereed journals uses them in congruence with the language-neutral definition agreed upon. The meanings of words, though, are different. They are not given, and they are not subject to laws which would place them outside the control of the members of the discourse community, who use them for their communication. In the course of communication, people may find it useful to negotiate content by paraphrasing the meanings of the lexical items they are using. Some of these negotiations may be so successful that they are taken up by subsequent contributions to the discourse. This is how meanings can change. Meanings, we have to remind ourselves, are strictly contingent. We have no way of predicting what a given lexical item may mean next year. This is not how cognitive linguists look at meanings. For them, the meaning of an utterance is its representation in the mind, a representation that is outside of the discourse and not affected by it. It is the result of the processes carried out, without us being aware of it, by the language mechanism we all have in our minds.
6
What is meaning?
In the past, many linguists had such a respect for meaning that they were careful to avoid the issue wherever possible. Traditionally, language study has had a strong focus on grammar. Grammar is a land of apparent law and order, in which every part ultimately finds its pigeonhole, if one sorts the elements for long enough. Words and their meanings, however, behave in a disorderly fashion. Words are ambiguous and fuzzy, as any glance at a dictionary entry will confirm. Most words, particularly the more frequent ones, come with more than one ‘sense’, and there is little agreement between dictionaries as to how many senses a given word has, or how these senses are distinguished. This may be one reason why many linguists in the twentieth century had so little patience with them. Mental concepts, on the other hand, were believed to be failsafe in this respect. Each concept was thought to represent one ‘sense’, clearly defined and unambiguously related to other concepts (such as hypernym–hyponym, part–whole, perhaps even synonym–antonym). To many linguists it therefore seemed prudent to leave the words of our natural languages to their poor cousins, the lexicographers. But this has not prevented the linguists from criticising the lexicographers for the inconsistencies that abound in even the best dictionaries, whenever the makers of our dictionaries try to make sense of their words. For corpus linguistics, a fairly recent paradigm in language studies that has evolved over the last forty years, this picture is no longer true. Corpus linguists are not concerned with mental concepts. For them, the meaning of a lexical item can only be studied in real language data, in the texts in which it occurs, in the contexts in which it is embedded. Indeed, corpus linguists maintain that the meaning of a word (or, better, a lexical item, regardless of how many words it consists of) in a given discourse (or the corpus representing that discourse) corresponds to the ways it is being used. Once words are embedded in contexts, people are only rarely aware of any ambiguity or fuzziness. If we read a text we do not usually have the problem of not knowing what the words are supposed to mean. Only if we look at these strings of alphabetic characters, with spaces before and after them, in isolation, do we start to wonder whether bank means ‘river edge’ or ‘financial institution’. Guided by the contexts in which 73
74
Meaning, Discourse and Society
they occur, we can even make sense of unfamiliar words. When the word globalisation became more frequent in the Guardian, jumping from 19 to 48 annual occurrences in the year 1993, the contexts in which this word was embedded can be read as paraphrases of its meaning, as in these citations: There is a certain globalisation going on in the world right now, the result of technological advance, the fall of communism, the rise of business in Asia The globalisation of market forces has already undermined local and regional ways of life in many parts of the world other trade initiatives that will increase globalisation and threaten the environment and future generations’ quality of life McDonald’s is more than just an horrific example of globalisation, or how to make everyone in the world eat the same soggy hamburger. The globalisation of business and finance means that we need new policies for international co-operation not just in Europe, but worldwide GLOBALISATION has rapidly become a fashionable concept among social scientists and in business and management literature. He said that the globalisation and de-industrialisation of the US economy had devastated the poor black working class.
For corpus linguists, the meaning of globalisation is given by the entirety of the contexts in which the word is embedded. Not so for cognitive linguists. For them, meaning is not in the discourse; rather, it is in people’s heads. They view traditional linguists, the philologists for instance, as sitting in a dark cave looking in one direction only, toward the back wall of the cave. Behind them is an open fire providing light, and between the fire and where they are sitting there is a catwalk on which the mental concepts move, casting lexical shadows on the wall. These shadows are all that the traditional linguists see. As long as they stay in this position, they take the verbal silhouettes for the stuff meaning consists of. If they only could turn around, they would be confronted with something more real, with mental concepts or cognitive representations. Regardless of whether these things are metaphysical Platonic forms, i.e. what Plato calls eidos or idea, or whether they are just models of the ‘real’ things, accepting them in lieu of words would be seen as a step in the right direction, a step cognitive linguists claim they have taken. They believe they are the first to have realised that the meaning of a word is its mental representation, the cognitive concept for which it stands. Progress, however, may not end there. Neurolinguists are already trying to identify the specific neurons in our brains that fire once the concept is thought. Thus, we are told, we can now hope that in a not-too-distant future the last obstacle to truth will be removed, namely the divide between brain sciences and mind sciences. Then we will have grasped what is really out there in the uncharted territories in our heads: the neurons representing mental concepts. In what follows, I will explore the key element of cognitive linguistics, the mental/cognitive concept/representation as the embodiment of meaning. To the
What is meaning?
75
extent that there is a common philosophical and theoretical basis to what is now called cognitive linguistics, including Cognitive Linguistics, we will find it in the cognitive sciences, as they have developed after the demise of behaviourism, and in the philosophy of mind. Many of my arguments refer to these foundations rather than to specific contemporary schools of cognitive linguistics, of which there are many. Because cognitive linguistics is far from being a uniform discipline, it will always be possible for one school or another to maintain that my charges do not apply to them. I am more concerned with the basic idea driving cognitive linguistics, namely that we have to look for meaning in the mind, than with individual schools. My goal is to show that this is a hopeless enterprise. These are some of the reasons: The mind is, as Richard Rorty calls it, a ‘blurred’ construct. ‘We do not have any data about the nature of the mental. In particular, we do not have any knowledge of what it is to have a mind by looking inside and inspecting our own’ (Rorty 1982: 325). There is no scientific evidence for an entity called the mind. All we are left with are models; e.g. the mind as the software in relationship to the brain as the hardware. The only content that can be negotiated between people is symbolic, content that is, or can be, verbalised. If you want to know what is in my head I will have to tell you. There is no non-linguistic access to my mental content. All public knowledge about mental representations is knowledge that has to be available in the form of natural language in order to be exchanged and shared. Some cognitive linguists draw the main dividing line between one-level semantics and two-level semantics. Those who subscribe to one-level semantics, for instance Jerry Fodor and Ray Jackendoff, deny that there is a systematic difference between what is being said and the mental representation of what is being said. According to this school, a natural language sentence such as: The globalisation of market forces has already undermined local and regional ways of life in many parts of the world. is as rich in meaning as its mental representation. The meaning of the word globalisation corresponds directly to the mental concept ‘globalisation’. Two-level semantics, on the other hand, holds that mental representations are richer and more specific, and certainly not isomorphic with the meaning expressed in natural language utterances. Stephen Levinson seems to make a distinction between two-level semanticists like himself, and ‘Cognitive Linguists’, a label he applies apparently only to one-level semanticists (Levinson 1997: 24). There is a multitude of views as to the nature of the mental concept, and there are different terminologies. Looking at these texts from the
76
Meaning, Discourse and Society
perspective of discourse studies, I find myself unable to distinguish between concepts and representations, and between what is called mental and what is called cognitive. Mental/cognitive concepts and/or representations all seem to refer to more or less the same idea. However, much of what I present here has been put forward in irreconcilably different accounts. For instance, Dan Sperber and Deirdre Wilson, the chief exponents of (cognitive) Relevance Theory, identify three views as to the number of concepts: there could be fewer concepts than there are words (the compositional view, argued by Anna Wierzbicka and also Stephen Levinson, among others); there could roughly the same number (a view to which Jerry Fodor and Steven Pinker seem to subscribe); or there could be infinitely more concepts than words (as Sperber and Wilson themselves argue (Sperber and Wilson 1998: 186f.)). Who is right? As long as there is no incontrovertible evidence of a mental concept, as long as mental concepts remain as elusive as Iraqi weapons of mass destruction, anyone’s guess is as good as anyone else’s. The cognitive landscape invites all kinds of sectarianism. My main claim is that we know too little about what is going on in our heads to find uncontroversial answers concerning the connection between thinking and language. Dreaming is something that happens to us. We do not consciously design our dreams. Thinking, on the other hand, is a mental activity requiring agency, as long as we believe we are in charge of our thoughts. (There was a time when people used to say ‘methinks’, perhaps disclaiming responsibility for their thoughts.) But no one has ever been able to show us what a thought is. Thinking, for me at least, involves consciousness, and consciousness involves intentionality. But the nature of thought is enigmatic. I find myself unable to establish at what point a thought becomes a thought. Is there something before something becomes a thought? What is the nature of a thought once it has become one? Are thoughts symbolic? If they have content, does this content also have a form? Can there be such a thing as pure content? The content of a thought is not fixed until the moment it is turned into language, whether it is internal speech, or a report to others. If there is more to a thought than what can be verbalised, how could we access this extralinguistic content? Is there a way to communicate it? Cognitive linguists, like all linguists, have to start with linguistic expressions. But they contend (and who could possibly argue?) that a speaker’s utterance must be caused by a thought, and that its intended destination, in the hearer’s head, is again a thought. Can we learn to travel that route? To me it seems there is no easy way to resolve these matters. The way I see it, the mental/conceptual concepts/ representations I find in cognitive accounts are no more, at their very best, than a duplication (and sometimes even a triplication) of what we can find out about the meaning of a lexical item, about the meaning of what has been said. Cognitivists are ‘interested ultimately’, in the words of Ray Jackendoff, ‘in the manner in which language is embodied in the human brain’. They are not
What is meaning?
77
interested in language as ‘an abstract phenomenon or a social artifact’. There ‘might be properties that language has because of the social context into which it is embedded’. But there are also ‘some important properties … that can be effectively studied without taking account of social factors’. Is this ‘mental stance’ a stance concerning the brain or the mind? While Jackendoff calls language a ‘mental organ’, his ultimate goal seems to be to locate it in the brain. ‘Language, vision, proprioception, and motor control’ all are ‘instantiated by neurons of basically similar design’. Thus, in the long run, we should study language as a ‘physical organ’ (all quotes from Jackendoff 1997: 2ff., Jackendoff’s emphasis). Should we assume, then, that mental concepts or cognitive representations have a physical reality? And will our quest for meaning have come to a conclusion once we have identified the neurons correlating with a linguistic expression? We use language firstly to interact with people and secondly to give our linguistic and non-linguistic interactions a meaning. Our societies are much more complex than those of our closest relatives, the non-human primates. This complexity requires not only a division of labour, but also distributed knowledge. We must negotiate who carries out which task, and we must provide the necessary knowledge. Language enables us to trade content. For people like me who are not interested in the language system as such but in the discourse, in what is actually said, language is, first of all, a social phenomenon. Language is public. Language is observable. Language means. Symbolic content does not exist as such in (spurious) minds. It must be created. It is the unpredictable result of social interaction, of negotiation between the members of a discourse community. If we find symbolic content outside of language, we need language to tell us what it means. No pictogram, not even the chevron, is self-explanatory. Dogs do not understand them. The person or people who invented them, and those who made them popular, must at some point have told the other people their meaning. Meaning always presupposes society. It can only be assigned in a discourse. What something that has been said means is not a matter of speculation but of negotiation. When someone has said something, we can discuss its meaning, whether we are its addressees, or have learnt about it in some other way. We may even ask if what we heard was really a linguistic utterance or just someone clearing their throat. We may not necessarily agree. It can be a discussion without a conclusion. But beyond our various interpretations of this statement, concordant or discordant as they may be, there is nothing to be discovered about something someone has said and its meaning. Meaning does not correspond to a physical reality, unless through the agreed use of an arbitrary sign. To know which neural processes and which hormonal outpourings lead to a statement, and which neural processes and hormonal outpourings it triggers in the people who are addressed, does not make us understand the statement any better. Neither does it support our quest for
78
Meaning, Discourse and Society
meaning if we relate what is said to a mental representation that will always remain outside our reach. Meaning, just like language, is public. It belongs to the sphere of social interaction, not to the realm of mental processes. This is, however, not the way cognitive linguists look at meaning. The German linguist Gisela Harras, who favours two-level semantics, distinguishing ‘semantic content’ from cognitive or mental concepts, gives an example intended to persuade us that a linguistic utterance does not tell the full story. She compares the two sentences ‘Open the bottle’ and ‘Open the washing machine’ and concludes that a verb such as open can convey an almost unlimited number of concepts (Harras 2000). The semantics of the verb, she maintains, do not tell us what open means when applied to different things. The bottle may have a screw top or a cork. The plumber opens the washing machine in a different way from the normal user. When we, as addressees, have to interpret the different utterances containing the verb open, we have to go beyond its semantic content, Harras says. The hearer has to apply cognitive mechanisms that will tell them which of the many concepts expressed by the verb open the speaker intended. For Dan Sperber and Deirdre Wilson, to whom Harras is indebted for her example, these concepts seem to be individual rather than universal: A concept, as we understand the term, is an enduring elementary mental structure … It is arguable that each of us has ineffable concepts [e.g. of a special kind of pain] – perhaps a great many of them … For the time being, we will restrict ourselves to effable concepts: concepts that can be part of the content of communicable thought … [T]here are a great many stable and effable mental concepts that do not map onto words. (Sperber and Wilson 1998: 189)
How are we to understand these strange mental concepts which are effable, but do not map onto words? For Sperber and Wilson, it is probably safe to assume, there are simple cases in which a mental concept corresponds to a single word in isolation, i.e. a word uncontaminated by the context. But they are much more interested in concepts for which the corresponding lexical item is no more than a highly ambiguous shorthand. Thus many people in the Western world have a mental concept for the act of uncorking a wine bottle, and they will translate the request ‘Could you please open this bottle’ into the unambiguous mental concept. To me, it seems quite unnecessary to resort to a concept which, according to Sperber and Wilson, does not map onto a word (but what about uncork?). Let us assume the bottle referred to in the utterance ‘Open the bottle!’ is a standard bottle of wine back in 1998. Then the cognitive representation of the verb open would be the concept of ‘uncorking’. If we accept this, the question must be allowed of how this mental concept differs systematically from the natural language expression. The concept ‘uncork’ is not only effable; it also happens to map
What is meaning?
79
miraculously onto the word uncork. But how do I move on from the rather general verb open to the exact concept ‘uncork’? Do I have to rely on cognitive mechanisms that make me understand that the speaker intended ‘uncork’ when they said open? A common answer would be that most wine drinkers have observed bottles being uncorked often enough. But how do they know that what they saw was ‘uncorking’, and that it is one of the many activities which can be referred to as ‘opening’? (Nobody would call ‘lifting a rag placed on a bucket of water’ opening the bucket.) The wine drinkers among us know that opening a bottle of wine normally means (at least still commonly in 1998) ‘uncorking’, and they are acquainted with the details, because they have encountered texts such as this, even if perhaps not as explicit. (www.ehow.com/how_1715_openwine-bottle.html): How to open a bottle: It’s actually pretty simple to open a bottle of wine. These steps are for a double-action, or wing, corkscrew, which has two arms (or wings) that help lever the cork out of the bottle. Remove the top of the lead or foil capsule by cutting around the rim of the bottle with the sharp point of the corkscrew. The arms of the cork-screw will have to be raised for this step. You can also make a slit in the foil and remove the whole capsule before beginning. Stand the bottle on a flat, hard surface at mid-chest level or lower. Lower the arms of the corkscrew. Holding the corkscrew as vertically and straight as possible, place the sharp end directly into the middle of the cork. Securely grasp the top of the bottle and the lower end of the corkscrew with one hand. With the other hand, begin turning the handle of the corkscrew clock-wise, applying an even, constant downward pressure into the cork. As the corkscrew goes into the cork, its arms will begin to rise. Apply more pressure if the corkscrew will not penetrate the cork. Keep turning the handle until the arms of the corkscrew are completely raised and the screw is well into the cork. With one hand on each arm of the corkscrew, press the arms down. This will lift the cork out of the bottle. Wrap your hand around the base of the corkscrew and lift straight up. Remove the foil, if necessary. Twist the cork off the corkscrew. Wipe the rim of the bottle with a clean, damp towel before serving, to remove any stray pieces of cork and, more important, any lead left by the foil.
It is rather unlikely that people learn a concept like ‘open’ in the sense of ‘uncork’ in this way. Normally someone who has never before seen a bottle of wine opened will be told, in a deictic act, by someone in a gathering that this is what they are observing, and from then on he or she will know what it means to open a bottle even if they never have learnt the meaning of the verb uncork. The important thing is that verbalisation is always possible. Meaning is symbolic content. We find it only in the discourse. The lexical item open is such a sign. How do we know what it stands for? All we have learnt about the meanings of lexical items we have learnt
80
Meaning, Discourse and Society
from other people’s contributions to the discourse. This is how we learn how to use words when we grow up. Our carers or, a bit later, our peers, explain what they mean. If new lexical items are introduced into the discourse, the text introducing them has to explain them, has to paraphrase them. If I have not been told what opening a bottle means, no cognitive mechanism will help me to understand the speaker’s intentions. However, if I know the meaning of open a bottle of wine, I do not need such a mechanism. Why do we not readily accept such a simple solution? One reason, I think, is that linguists have for a long time attributed too much importance to the meaning of a single word in isolation. We grow up in the belief that words are the core elements of language, and that their meanings are registered in dictionaries. Indeed, there we are confronted with the apparently incontrovertible evidence that words, frequent words like open, are fuzzy, polysemous or ambiguous. Much of the endeavour of cognitive linguistics is directed at disambiguation, at the resolution of ambiguity. For we feel that our thoughts, unlike words, are unambiguous. But language is not more ambiguous than thinking. When we are confronted with a text of common length, we rarely doubt its meaning. Once we give up the belief that single words viewed in isolation are the core elements, the ambiguity starts disappearing. All we have to do is to replace the linguistic construct ‘word’ by the linguistic construct ‘unit of meaning’, defined as a node word plus all the words in its immediate context that make it unambiguous. Instead of wondering about the polysemy of open, we now have to deal with a unit of meaning, a lexical item open a bottle of wine, which is fairly monosemous. If our speaker tells us ‘Last night I opened a bottle of wine’, we still cannot be absolutely sure that she uncorked it. She might have unscrewed the top, if she is the kind of person to have wine with a screw top. Unless she tells us, we will never know. No cognitive mechanism will provide this information. Two-level semanticists believe that a natural language expression is less rich than the mental representation it conveys. This mental representation is identical with the speaker’s intentions. The hearer has to employ a mental mechanism enabling them to reconstruct the speaker’s intentions in full, on the basis of an utterance that only gives a rough indication of them. This opens the stage for Herbert Paul Grice and his highly influential Studies in the Way of Words (1989). Grice’s goal was to discover the mechanisms, or maxims, that speaker and hearer must follow for the hearer to capture the speaker’s intentions. There are four key maxims: (p. 1) Do not say what you believe to be false; (p. 2) Make your contribution as informative as is required for the current purposes of the exchange; (p. 3) Be relevant; and (p. 4) Avoid obscurity of expression, avoid ambiguity, be brief, be orderly. We should not read these maxims as being based on observed evidence or as moral injunctions. Language philosopher Kent Bach believes that ‘[a]lthough Grice presented them in the
What is meaning?
81
form of guidelines for how to communicate successfully, I think they are better construed as presumptions about utterances, presumptions that we as listeners rely on and as speakers exploit’ (Bach 2005). ‘Following Grice’, says Gisela Harras, ‘an act of communication can be defined as follows: A speaker S communicates something to hearer H, if [if and only if] S intends H to react in a certain way r S intends H to recognize (i) S intends H to react in a way r by means of recognition of (i)’ Harras 2000: 21–2 Speaker and hearer have to co-operate, says Grice, in unearthing the speaker’s intentions. Some might find this a trivial statement. But Harras reminds us that ‘Grice himself and legions of linguists consider this principle and the maxims to be a powerful instrument for explaining how people convey and understand relevant information.’ Can we really hope that Grice’s maxims will allow us to interpret semantically underdetermined utterances ‘in accordance to the speaker’s meaning’ (Harras 2000: 22)? If the utterance is inconclusive, then how do we learn about a speaker’s intentions? Gloria Origgi and Dan Sperber’s point of departure, in their article ‘Evolution, communication and the proper function of language’ (2000), is Relevance Theory, the radicalisation of the Gricean maxims that has become popular in some linguistic circles. Their paper begins by reviewing Ruth Millikan’s anti-Gricean stance, first put forward in her book Language, Thought and other Biological Categories (2004), namely that only the text matters, not the speaker’s intentions. This is rejected by Origgi and Sperber. For them, only the speaker’s intentions are relevant: ‘What makes human communication possible at all, however, is human virtuosity in attributing intentions to one another’ (Origgi and Sperber: 8). They reject Millikan’s position that, in their words, ‘[a]ll the possible contextual meanings of a linguistic device … must be conventionally associated with this [textual] device’ (Origgi and Sperber 2000: 8). But according to Origgi and Sperber, this cannot be the case, because the linguistic analysis would have to be far too complex. If all readings were acceptable, then, they insist, we would end up with a ‘massive ambiguity’. The expression ‘Eat!’, for example, could be interpreted in many ways, as a command, as a request, as encouragement, or as advice. It could also be understood metaphorically or as irony. The approaches of Grice, of Sperber and Wilson, and of Origgi and Sperber, on the other hand, can resolve this ambiguity. They take the ‘linguistic’ meaning of a text only as a clue, without attempting to specify its different readings. It is this clue, indeterminate by itself, that lets us recognise the meaning intended by the speaker. They believe that such a model is superior to the one offered by Millikan. It can do without the
82
Meaning, Discourse and Society
kind of in-depth linguistic analysis that Millikan’s proposal would require. The kind of fine-grained linguistic analysis necessary would be impossible in any case, for ‘the more massive the ambiguity implied by the theory, the less plausible that human minds can deal with it’ (Origgi and Sperber 2000: 9). All that the text gives us is a trigger: ‘Linguistic comprehension involves, at an intermediate and largely unconscious level, the decoding of linguistic stimuli that are then used as evidence by the hearer, together with the context, to arrive inferentially at the speaker’s meaning’ (Origgi and Sperber 2000: 15). Would that not mean that the linguistic utterance is just the key for reading the mind of the speaker, in the same vein in which a medium commonly touches a personal possession of the deceased to facilitate the contact with the immortal soul? This is exactly what Origgi and Sperber mean: ‘There is a growing body of evidence and arguments tending to establish that a mindreading ability is an essential ingredient of human cognition … What are the relationships between mindreading and the language faculty? Millikan argues that there is no such thing, whereas Grice and post-Griceans assume that linguistic communication involves a form of mindreading where, by speaking, the speaker helps the reader to read her mind’ (Origgi and Sperber 2000: 16). Not all cognitive linguists focus on the speaker’s intentions. Some, like Stephen Levinson, though close to the two-level semantics camp, simply want to investigate the relationship between our thoughts, i.e. the representations we have in our minds, and the utterances expressing them. As it is impossible to access mental representations directly, they employ a more indirect method of comparing languages. What is common to all of them might well be as close as we get to the language of thought. However, this procedure does not tell us everything. If there are languages that need to distinguish between singular and plural, like English, and other languages that do without this distinction, like Chinese, which is closer to mentalese? Stephen Levinson seems sure that an answer can be found. The question he sets out to answer is: ‘Are the representations that underlie linguistic meaning (i.e. semantic representations, SRs) the same kind of representations that underlie non-linguistic thinking (i.e. conceptual representations, CRs)?’ He is aware that ‘the notion of “representation” raises a caveat: different scholars have deeply different ideas about what kind of a thing a representation of a thought or a meaning might be’ (Levinson 1997: 15). My concerns are slightly different. Why could we not just compare a natural language utterance, focusing on its meaning, with a (corresponding?) thought? Why do we, instead, have to compare a representation of this utterance with a representation of the (respective?) thought? In which way do an utterance and its representation, or a thought and its representation, differ? Are they at least functionally isomorphic? If they are not, how can we be sure the representation properly represents the utterance, or the thought, respectively? If they are functionally isomorphic, why can we
What is meaning?
83
not work directly on the relationship between the utterance and the thought? The reason is simple. Before thoughts are somehow turned into language they remain ineffable, and therefore we neither exchange nor share them. But if thoughts are ineffable, then there is not much that could persuade us that a conceptual representation of a thought is trustworthy. It is this violation of Ockham’s razor that we encounter time and again in cognitive linguistics. Instead of investigating what is readily accessible in a brute fact that no one could contest, namely the utterance as such, we are fed with arbitrary constructs, whether they come as semantic or conceptual representations, or even as invariably elusive thoughts themselves. Stephen Levinson works hard to make us understand what he sees as the difference between semantic and conceptual representations, presenting cases in which, according to him, what is said is not what is thought. Colour adjectives always come in handy in this kind of research. In Tzeltal, we are told, there is no word for ‘blue’, and none for ‘green’, only one (yax) for both of these colours (‘grue’). But if asked ‘yax like what?’ the Tzeltal people are perfectly able to say ‘yax as the sky’, or ‘yax as the grass’. This shows that mentally they are capable of distinguishing the two thoughts ‘blue’ and ‘green’. Levinson infers that ‘generally speaking, our thought is specific, but its linguistic expression is often necessarily general, non-specific, even imprecise’ (Levinson 1997: 16f.). He stops short of saying that ‘blue’ is a conceptual representation, i.e. a lexical item of our language of thought. But is that not the gist of his claim? His arguments do not convince me. Italians have several words for ‘blue’ while my native German has, like English, only one. Italians have to decide if what they see that seems blue to me is azurro, or blu, or livido. Why should we not expect the Tzeltal people to think in Italian categories rather than in English ones? Why should it be, of all possible colour categorisations, the English distinctions of blue and green which are more or less identical with everyone’s mental concepts? My computer screen, I have been told, can represent zillions of colours. Why are these not the colours my thoughts are about? Why do even blind people talk about red tomatoes, blue skies and green lawns? We see, or believe we see, or form an image of, the colours we are trained to associate with things. In other cultures, blackberries might not be seen as black, red wine not as red, and green herrings not as green. Our thoughts replicate, permute and carry further what we are told in the discourse, always making us believe that what we have been trained to see is reality. Another example presented by Levinson is the tu/vous distinction found in French but not English. But here he interestingly argues the other way around. If an English person thinks ‘You are mistaken’, a French person may have the ‘corresponding thought, but she’ll have to add the distinction between tu and vous if she is to voice it’. This tells us two things about Levinson’s conceptual representations. They are surprisingly isomorphic with the English language.
84
Meaning, Discourse and Society
Speaking English, I can harbour ‘the thought “You are mistaken” and speak my mind directly’ (Levinson 1997: 21). The French, however, have a serious problem. They have to add the tu/vous distinction to their conceptual representations addressing single persons, to turn them into semantic representations, a distinction which is apparently not in the language of thought but only in their natural language. This tells us something important about mentalese. It distinguishes blue from green, just like English but not like Tzeltal, but it does not distinguish between tu and vous, again just like English, but not like French. Not that English is perfect: Levinson points out that it does not distinguish between ‘the female sibling of my mother’ and ‘the female sibling of my father’. Both concepts are covered by the word ‘aunt’, while many other languages (and, supposedly, mentalese) have two different words (Levinson 1997: 16). Levinson knows that this is treacherous territory, and he admits: ‘But it is also possible that as a result of those acquired habits of language production [different language communities] do indeed just entertain systematically different thoughts’ (Levinson 1997: 22). So what is his conclusion? He advocates the stance that mental concepts, including colour adjectives, are decomposable. We can view more complex mental concepts as being composed of a limited number of basic semantic bricks, a stance similar to that of Steven Pinker, Anna Wierzbicka and many other cognitive linguists. These ‘atoms’, as he calls them, are indeed universal and innate, while the more complex representations can be specific to a language and culture. If you do not subscribe to this view, and if you are a one-level semanticist for whom semantic representations are isomorphic with conceptual representations, he argues, ‘then it seems you must reject the thesis of the psychic unity of mankind and adopt radical Whorfianism’ (Levinson 1997: 26). But once we abide by a two-level semantics and the decomposition stance, the dogma of a universal language of thought can be saved: ‘To the (perhaps large) extent that in thinking and speaking individuals use different concepts from composition of universal atomic ones, and continue to have access to that decomposition, we may indeed talk of universal representational systems’ (Levinson 1997: 28). He is aware, however, that there are other ways to look at the evidence. Discussing the representation of space, he admits: ‘[H]umans from different cultural groups might talk about, think about, visualize and gesture about the same spatial arrays utilizing distinct concepts and distinct processes.’ Language and conceptual representations may be more closely connected than initially thought, for ‘to talk about what we see and feel, we must be able to “see” and “feel” in terms that we can talk about’ (Levinson 1997: 39). It is this important concession that makes Levinson a cognitive linguist worth listening to. There is another thought in Levinson’s article worth pondering. Directed at those linguists Levinson calls ‘Cognitive Linguists’, such as Jerry Fodor and Ronald Langacker, i.e. those people believing in a one-level semantics where
What is meaning?
85
the semantic representation of an utterance is thought to be isomorphic with its conceptual representation, Levinson states: ‘The arguments [proffered by two-level semanticists like himself] also happen to dispose of an extreme kind of relativism, in which it is held that we think in language, that thinking is actually subvocal speaking, “inner speech” or the like’ (Levinson 1997: 24). Relativism for Levinson is when our thoughts are determined to a large degree by the language we speak, by the discourse that we are exposed to, by the ‘reality’ that this discourse keeps constructing. For many cognitive linguists, relativism is a thing from hell, depriving us from all our cherished certainties. Relativism means that your terrorist can be my freedom fighter, a terrible thought. This condemnation of ‘relativism’, often found among language philosophers, resonates with John Searle’s denunciation of ‘social constructionism, pragmatism, deconstructionism, relativism, postmodernism’. As Searle sees it, these views owe their popularity to ‘a basic urge to power’ on the part of academic teachers who foist them upon their students (Searle 1998: 15ff.). Instead of telling them how things are in reality, they only want to promote their respective agendas. Would Levinson have us believe that our language is more slanted than our thoughts? In which sense are Levinson’s semantic representations less real than what he calls conceptual representations? Is it their referential relationship to the discourse-external ‘real world’ out there? Or is it the underlying conviction that conceptual representations are more real because they are somehow embossed on brain tissue, and that therefore they are more than theoretical constructs and cannot be dismissed as linguistic whims, such as the tu/vous distinction? As promised at the outset of his article, Levinson comes to the conclusion that semantic representations cannot be identical to conceptual representations, and suggests that further empirical investigation is needed to ascertain their relationship. Apparently he is not bothered by what he calls the ‘big stumbling block’: how do we investigate the conceptual content of the ‘language of thought’ independently of language? (Levinson 1997: 29). I do not want to be misunderstood. It may make sense to talk about minds, to construct them as real in the continual negotiations we carry on in our discourses. It does not necessarily matter that there are no ‘brute’ facts supporting their existence. All that is asked for is that minds, as arbitrarily constructed objects of our discourses, enable us to engage collectively with our problems better than if we had not constructed them. We all like to think in terms of cause and effect. Therefore it make us feel good to have a mind to which we can point when we want to explain someone’s behaviour. But the mind is as indeterminable, as invisible and as elusive as the soul. Neither the French nor the German language have an unambiguous word for this strange object. The German word normally used is Geist. But Geist, often rendered in English as ‘spirit’, is not
86
Meaning, Discourse and Society
necessarily monadic. It can equally well be a collective entity, or an entity not residing in human bodies at all, e.g. Hegel’s Weltgeist (‘world spirit’). We also find as equivalent Verstand (i.e. our reasoning faculty, normally in opposition to Gefühl [‘feeling’]). English–French dictionaries normally suggest esprit as the French equivalent. But esprit is much more than mind. Out of the six main senses given for esprit in the 1970 Petit Robert, only sense IV corresponds more or less to mind: ‘La réalité pensante’, the first sub-sense being described as ‘Le principe pensante en général, oppose à l’objet de pensée, à la matière’, and as as an example we are given the phrase ‘l’esprit humain’. But this phrase normally refers to humankind in general, as in ‘la civilisation et l’esprit humain’. The second sub-sense is more closely related to the English mind: ‘Principe de la vie psychique, tant affective qu’intellectuelle, chez un individue.’ As synonyms, we are given âme, conscience, moi. In the other sub-senses of sense IV, esprit is described as synonymous with caractère; entendement, intellect, intelligence, pensée, raison; cerveau, cervelle, tête. The key citation given for this sub-sense reads: ‘L’esprit est toujours la dupe du coeur’. I wonder if we would find in English ‘The mind is always betrayed by the heart’. This is not the way the English would normally think of the mind. Google can find only eight occurrences that would roughly fit the bill: ‘Sometimes the heart betrays the mind’. What I have said about the mind as a discourse object applies also to mental/ cognitive concepts/representations. If they help us to solve the riddle of intentionality, of consciousness, of knowing what an utterance is about, I have no problem accepting them. Often when I complain that cognitive linguists so far have failed to convince me that it is useful to talk about mental concepts, the name of Ronald Langacker is mentioned as one of the Cognitive Linguists who bases his conceptualisation of mental processes involving language on empirical evidence. Langacker would not be a Cognitive Linguist if he did not place meaning quite firmly into the realm of the mind. For him, ‘[m]eaning is equated with conceptualisation. Linguistic semantics must therefore attempt the structural analysis and explicit description of abstract entities like thoughts and concepts … Because conceptualisation resides in cognitive processing, our ultimate objective must be to characterize the types of cognitive events whose occurrence constitutes a given mental experience’ (Langacker 2002: 2). But he keeps an open mind: ‘Could it be, for example, that language resides entirely in individual minds? … At the other extreme, it might be maintained – perhaps from a strictly Platonist or interactionist perspective – that language has no cognitive representation whatsoever’ (Langacker 1997: 229). It is through: interaction and mutual accommodation that the conventional patterns of activity we call ‘language’ arise, sustain themselves, and undergo continual adjustment. The study of its interactive basis is therefore essential to a full account of the acquisition, structure, and use of language. I insist, however, that it is not equivalent to such an account, which must
What is meaning?
87
also encompass the specific mental adjustments that enable speakers to communicate in a conventionally sanctioned manner. To focus exclusively on the interactions per se would be as pointless as ignoring them altogether – either option omits an essential component of language, neither of which can be well understood without the other. (Langacker 1997: 233–4)
Ultimately, however, it is meaning we are concerned with. Once conventions, mutually agreed rules, are obeyed, we must not shy away from looking into people’s heads: ‘The meanings of linguistic expressions are conceptualizations, and the relevant conceptualizers are the speaker and the addressee’ (p. 242). Meaning, then, is what we have to look out for in the depths of the mind, the way an ‘expression is actually comprehended – its full, richly detailed contextual understanding’ (p. 235). ‘The immediate goal of semantics, for example, is to identify conceptual components and elucidate the structure of the complex conceptualizations evoked by linguistic expressions’ (p. 240). This is about as far as Langacker takes us. What we are not given is an illustration of such conceptualisations or the conceptual components they consist of. What is a ‘full, richly detailed, contextual understanding’ of an utterance? This is what we are not told. No example of a ‘cognitive representation’ is given. Instead, we find the escape route habitually taken by cognitive scientists when the reader expects a glimpse of mental activity: the sudden jump from the mind to the brain: Now, I do talk about mental representations, but the term does not refer to complex depictive entities stored as such. It merely indicates the occurrence of neurological adjustments, notably changes in the strength of synaptic connections that influence subsequent processing and facilitate the emergence of patterns of activity constitutive of particular kinds of mental experience. A speaker obviously has no introspective access to these adjustments, which do not in any case resemble the emergent experience they engender. In fact, even that experience lies beyond our reach, in the sense that we have no conscious access to the character of the cognitive processing that constitutes it. Although to some extent we monitor our experience, and may even attempt to analyse it, for the most part we merely have it. (Langacker 1997: 239)
Thus, I fear, the ‘full, richly detailed, contextual understanding’ of an utterance is an experience that we have but cannot really be aware of. No introspection allows us to observe our mental states, our own intentionality. There is only one escape alley: mental representations can best be described as neurological adjustments. The experience itself remains ineffable. As their last resort, cognitive linguists of all feathers seem to agree that it is only the neural sciences that can enlighten us about the ‘true’ meaning of words, phrases and utterances.
7
Where should we look for meaning?
Cognitive linguistics, as I see it, has failed to come up with a convincing theory about the mind being the seat of meaning. Similarly, the philosophy of mind has been unable to present us with a model of the mind free of internal contradictions. This leaves us puzzled. There are four options concerning the seat of meaning that have been explored, to a greater or lesser extent, namely: a) the individual intentional mind, the mind that understands meaning; b) the individual computational mind/brain, i.e. the mind enclosed in the hardware environment of the brain, carrying out programs for processing language; c) the brain where language becomes reality; d) the discourse as the collective mind. In this last chapter of this part of the book, the part devoted to the cognitive conception of the mind as the seat of meaning, I want to summarise the two core stances, as I see them, namely the one of the intentional mind and the one of the computational mind. I will then explore a more recent stance, focusing on the brain as an object of science rather than on the mind as an object of speculation. Finally, I will introduce my own stance, that of the discourse as the collective mind of a discourse community, the stance to which the second part of this book will be devoted. Option a: The individual intentional mind People have intentionality. They experience what is going on inside them and around them. They think. They feel. They remember. They make plans and go about putting them into effect. We experience ourselves as agents in charge of our lives. We have learned to ascribe intentions to what we do, thus differentiating between processes in which we are involved, like hearing or seeing, and actions like listening or watching. We can testify to ourselves and others as to how we perceive ourselves and the natural and social worlds surrounding us. Taken altogether this is the ‘aboutness’ that intentionality amounts to. But once we want to inspect our own intentionality, it withdraws, becoming ineffable, 88
Where should we look for meaning?
89
renderable only clumsily in reports, in testimony that must, of necessity, remain hearsay, because direct access is impossible. As soon as we turn our experiences into texts, contributions to a discourse, they are moulded by the ways we have learnt to talk about our mental states. We know what is normally said, and choose from this repository what seems to be more or less adequate. We construct the intentions we ascribe to our actions or utterances, on the spot, after the act, when we are asked (by ourselves or others) to account for them. We say things like: ‘And then I decided to apply for this job’. But what exactly does this activity called decide consist of? How did we go about it? Was our decision rational, or was it driven by beliefs, attitudes and desires? To what extent were we aware of them? Should we perhaps agree with those neuroscientists who claim to have proven that the brain has already effectuated our ‘decision’ nanoseconds before we think we have taken it? Is our intentionality, if not an illusion, just an expendable feature of human beings that ‘supervenes’, in the terminology of the philosophy of mind, on the neural processes without actually being ‘causally’ connected to them? Not all philosophers of mind these days prefer the brain to the mind. Ruth Millikan, for instance, a former student of Wilfrid Sellars (and thus endeavouring to bridge the gap between the philosophy of mind and analytic philosophy), defends the notion of intentionality and pleads for first-person experiences, and not innateness, as the precondition for mental concepts to become relevant: ‘[A]ll concepts, including logical concepts, are tested for their very having of content through ongoing experience … What can be gained through conceptual analysis is then only what has previously been inductively acquired through experience’ (Millikan 2004: 1). For her, intentionality is a specifically human function we owe to evolution, as an adaptation that makes human beings fitter for survival. As she sees it, the intentionality both of language and of thought concerns the way these serve their proper functions: ‘Speaking extremely roughly at first, proper function is survival value. In the simplest instances, the proper function of an item is what it has been doing, better, what the predecessors from which it was reproduced or copied were doing, that helped account for continued reproduction or copying and hence for its existence’ (Millikan 2005). A person’s intentionality, or the experiences they have, are not accessible from the outside. They are not even accessible to the introspection of the person themselves. A person has it. She or he is conscious of her/himself, has certain intentions, has plans, has moods, and feelings like fears and desires. A person experiences her/himself as being in an intentional state. This intentionality is real, claims Millikan. It is a result of Darwinian adaptation, ensuring the survival of those who had it. Another perspective is equally possible. We can view intentionality from a social perspective, as opposed to an individual one. Intentionality then is what is being said in the discourse about how people experience themselves, what is
90
Meaning, Discourse and Society
said about the motifs underlying their action, for instance their utterance. We can view a person’s intentionality as something being constructed collaboratively by negotiations between the members of a discourse community, which can involve the person in question, but does not necessarily have to. Our discourse is full of such constructions. We continually exchange assumptions about the motivations of our family members, our friends, our employers and our politicians. The anthropological linguist Alessandro Duranti doubts that individual intentionality is more than a social construct which we happen to find in some societies but is absent from others. He asserts that in Samoan, for instance, there is no good equivalent for the concept of a individual’s intention. The word loto commonly used in translations is not the same; expressions including loto: i) do not refer to actions that are planned but instead tend to coincide with dispositions or inclinations; ii) do not express ways of being of which a person is necessarily conscious; iii) do not display a recurring set of beliefs that can be attached to or are implied by these states; iv) tend to have an affective meaning; and v) tend to be embodied attitudes or practices (they are usually expressed through and interpreted as a combination of verbal and kinesic behaviour). (Duranti 2006: 34, emphasis in the original text)
To be aware of ourselves, of our actions and motivations, to feel ourselves, to have moods and emotions, to interpret the behaviour and facial features of other people as signs signalling their awareness and feelings, to make sense of our environments, to interact with other people in a meaningful way; all that taken together is what we call intentionality. We believe we know it is there; yet it remains an intractable enigma. ‘A rose is a rose is a rose’, said Gertrude Stein: it does not go any deeper. The same seems to go for consciousness and for intentionality.
Option b: The individual computational mind/brain Besides the intentional mind, there is also the option of a mind stripped of its intentionality, the mind as a language-processing mechanism that functions nicely without necessarily being aware of what language input and output is about. This option is based on the computational model of the mind. The brain provides the hardware and the hard-wired circuits. The mind is the sum of all the programs running in the brain. I will therefore talk about the mind/brain. This option suggests a mind that does not know what its thoughts are about, or if it knows, this is an irrelevant, supervenient feature not necessary for its working. In the 1970s and 1980s, John Searle used to make a fine distinction between ‘intentional states’ and ‘understanding’. For him, a mental mechanism can be in an intentional state without understanding what this mental state is
Where should we look for meaning?
91
about. He is invariably opposed to the view he characterises as ‘strong A[rtificial] I[ntelligence]’, namely that ‘the computer is not merely a tool in the study of the mind, rather the appropriately programmed computer really is a mind in the sense that computers given the right programs can be literally said to understand and have other cognitive states’ (Searle 1980: 417). The relevant word is understand. For Searle, computers can have intentional states, just like minds/brains. But understanding (in the literal sense) is something only people are capable of. Understanding is knowing what something symbolic, a sign, stands for: ‘Newell and Simon (1963) write that the kind of cognition they claim for computers is exactly the same as for human beings. I like the straightforwardness of this claim, and it is the sort of claim I will be considering. I will argue that in the literal sense the programmed computer understands what the car and the adding machine understand, namely, exactly nothing’ (Searle 1980). For Searle, my reading a text will put my mind/brain into an intentional state. But it seems that, for him, this intentional state is different from the mental representation I have of this text, or from knowing what the intentional state stands for. For he asks the rhetorical question: ‘How can that state of my brain – consisting in such things as configurations of neurons and synaptic connections, activated by neurotransmitters – stand for anything?’ (Searle 1998: 90). For us conscious human beings, texts may well have a meaning. For us, language consists of signs that can be interpreted. For us, language is symbolic. The intentional state of my mind/ brain, due to my having read the text, is, for Searle, not symbolic. I do not need to be conscious to have my mind process the text properly. Consciousness, it seems, has become for the Searle of the 1990s an expendable concept. In his 1992 book The Rediscovery of the Mind, Searle still espoused the importance of consciousness: ‘In one way or other, all other mental notions – such as intentionality, subjectivity, mental causation, intelligence – can only be fully understood as mental by way of their relations to consciousness’ (p. 84). To frame the mind on the model of the computer, he then said, will not get us anywhere. There has to be someone or something, a homunculus inside us whose task is to make sense of the computations our mental mechanism carries out: Typical homunculus questions in cognitive science are such as the following: ‘How does the visual system compute shape from shading; how does it compute object distance from size of retinal image?’ A parallel question would be: ‘How do nails compute the distance they are to travel in the board from the impact of the hammer and the density of the wood?’ And the answer is the same in both sorts of cases: If we are talking about how the system works intrinsically, neither nails nor visual systems compute anything. We as outside homunculi might describe them computationally, and it is often useful to do so. (Searle 1992)
There are no homunculi. So something has to be wrong with the computer as the model of the mind. This is a thought shared by Daniel Dennett. In his article ‘Reflections on language and mind’, he ridicules Fodor’s claim that the
92
Meaning, Discourse and Society
language of thought ‘does not itself require interpretation by the brain that speaks it’, because ‘it creates an artifactual problem about the ‘access’ of consciousness’: Unconscious cognitive processes granted on all sides, and if it is conducted in Mentalese … getting some content translated into Mentalese cannot be sufficient for getting it into consciousness, even if it is sufficient for getting it understood. There must be some further translation or transduction, into an even more central arena than Central Processing, into some extra system … Beyond understanding lies conscious appreciation. (Dennett 1998: 287)
Understanding, for Dennett, is more than responding to a stimulus with the proper response. That, he says, is not enough. There also has to be conscious appreciation, something the computer cannot provide. For the computational model to work, we would need on top of it, something like a ‘Central Understander … an ominously wise overseer-homunculus who handles the hard cases in the workshop of consciousness’. That, Dennett tells us, is the wrong approach. For him, cognition (or thinking) is a decentralised process, involving any number of modules interacting with each other in a self-organising ‘unsupervised competition’ (all quotes from Dennett 1998: 287–8). It is this dialogue between modules that creates the phenomenon of intentionality. Dennett’s approach has something compelling: the mind as an arena where an indefinite number of modules interact with each other leading to unpredictable contingent results. Its disadvantage is that it is only a model. It may show us how we can understand the emergence of intentionality. It does not allow us to observe it. The eliminativists, those philosophers of mind or cognitivists for whom intentionality is, at best, a supervenient epiphenomenon, like Stephen Stich, argue ‘that cognitive psychology does not (or, in any case, should not) taxonomize mental states by their semantic properties at all, since attribution of psychological states by content is sensitive to factors that render it problematic in the context of a scientific psychology’. Stich has therefore proposed in his 1983 book From Folk Psychology to Cognitive Science (Stich 1983) a syntactic theory of the mind, in which the semantic properties of mental states play no explanatory role. Just as we can imagine a computer that can translate a text from Chinese into English and preserve the content, we can envisage a mind that carries out syntactic instructions, turning, for instance, a sentence in the passive voice into an active one correctly, preserving meaning, without the need for an overseer to know what is going on. If we take the computational model of the mind seriously, then we have to accept that mental language processing does not depend on intentionality (as long as we disagree with Searle and maintain that an intentional state presupposes consciousness). Computers are not aware of the meaning of the entities (words, phrases, sentences, texts) they are processing. Machine translation using the
Where should we look for meaning?
93
conceptual interlingua approach is a process functionally equivalent to a mechanical calculator. The processed content can be interpreted as signs or symbols only by us, the observers, not by the computer. We can interpret them, we can discuss their meaning. The computer only computes them, just like an adding machine. It has no agency; it is not in a position to decide between two possible translations of a linguistic sign. For us, it is a sign because we have a consensus that it stands for something else. We may not quite agree what it means, and we may have to negotiate its meaning, that is we can discuss whether the words sorrow and grief are largely synonymous or mean quite different things. While mentalists like Ruth Millikan, who acknowledge the role of intentionality, also seem to favour a Darwinian perspective, Noam Chomsky and Jerry Fodor both harbour doubts that our uniquely human language faculty can be explained by evolutionary adaptation. For Fodor, the buck of understanding stops with the language of thought, and he does not tell us whether people are aware of their mental representations. For Chomsky, language in the narrow sense (FLN: faculty of language: narrow) is syntactic computation, as in the case of recursion, for instance. Meaning is not mentioned in his papers on evolutionary aspects. And while concepts are mentioned in the 2002 paper ‘The faculty of language: what is it, who has it, and how did it evolve?’, co-authored with Marc Hauser and Tecumseh Fitch, they occur as part of the FLN somehow linked to a ‘(formal) semantic system’: ‘[A] key component of FLN is a computational system (narrow syntax) that generates internal representations and maps them into the sensory-motor interface by the phonological system, and into the conceptual-intentional interface by the (formal) semantic system’ (Chomsky, Hauser and Fitch 2002: 1571). However, as part of the language faculty in the broad sense (FLB), concepts seem to be the kind of mental representations that apes and birds have in common with human beings: ‘A wide variety of studies indicate that nonhuman mammals and birds have rich conceptual representations’ (Chomsky, Hauser and Fitch 2002: 1575). In both cases, intentionality plays no role. In the case of FLB, these concepts are definitely not signs, and they do not correspond to signs. They do not stand for something else. A little bird watching an approaching snake does not initiate an act of interpretation of this sign. It does not ponder whether this snake is a threat or not. The image of the snake triggers the escape mechanism. If we could ask the bird ten minutes later why it flew off, it would respond ‘Did I really?’ Neither should we see the conceptual representations in the narrow sense of the language faculty as signs or symbols in need of interpretation. The FLN representations are generated by a computer mechanism and mapped onto semantic formulae that will be algorithmically processed, as ‘uninterpreted symbols’. The view proffered in the Chomsky, Hauser and Fitch 2002 paper has been rejected in a paper by Ray Jackendoff and Steven Pinker (2005). In a subsequent response by Fitch, Hauser and Chomsky,
94
Meaning, Discourse and Society
the authors make it clear that they have problems in accepting the language faculty in the narrow sense as the result of an adaptational process: ‘The FLN/ FLB distinction, we hoped, would help the field to see that there is no incompatibility between the hypotheses that FLB is an adaptation that shares much with animals, and that the mechanism(s) underlying FLN might be quite unique’ (Fitch, Hauser and Chomsky 2005:183). For Europeans, it is strange to observe how ‘adaptation’ has become such a controversial concept in America. In the Fitch et al. paper we find this telling statement: ‘The term “adaptation” conceals a conceptual minefield, long recognized as such by practicing evolutionary biologists’ (p. 184). While the authors are ready to admit that the language faculty in the broad sense, that is the ability to communicate, can well be explained in terms of evolutionary adaptation, they are keen to exempt at least some core FLN components from being contingent results of environmental adaptation: As [neural, developmental, evolutionary and molecular biology] progresses, biolinguistics will help to drive an increasingly rich understanding of human and animal cognition, and to a broad comparative approach to understanding the many shared cognitive mechanisms that are part of the human language faculty. Ultimately, we think it is likely that some bona fide components of FLN – mechanisms that are uniquely human and unique to language – will be isolated and will withstand concerted attempts to reject them by empirical research. (Fitch et al. 2005: 206)
I find it hard to envisage a time when ‘biolinguists’ will have revealed to us the nature of the computational mechanisms of the mind/brain that enable us to turn thoughts into language and language into thoughts. Jerry Fodor, as we have seen above (Chapter 3), would fully support the claims of Fitch, Hauser and Chomsky. He, too, rejects the idea that our faculties can be reduced to adaptation. For him and supposedly also for the three authors, ‘the alternative possibility to Darwin’s is that the direction of phenotypic change is very largely determined by endogenous variables’. In their view, at least the language faculty in the narrow sense had been dormant all the time in the genes of our ancestors, only to be awakened at the right stimulus. This is a story that reminds us very much of the concepts of ‘carburettor’ and ‘bureaucracy’ which were, according to Chomsky and Fodor, already somehow latent in our genetic outfit, even when there was no carburettor and no bureaucracy around in our environment. The computational model of the mind turns human beings into zombies. However much they continue to believe in their own consciousness and their intentionality, inside of them we find a mechanism, operating on ‘endogenic variables’, dormant computer programs that were activated when the need for FLN mechanisms arose. This language faculty enables us to generate one new sentence after another, but we have no control over their content, and apparently their content does not really matter. What counts is that they can all be seen as strings of words put together according to pre-programmed instructions.
Where should we look for meaning?
95
The kind of cognitive linguistics based on the computational model of the mind tells us that it does not really matter what we say, as long as it is ruleconformant. It deals with language by transforming utterances into ‘mental representations’ without telling us what these representations mean to whom. Just like Saussure’s langue-linguistics, it only takes account of the synchronic dimension of language. The mental concepts are seen as immutable, permanent and context independent, just like the concepts we encountered above in terminology and in artificial intelligence. Nothing that is being said in the discourse can have an impact on them. The discourse in which we talk about carburettors and bureaucracy apparently does not make us any wiser. Our minds are monadic entities, perhaps accepting stimuli from their environments but not embracing content as their input. The members of the society we live in can exchange and share content as much as they want; whatever is happening is not the consequence of their symbolic interactions, but of the innate programs that are switched on or switched off by unidentified triggers. This is, alas, not the reason why the computational model of the mind has become less referred to in recent years. Its disadvantage had always been that it was just a model, not the real thing. In some ways, the computer was still seen to be different from the mind, and there was much disagreement about which features of the computer actually corresponded directly to mental features. Over the years, a new paradigm began to emerge, the paradigm of the neural sciences. Ever more advanced technology made it possible to learn fascinating things about the brain itself, so that the computer as its model became expendable. The mental concepts may still have been constructs of the cognitive linguists, more model than reality. But once we learnt to observe directly which neurons would fire and which synaptic connections were made when a word was said or heard, then the reality of the corresponding concept would have been established, would it not?
Option c: The brain where language becomes reality In 2005, John Searle, in his unreservedly panegyrical review of Christof Koch’s book The Quest for Consciousness, summed up the new dogma: [T]he inner world of thoughts and concepts is forever hidden from consciousness, as is the external, physical world, including the body. One consequence of this hypothesis is that many aspects of high-level cognition, such as decision-making, planning, and creativity, are beyond the pale of awareness. These operations are carried out by the non-conscious homunculus residing in the front of the forebrain, receiving information from the sensory regions in the back, and relaying its output to the motor system. A further consequence is that you are not directly conscious of your thoughts. You are conscious only of a re-representation of these in terms of sensory qualities, particularly visual imagery, and inner speech. (Searle 2005: 37)
96
Meaning, Discourse and Society
For the reformed Searle of today, this is a timely confirmation that traditional intentionality, the problem that troubled him for such a long time, an intentionality tied to consciousness that gives us the power to reflect about our thoughts, an intentionality that smacks of relativism, has now finally been revealed as the phlogiston of the philosophy of mind. Reflecting on what we hear and what we say, on what we read and what we write, does not get us anywhere. The homunculus conundrum that he struggled to resolve ceases to be a problem once we reveal intentionality to be but a figment of our imagination. Then we can rest assured that all our experiencing is grounded in strictly material neural processes and therefore the embodiment of realism. The mind–body dualism that has worried people ever since Descartes has finally been replaced by scientific monism. Once we have a brain scan of someone’s cognition, all discussion must come to an end. In a similar vein, one of Britain’s leading art historians, Professor Emeritus John Onians, has now begun working in the truly innovative field of neuroarthistory, and published, in 2005, a book with this title (Onians 2005). On the web, I found this quotation by Onians: The most interesting aspect of neuroarthistory is the way it enables us to get inside the minds of people who either could not or did not write about their work. We can understand much about the visual and motor preferences of people separated from us by thousands of miles or thousands of years. (http://sciatnight.com/2006/09/ 06/neuroarthistory-whats-next/)
It seems that these days a painter’s brain scan will tell us all about their paintings, just as an author’s brain scan can easily replace their book. A brain scan will reveal and explain why Florentine Renaissance artists paid so much more attention to clear lines than their Venetian counterparts. After all, they were living in a town defined by narrow streets and high houses with clearly marked fronts. In Venice, on the other hand, the water of the canals reflected the sunlight and coloured them in ever-changing hues, while blurring the lines. In a review of Onians’ book, published in ScienceDaily, 7 September 2006, Onians’ claim is summarised in these words: Neuroarthistory can also explain why Florentine painters made more use of line and Venetian painters more of colour. The reason is that ‘neural plasticity’ ensured that passive exposure to different natural and manmade environments caused the formation of different visual preference.
I wonder if we really have to look at brain scans from painters working in Florence and Venice to notice the difference between disegno and colorito. After all, Renaissance artists from these two towns were fully aware of this ongoing discussion and had to position themselves accordingly if they wanted to sell their paintings. I do not doubt for a moment that continually exercising the brain in certain ways results in a strengthening of synaptic connections between the
Where should we look for meaning?
97
neurons in the brain area involved. Yet I am equally convinced that a Florentine painter could easily have ‘faked’ the Venetian style, and the other way around. Hans van Meegeren’s Vermeer forgeries were good enough to fool the curators of the Amsterdam Rijksmuseum. Should we believe that Meegeren’s brain scan would have borne a distinct similarity to that of Vermeer? John Onians’ analysis of brain scans has not revealed a single aspect of the paintings, their production or reception that is new or noteworthy, that has not been pointed out and discussed time and again in the art history discourse. In our new age, neurons are even seen to provide forensic evidence. Truth is in the brain scan. For the first time a suspect, the 24-year-old Aditi Sharma, has been convicted (to life) on the evidence of a brain scan for the murder of her fiancé. She had submitted to a Brain Electrical Oscillations Signature test (BEOS): After placing 32 electrodes on Ms. Sharma’s head, investigators said, they read aloud their version of events, speaking in the first person (‘I bought arsenic’; ‘I met Udit at McDonald’s’), along with neutral statements like ‘The sky is blue,’ which help the software distinguish memories from normal cognition. For an hour, Ms. Sharma said nothing. But the relevant nooks of her brain where memories are thought to be stored buzzed when the crime was recounted, according to Mr. Joseph, the state investigator. The judge endorsed Mr. Joseph’s assertion that the scans were proof of ‘experiential knowledge’ of having committed the murder, rather than just having heard about it. (Giridharadas 2008)
While even the technology-friendly New York Times remains (for the time being) staunchly sceptical, it might well be that we are witnessing a revolution in our legal systems: ‘Law enforcement officials from several countries, including Israel and Singapore, have shown interest in the brain-scanning technology and have visited government labs that use it in interrogations, Indian officials said’ (Giridharadas 2008). Should we really look at our brains as mechanisms, complex as they might be, that are documenting reality as it really is? Can I distinguish between the neurons that fire when I hear the sentence ‘I bought arsenic’ and the neurons firing on hearing the sentence ‘I bought strychnine’? More important perhaps is the question whether it is imaginable that we can objectively read a brain scan without engaging in an act of interpretation whose outcome is open? If we cannot agree what we can see on a painting, or on an X-ray, why should people agree what is on a brain scan? Therefore I am not so sure that there is much to be learnt from the observation of brain activity with the newest technological equipment. In this section, I will attempt to explore the reasons why cognitive linguists and former philosophers of mind seem so keen to devolve their responsibilities to brain experts. The new paradigm that is about to replace this old cognitive programme seems, once again, to unmask the intentional mind as a cherished illusion. Instead we are offered the intentional states of a non-conscious homunculus
98
Meaning, Discourse and Society
residing in our forebrain. We are called upon to accept as the true meaning of an utterance the neural activity it causes in cerebral tissue. Therefore it should not make us wonder that Dietmar Roehm (2004) calls it ‘successful language comprehension’ if it is technically possible to map an ‘event-related brain potential’ unambiguously to a language event. The promise is that neurolinguistics will answer the questions that non-intentional cognitive linguistics has invented. The cognitive sciences and the brain sciences have been two separate worlds. There is still no common language for neuroscientists and philosophers of the mind, and it is increasingly doubtful whether there will ever be one. Steven Rose, in his recent book The 21st Century Brain, has compared Yadin Dudai’s (1989) seminal Neurobiology of Learning and Memory with the equally influential Essentials of Human Memory, written by Alan Baddeley (1999), a leading cognitive psychologist, and concludes: ‘It is hard to believe the two authors are studying the same phenomenon, and the list of references at the end of each is virtually non-overlapping’ (Rose 2005: 212). Brain and mind are as irreconcilable as ever. Neuroscientists and cognitivists live in different universes. Just as the cognitive sciences are trapped by substituting the computer for the mind they claim to investigate, the neural sciences fail to make a connection between brain activity evidence and the content of representations. In his explorations of the links between the brain memory and the mind memory, Rose has to admit, time and again: ‘No one knows’. The two positions remain, at least for neuroscientists of Rose’s standing, irreconcilable (Rose 2005: 212). Do we make our own decisions, or does the brain decide for us? Does the brain process our linguistic intake, or do we understand texts? Where little is known, all answers seem possible. Claims remain unfalsifiable. This is, as I see it, very much the situation with cognitive linguistics today. We just do not know how language impacts on thinking, or how our thoughts are turned into language. All we hypothesise about the nature of thinking, its universality, its relationship to language, its relationship to neural processes, is largely uncorroborated by empirical data and shareable reasoning. Some analysts insist now that this sorry state of affairs is about to change. This, they say, is due to the increasing collaboration between experts from both fields. George Lakoff, the Berkeley professor who has made conceptual metaphors of the kind ‘life is a journey’ or ‘love is war’ a household name, and who has for a long time seen the mind as an embodiment of the body (for example, in his book co-authored with Mark Johnson Philosophy in the Flesh (1999)), has recently joined forces with Vittorio Gallese, professor at the Department of Neuroscience at the University of Parma, in a common endeavour to relate brain activity to what we believe to know. The result is their paper ‘The brain’s concepts: the role of the sensory-motor system in conceptual knowledge’ (Gallese and Lakoff 2005). Their first sentence reads:
Where should we look for meaning?
99
Concepts are the elementary units of reason and of linguistic meaning. They are conventional and relatively stable. As such, they must somehow be the result of neural activity in the brain. The questions are: Where? And How?
The one key problem is hidden in the word result. Is it, as John Onians tried to make us believe, the mental perception of the environment in which we find ourselves that moulds the brain, or is it the brain that moulds what we experience? Or does it have to be a causal relationship at all? What was wrong with the computational theory of the mind was, Gallese and Lakoff maintain, ‘the propensity to analyse concepts on the basis of formal abstract models, totally unrelated to the life of the body’ (p. 455). The second, more important, issue is the nature of ‘linguistic meaning’. A linguistic expression, for instance a lexical item or a word, stands for something else, a certain content. That is what makes it a sign. What it stands for is something that people have come to agree upon, by way of explicit or implicit negotiation, unless we happen to believe that we were born with a mechanism that makes the mental concept of ‘carburettor’ or ‘bureaucracy’ available once we are exposed to the right stimulus. Meaning is, of necessity, symbolic. Something that is not a sign, not a symbol, does not mean anything. Even when we ask ourselves about the meaning of life, we look at life as standing for, or realising, something else, for instance a test determining our place in life after death. Gallese and Lakoff acknowledge that this is how concepts were seen in the old days, but believe that now those days are past: Thus, following the line of arguments of early cognitivsm, concepts are symbolic representations by nature, and as thinking, they can be reduced to symbolic (not neural) computation. We will propose a radically different view. We will argue that conceptual knowledge is embodied, that is, it is mapped within our sensory-motor system. (p. 456)
It should be noted that in their view of cognitive linguistics, Gallese and Lakoff refer to a computational model of the mind like that of the early Fodor, in which ‘symbolic computation’ is seen as ‘a functional system whose processes can be described in terms of manipulations of abstract symbols according to a set of formal abstract rules’, treating them not as symbols at all, but as the elements of a rule-based process, like a game of chess. This has been the trick of editing intentionality out of meaning, by reducing understanding to an algorithmic (often called syntactic) computation. Even though Gallese and Lakoff initially claimed that concepts are the result of neural activity, they are now ready to accept that the outside world cannot be neglected. Thus they call their framework an ‘interactionist theory of meaning’, because it is ‘structured by our constant encounter and interaction with the world via our bodies and brains’ (p. 456). Apparently it is only structured by the non-symbolic environment of brains, and not by the content of the discourse. The concept chosen by Gallese and Lakoff for their enquiry is ‘grasp’: ‘We will show that the information structure needed to
100
Meaning, Discourse and Society
characterise the conceptual structure of grasp is available at the neural level in the sensory-motor system’ (p. 457). This is the information structure: ‘To understand the meaning of the concept grasp, one must at least be able to imagine oneself or someone else grasping an object. Imagination is mental simulation’ (p. 458, emphasis in the original). But how is the mental act of imagination related to brain activity? It seems there is a verbal spell that does it: make use of: All actions, perceptions, and simulations make use of neural parameters and their values. For example, the action of reaching for an object makes use of the neural parameter of direction; the action of grasping an object makes use of the neural parameter of force. So do the concepts of reaching and grasping. Such neural parameterisation is pervasive and imposes a hierarchical structure on the brain: The same parameter values that characterise the internal structure of actions and simulations of actions also characterise the internal structure of action concepts. (p. 458)
Normally it is people, agents, who make use of things. Not so here. Instead we find actions being equipped with agency. It is now the action that decides if it makes use of a given neural parameter or not. The neural parameters an action has made use of characterise the internal structure of this action, imagined or real, and they are mapped onto the internal structure of the corresponding (mental? or neural?) concepts. This is how neural parameters are defined: The firing frequency over that collection of neurons is a neural parameter and the mutually exclusive low, medium, and high firing frequencies are values of that neural parameter. (p. 465)
It seems that the firing of neurons characterises the internal structure of actions, and subsequently of the corresponding concepts. The cluster of neurons firing when I reach out for something and those when I grasp something are, or represent, the parameter of direction or of force. Perhaps we can interpret these parameters as the semantic primitives, the basic building blocks of more complex concepts, that certain cognitivists (for instance Wierzbicka) believe to be the innate equipment of our minds. At the end of their project, Gallese and Lakoff are able to present the neural concept of ‘grasp’. However, what the reader is given is not a brain scan in which the relevant neuron clusters are highlighted, but a description that is not so dissimilar to the ones we were confronted with in the days of generative semantics: The example of grasp The example we have been using all the way through is the concept grasp. Here is what a schema for grasp might look like in this theory. The parameters divide up in the following ways: The grasp schema. The role parameters: agent, object, object location, and the action itself.
Where should we look for meaning?
101
The phase parameters: initial condition, starting phase, central phase, purpose condition, ending phase, final state. 3. The manner parameter. 4. The parameter values (and constraints on them). The various parameters can be described as follows. Agent: An individual. Object: A physical entity with parameters: size, shape, mass, degree of fragility, and so on. Initial condition: Object location: within peri-personal space. Starting phase: Reaching, with direction: toward object location; opening effector. Central phase: Closing effector, with force: a function of fragility and mass. Purpose condition: Effector encloses object, with manner (a grip determined by parameter values and situational conditions). Final state: Agent in-control-of object. (p. 467)
The neural parameters listed here (which we are supposed to read as [equivalent to] the relevant neuron clusters) look surprisingly similar to the categories found in more traditional models of cognitive linguistics. But cognitive concepts are signs; they are symbolic; they stand for something else, and that is why they need to be interpreted, either by a conscious homunculus, or processed as uninterpreted symbols by a homunculus without intentionality. Neural concepts, however, seem to be different. They are not signs standing for something else. They are like Gertrude Stein’s rose which is ‘a rose is a rose’. The word object we find above in the sketch of the grasp-schema is only a name given to a parameter, and this parameter is not symbolic but a cluster of neurons. This should give the reader a pretty clear idea of how a grasp schema is structured in terms of neural parameters and values of the sort we described in the sections above on neuroscience. Note that we have written down symbols (e.g., final state) as our notation for functional clusters. This does not mean that we take functional clusters themselves to be symbolic. The symbols are only our names for functional clusters, which, as we have seen, function from a computational point of view as neurally realised units. (p. 467)
Is such a ‘computational point of view’ acceptable to Gallese and Lakoff? Their conclusion seems to be an outright rejection. This is how they sum up their neural approach to language: Another major consequence concerns language. Language makes use of concepts. Concepts are what words, morphemes, and grammatical constructions express. Indeed, the expression of concepts is primarily what language is about. If we are right, then: Language makes direct use of the same brain structures used in perception and action. Language is not completely a human innovation. There is no such thing as a ‘language module.’
102
Meaning, Discourse and Society Grammar resides in the neural connections between concepts and their expression via phonology. The semantics of grammar is constituted by cogs – structuring circuits used in the sensory motor system. Neither semantics nor grammar is modality-neutral. Neither semantics nor grammar is symbolic, in the sense of the theory of formal systems, which consists of rules for manipulating disembodied meaningless symbols. (Gallese and Lakoff 2005: 473)
For me, the key statement here is that ‘semantics is [not] symbolic’. Gallese and Lakoff’s neural concepts, while being expressed by the words of natural languages, are not symbolic; they are not signs standing for something else. The concept ‘grasp’ stands for itself, namely for the cluster of neurons firing when the brain is activated by the corresponding phonemes, or by the perception of a ‘grasp’ action, or by the imagination of a ‘grasp’ action. But who, or what, is here imagining something? ‘Imagination is mental simulation’, we read above (p. 458). Apparently Gallese and Lakoff cannot entirely discharge the mind as easily as they would like to. They still need it, and the mind they need has to be an intentional mind. A computational mind can hardly be said to imagine anything. The same, I think, is true of the brain. No brain scan with highlighted neuron clusters will ever tell us anything about the word grasp, or the concept ‘grasp’. Neurons are neurons are neurons. Even if neurons were material signs standing for something else, as a red traffic light stands for an instruction to stop, then we have to be told, by the use of language, what it stands for. Just as the ‘mental representation’ of a word or an utterance cannot be communicated other than as a linguistic representation of it, that is as a paraphrase of the word or utterance, the addition of the ‘neural representation’ needs to be expressed in natural language terms, as we have seen above. The only way to talk about mental or neural concepts is to translate them back into ordinary language expressions. As long as mental or neural content cannot be accessed directly, but only in verbal form, cognitive linguistics and neurolinguistics gives us little more than a duplication or triplication of the original expression. Words are signs. Language is, of necessity, symbolic. It seems that the new brain sciences have so successfully intoxicated eminent philosophers such as Daniel Dennett and John Searle, that they have forgotten this synthetic truth, namely that what we call language is a mode for communicating symbolic content, that is content whose meaning has to be interpreted. Peter Hacker, the highly respected Wittgenstein interpreter is also focusing his attention on the new brain sciences, though with a critical perspective. He has joined forces with Maxwell Bennett, an accomplished neuroscientist. Together they wrote the book Philosophical Foundations of Neuroscience (Bennett and Hacker 2003),
Where should we look for meaning?
103
in which they explore the philosophical foundations of this discipline which has rather suddenly become too attractive. What divides Bennett and Hacker from Dennett and Searle is the issue of agency (closely correlated to intentionality), which is the (apparent) ability of human beings to be in control of their actions. If there is agency, then this is what defines a person as a holistic entity, but it is not a feature of the brain. Bennett and Hacker’s book was strongly criticised by Searle and Dennett, who, though not in agreement on other matters, both reject agency as an illusion. In the fall of 2004 the American Philosophical Association invited the two authors and their two critics to a meeting for an exchange of their arguments. They subsequently agreed to turn their statements into written versions for publication as a book. The result is the fascinating small volume Neuroscience and Philosophy: Brain, Mind and Language (Bennett et al. 2007), with an introduction and conclusion by Daniel Robinson. In the first part, Bennett and Hacker put forward their arguments. Then there are rebuttals by Dennett and Searle, and finally Bennett and Hacker reply to their criticisms. Taken together, it is an excellent example of what academic discourse should be, namely plurivocal. While the contestants themselves make few concessions, the coerced collaboration results in new paraphrases of existing arguments. The permutation and reformulation of arguments is equivalent to a hermeneutic reinterpretation of the stances taken previously by the discourse partners. In my discussion of the discourse, I will first discuss Bennett and Hacker and then Dennett and Searle. My focus is, of course, on language: what it means to say that a linguistic expression has a meaning, and how this meaning can be tracked down. Unfortunately, language as such, even though awarded a place in the subtitle, does not make much of an appearance in this little book. Therefore we will have to content ourselves with the mental or neural representations of verbal, i.e. symbolic, content. As we shall see, the four authors all touch upon the issues that I have discussed in connection with Gallese and Lakoff’s article. Noteworthy in all the contributions is the absence of society. It seems that minds and brains are monadic entities which function independently from society’s input. But this is an issue I will revisit at a later point. For Bennett and Hacker, the key issue is ‘[t]he relations between the mind and the brain, and between the psychological and the behavioural’, and they are worried about the ‘persistent ascription of psychological attributes to the brain’ that they find in contemporary neuroscientific literature (Bennett and Hacker 2007: 6–7). Such entanglements are bound to ensue when terms used for mental concepts such as imagination, simulation, agent, object, symbol (as we found them in Gallese and Lakoff’s paper) are applied to the description of neural states. For Bennett and Hacker this constitutes either a confusing homonymy, or figurative or metaphorical usage, or these terms are used ‘as analogical
104
Meaning, Discourse and Society
extensions of ordinary expressions’, and they conclude that with ‘regard to the misuse of the psychological vocabulary’ involved in ascribing psychological predicates to the brain, all the evidence points to the fact that neuroscientists are not using these terms in a special sense (this usage here marked by the asterisk) (pp. 24–25). As an example they turn to Semir Zeki (2000: 547): ‘When Semir Zeki remarks that the acquisition of knowledge is a “primordial function of the brain”, he means knowledge (not [a homonym] knowledge*) – otherwise he would not think that it is the task of future neuroscience to solve the problem of epistemology (but only, presumably, of epistemology*)’. The same is true for representation and symbolic: ‘If “representation” does not mean what it ordinarily does, if “symbolic” has nothing to do with symbols, then it may indeed be innocuous to speak of there being internal, symbolic representations of the brain’ (Bennett et al. 2007: 26). Such neural ‘representations*’, however, can only be causally related to content in the way a speech event is related to the tape on which it is recorded, while mental representations are ‘symbols or symbol systems with a syntax and meaning determined by convention’ (p. 27). This is as close as Bennett and Hacker ever get to the notion that normally the presence of a symbol suggests that there is more than one person to be considered: someone has to set it up, someone has to recognise it as a symbol, and both have to agree on the meaning. But, as Bennett and Hacker find, neuroscientists like to talk about symbols of the brain: But one cannot go on to say, as Frisby (1980: 8) does, that ‘there must be a symbolic description in the brain of the outside world, a description cast in symbols which stand for the various aspects of the world of which sight makes us aware.’ For this use of ‘symbol’ is evidently semantic. For while smoke means fire, in as much as it is a sign of fire (an inductively correlated indication), it is not a sign for fire. Smoke rising from a distant hillside is not a description of fire cast in symbols, and the firing of neurons in the ‘visual’ striate cortex is not a symbolic description of objects in the visual field … The firing of cells … do not stand for anything, they are not symbols, and they do not describe anything. (Bennett et al. 2007: 28, emphasis in original)
Even if the technology for observing the firing of neurons in all detail were available, this would not help us to interpret what has been said. We might correlate the magnetic data on a tape or the laser-readable data on a CD with the sound that is recorded. But it is the spoken text that we want to interpret, not the substance on which it is recorded. Daniel Dennett, unsurprisingly, disagrees. But first he points out that much else of what Bennett and Hacker say takes up criticisms that he, Dennett, voiced decades earlier, and he complains: ‘We have so much in common, and yet Bennett and Hacker are utterly dismissive of my work’ (Bennett et al. 2007: 76). This is perhaps not so mind-boggling. For he commits what Bennett and Hacker call a categorial error: using mind terms for brain stuff: ‘It is an empirical fact, and a surprising one, that our brains – more particularly, parts of our
Where should we look for meaning?
105
brains – engage in processes that are strikingly like guessing, deciding, believing, jumping to conclusions etc.’ (p. 86). Dennett’s master stroke has always been to deny the abyss between mind and brain. Therefore, he says, we are right to use these terms as we want. Hacker ‘discovers this ubiquitous use of intentional terms in neuroscience, and he’s shocked, I tell you, shocked! He doesn’t know the half of it. It is not just neuroscientists; it is computer scientists (and not just in AI), cognitive ethologists, cell biologists, evolutionary scientists … all falling blithely in with the game, teaching their students to think and talk this way, a linguistic pandemic’. But, he insists, it is legitimate to use these terms, for once we take the concept of a person apart we find that inside this person there are a possibly very large number of simple little mechanisms equivalent to machines: The idea is that when we engineer a complex system (or reverse engineer a biological system like a person or a person’s brain), we can make progress by breaking down the whole wonderful person into subpersons of sorts agentlike systems that have part of the prowess of a person, and then these homunculi can be broken down further into still simpler, less personlike agents, and so forth – a finite, not infinite regress that bottoms out when we reach agents so stupid that they can be replaced by a machine. (p. 88)
Can we be sure, then, that mental and neural phenomena are one and the same? Not quite, it seems. In his opening moves, when still insisting that he had arrived at Bennett and Hacker’s conclusions decades earlier, he quotes from his own book Content and Consciousness (Dennett 1969: 91): An analysis of our ordinary way of speaking about pains shows that no events or processes could be discovered in the brain that would exhibit the characteristics of the putative ‘mental phenomena’ of pain, because talk of pains is essentially nonmechanical, and the events and processes of the brain are essentially mechanical.
For me this means that the meaning of a sentence like ‘I am in pain.’ cannot be reduced to clusters of firing neurons. According to Dennett (1969), a person can interpret the sentence. A brain cannot, because it is a mechanical device. But according to Dennett (2007), a person is no more than the array of tiny computational devices of which the brain consists. Daniel Dennett, I have no doubt, will find it easy to reconcile the two positions. I don’t think I could. John Searle’s opening move is similar to Dennett’s: he addresses issues of agreement and also claims that he has ‘made these points myself in a number of writings’ (Bennett et al. 2007: 98). He concedes that Bennett and Hacker are ‘right to point out that one’s normal relation to one’s own experiences is not epistemic’ (p. 97). This means that we cannot have direct access to first-person experiences. All we can have is (our own) testimony of them. For Searle, such first-person experiences (technically: qualia) are ‘co-extensive’ with conscious states. These states exist in the brain. They are realised as higher level or system features. For examples, conscious thoughts about our grandmothers are processes occurring in the brain, but, as far as we know, no single neuron can cause and realize thoughts about a
106
Meaning, Discourse and Society
grandmother. Consciousness is a feature of the brain at a level higher than that of individual neurons. (p. 99)
He then quotes Bennett and Hacker to show where the disagreement lies: ‘They also say, “we part company with Searle … when he claims that mental phenomena are caused by neurophysiological processes in the brain and are themselves features of the brain”’ (p. 99). Searle, as we have seen above, is not really interested in people’s intentionality. For him, consciousness is at best a supervenient, secondary phenomenon without a real function. Searle distinguishes between intentionality (which he considers as something that defies scientific observation) and intentional states, which can be ascribed to computers and can be investigated. Thus he seems here to distinguish between consciousness, as a re-representation of what is actually the case, and conscious states in which brain are. We do not have to be conscious of the conscious states of our brain, but if we are, this consciousness is not about these states themselves, taken as representations of content, but a re-representation of this representation. For the brain to be in a conscious state, it takes, according to Searle, more than a single neuron. He does not tell us, however, how big the neuron cluster has to be to jump from the single neuron level to the level of a conscious state. In their response, Bennett and Hacker finally let society into the arena (as they had already done in their 2003 book Philosophical Foundations of Neuroscience [PFN]), even though still through the back door, and without a proper introduction. It happens while they point out the difference between neural clusters and books, both of which can be said to record content: We criticized J. Z. Young [1978] for holding, as many neuroscientists do, that the brain contains knowledge and information ‘just as knowledge and information can be recorded in books or computers.’ Professor Dennett avers that we did nothing to establish that there is no concept of knowledge or of information such that it cannot be said or encoded in both books and brains. In fact we did discuss this (PFN 152f.) But we shall explain again. A code is a system of encrypting and information transmission parasitic on language. A code is not a language. … Knowledge is not encoded in books, unless they are written in code. There is a code only if encoders and intended decoders agree on encoding conventions. In this sense there isn’t, and couldn’t be, a neural code. In the sense in which a book contains information, the brain contains none. (Bennett et al. 2007, p. 137)
What is said about encoders and decoders also (and even more) applies to language users. Knowledge is only knowledge if it can be accessed, if it can be ‘read’. But the neuron clusters are not there to be read, and no technological breakthrough will ever allow us to read someone’s thoughts as they are stored in that person’s brain. This brief outline of Neuroscience and Philosophy cannot do justice to the elaborate argumentation used by the four authors. My objective is different: I want to argue that the neurosciences cannot provide the answer to the conundrum of meaning. No matter how much we zoom in on the neuronal tissue,
Where should we look for meaning?
107
or on how exactly we can determine which synaptic connections are charged when clusters of neurons are firing, none of these investigations will enlighten us about the meaning of grandmother. The brain is not symbolic. It is related to brain-external stuff by physical or chemical, in any case causal and material features. It can store symbolic content, but it does not interpret it. The cognitive sciences, on the other hand, have failed to show us the reality of the mind. They have failed to cope with intentionality and consciousness. They have been unable to show us what a mental representation of an utterance or the mental concept corresponding to a lexical item would look like. In spite of the extremely complicated models and apparatuses that they have developed, all they have come up with are natural language paraphrases of natural language utterances. These paraphrases, however, contain nothing that has not already been said in the discourse. It is perhaps time to move on.
Option d: The discourse as the collective mind Cognitive linguistics has failed because it was not able to provide shareable data, in the sense of brute facts, for which there would have to be a consensus at least within the cognitive linguistics community. Instead we find a range of more or less carefully constructed models and hypothetical statements about the nature of mental concepts. All attempts to assign to the mind some kind of ontological status have failed. The mind as an interface between a person’s brain and that person’s behaviour or agency may be a useful conceptualisation but is never more than just that: an arbitrary construct. Yet if the mind is nothing but a construct, this also holds for the language of thought, for mentalese. There is no way to decide whose construct of a mental concept is more in line with brute facts, because of a lack of such facts. The mind has not become the philosopher’s stone of the alchemy of meaning. The brain sciences certainly can lay more claim to scientificity than the cognitive sciences. With the development of a rich technology that can detect, count, measure, analyse and make visible neural phenomena, there is an abundance of brute indefeasible facts. But the brain is part of the body. Material phenomena, such as neural phenomena, are caused by material phenomena and can cause material phenomena. We may be able to determine the neuron clusters that fire whenever the word grandmother is spoken. We may even be able to show which neuron clusters fire when people think of grandmothers in general as opposed to their own grandmother. But words, and any kind of imaginable mental concepts, are symbolic; they stand for something. Whatever they may stand for, they do not stand for clusters of neurons. There can be only one consequence. People interested in meaning have to abandon the search for the mind and for mental concepts. They also have to give
108
Meaning, Discourse and Society
up their hope that others like the neuroscientists have the answer. Is there an alternative? Lexical items, text segments and whole texts mean. They are meaningful for people, people endowed with their inalienable intentionality. But this intentionality is inaccessible. This must not worry us, though. People are not monadic entities whose interactions do not reflect on their nature. Even Christian hermits need other people to recognise them as hermits. Even Robinson Crusoe (or his author) had to invent a Friday to keep him intact as a person. The many Robinson Crusoes who had no narrators, and whose diaries (if there were any) were never found, made no impact on our discourse. They left no traces of knowledge behind. When we watch a play, we are not interested in the actor’s intentionality, not in their feelings or their world-views. We watch a play to find out about the people in the play. These people behave in a way that we can make sense of, they say things that mean something to us; they interact with their environment in a way we can interpret. What comes over as their intentionality is the way they (or more correctly, the actors enacting them) present themselves. What really makes a person a person is not how they ‘feel’ themselves, but how they appear to us. It is, then, our faculty of empathy that gives us the illusion of their intentionality, of seeing them as they are. It is our understanding of other people’s intentionality that matters. This is what we can talk about. And this is what makes us develop our own intentionality. A feral child, a child growing up without communicating verbally with other people, without sharing and exchanging symbolic content, will never develop a sense of their identity, of their individuality, of themselves as a unique person, defined by their feelings, their beliefs and their attitudes. Without being part of a network of symbolic content, without being part of society, there is no intentional mind. Our intentionality, our consciousness, comes into existence by being a part of society, a member of a discourse community. We develop thoughts, feelings, ideas, beliefs and attitudes in collaboration with others. Our intentionality is part of the collective intentionality of the discourse community to which we belong. Our mind is part of a collective mind. The mind, the brain, our individual intentionality and our individual consciousness will forever remain inaccessible. The collective mind, however, is real. It is the discourse. The discourse is, unlike the mind, an object that can be inspected and investigated. In this sense, the discourse is more real than the reality we normally take for granted. We can never be absolutely sure if there are angels or dragons or Iraqi weapons of mass destruction. Nothing in the reality out there tells us whether someone is a freedom fighter or a terrorist. Nothing tells us if there is such a thing as quarks. But our contributions to the discourse are real. They can be made accessible. They can be shared. They are there to be interpreted. The reality we take for granted is no more and no less the reality we
Where should we look for meaning?
109
have constructed for ourselves in the discourse to which we contribute. This discourse is our collective mind. Over the past four hundred years, people in the Western world have constructed a reality in which each person is a monadic autonomous entity fully in charge of themselves. But before that time and in other cultures, other realities were constructed, realities in which the individual person was always a part of the society in which they lived. Once we accept such a view as an alternative to the dominant paradigm, we can come to terms with issues such as symbolic content, intentionality, the mind, and, last but not least, meaning. I will explore this alternative in the next part.
Part II
Discourse and society
8
Language as discourse
Children learn to speak as they grow up among people who are already speaking. These people are normally their parents or other carers, or other people with whom they are brought into contact, and, of course, their peers. It is the interaction with other people that lets children learn the language(s) they encounter in their environment. It is not more normal for children to learn just one language than several. Multilingualism is the norm in many societies and situations, and this is not a new phenomenon. A child in Hong Kong will have few problems communicating with their mother in English, their father in Mandarin, their nanny in Malay and their peers in Cantonese. How they do it we hardly know. Their linguistic behaviour in given situations, and their own reports, are our only sources of information. We know that some children learn languages faster than others, and there is an abundance of literature detailing the ways in which progress is made in the course of language acquisition. We also know that the faculty for learning new languages without particular effort wanes during puberty. What we do not know, in spite of Noam Chomsky and Jerry Fodor, is how this language faculty works. So far, all the models representing it have had to be withdrawn at some point. What we call language in this sense is the children’s interactional behaviour that lets them deal with symbolic content. They learn the languages they encounter, including accents, dialects, registers, and levels of sophistication. The verbal skills of a child growing up in a deprived household, in which family members have other things to worry about than the niceties of a good education, will compare badly to those of children whose parents take great care to develop their eloquence. While the faculty for learning language (whether spoken, signed or written) is certainly innate, they learn what they encounter. There is no innate universal grammar, and no innate repository of mental concepts waiting to be triggered. It is like learning to participate in a game of baseball or football. Some elements can be gleaned from observation, others just seem to make sense and are the result of trial and error, and some things have to be told. There is a difference, though. In sports, rules have to be obeyed. If they are violated, there is no game, or it becomes a different game. We employ referees or umpires to make sure the rules are observed. Apart from teachers, editors and perhaps a few 113
114
Meaning, Discourse and Society
other professions, no one in particular is there to punish us for rule violations in language. Indeed, for spoken language, when it comes to grammar, not very much seems to be regulated. In a foreign environment, friendly local people will try hard to make sense of the few ill-applied words I speak. We normally get away with all kinds of unruly language. The language mechanism we carry around in our minds is not what I am interested in. For me, language is what takes place between people, not inside them. Using language is exchanging and sharing content. People are social beings. They grow intellectually if they are integrated in communities. In isolation, their minds tend to deteriorate. It is not the mode language takes that nourishes their thinking. Whether spoken, written or signed, what counts is the content, not the form of what is said. Human beings are unique in this ability to relate symbolic content, for instance the story that Freddy’s dog has just had a litter of three pups, two of them quite white and one of them with black paws. Freddy may live in the next village, and the man who tells us the story has perhaps not seen the pups himself, but has heard about it from Freddy’s wife, whom he has met shopping. Some animals are said to communicate just like human beings. However, I do not believe that any dolphin or any chimp will ever tell us a story about Freddy’s dog. We can look at the story about Freddy’s dog as a text. This text can be investigated from two points of view. We can look at its form, assigning to it a phonetic, syntactic and lexical description. We can also look at its meaning. We can assign to it a paraphrase. If the original text was in English, it can be paraphrased in Cantonese (which is the same as being translated into Cantonese), for those people in Hong Kong who do not speak English. But it will not be possible to reconstruct the original form from the paraphrase or the translation. Computers can be programmed to deliver a thorough description of its structure, a sentence or text parse, providing us with a full description of its form. Such a parse can be reverseengineered, to reconstruct the original form. Paraphrases and translations (at least as long as they are carried out by humans), on the other hand, are arbitrary acts involving intentionality and their outcome is not predictable. Every paraphrase, every translation is also always an interpretation. Interpretations are conscious acts. Their results are, unlike the results of computational processes, contingent. Meaning is what people have said about a word, a phrase, a text segment or a text. They will tell us the meaning of globalisation or grandmother or intelligent design, or what Hamlet’s famous monologue is ‘really’ about or how Shakespeare’s contemporaries understood his play Hamlet. Meaning is in the discourse. It is not the reality out there that tells us if there is such a thing as globalisation or intelligent design, and there is no strict method that will tell us what any text (segment) means. Otherwise we would need only one translation of Shakespeare’s plays in German. There are dozens, because every translator reads Shakespeare differently. Over the last fifty years, the word discourse has come to mean many things, so it might be useful to try to explain how I normally use this word. For me,
Language as discourse
115
the discourse, and the texts (utterances) that constitute it, are the only firm ground, the only reality we have. ‘We’, in this sense, is not only the community of linguists. It is also all the members of society reflecting on themselves, their relationships with other people, and on their grasp of reality. In this part of the book I will show that the raw reality out there, unmediated by the discourse, including the brute facts of which it consists, is never available to us. It is as inaccessible to us as the working of our individual minds. The reality out there includes the stuff we see and hear and feel, including the sky above us, the ground on which we stand, all the props, for instance apples and pears, that we have learned to distinguish, the temperature of the weather, our own bodies, the people around us, Iraqi weapons of mass destruction, angels and unicorns, prime number pairs, the discipline of linguistics and also society itself. For those without language all this is no more than stuff. For them there are no apples or pears as objects of reality, because they have no expressions, no lexical items, no language signs that would enable them to distinguish between the two. They may have non-symbolic concepts of them, making them prefer one over the other, but they would not be aware of these concepts. People who are not taking part in any discourse might still feel hungry, or cold, or even sad, but they would not be conscious of their states. We would not find them enjoying the beauty and sublimity of the Lake District, because beauty and sublimity would mean nothing to them. They would have no sense of identity, and they would not be aware of their thoughts. It is the discourse that makes reality available to us. This reality, however, is not the unstructured reality out there; it is a reality that we, the members of a discourse community, have constructed for ourselves. In many ways, it will not be totally at odds with the external reality. What we feel, what we do and find we cannot do, will determine what we make of the utterances we hear from other members. Those we agree with, those that we can repeat, in whatever form, would be those we find not in contradiction with what we take to be our non-symbolic, unmediated existence. But we must remember that these feelings, dispositions and drives are not accessible. Neither can we be sure to what extent they are really non-symbolic, ‘in the flesh’, as George Lakoff would say, unmediated by the discourse into which we are drawn. To feel the beauty of Bach’s fugues, intellectual puzzles more than hormone-arousing music; to prefer source water over Coca Cola; to want to do mountain-climbing, owe more to the discourse of middle-class complacency than to naturally occurring hormonal variations. What will be successful in the discourse is something we cannot predict. This is not because there are discourse-external factors determining how the discourse will develop. The discourse at large is irreducible to discourse-external laws. It is as unruly as the weather. What has led to what, can only be determined after it has taken place. We can measure the success of an utterance by the traces it leaves in subsequent utterances. We cannot predict whether the people taking global
116
Meaning, Discourse and Society
warming seriously, or their critics, will prevail. It might well be that the last person standing will still insist that global warming was caused by sunspots and nothing else. As to the role our non-mediated existence, or the reality out there, plays in the way the discourse evolves, we can only speculate. It is not something we will ever find out. What we have in our heads will be forever hidden from us. All we know is knowledge that is publicly known, knowledge that has been, at some point, contributed to the discourse, by other people or by ourselves. It is this knowledge alone that constitutes our reality, the only reality we can talk about. All we have is the reality of texts, the meaning they express, the knowledge they contain. The texts are real. Not all texts that have been entered into the discourse will be accessible. Many will be lost or forgotten, or will have just disappeared. But the remaining texts are there for us to discuss them, to form an opinion on them, to reflect on them, and to react to them by contributing more texts. These texts form the accessible part of the discourse. It is discourse that constitutes our only unquestionable reality. All we can negotiate is discourse itself, not the world outside it. I distinguish between discourse at large and smaller, special discourses which are subsets of it. The discourse at large, in its widest extent, consists of all spoken, written or signed utterances from the time when people started using language, in any dialect or language, as long as they had an audience. What people signed to themselves when they were not observed by anyone, what they wrote in diaries that were burnt before they were read by anyone else, the speeches people gave to imagined audiences while hiking in the solitude of dense forests, do not count as a contribution to the discourse, because such utterances forego the chance to have an impact on subsequent contributions to the discourse. The full discourse at large is not available to us. Most of what has been said over time has long been lost. The discourse at large which is, in principle, available for research is a repository, an archive, of those texts that exist in recorded form. In most cases they have been written. Some spoken language texts exist as recordings on audiotape or other mediums. Signed texts can be preserved, for instance on DVDs. Memorised, or remembered, texts in the heads of people do not count, as we have no direct access to them. Recordings of recitals of such memorised texts, however, are a part of the accessible discourse at large. This discourse is a live, active archive, and a never-ending stream of new texts is continually added to it. Even the discourse at large cannot be a sensible object of research. It is a bit like the map described in Jorge Luis Borges’ one-paragraph story On Exactitude in Science, a map of a province or an empire which is, in the end, made so detailed that it becomes isomorphic, in size and content, with the represented surface itself. What we can study can only ever be a tiny selection of discourse. But we have to define what this selection is in such a way that we can be sure for each
Language as discourse
117
text whether it belongs to this particular discourse or not. We might be interested only in texts written in English between 1 January 1901 and 31 December 2000. But much more is needed. We have to define what counts as a text written in English. We can include texts written by non-native speakers, or we can define a level of linguistic competence that has to be satisfied. We have to decide if we want texts only written by native speakers in English-speaking countries, or if we also want to include the first English tests written, for instance, by Chinese beginners. On the other hand, we can also define a discourse consisting of Shakespeare’s works, regardless of the language in which they are written. We may have difficulties deciding what counts as a translation, and what is more a re-creation. We can define a special discourse on global warming. We can say that we want to include in this discourse all newspaper articles written in English in which the phrase global warming occurs at least once. Should we also include texts which do not feature this phrase but mention climate change, or the greenhouse effect? Another discourse we might be interested in is that consisting of texts, spoken but recorded or transcribed, of Alzheimer patients. It is, first of all, up to the researcher(s) how to define their discourse. As I have shown, defining parameters can be either text-internal (e.g. language, or the occurrence of certain lexical items), or text-external, like parameters of space or time situation, or categories describing people. Often it will not be feasible to compile a corpus of all the texts belonging to a discourse. Then it is necessary to define in which way the texts included in such a corpus have been sampled. The reason why we should not just collect the texts we find interesting for a particular research question, but take great care in defining our research object, is that we want to give the other members of our peer communities the opportunity to agree or disagree with our findings in a principled way. The discourse is what provides the common undisputable basis for the negotiations. Setting up the parameters defining a discourse or a sample thereof will reflect our prejudices. As I will show below in Chapter 15, prejudices are unavoidable. Our prejudices sum up the knowledge we have, as an individual or as a group. By talking about them we become aware of them. This allows us to achieve a higher level of reflection. But as all discussion of content can be resumed at any moment, all the reality we have jointly constructed is generally only provisional. The discourse is open-ended. The consequence is that no interpretation will ever be the final one. This is what I will show in Chapter 16. Some may regard this absence of finitude regrettable, as it seems to cut the ground from beneath our feet. But it also ensures that the discourse remains open to innovation, to new ways to talk about the world, by recombining, permuting and reformulating what has been said before. Unless the discourse were open to unpredictable development, we would not be able to reconstruct the reality that we encounter in the discourse. It is the collaborative act of interpreting what we find, by accepting it, rejecting it or modifying it, that puts us in charge of reality.
118
Meaning, Discourse and Society
So far I have talked about special discourses mostly from the standpoint of the observer(s). We want to find out what has been said in certain texts at a certain time about certain things. But the people who are contributing their texts to the discourse are not ants or chimpanzees whose behaviour is being analysed by the appropriate experts and who themselves have no say in their findings. The people whose texts we analyse can talk back; they are in a position to discuss with us the interpretation of texts that, we, the discourse experts, suggest. Unless we are dealing with a discourse of the distant past or places far away, we should engage in a discourse with them about their discourse. For any discourse is self-referential. People talk about what people say. They know in a given situation what should be said and what not. The Tories know what Tories are allowed to say and what not. They have their own ideas of what belongs in a given discourse and what not. Therefore we, the observers, and they, the observed, will have to enter a reciprocal discourse; the observers will have to collaborate with the observed in the interpretation of their discourse. For each special discourse is only a segment of the discourse at large, and each text in a special discourse is intertextually linked to texts outside. There is no vantage point outside a discourse that can give us an undistorted, unprejudiced view on what is being said in a discourse. None of us will ever be able to escape the prison of the discourse at large. Just as the content of a discourse is a matter of interpretation, discourse boundaries are always arbitrary and can be redrawn. The special discourses thus are constructs of the discourse at large, the discourse where the observed and their observers meet. The way we often talk about these discourses implies that we see ourselves as participants in many of them, often at the same time. I say something in a discourse made up of the people in a classroom, and I say it as a member of the staff of an English department, and I also say it as one of the German academics who is teaching at British universities. What I say belongs to the discourse of linguistics, but also, I hope, to that of social studies. It is, more specifically, part of the discourse of a social constructionist stance. It belongs to a discourse of a given date, and in a given place, a discourse of people in their early sixties, of the male sex. Indeed, what I say can be pigeonholed in a nearly infinite number of special discourses. An interesting question is in what way can my definition of discourse be related to the ‘discursive formations’ at the centre of Michel Foucault’s quest for knowledge and meaning. Foucault did not see himself as a linguist. In France, the sixties were the heyday of a linguistics focusing on the language system, a kind of Saussurean structuralism dressed up as Chomskyan syntax, that owed its clout to Claud Levi-Strauss. This was a paradigm that Foucault thought inadequate to deal with the meaning of the discourse (‘parole’ in Saussure’s, and ‘performance’ in Chomsky’s terms). Like Claude Lévi-Strauss, he did not distinguish it from Chomsky’s universalist theory of language. Both were the linguistics of the language system. Even though Foucault purposefully disregarded the linguistic form of his discourses, and insisted on calling his object of investigation
Language as discourse
119
‘discursive formations’, he saw them as real objects, as objects comparable to the artefacts dug up by archaeologists. For Foucault, a discursive formation is an ‘objective’ entity, something that is defined by discourse-external givens. Whether a given text belongs to a given discourse is not something that has to be negotiated. It depends on its being covered by the rules underlying this discursive formation. In his view, there has to be a set of scientifically discoverable regularities: If I succeed in showing, as I shall do shortly, that the law of such a series [a group of acts of formulations] is precisely what I have called so far a discursive formation, if I succeed in showing that this discursive formation really is the principle of dispersion and redistribution, not of formulations, not of propositions, but of statements (in the sense in which I used this word) the term discourse can be defined as the group of statements that belong to a single system of formation. (Foucault 1972: 107, emphasis in the original)
This is repeated some pages later: ‘Now, what has been described as discursive formations are, strictly speaking, groups of statements’ (Foucault 1972: 115). What binds them together had already been explained. It is a law: in order to find out what binds the statements of nineteenth-century doctors together, ‘we must first discover the law operating behind all these diverse statements’ (Foucault 1972: 50). The identity of the discourse is established by a law, by regularities, by a set of rules operating in the texts belonging to a given discourse. This is a theme repeated countless times in his Archaeology of Knowledge. Another example is: ‘[The] ‘discursive practice’ … is a body of anonymous, historical rules, always determined in the time and the space that have defined a given period, and for a given social, economic, geographical, or linguistic area, the conditions of operation of the enunciative function’ (Foucault 1972: 117). The final goal of the discourse researcher is to write the archaeology of a given discourse. ‘Archaeology … is to uncover the regularity of a discursive practice’ (Foucault 1972: 144f.). For Foucault, a special discourse is a real-world entity whose identity rests on laws, rules and regularities. The question he does not address in a straightforward way is the nature of these principles. Are they, as the word law would imply, comparable to the immutable laws of nature, such as the famous second law of thermodynamics? Are they rules instituted by social convention, either explicitly or implicitly? Are they maxims each member of a discourse society has to follow out of a moral obligation? ‘The rules of the formation [of a text]’, he says, ‘operate not only in the mind or consciousness of individuals, but in the discourse itself; they operate … on all individuals who undertake to speak in this discursive field’ (Foucault 1972: 63). For me, this sentence is more an obfuscation than a clarification. Foucault was writing at a time in which, in his words, a ‘discourse rupture’ took place. In French intellectual circles, particularly in the Tel Quel grouping, the fascination with structuralism was increasingly clashing with the rediscovered charms of hermeneutics and the kick-off of post-structuralism. For reasons
120
Meaning, Discourse and Society
for which we perhaps have to search in Foucault’s biography, he was vitally attracted to the concept of order in the chaos of reality, to the quest for some Archimedean point which would distinguish necessity from contingency. Linguistics, as he understood it, had nothing to contribute to the interpretation of the discourses of the past in which he was interested. Yet he also retains the structuralist obsession with the concept of an abstract system, a system existing independently of the texts that make up a discursive formation. While the focal point of his enterprise is the interpretation of the thought archives of the past, he does not establish them as emanations of la parole, but of a system based on rules. Thus Foucault’s approach leads him, in the end, into a cul-de-sac. He is, in all his writing, unable to give us a clear example of what the laws, rules, and regularities are that establish the identity of a discursive formation. Many years ago, Dietrich Busse and I claimed that the texts belonging to a given discourse have to deal with the objects, topics, epistemic complexes or concepts of the study to be carried out, that they are therefore semantically related and share a set of statements, communicative situations, functions and purposes (Busse and Teubert 1994: 14). Today, I would be more reluctant. All these conditions are themselves constructs made by the discourse observers. They decide what is semantically related, what is a communicative situation, what functions and purposes can be ascribed to texts. But it is also a discourse, namely the discourse of discourse observers, that establishes and constructs the discourse objects ‘discourse observer’ and ‘discourse contributor’ and consequently also ‘discourse community’. Even ‘discourse’ itself is an object of a discourse, constructed by that very discourse. A discourse, I would say today, is not generated by laws; it is contingent, and so is any interpretation of it. It is an arbitrary decision to conclude that the intertextual links obtaining between texts are strong enough to justify assembling these texts in a special discourse, while leaving other texts outside. There is no law to which we can appeal for an objective decision. Discourses do not exist as such. Paradox as it may sound, I would like to argue that it is discourse that creates discourse, in an autopoietic, bootstrapping fashion. A few words also need to be said about how my view of the discourse compares to that we find in the relatively recent, but certainly very active, field of critical discourse analysis. CDA is loosely based on Michael Halliday’s paradigm of Systemic Functional Linguistics, the dominant backbone of the kind of applied linguistics we find in all the university English departments in Britain and all over the world, wherever English Studies are moulded by British rather than American academia. At the heart of CDA is the relationship between the discourse and society. While I admire the plurivocality and open-minded dynamics of this approach, I take issue with the way it describes the relationship between discourse and society. For the doyen of CDA, Norman Fairclough, it is society that determines discourse. Discourse is ‘social practice determined by social structures’, ‘actual
Language as discourse
121
discourse is constituted by socially constituted orders of discourse’, and ‘orders of discourse are ideologically shaped by power relations in social institutions and in society as a whole’ (Fairclough 1989: 17). Even though he accepts that discourse also ‘has effects upon social structures’, it is clearly society that is in the driver’s seat. We are to assume that there are social structures which somehow pre-date discourse. It is a perspective of society reminding of the way biologists look at the social lives of ants or bees, a perspective from the outside. For us, whether linguists or just lay members of discourse communities, such a perspective is not viable. While we can talk about queens in bee hives and alpha males in primate troops, I am not sure if these queens and alpha males are aware of their status. Should we really expect drones to be depressed because they are not given a share of the work to be done and exhaust themselves in their reproductive activity? For me, it is the discourse that constructs social structures, not the other way around. There is no vantage point outside of the discourse from which an observer could describe the way a group of people is structured. Foucault’s ‘orders of discourse,’ too, are not possible without a discourse at large in which these orders must be negotiated. I agree, with Foucault and critical discourse analysts, that much of what we say is the result of ‘ideological’ positions that have been passed on to us by authorities. But I would not be aware of authority if it had not been verbally communicated. Neither bees nor apes reflect consciously on the hierarchies in which they find themselves. Hierarchies have to be discussed in a discourse in order to become an object of this discourse, and only if they are discourse objects can we make ourselves aware of them. The way I see the relationship between discourse and society puts me, I am afraid, very much in a minority position. Even one of the most outspoken critics of CDA, the linguist Henry Widdowson, leaves no doubt that ‘[y]ou cannot study discourse in disregard of social factors and I do not know of anybody who claims you can’ (Widdowson 1995: 159). What he argues is that ‘discourse analysis has made statements about social attitudes and beliefs, the exercise of power, the influence of ideology, and so on, with scant reference to the linguistic data; and another tradition has made statements about the specifics of language in use without paying much attention to social factors’ (p. 159). While I am not aware of any social factors which are not first of all discourse constructs, I fully agree with Widdowson’s admonishment that only real language data, i.e. the reality of the discourse, can be the evidence on which the peer community can agree. This is also Michael Toolan’s position: CDA ‘needs to be more critical and more demanding in the text linguistics it uses [and] it must strive for greater thoroughness and strength of evidence in its presentation’ (Toolan 1997: 101). For CDA, discourse is a social practice. This is not my view. As I will try to show in this part of my book, it makes sense to turn the relationship around and to view society as a discourse construct. There are people, and sometime these people interact. They do things together. They help each other. They fight each
122
Meaning, Discourse and Society
other. So far this is not different from a troop of primates. What makes up society is not the people themselves but their interactions; it is what is happening between people. Society becomes an object of observation and reflection at the very moment when people’s interaction becomes symbolic, when acts carried out have acquired a meaning, a meaning which has been discussed, interpreted and reflected upon by the people and a certain consensus (which need not be comprehensive) has been reached. Society is the sense that a group of people make of themselves. An interaction becomes symbolic only when people assign a meaning to it. Throngs of people on the street can mean many things: people are trying to get hold of something considered valuable but scarce (when in Britain they would queue); a charismatic preacher is elating the spirits of his followers; a procession of students are on their way to their degree congregation; millionaires are demonstrating for tax cuts, or the beginning of the world revolution. A society is what remains, for the members of a society and for their observers, once one disregards the people’s individual intentionalities and focuses only on their shared, collective intentionality. While a person’s intentionality is something that cannot be observed, the collective intentionality of a group is accessible. It is expressed in that group’s discourse. The relationships obtaining between the members of a group exist in, or can be inferred from, the texts that make up their discourse. Without discourse, there are no kinship relations, there are neither friends nor foes, there are no kings and no subjects, there is no property, there is no freedom, and there are perhaps not even genders. I obey the orders of my line manager because I have been told that I will lose my job if I disobey her. I know that there are Iraqi weapons of mass destruction because it has been continually asserted in the media I consume, as long as I have not heard anything to the contrary. The members of a chimpanzee community are not aware of a power structure. Only to her human observers does it seem that that a female group member is deliberately grooming an alpha male because she wants to move closer to the top. She herself is unaware of her motives, if she ever had any. She is also immune to the ideology that chimpanzees are ‘better’ than gorillas, because there is no discourse in which such an ideology could be formulated. Her behaviour, as well as that of her fellow group members, is non-symbolic. Only to us, to her observers outside of her discourse, do her interactions make sense. Once there is language, once there is a discourse, ideology becomes unavoidable. If the world outside is not (directly) accessible by us, then we cannot tell whether the world we construct in our discourse properly represents this reality, including our social reality, the reality in which our symbolical interactions take place. Jürgen Habermas, who is often quoted in CDA texts, accepts Karl-Otto Apel’s view that we should look at the discourse as some kind of metainstitution on which all other social institutions depend, for social activity consists solely in ordinary language communication. (‘Es hat einen guten Sinn,
Language as discourse
123
Sprache als eine Art Metainstitution aufzufassen, von der alle gesellschaftlichen Instituionen abhängen, denn soziales Handeln konstituiert sich allein in umgangssprachlicher Kommunikation’ (Habermas 1971: 53)). However, he continues: ‘But the discourse is also a medium of domination and social power. It serves to legitimise relations of organised force. To the extent that the legitimisations fail to spell out the power-based relationship which they enable, even though it is contained in these legitimisations, the discourse is ideological, as well.’ Because we cannot see through the power-base relationships. Habermas claims, the discourse represents reality in a systematically distorted way, and the consensus to which sense negotiation leads is forced. Instead of communication, there is, for him, only pseudo-communication. He does not tell us, though, how we would recognise a non-distorted discourse. He insists that it is due to the distortion of the discourse that we end up having a biased view of reality. But is this reality really available for our scrutiny? I do not want to be misunderstood. Of course I accept as a fact that there is a social reality which I encounter in my interactions, that there are people who put themselves in charge of me, that equality before the law is an illusion, that there is plenty of torture even in the so-called civilised world, and that women are still in many respects oppressed. That these are facts is shared opinion of many of the discourses in which I participate. There may be other discourses in which these opinions are not considered mainstream. But it is my own personal decision to choose the former and to shun the latter. Taking a moral stance means to choose, on private grounds, between available options. Different discourses construct different realities. None of these ‘realities’ can claim to be the only one. They do represent different ideologies. But they never represent reality as it ‘really’ is.
9
Society presupposes language, and language presupposes society
Language as an essential human feature Society presupposes people interacting with each other, and it presupposes symbolic content. For what turns behaviour into interaction is that a meaning is assigned to it. A congregation of people on the town square is more than the accumulation of each participant’s meaningless behaviour; it can be interpreted as a political demonstration, a religious ritual, a celebration, a congestion of shoppers, or the beginning of world revolution. What it means has to be negotiated by the people involved in the interaction. It is not the sociologist or the anthropologist who has the last word. This is what sets a society apart from a pack of dogs or a colony of ants. Their behaviour becomes an interaction only by the grace of the observer. Unlike dogs and ants, people can talk back to their observers. When linguists want to find out what a verbal utterance or any other interaction means, they have to ask the people. That meaning is constructed by the people, and not by those who observe them, is something linguists, sociologists and anthropologists tend to forget quite easily. Verbal communication is a prerequisite of society. That does not mean we cannot survive in a social situation without speaking the other people’s language. An anthropologist has good chances of surviving within a monolingual tribe somewhere in an uncharted valley of Papua New Guinea. Gesturing will go a long way. Together, they will eventually move on from indexical or deictic gestures to symbolic gestures, and they will soon also develop an incremental common vocabulary. For integrationist linguistics, a fairly recent approach to looking at acts of communication as being embedded in a given situation, this means that there is no strict borderline between non-symbolic behaviour and symbolic interaction. For integrationists, language as strictly symbolic representation is only one form of interaction and communication. In their view, it is impossible to distinguish clearly between the symbolic and non-symbolic elements of a spoken utterance. They would also reject the notion that the signs of which language consists – words, phrases, sentences, etc., are ready-made artefacts that are there for us to use. What a word means, they would say, is the result of an ad hoc negotiation between speaker and hearer. Whenever people communicate, they turn words into signs. This stance is, at first view, attractive; it explains why single words in isolation are often so fuzzy and ambiguous. 124
The language–society link
125
Yet it does not fully convince me. That there is a convention linking the acoustic sequence
to a concept ‘chair’ is something speakers have normally learned before they make use of it. Of course the concept is vague enough. There is the prototypical chair, the implement in our room referred to as chair by our mother; there is the chair we came across in a young children’s picture dictionary; and there are other chairs, easy-chairs, wheelchairs, office chairs, barstools and a wide range of borderline cases. What is normally called a chair, and what is not, has to be learned. Of course, we acquire the first crumbs of language by trying to interpret what initially might appear to us to be non-symbolic behaviour as symbolic interaction. This is how our anthropologist and her tribe in Papua New Guinea will begin to get on with each other. While they will not exchange verbal utterances, they will communicate by gesturing and making facial expressions to indicate their (friendly or aggressive) mood, and by pointing to nearby objects. If our anthropologist cannot find a translator, and if she has the patience of Robinson Crusoe, she and the members of the tribe she is visiting will eventually agree on some kind of common language, and only then will they be able to exchange narratives, like the story about a dog’s newly born pups. Narratives are normally about things outside of one’s immediate environment. Ostentation does not help with them. Indeed, Marcel Marceau’s fame rests exactly on his art of taking nonlinguistic communication to levels we normal people cannot hope to attain, to turn it into symbolic communication and to narrate stories as if he were using language. Still, integrationist linguistics is a much-needed counterweight to mainstream linguistics. It rejects the key notion of twentieth-century mainstream linguistics, namely a stable, possibly innate language system that lets us predict what can and can’t be said. Instead it aims at developing a linguistics that is able to handle la parole. The concept of a language system as such is regarded as vacuous. Only utterances we find in la parole, in the discourse, have a meaning. It is a pleasure to read how Roy Harris, the key proponent of integrationist linguistics, takes apart prominent linguistic top-down paradigms of the twentieth century, like that of Noam Chomsky, until they fall to pieces, for instance in his book The Language Machine: A language is no longer envisaged as needing human language users at all, or a human brain to house it, as long as it has a machine to generate it; and its sentences serve no social or communicational purpose of any kind. (Harris 1987: 74)
The only problem I have with his all-encompassing, integrationist approach is that it does not sufficiently distinguish between symbolic social interaction and non-symbolic behaviour. I readily admit that it is a contingent distinction, an arbitrary decision as the result of an interpretation of evidence at hand. Normally people interact perfectly well without asking themselves which elements of their interaction are symbolic and which are not. When I produce a spoken utterance, I do not consciously distinguish between non-symbolic
126
Meaning, Discourse and Society
features (e.g. my ineradicable German accent) and features that may contribute to meaning (such as my intonation). While my accent is a symptom betraying my background, my intonation can add a certain interpretation to what I say, and can therefore be viewed as symbolic. People involved in an interaction often have better things to do than to reflect on which part of what is happening is symbolic. Without being prompted, members of oral societies seem not to be aware that their (spoken) acts of communication can be described as symbolic. It takes literacy to form the idea that there are things such as language and words, and that words are signs standing not for their acoustic value but for something else. Symbolic interaction presupposes the arbitrariness of signs. Arbitrariness does not imply that once the meaning of a sign (a lexical item) has been established it cannot change again. It only suggests that the meaning of signs is contingent and subject to a never-ceasing negotiation between the people involved. Take this example: ‘“It’s a stated policy of this government to have regime change (in Iraq). And it hasn’t changed”, President Bush said at a press conference last week. “Regime change” is the recent euphemism for overthrowing a government by covert or overt means.’ (The Gazette, Colorado Springs, 13 July 2002). The text argues that the lexical item regime change has changed its meaning and is now used as a euphemism for military intervention, and we are to understand that this semantic change should not be easily accepted. Discourses are full of contributions of members proposing or resisting a change to the meaning. It is this social activity of negotiation that forces us to develop a capacity of interpretation, taking us beyond merely responding to stimuli and making us reflect on what is happening. It makes us conscious that something is going on. The collective act of negotiation causes the individual mind to develop an intentional state, to become aware that an utterance is more than just a reflex to a stimulus, that utterances are about something. When in humankind’s dawn a member of the tribe encountered a lion, their response would not be just to run away but also the ejaculation of a loud shriek, a sign for the people around to join in the flight. The ‘meaning’ of this shriek, however, was not arbitrary, was not the result of negotiations in which different people offered different interpretations. The shriek was no more than an instance of automated behaviour. But the integrationist linguist Michael Toolan has plausibly explained that an increasing complexity of the social group and its interactions made it desirable to distinguish between different shrieks, for instance those for running away from the lion and those for chasing the wild boar a hunter saw in the distance (cf. Toolan 1996: 106 ff.). At that point, possible ambiguities had to be recognised and resolved. This is how automated behaviour became symbolic interaction. When we travel abroad, we find that even the meaning of a red traffic light at a road intersection is a sign that can be interpreted in different ways. Normally, we
The language–society link
127
take a red light as a stimulus to which we more or less unconsciously respond. In some countries, such automatic reflexes might get us into trouble. Travel guides often provide the necessary knowledge. They try to make us believe that a red light in Italy means that you do not have to stop but should drive on more carefully. The line between uninterpreted behaviour and symbolic interaction, often hardly visible and poorly defined, confines my remit as a communication researcher. Below, in Chapter 10, I will return to integrationist linguistics, which, as I see it, does not accept this distinction. Language and society We talk about society when we interpret the behaviour of a group of people as symbolic, that is meaningful, interaction. This begs the question of whether the people themselves are part of this society. It is a question that is not as absurd as it appears. Where do we have to look for the meaning of people’s interactions? How is meaning assigned to a congregation in the town square? As I have shown in the previous part, for cognitive linguists meaning resides in the individual minds of the people. The accumulation of all these minds will tell us what the congregation means. This is the position of the German sociologist Max Weber. For him, a society is the aggregate of the intentionalities of all its members. Once we know what each person thinks we will know everything about this society. ‘We can accomplish something which is never attainable in the natural sciences, namely the subjective understanding of the action of the component individuals’ (Weber 1968: 15). But the insurmountable problem with this theory is that people’s thoughts are not accessible to observation. For linguistics, Weber’s approach would mean a multiplication of minds in need of investigation. As long as we accept the view that minds are not up for inspection, it is a view we must reject. Fortunately, sociology offers other views, as well. The French sociologist Emile Durkheim has an intermediate position. For him, society is ‘more than the sum of the individuals who comprise it, and this “something more” is the distinctive domain of the science of sociology’ (Hughes, Martin and Sharrow 1995: 160). More precisely, he explicitly states, in his letter on ‘The psychological conception of society’ (Durkheim 1982: 40) ‘That the content of social life cannot be explained by purely psychological factors, namely by the states of the individual consciousness, seems to us as plain as can be.’ Durkheim stresses that for a social fact to come about, there must be a ‘plurality of consciousnesses’: ‘[S]everal individuals at the very least must have interacted together and the resulting combination must have given rise to some new production’ (p. 45). Durkheim, though, does not show how we can distinguish between symbolic content present in the individual mind and that present in society at large. As we
128
Meaning, Discourse and Society
have seen above (Chapter 1), this is essentially also true for the position of Ferdinand de Saussure’s Course in General Linguistics, at least as far as la langue is concerned, i.e. the system that the (French) language constitutes. He calls it a ‘social product’ which is located ‘potentially in every brain, or more exactly in the brains of a group of individuals; for the language is never complete in any single individual, but exists perfectly only in the collectivity’ (Saussure 1983: 13 [30]). In Saussure’s view, language is both a mental and a social phenomenon. It might well be that this dichotomy was not seen as something relevant then. For Saussure it was the system as such that mattered, the web of more or less obvious regularities that guide our lexical and grammatical choices and which exist whether we make use of them or not, just as the laws of mathematics would still be the same if no one was aware of them. La langue is a system that, for Saussure, exists in the same celestial sphere as mathematics, an object of study that does not require the researcher to learn about the build-up of the mind or of society. For Saussure, la langue is what justifies the existence of an academic discipline of linguistics; it is la linguistique propre. Noam Chomsky has been more outspoken about the location of the language system, as he sees it. For him it is only an abstraction, a theoretical model reflecting the properties of our language faculty, the ‘language organ’ we are all born with. His language system is not language specific, as Saussure’s langue, but universal, actually the system covering the language of thought, or what he used to call in the days of Generative Transformational Grammar the ‘deep structure’. As an organ, it is situated in the brain, as one of a number of modules of the mind. My own interest is neither la langue nor Chomsky’s universal grammar. They are arbitrary constructs, and anything can be said about them. My focus is language that exists in the form of data; namely that which has been said or written. These are the verbal utterances (and their symbolic content) that make up a discourse, not Saussure’s langue, but his parole. What Saussure calls la parole is, I believe, much the same as what I call a discourse, the entirety of all the (extant and accessible) texts (verbal utterances) entered by the members of a specific discourse community. Unlike the language system or la langue, this data can be made available for inspection by the members of the discourse community, and by their observers, linguists, anthropologists or sociologists. Unlike the language system, and unlike mental representations, real language data can be shared by linguists and they can be negotiated and interpreted. The utterances of la parole cannot be researched or described in isolation. This may have been the reason why Saussure never made good on his promise to deliver a theory of la parole. It would have coerced him into tackling the diachronic dimension of the discourse. An utterance is an event occurring in a given place and at a given time. In the case of an oral event, there is normally
The language–society link
129
one situation in which speaker and hearers are participating. In the case of a written utterance, there is a text. In both cases, the utterance means something. This is why people crave utterances. Only linguists would be interested in an utterance because it contains an accusative with an infinitive or a split infinitive. To be able to interpret a spoken utterance we have to find out about the speech situation in which it has taken place, including its environment, about the people involved, and about preceding speech events or texts, where relevant. In order to understand a (written) text, we have to know to which other texts our text is linked, by explicit or covert references, or by the traces it leaves in subsequent texts. This would have to be the remit of a linguistics concerned with la parole. Discourse, the (spoken and written) texts making up a discourse, takes place in a discourse community and therefore is a social phenomenon. If we discount the cognitive perspective, if we take away the individuals that make up society, what remains of language, and what remains of society? The German sociologist Niklas Luhmann probably has the most extreme view of society. For him, a radical constructivist originally in the tradition of Humberto Maturana and Francisco Varela, but challenging their cognitivist perspective, it is a structure of and for the interactions between people, between human beings with autonomous minds, who have a sense of self-awareness and intentionality. This implies that the people themselves and their consciousness are not a part of society as he sees it. A person’s self-awareness, their intentionality, their beliefs and attitudes, including their understanding of what the others say, lies outside of society. Society is no more than what takes place between people. To exclude the members of society from a structural analysis of society is, at least at first glance, a startling thing to do. What do we gain from saying that society is what is left of a (large) group of people after you have taken away their intentionalities? People’s utterances, the texts they contribute to the discourse, assign an interpretation to their verbal and non-verbal interactions. This negotiation of meaning is a collaborative activity. We can discuss what our own actions or texts mean, and what other people’s actions or texts mean. We can endorse, reject or modify what we are told. The result will be one interpretation agreed by all, or several competing interpretations. As long as we do not bother what people do, their behaviour, their actions remain meaningless. Only if Tom tells us he is looking for his contact lens can we understand why he is on his knees and making strange movements with his head. Whether Tom tells the truth or lies, whether we accept his assignment of meaning to his behaviour or not is something we cannot know. For the observer, there is only the symbolic, often verbal, behaviour to work on. For Niklas Luhmann, it is this symbolic behaviour that constitutes what he calls ‘communication’ and what I call the discourse. As there is no text which does not refer to an action or previous texts,
130
Meaning, Discourse and Society
‘communication’ comprises all texts of a discourse community in their diachronic order. Together, they generate what he calls ‘sense’, or what is normally called intentionality; in this case, not the intentionality of an individual mind but that of the discourse, or, as Luhmann calls it, ‘communication’. It is this collective intentionality, this ‘sense’ that constructs and defines society. This is why society presupposes language. Niklas Luhmann himself calls his proposal ‘a radically anti-humanistic, a radically anti-regionalistic and a radically constructionist concept of society’ (Luhmann 1998: 35). It is this collective intentionality of the discourse, not the intentionality of the individual mind, that observers can inspect. The important question, however, the question left open by Luhmann, is the extent of the discourse. Are the discourse observers part of the discourse community, and are their reports part of the discourse? Or does Luhmann see himself (and, similarly, all other observers) at a vantage point overlooking the discourse? The biologist who investigates the goings-on in colonies of ants, is placed outside. She does not communicate with them, nor does she expect the ants to realise that they are ants and that they are collaborating in keeping their ant colony going. She is not concerned with their minds. Luhmann’s observer of a discourse that goes on in a community is also not interested in people’s minds. For her, the discourse is data, data that can be shared, and on which certain investigative procedures (such as counting and measuring) can be carried out by following clearly defined methods yielding always the same results on the same data sets. Such a perspective is comparable to that of ant researchers or other biologists. But that is not all. Ant colonies are not societies. What their communication lacks is ‘sense’. Luhmann does not tell us if this ‘sense’, this collective intentionality, is only ascribed to the discourse by the observer (like in the case of the ants), or if it is something that distinguishes a discourse community from an ant colony. For Niklas Luhmann, the negotiations within the discourse generate ‘sense’, and this sense defines society. The people and their individual minds do not form a part of the discourse. This is not our mainstream, ‘humanist’ view that asks us to look ‘for explanations of the social world (either in terms of what individual people do and feel or in terms of groups, classes or societies)’, but a view that ‘looks into the linguistic space in which they move with other people’, as the social constructionist Vivien Burr calls it (Burr 1995: 40). Luhmann’s ‘society is a system that is closed in terms of communication. It generates communication through communication. Its dynamics consists of the effects communication [i.e. a contribution to the discourse] has upon communication [i.e. previous contributions]’ (Luhmann 1998: 95). Niklas Luhmann is, of course, so deeply immersed in literacy that he is not aware that his concept of ‘communication’ does not hold for oral communities in which what is said is taken to be the expression of the speaker’s intentions,
The language–society link
131
and is linked to a ‘reality’ shared by speaker and hearers. The spoken language of the primordial speech situation (covered in the next chapter) cannot easily be seen as dislocated from the people and the environment involved in the situation. For Luhmann, ‘communication’ is the discourse as the repository of all the texts that are part of it. For the observer, they must be accessible. They must exist in writing, or as a transcript, or at least in some other recorded form. This is also the case for Jacques Derrida, when he talks about the ‘absence of the transcendental signified’: in any new contribution to the discourse, a signifiant does not refer to some discourse-external signifiée but to previous occurrences of this signifiant. ‘From the moment that there is meaning there are nothing but signs’ (Derrida 1974: 50). It is this view bred on literacy that enables us to see the discourse as a closed system. Admittedly, this closedness is, for Luhmann, relative. The discourse is linked, he says, through ‘structural couplings’ (a term he has borrowed from Humberto Maturana) to its environment, that is to the individual, communicating minds. Luhmann’s discourse as a closed system is autopoietic: it reproduces itself, by ever new permutations and recombinations of already existing discourse matter. However, without innovation coming from the outside, it would soon run into loops, into dead motion. It would replicate itself in the form of identical clones. It would only generate noise, not ‘information’. Information, as Luhmann sees it, has to come from the outside. ‘The system refers to its environment through information’ (Luhmann 1998). Information is the difference between what is being said now and what has been said before. Only the fact that something said now is not an identical clone but a mutation, a recombination, a reformulation of something said before, turns it into information. A simple replication does not change the system. If I only repeat what has been said before, without varying it, I produce noise, but I do not generate ‘information’. Information is what has an impact on a system and distinguishes it from inefficacious noise which does not cause a relevant change. Therefore we should call only those new variants information which themselves also provoke new mutations. When new variants, recombinations and reformulations of what has been there before, are repeated in subsequent texts, again as slightly varied but recognisable copies, or as long as there are other detectable responses to them, they have an impact on the discourse. This can be viewed as a random, contingent process that does not presuppose individual minds as conscious, planning agents. From this perspective, the discourse is not only autopoietic; it is also self-referential. All new emerging content refers to previous content without just replicating it. We can view it as a continually evolving system of texts, text segments and lexical items, in which all new material is linked to previous material, which it repeats and varies at the same time. The discourse, as I have described it here, can be studied without reference to the outside world or to the minds of the members of the discourse community, in
132
Meaning, Discourse and Society
the same way as we could study the gradual mutations of the genome of a given species. If we want to investigate the change of meaning of a noun such as globalisation, our study of successive recontextualisations of this word in new discourse material can be compared to an investigation of the successive mutations that a given gene sequence undergoes from one generation to the next. Some of these changes will cause further mutations. Only these are relevant. The mutations that are subsequently only replicated, but do not engender further change, have no impact. But by processing discourse data in such a scientific, ‘objective’ way, we will not gain access to the ‘sense’ generated in the discourse. The findings need to be interpreted. In the case of the biologists observing ants, it is their discourse community of peers with whom they will discuss possible interpretations. The ants will not be asked. They are not in the habit of assigning interpretations to their behaviour. But if we want to interpret the data we have for a given discourse, we have to deal with the ‘sense’ that has been generated in this discourse through the contributions of the members of the discourse community. People talk about what has been said in the discourse. Indeed, a substantial part of any discourse consists of verbal interactions in which people collaborate in their endeavour to make sense of their own and other people’s contributions. Google lists 200,000 occurrences of ‘what I wanted to say is’. As I see it, it is this sense, it is the discourse-internal interpretations of the discourse itself, this element of selfreferentiality, that constitutes what Luhmann calls ‘sense’. To capture it, the observer has to leave the confines of a strict methodology, the essential feature of the natural sciences, behind. The findings are what the community of discourse observers have to agree on. But an interpretation of the discourse community’s own interpretation is always an intentional act, and its outcome is unpredictable and contingent. Any interpretation from the outside is subject to being reinterpreted by the discourse community itself. People, unlike ants, do talk back. Only if the observed and the observers merge their discourses does ‘sense’ become tangible. Through observing and interpreting itself, society creates itself as society; it becomes ‘society’s society’, ‘die Gesellschaft der Gesellschaft’, i.e. the title of Luhmann’s magnum opus. However, scientific method as such only assembles the necessary discourse evidence; it will not acquaint us with its ‘sense’, with the collaborative intentionality of a discourse community. This is a problem. Blumer’s symbolic interactionism Only people organise themselves as a society. In order to do so they have to rely on language. For only language makes it possible to conceptualise behaviour as a social practice. Baptisms, weddings, birthday celebrations, excursions, all collaboration, the division of labour, anti-war rallies, religious services; these are social constructs, and therefore they need a verbal expression to be
The language–society link
133
understood. Without a word for work there would not be the social construct ‘work’, nor would there be, without the pertinent words, constructs such as property or justice or even society (cf. Teubert 2007). For only human beings can be aware (and reflect the fact) that they are members of society. They can only be aware of it because society is something people talk about. This is, of course, anything but a new insight. That it is meaning which constitutes society is the essence of Herbert Blumer’s theory of symbolic interactionism. Blumer is a younger colleague of George Herbert Mead, whose aim was to fuse the pragmatism of Charles Peirce, William James and John Dewey with a behaviourism concerned with social interaction. In Blumer’s view, [t]he first premise is that human beings act towards things on the basis of the meanings that the things have for them … The second premise is that the meaning of such things is derived from, or arises out of, the social interaction that one has with one’s fellows. The third premise is that those meanings are handled in and modified through, an interactive process used by the person in dealing with the things he encounters … Thus, symbolic interactionism sees meanings as social products, as creations that are formed in and through the defining activities of people as they interact. (Blumer 1969: 2/5)
For Blumer, the salient behavioural feature that distinguishes humans from animals is the act of interpretation, of making sense of our environment and the people we encounter there. When someone reacts to a kick by kicking back without deliberation or intention, the event of kicking back is not symbolic: ‘Non-symbolic interaction takes place when one responds directly to the action of another without interpreting the action’ (p. 8). Where I disagree with Blumer, though, is that for him interpretation is not a collaborative action but the work of an individual mind, even if this mind engages in an internal dialogue with itself. Thus Blumer, like so many scholars before him, reifies the individual mind at the expense of the discourse as the collective mind: ‘The human being is seen as “social” … in the sense of an organism that engages in social interaction with itself by making indications to itself and responding to such indications’ (p. 14, my emphasis). Of course I do not deny that we tend to argue with ourselves. But the arguments we use are arguments that we are borrowing from and have encountered in real dialogic situations. Unless we have been involved in a collaborative discourse, there would be nothing we could discuss with ourselves. It is true that Mead had voiced the same idea as Blumer: ‘We have seen that the process or activity of thinking is a conversation of the individual between himself and the generalized’ (Mead 1934: 224n). But he moved on from that position. George Cronk describes Mead’s stance in the Dictionary of Literary Biography: In self-consciousness the individual appears as an object within his or her own experience but does so only on the basis of social relations and interactions, only by means of his experiential transactions with other individuals in an organized social environment. (Cronk 2003: 224)
134
Meaning, Discourse and Society
Still it remains Blumer’s achievement to have highlighted the distinction between responding to and interpreting a stimulus. Ants communicate by responding to pheromones. But they are not aware that they are communicating. They form colonies, but not societies. Human society is a construct of a selfreferential, interpretive discourse. We become members of society by learning to respond to utterances, to speak ourselves, and to reflect on what others say and on what we ought to say in a discourse. This is a skill we cannot learn on our own. Feral children who grow up in isolation, outside of society, do not learn to speak. They will learn to respond to environmental stimuli. But symbolic interaction, the exchange of symbolic content, goes beyond stimuli and responses. Symbolic content offers the possibility of reflection. Ants react to their environment; they do not ponder what to do next. It is through the interpretation of symbolic content that we learn to exercise the art of reflection. While it is language that enables us to move on from responding to a stimulus to interpreting it, this does not mean that people must necessarily be aware of the fact that they are using language in this way, or that they must be conscious that they are carrying out an act of interpretation. Reflection, as I will show below, is not yet fully part of the primordial speech situation. Reflection in the sense that we are aware that a text constitutes symbolic content, that we consciously inquire about the meaning of a linguistic utterance, seems to exist much more in literate societies. Without language, there is no interpretation. My second objection to Blumer’s approach is therefore his total disregard for the working of the discourse. Blumer fails to address the indispensable role of language, of la parole, of the discourse, without which there would be no symbolic interaction. For him, interactionism is an empirical science, and his goal is the ‘examination of the empirical social world’ (Blumer 1969: 34). This goal is at odds with his own key axiom that it is meaning alone that creates a shared reality. Only the discourse turns the stuff or matter of the world out there into discrete, identifiable objects. We first need to construct our reality by collaborating in the discourse, before this reality can become the object of symbolic interaction. Whenever we interpret behaviour as meaningful, or when we discuss the meaning of a text, we do not do it for ourselves but with the goal of finding others to agree with us. Whenever we engage in an act of interpretation, we take into account what we expect the other members of our discourse community to say and try to pre-empt their objections. This was already seen by Valentin Volosinov and Mikhail Bakhtin in their book Marxism and the Philosophy of Language: ‘[T]he psyche must be derived from ideology. Psychology must be grounded in ideological science. Speech had first to come into being and develop in the process of the social intercourse of organisms so that afterward it could enter within the organism and become inner speech’ (Volosinov 1986: 39) and ‘Any utterance … makes response to something and is calculated to be responded to in turn’ (p. 72). Our aim is to have other people accept our interpretation.
The language–society link
135
We may read a poem for our own pleasure. The act of reading it will be accompanied by a first-person experience. It will cause what we will describe, if asked, as feelings, or as a mood, or it will give rise to ideas, or let us remember things past. But such a first-person experience is not an interpretation. It only becomes an interpretation if we give it a symbolic form, if we turn it into a text, if we give testimony of our first-person experience. The first-person experience can be seen as something taking place outside of the discourse, inside my individual mind. By turning it into an interpretation, we enter it into the discourse. At that very moment, it ceases to be discourse-external, and instead becomes symbolic content that can be negotiated. The link, however, between the ‘raw’ experience and the testimony of it is tenuous and not accessible to scrutiny. Interpretation is not something a monadic mind does for itself; it takes place within a discourse. Normally I would not interpret a text for myself. If I believe no one I know would be interested in what I think about a particular poem, I would not bother to analyse it and to give account of what I think it means. Even if I do, my interpretation will remain inconsequential until I let other people know. As long as a first-person experience is not converted into language, it is not symbolic. It does not have a meaning. It is not part of the discourse. First-person experiences remain mute and meaningless unless they are turned into texts. When someone refuses to give us testimony of their first-person experiences, they remain mute and will have no impact on us or on society. Society will not be affected by them. Interpretations only count if they are made public and can become an object of negotiation. However, there is no need for a discourse community to agree on any interpretation. Even if (or perhaps: especially when) an interpretation causes confrontation, it will have an impact. Interpretations are always provisional. It is this limitless procession from interpretation to interpretation that generates ‘sense’ in Luhmann’s sense. Once interpretations only repeat each other, all innovation ceases. In order to generate ‘sense’, the discourse has to remain plurivocal, encouraging new things to be said. Only the divergence of interpretations engenders new interpretations. Interpretations need to be public. From the individual mind to the discourse as the collective mind In Luhmann’s model of the discourse (what he calls ‘communication’), there is no room, and no need, for individual intentionality. There is ‘sense’, but this sense is what defines society, and it does not matter if we find it in the minds of the people or not. Luhmann’s ‘sense’ is society’s awareness of itself. It is the never-ending effort to interpret and reinterpret what is already there, new signifiants referring to previous signifiants but never referring to discourseexternal signifiées, no matter how far back we go. If we look at the discourse in this way, we can view it as a distributed collective mind, and collaborative interpretation as the creation of ‘sense’.
136
Meaning, Discourse and Society
However, we seem to be innately primed to regard intentionality, this awareness of aboutness, as something very personal and inherently individual. Intentionality, we contend, presupposes consciousness, and the only way we have learned to think about consciousness is to regard it as a unique first-person experience. The idea that we can view the evolution of a discourse as the emergence of a collective intentionality seems at first sight quite strange. It feels counterintuitive to assume that a society could have a collective agency that is different from the aggregate of individual agencies . But as Colin McGinn has said, neither the cognitive sciences nor the philosophy of mind can explain the nature of consciousness (McGinn 1991). Therefore we have to accept that we gain nothing from drawing on the concept of individual consciousness when we discuss intentionality. The texts comprising a discourse are certainly about something. Outside of primordial speech situations (see below, Chapter 10), they are not a mirror of a shared reality outside the discourse. Rather, they refer to things that have been previously said. ‘Globalisation’ is not something that I can show my companion by pointing to an object we both see in our environment. When I mention globalisation in an utterance, my companion will see it as an uptake of former occurrences of that word and of the contexts in which they occurred. The last three hundred years of Western intellectual history make us dread the idea that we should trade in our cherished first-person experiences for a discourse representing a collective mind. Surely it would mean that being aware of ourselves is just an illusion? But this is not the case. Our individuality only exists in exercising our otherness inside a discourse community. What makes up our individuality is the result of accepting, modifying and rejecting symbolic input from the outside. The way in which reading a poem goes along with an individual first-person experience is primed in confrontation with others. We are a lot less original in our thoughts than we would like to have it. Our individuality is forced upon us in our very early years of language acquisition. It is the symbolic interaction of our carers and our peers with us that generates it. Feral children are not aware of their personhood. They relate to their environment without the mediation of the symbolic, without language. They are at the mercy of it. The only way to control the environment is to take it into the discourse. Once we abandon the quest for the intentionality of the individual mind and content ourselves with the aboutness of the collective mind, i.e. the discourse, it immediately makes sense to no longer insist that what is said must refer to a discourse-external reality. The successful permutations of the discourse are sufficient stuff to work on. They are neither the predictable results of ruledriven mechanisms, nor caused by the evolution of our environment. They just happen. One way to make sense of the discourse is to look at it as something emerging from the collaboration of individual minds. These minds in turn are continually re-creating themselves through the symbolic input they receive.
The language–society link
137
For me, all the other individual minds are only accessible as contributors to the discourse, just as I am for them. Only the discourse tells us about other people. The other way to look at the discourse is to look at it as an emergent phenomenon, resulting from an aleatory, autopoietic and self-referential mechanism. From this perspective we can safely disregard the intentionalities of the voices represented in it. Each new text is then no more than a recombination of elements that were already there, a reformulation of a previous formulation, a permutation of an earlier order. It creates ‘sense’ if it leaves an impact on the subsequent discourse. Otherwise it is no more than ‘noise’. From such a perspective, all new texts are as contingent and unpredictable as the next poem of the national poet laureate. Of course, the discourse mechanism has to exhibit a certain level of complexity to avoid running into straightforward replication, into interminable loops of repetition of identical text sequences. Take, as an example, Joseph Weizenbaum’s Eliza, a computer program impersonating a psychotherapist, which was apparently engaging enough for Robin Dempsey, a character in David Lodge’s novel Small World, to keep him addicted (p. 243): ‘Hello’, Robin typed on the keyboard. ‘Hello, how are you feeling’, replied Eliza, the printed words appearing instantly on the VDU. ‘I’m feeling terrible’, Robin replied. ‘What do you mean by that precisely?’ ‘I’m feeling depressed’ ‘Do you have any friends?’ ‘Not really.’ ‘Why is that?’ Robin Dempsey paused for some considerable time before he typed in his reply. ‘I don’t know.’ ‘Do you really not know or are you just avoiding the question?’
For Robin, the fact that Eliza, as he indeed is aware, is only a computer program, does not keep him from becoming hooked on these exchanges. The illusion that Eliza’s questions help him to interpret the situation in which he finds himself is strong enough to keep him going. Just as Eliza is nothing but a computer program, it is easy enough to imagine a complementary program impersonating a therapist’s client, in this case Robin. Given sufficient complexity, both programs together can generate a discourse that will run for a long time before beginning to replicate itself, i.e. running in a loop, even though the programs form a closed system, not coupled to any computer-external environment, such as individual minds. Once the programs are good enough to imitate authentic language, it will be more or less impossible to distinguish their discourse from one produced by ‘real’ individual minds. This discourse will certainly be about something, and it does not matter that the computers involved
138
Meaning, Discourse and Society
lack intentionality. My point is that we can analyse any discourse, not just a computer-generated discourse, as such an autopoietic, entirely self-referential system. ‘Communications [i.e. contributions to the discourse] generate, if reproduced autopoietically through recursive repetitions, an emergent reality sui generis’, says Luhmann. ‘There is no communication from one mind to another which is not socially [i.e. discursively] mediated, and there is no communication from individual to society … Only society [i.e. the discourse] can communicate’ (Luhmann 1998: 105). The fact that a discourse is machine generated does not make it less interpretable. The analysis of Robin’s dialogue will tell us that there is a lexical item feeling terrible, and that this lexical item can be paraphrased by feeling depressed. It is impossible to distinguish such an aleatory mechanism from what we may call intelligent design. It also demonstrates that for making sense of the discourse there is no need to presuppose individual intentional minds. To analyse the discourse as a collective mind does not force us to look at the people taking part in it. As we shall see further below, this view of the discourse is closely related to Hans-Georg Gadamer’s and even more to Paul Ricoeur’s views and their postSchleiermacherian conception of hermeneutics. Meaning is an emergent phenomenon of the discourse, and it can be investigated without resorting to the minds of the people who have contributed to it. This is, however, only one of the two perspectives we have to assume whenever we want to make sense of what has been said. Our joint effort to collaboratively make sense of symbolic content is not what could be described as an autopoietic process, a process taking place within an uninhabited discourse space. It is not a process but an activity presupposing conscious agents. If we postulate a collective mind, we must also accept distributed minds, unfathomable as they may be. Interpretation is an act. People engaging in this act are exercising their minds. Each contributor has their say. The result of our collaboration will reflect our individual contributions. As long as we try to observe them in statu nascendi, before they have been contributed, we are bound to fail. As long as they are not verbalised, they are not symbolic and do not have a meaning. Interpretation that has an impact is an interpretation entered into the discourse. It is not a monadic but a dialogic activity. Mihail Bakhtin has made us aware that a discourse is never monologic. Only the ensemble of different voices, discourse as heteroglossia, generates sense. The step we have to take is to substitute the collective mind for the individual mind. Sense making involves negotiation; it is, by definition, a collaborative activity. François Cooren has shown in his analysis of boardroom discourse that the outcomes of such negotiations are never entirely predictable, and that they cannot be reduced to the mechanisms of individual minds (Cooren 2004). The result is more than just the addition of the individual mental states; it is something that was not there when the meeting began. However, the study of what is now termed ‘collaborative cognition’ is
The language–society link
139
nothing new. Many decades ago, Jean Piaget and Lev Vygotsky looked at the way children acquire knowledge, and they emphasised the role of symbolic interaction. I will deal with the collaborative construction of reality in more detail below in Chapter 11. How does Luhmann’s communication system compare to the main linguistic paradigm of the twentieth century, language reduced to a language system? The key difference, I believe, is that for Luhmann it makes no sense to try to describe a stable, rule-based mechanism on the basis of which we could say what can be said tomorrow and what not. His model, as I see it, is that of an emergent, self-organising and contingent system in which there are no rules that would have permanency. Saussure’s system of la langue does not allow for change, and Chomsky’s universal grammar is engraved in the language organ. Both systems are defined from the outside. Luhmann’s system, on the other hand, is autopoietic, continually re-creating itself in unpredictable ways and feeding on the texts it generates. While for Saussure and Chomsky there would still be a language system even if no one used it, for Luhmann only the discourse as the incremental entirety of texts constitutes the system. His system is co-extensive with la parole, with the discourse as data, and not reducible to a list of rules and a list of elements.
10
A closer look at oral societies
It is the primordial speech situation which we have to investigate if we want to find out what distinguishes spoken from written language. For me, the typical primordial speech situation is an informal gathering in which participants act, each participant contributing their individual intentions to an emergent collective interaction. This interaction is, to a greater or lesser extent, symbolic, in the sense that it includes verbal communication concerning the behaviour of participants, communication which negotiates, explains and reflects what otherwise would be just behaviour. Primordial speech situations take place in settings shared by participants and encompass the objects which make up this setting. They let their participants share a reality available to all of them. It does not matter so much how ‘real’ this reality is in a more philosophical sense. Certainly this ‘shared reality’ is already constructed and endorsed through endless negotiations that have taken place in an endless row of past speech situations. In times immemorial, there may have occurred a situation in which the participants found it fit to ‘carve out’ of the stuff that constructs the ‘reality’ an assemblage they agreed on calling a chair. In subsequent situations, they used this label chair for similar assemblages serving the same function. In a current situation the question would hardly arise whether some concrete object could be called a chair; but if it does, for instance with an ambiguous postmodern piece of furniture, negotiations will be collaboratively conducted leading to some degree of agreement among participants as to whether the object should be given that name. In primordial speech situations, the participants interpret collaboratively their object-oriented behaviour (which includes the objects affected by the behaviour). They link an object of a ‘shared reality’ to a lexical item which obtains its reference through this link, and, equally importantly, its meaning through the paraphrases and definitions by those participating in the situation. Henceforth it is possible to talk about chairs even if there are no chairs within the shared reality of the situation. In the absence of the object in the situation, it is only meaning which matters. When chairs are discussed, references can be made to everything that has been said about chairs in previous speech situations. This is the moment when the object of our ‘shared reality’ becomes an object of the discourse. This is how meaning can replace reference. 140
A closer look at oral societies
141
While it can happen in speech, it is an inevitable characteristic of written language. For good reasons, we can distinguish between the ontogenetic and phylogenetic aspects of meaning negotiation. A child will accept whatever her carer tells her is a chair, regardless of whether the carer uses deictic acts (pointing at a chair in their environment) or paraphrases it (for instance as something to sit on). The carer herself has learned the meaning of chair in much the same way, and she will have come across many chair-pointings and chair-talks during her life. At every new ontogenetic instantiation of the word chair, the phylogenetic history, the chair-discourse in its diachronic dimension (in the form of both ostentation and paraphrase), will inform the ontogenetic act. Once introduced to the word chair, the child will be exposed to innumerable instances of chair-pointing and chair-talk. Very soon chair-pointing will lose its semantic function and only serve a demonstrative purpose. Thus each typical primordial speech situation profits to a large extent from the shared memories of previous primordial speech situations. If there is conflicting testimony concerning such memories among the participants, the meaning of those lexical items involved will have to be renegotiated. Indeed, almost all persons, objects, settings, actions, states and processes (for all of which I will use ‘objects’ as an umbrella term) mentioned in an informal conversation will have been introduced, by deictic acts and/or by paraphrases, in previous conversations in which the same or different people have participated. The phylogenetic dimension of meaning assignment will normally guarantee sufficient agreement. Only if discourse objects unfamiliar to some or all participants are introduced does their meaning have to be collaboratively interpreted, by ostentation or by paraphrases. Such interpretations are always provisional. Nothing will keep people from renegotiating the meaning of any object discussed previously in a primordial speech situation. For instance, there may have been a time when all artefacts made to sit on were called chairs. Unhappy with this state, someone might successfully suggest in a speech situation that chairs should be distinguished from stools, and some weeks later introduce benches as objects, and so on. The reference to objects in a shared reality would normally precede negotiation of their meaning. Once done, however, we can talk about chairs, stools and benches as discourse objects in situations where we find none of them. What I have called here deictic acts, ostentation or pointing to objects, should not be understood as reference in the same way as it is discussed in analytic philosophy as the relationship obtaining between a lexical item and an object in the discourse-external reality, i.e. Gottlob Frege’s Bedeutung or reference in the sense of the early Hilary Putnam. For me, the environment of a primordial speech situation, the people engaging in the discourse, and the stuff within the sensory perception of each of them, is to be seen as an extension of discourse. As such,
142
Meaning, Discourse and Society
these people and the stuff are discourse objects by virtue of earlier speech situations. They are the elements of a reality that has been constructed collaboratively in these situations. As I see it, a deictic act creates a discourse object; it does not establish a link between a language expression and an object of a discourse-external reality. We are primed to believe that language is a mirror of the reality out there, because in archetypal speech situations (those in which an object is discussed for the first time) we would expect that many of the objects discussed are ‘real’ objects actually present in that situation. For me, any reality outside of the discourse is without relevance. I like to look at it as nothing more than unstructured stuff. Deictic acts construct, out of this stuff, objects within a discursively constructed, shared reality consisting of ‘discourse objects’. Before anyone has introduced us to the object ‘chair’, it is no more than stuff. We may even sit on it, as we often sit on many other things, without conceptualising what we happen to be sitting on. But by labelling it chair and paraphrasing the meaning of this lexical item, we turn stuff into a discourse object. Similarly, other objects, including abstract objects such as ‘marriage’ or properties, states, processes and actions, sometimes even concrete virtual objects such as ‘dragons’, are also constructed. They populate our shared reality in the same way as concrete objects that can be seen, heard or smelled. Together, these objects, abstract or concrete, natural or artefactual, properties, states, processes and actions, make up this entirely discourse-internal reality. There are, in a primordial speech situation, also objects which belong to the lifeworld of only one, or some of the participants. As long as they are not contested by other participants, they can become an implicit part of their shared reality. If Charles invites Tom: ‘Why don’t you come home with me and have dinner with us; then my wife Mary will have a chance to get to know you’, Mary becomes part of their shared reality. We normally interact on the basis that these secondary objects are as ‘real’ as the objects present in the shared reality. The objects of our shared reality become discourse objects, because the shared reality of a primordial speech situation is an extension of the discourse. However, not all talk in primordial speech situations is about the construction of concrete objects, present or absent, as discourse objects. Much of what is said has to do with the construction of interpersonal relationships with other participants of primordial speech situations. For each participant, all other participants also acquire their status as individuals, their identity, only through the discourse. Constructed as discourse objects, they are extensions of the discourse, and we deal with them as such. The ‘real’ people outside of the discourse do not matter. The objects of our shared reality, abstract or concrete, natural or artefactual, properties, states, processes, actions and people, are, as discourse objects and as lexical labels, symbolic. We take the Mary we encounter as a participant in a
A closer look at oral societies
143
primordial speech situation to be real. But to the extent that we take this situation to be an extension of the discourse, she is also a discourse object. Whatever she says, whatever is said to her, and whatever is said about her is no longer the discourse-external Mary, but Mary the discourse object. As such, she can be dealt with. The real Mary is as inaccessible as the people we find on photos where everyone has forgotten who they were. Mary the discourse object, Mary the lexical item Mary, is symbolic. She is a sign standing for something else. She may stand for the Mary we think we know, the ‘real’ Mary. Or she may stand for the Mary we encountered in former primordial speech situations. Or she may stand for the Mary we have encountered in Tom’s stories about his wife. All discourse objects are symbolic. But people engaged in primordial speech situations normally do not see their utterances as symbolic. They do not normally ask themselves what the discourse objects mentioned stand for, what the lexical items corresponding to the discourse objects mean. For them, an utterance is the expression of the speaker’s intentions, not a text with a meaning. I will come back to this important aspect later. In primordial speech situations, people believe that what is being said are the speaker’s intentions. This is what hearers are interested in. They do not want to know what an utterance means. This difference corresponds to the dichotomy between ‘concepts’ and ‘signs’ that Claude Lévi-Strauss discusses in his book The Savage Mind (1966: 18–20, et passim). Concepts, he says, ‘aim to be wholly transparent with respect to reality’, while ‘[s]igns, in Peirce’s vigorous phrase, “address somebody”’ (p. 20). Concepts do not distinguish between objects of a shared reality and ‘real’ objects, objects pre-existing in the discourse-external reality. Concepts are perceived not as results of a negotiation between the members of a discourse community, but as self-evident, due to their presumed transparency. The monadic mind can form concepts. But it cannot establish signs. What a sign stands for has to be agreed upon by speaker and hearer. Signs presuppose a discourse community. To become aware of the fact that what we took to be conceptual images of the reality out there are indeed no more than discourse-internal arrangements, is what constitutes for Lévi-Strauss the irrevocable entry into culture, and the farewell to the lost ground of mythical reality. What Lévi-Strauss calls concepts are perhaps not so different from symptoms, as discussed above, in opposition to symbols. In order to learn that smoke can be a symptom of fire, I do not have to engage in the discourse. It is something I can find out for myself. In this sense, the world of oral societies is full of symptoms or concepts. Their shared realities are populated with meaningful objects: trees, landmarks, rainfalls, storms, the waxing and waning of the moon, ebb and flow, the seasons, the advent of a locust swarm and much, much more. These symptoms become signs inasmuch as they and their implications are surely discussed among themselves. But in an oral society people are hardly aware that
144
Meaning, Discourse and Society
they are dealing with meaning. It is much more like finding out the intentions of the symbolic object in question. They assure themselves that when the shadow of a given tree touches the boulder nearby for the first time in the year, they have to start sowing. But these objects and events remain symptoms insofar as their implication is thought to be a non-negotiable given by the deity. For people in an oral society, people to whom it does not matter what something said means, only what it tells us to do, there is little difference between the intentions of a speaker, the implications of a sacred event, and the meaning that is agreed by consent. Only literate societies distinguish symptoms from symbols. The meaning of a symptom is not arbitrary. A symbol presupposes that the person who ‘signs’ and the addressees of the ‘signing’ have entered an accord concerning its meaning. It seems to be something we only find among human beings. Perhaps it is this aspect of language that Noam Chomsky (together with Hauser and Fitch, see above, Chapter 8) deems to be uniquely human and not explainable as the result of evolutionary adaptation. At the point of transition from symptom to symbol we find the exegetic work of the shamans. Trees with their shadows, which had been concepts or symptoms in the sense that they evoked memories of past events, perhaps even by causing synaptic connections to become live and provoking largely non-symbolic behaviour, turned gradually into signs or symbols in the sense that they stood for something else which had to be explained by the experts. It is significant that the priest caste was the first to exploit the arrival of writing. With the arrival of literacy, speakers’ intentions and the meaning of a text become, as we shall discuss below in more detail, two separate things. Written texts are, as Paul Ricoeur calls it, ‘dislocated’ from their authors. We, as their readers, do not even know if there is an author, or, as in the case of Eliza, a computer program. We do not know whether the book Marxism and the Philosophy of Language was written, as it says on the cover, by V. N. Volosinov, or, as many experts assume, by Mihail Bakhtin. In a literate world, it is the (author-independent) meaning of a text that counts. In the case of an utterance in a primordial speech situation, we want to know what the speaker wants. Speaking and hearing are simultaneous, embedded in a shared setting; they are, as Roy Harris says, ‘cotemporal’ (Harris 1990: 48). The discourse-external reality is within sight of what is said. Constructing discourse objects by paraphrasing lexical items gives us our shared reality. This also takes place in the case of speech acts other than assertions. When I make a suggestion, or a plea, or a promise, when I complain or ask a question, my audience will understand my utterance because it refers to a shared reality. Whatever is unclear will have to be discussed. However, not all symbolic interaction taking place in primordial speech situations deals with the shared reality of the setting in which it takes place, or of previous situations. Normally a substantial part of the communication can be
A closer look at oral societies
145
classified as narration. Often, when we are not carrying out one of the marked speech acts, we tell stories. If I tell my friends that during the thunderstorm half an hour ago a ball of lightning slowly entered the room in which I was sitting, came in through the window, then slowly circled the lamp and finally left again through the same window, then this is something that they cannot ascertain for themselves. They probably have never encountered anything like it. There is nothing they can do to establish whether or not I spoke the truth. For them, it will be just a story, perhaps fascinating, but as irrelevant as the story of Troy was to Homer’s audience. Narratives allow us to construct events outside of our situational embedding, to communicate content whose truth cannot be experienced by the hearer. The people who may have listened to recitations of the Iliad in the centuries before it was written down were not the least interested in the historical truth of the events. For the audiences of recitations of the Iliad, or of the golden legends, truth was not the issue. All the medieval chronicles, too, mixed historical events liberally with fiction. The quality of the story, not its truth, was what mattered. Narratives deal with a world outside of the shared reality. How do they fit into primordial speech situations in which utterances are, as I said, linked to a shared reality? What makes them acceptable? My suggestion is that we should look at the evolutionary origin of speech to find an answer to this conundrum. Ever since Etienne Bonnot de Condillac, Jean Jacques Rousseau and Johann Gottfried Herder, the idea has been aired that the primordial form of language was not speech but gesture. It seems that George Herbert Mead came to the same conclusion independently. Today, with the confluence of sign language studies and primate research, there is no shortage of publications positing the evolutionary origin of our language faculty in gesturing. Arguments for this theory are found, for instance, in the article ‘Signs of the origin of syntax’ by Armstrong, Stokoe and Wilcox (1994). This study, which is complemented by a number of more or less critical commentaries by members of their peer group, concludes: The scenario we propose involves an evolutionary stage in which visible gestures, for all the reasons proposed above, take the lead with respect to flexibility of output and, critically, the elaboration of syntax. (p. 358)
Neurolinguistics, as much or as little as its findings may be relied on, has come to the support of this thesis. An example is the article ‘Mirror neurons and the evolution of embodied language’ by Fogassi and Ferrari (2007). There we read: A suggestive hypothesis would be that the ventral premotor cortex, endowed with the control of both hand and mouth actions, could have played a pivotal role in associating gestures with vocalizations, thus producing new motor representations. At this stage, the mirror-neuron system, because of its capacity to match the seen/heard gesture or
146
Meaning, Discourse and Society
vocalization with internal motor representations, allowed the observer/listener to assign a meaning to these new vocal–gesture combinations. (p. 140)
If indeed signing, gesturing, was the original form of communication, we could have here an even more appealing explanation for our acceptance of an unbreakable link between language and reality. What we can convey by gestures, what we can point to, what kinds of mental states can be symbolised by innate or acculturated facial expressions and by ‘body language’, will only be successful as long as we are part of a situation in which these gestures make sense, where there is a reality we can share. In the remote valley in Papua New Guinea, our anthropologist can still interact to a certain degree, despite not speaking the language, by gesturing. But she will not be able to narrate stories lying outside of the situation in which she is embedded. Nobody knows to what extent the gesturing of our ancestors included conventionalised signs. It seems impossible that the intention to hunt a boar could be expressed by gestures. But it seems doubtful that it actually comprised a range of signs sufficient to tell a story. Gestures, however, are not the only kind of non-vocal behaviour relevant for establishing interpersonal relationships. Grooming also links people. Grooming, however, is, according to Robin Dunbar (2004), not so different from gossip, and, gossip, he knows, makes up two-thirds of our conversations. Gossip is so prevalent among humans, he claims, because it is equivalent to the grooming that we find in the other primates. Grooming defines the relationship between the members of a troop. But when a troop has too many members to be groomed, it becomes too time-consuming and therefore inefficient. I can groom only one partner at a time; while with a single narrative I can reach a whole group of people. The transition from bodily to acoustic communication brought with it the chance to substitute the human voice for tactile contact. Whoever was better at gossiping had a bigger likelihood of passing on their genes to subsequent generations. For grooming and gossiping, it is less the content than the contact that counts. Yet content there must be. The more a story evokes a shared reality, the more plausible it is and the more it will please its audience. This does not mean that telling stories of gossip is more unidirectional than interactive. Grooming relies, to a certain extent, on reciprocity. If a group member gossips, this is an incitement to others to join in, to embellish or to complement the story, or to reciprocate with their own stories. Gossiping is a way to establish and reconfirm interpersonal relations. Perhaps, then, it is necessary to distinguish non-gossiping storytelling from gossip. It does not much matter if these stories are thought to be invented or true, or a blend of both; they have to be deceptively ‘real’ to convince us. It makes sense to assume that storytelling has always been a part of primordial speech situations. Once shared reality and storyline become indistinguishable we are
A closer look at oral societies
147
won over. There were only a few people who doubted the veracity of the weapons of mass destruction ploy that apparently vindicated the invasion of Iraq. A story so carefully embedded into our shared reality, told and retold evening after evening on television by reporters as familiar to us as our best friends, did not need additional verification. Veracity was not an issue. While in our modern, literate world we tend to take truth for granted once we read a story in print, primordial speech situations, whether they negotiate a shared reality or create comfort through gossip, do not normally touch on the issue of truth. We owe our modern insistence on truth not to oral communication but to the invention of writing. Storytelling means entertaining the group. Unlike grooming, it does not require a response. There is a third type of narration, the kind of story that generates and enhances group identity. As far as oral societies are concerned, myths are not distinguished from historical narratives. The participants of speech situations in which these stories are told do not necessarily subscribe to their veracity even if they may ‘believe’ in them, as one believes in religious content. Unlike interactional gossiping, these narratives derive their efficacy from repetition. They are a core part of the rituals that generate a sense of identity among the participants of a speech situation. Their key function is to draw a line between ‘us’ and ‘them’. In many oral societies, there are specially trained people, professionals, in charge of reciting these memorised narratives. Any kind of collaborative negotiation of their content is discouraged. In this way, they also help to sustain an established social hierarchy. As is the case with newspapers in a literate society, a strict divide is maintained between those issuing content and those consuming it. This is why it is mostly in these stories that we find elements without a basis in past shared realities. Dragons, deluges, deities, spirits and miracles of all kinds make their impact on beliefs and attitudes, not through plausibility but through repetition. The ‘shared reality’ is a relative concept. Even in the same setting people will never have the same view of the objects within everybody’s visible range. Because of their individual pasts, in particular their participation in past speech situations which are never co-extensive with those of other people, (or, once literacy has arrived, because of their individual reading experiences), people will be exposed to different segments of the discourse, and when talking about a discourse object, presuppose different paraphrases. But it is not the differences in these people’s first-person experiences that I am concerned with. We can only inspect the different ways people talk about things. Once these differences are perceived as relevant, they will be discussed. If everyone agrees to call the new seating equipment in a bar a stool, and not a chair, no negotiation will ensue. But some old-fashioned people might refuse to call them that, as stools are not supposed to have backrests. A healthy discourse is, almost by definition, plurivocal. A new level of reflection will only be attained if utterances clash
148
Meaning, Discourse and Society
with each other. This will lead to a negotiation whose result is something new, something that has never been said before, something that is unpredictable. It does not really matter if a compromise arises, or if there are still dissenting voices. In any case, the result is only provisional, and negotiations can be resumed any time and lead to different results. Yet what is important is that without negotiation there is no consciousness of a ‘shared reality’. It is the collaborative act of interpretation that allows us to reflect on our shared reality. Interpretation can break up the illusion that the discourse deals with reality, that what is said refers to the world outside. There was a time when communication consisted of nothing but primordial speech situations. They still play a prominent part in our literate societies today. Perhaps they are, though, less primordial than they used to be. It might be wise to assume a cline from more towards less primordial speech situations, that is from those situations we would expect to find in stone-age communities, to those that we experience day to day around the family table, in pubs or other pastime congregations, and, even further on, to those situations in which more abstract knowledge is discussed, such as boardroom meetings, seminars on gender issues, or panel discussions on big bang theories. It is the role of ‘shared reality’ that determines the position of a speech situation. Once we discount narration on the grounds that it is a kind oral grooming, then we find, at the one extreme, situations in which most of what is said refers to a shared reality, that is to the objects of the setting in which the situation takes place. On the other end we would find speech situations containing an increasing number of features typical of literacy. A university seminar session discussing a poem would hardly focus on potential links between words, phrases and other units of which the poem in question is made up, and their links to a shared reality. Rather, it would draw parallels between the poem and other poems, and it would look for interpretations of this poem, or other comparable poems, as found in the literature. The setting of the seminar situation and even the interpersonal relations between individuals taking part in this speech situation are much less relevant than the negotiation of the content itself. What has been achieved at the end of the session is an interpretation of discourse-internal content, not a link with the world outside. The essential characteristic of primordial speech situations is therefore the role of the shared reality. It gives us the illusion of prima facie knowledge. What we encounter in a shared reality will always be more than just testimony. In our literate world, we find ourselves in many speech situations dealing with stories whose ‘truth’ we can only evaluate on the basis of the meaning of the utterance taken as a text, not on the basis of a shared reality. When we, in the build-up towards the Iraq war, talk about weapons of mass destruction in a rather informal setting, this topic is neither narration in the sense of gossip (where it is the quality, not the truth of the story that counts) nor does it deal with
A closer look at oral societies
149
objects present as a shared reality. The Iraqi weapons of mass destruction are, at least for us normal news consumers, only discourse objects, and what we know about them, or the meaning of the corresponding lexical expression, never goes beyond what is said about them. The American Secretary of State Colin Powell, aware of the somewhat dubious state of discourse objects, wanted to elevate them, in his security council presentation, to objects one can point to in a shared reality. But eventually readers learnt there was nothing to point to. Yet most of the knowledge we accept as reliable in our literate societies tells us normal people about discourse objects and is far removed from a shared reality that exists for perhaps no more than a handful of experts, or it is a kind of entirely discourse-internal knowledge without even a figment of a shared reality, like our talk about globalisation. It is this kind of knowledge that distinguishes literate from oral societies. In many professions, we deal most of the time with knowledge existing only in text form. We have no way of tracing these texts back to a situation in which they may have been utterances accompanying and negotiating a reality shared by all present. Only primordial speech situations can provide a link between the shared reality and what we say. All other speech situations are, as much as they deal with ‘facts’, based on nothing but testimony. Once literacy has arrived, we encounter an ambiguous relationship with this new medium, ranging from scepticism if we see a text as a piece of narrative gossip, to the conviction that a text, because it exists in writing, must be true. It is the unavailability of an author that can make us doubt a text’s veracity. We, the readers, cannot test their reliability. But a text without an author can also be read as a revelation of arcane knowledge, as the incarnation of the non-symbolic world in the discourse. While scepticism about reliability can give rise to our critical and reflective powers, our unquestioning trust in the written word can foster the illusion of an unbreakable link with reality. Both aspects will be discussed in the following chapters.
11
Differences between oral and literate societies
As we have seen, we find speech situations in both oral and literate societies, and they differ accordingly. The invention of writing has changed the discourse at large to such an extent that we may well wonder if we really should subsume communication in oral and literate societies under the same umbrella term ‘language’. A discourse, this entirety of contributions made by the members of a discourse community, is, as far as I see it, more or less equivalent with Saussure’s concept of la parole, the manifestation of language. But what is a manifestation of language? How equivalent is the transitory acoustic phenomenon of a speech utterance to the material evidence of a text, carved in stone, written on parchment, or accessible on the internet, downloadable and printable as hardcopy? The spoken language of the primordial speech situation is more than just ‘text’. Text may be a vital part of symbolic interaction, but it is not its sole ingredient. As said above, the linguistic content of a spoken utterance cannot be separated from the setting in which it occurs, from the voice in which it is delivered, from the ostentation, the gestures and the facial expressions that accompany it, not even from the objects that make up the reality shared by the participants of the situation. All these ingredients extend the discourse and are inseparable from what is said. To my knowledge, it is only the relatively small community of integrationist linguists around Roy Harris who base their approach on this acknowledgement which throws the whole idea of a narrowly defined concept of language (and a narrowly defined linguistic discipline) into serious doubt. The members of genuinely oral societies are, under normal circumstances, not even aware of the fact that they are communicating verbally, just as we hardly ever reflect on the gestures we are using all the time. They certainly have no reason to discuss language. The anthropologist Jack Goody believes that most oral societies do not even have a word for what we call a word (Goody 1987: 274). Even the Greek and Latin words we take to be equivalent to this concept, logos and verbum, mean first of all speech, not this Western medieval construct of a concatenation of alphabetical characters between spaces. From the perspective of an oral society, spoken language is hardly seen as symbolic. Of course, from the observers’ perspective there can be no doubt that 150
Differences between oral and literate societies
151
saying chair evokes something else, namely an object within a shared reality, or a discourse object (in case there is no chair around). But would people in an oral society really understand our modern claim that lexical items are signs? Would they not insist that signs would have to be concrete, material objects? Spoken language is a transient acoustic phenomenon, seemingly immaterial. The idea that an utterance can have a meaning independently of the speaker’s intentions would be met with disbelief. Someone’s utterance is a part of their genuine behaviour; and how someone behaves is usually not seen as an expression of content. There is no ‘language’ that they could see as something separate from saying the word or from gestures and ostentation, or from the objects the situation is about. One of the reasons why it is so difficult to make sense of a transcript of a primordial speech situation is that it does not record these deictic acts. Spoken language becomes symbolic only in the observers’ perspective, a perspective immersed in the literacy of the anthropologist’s notes and transcripts. However, the interaction between the literate anthropologist and her illiterate community will have an effect on the perspective of those observed. An oral community, confronted for the first time with writing, is bound to grasp the possibility of a language as something material, as consisting of signs standing for something else. With the advent of literacy, there are two sometimes incompatible paradigms of language. Primordial speech situations still prevail, but they are complemented by situations presupposing literacy. To tell the two apart is not always easy. As Ian Hacking has shown (Hacking 1992), the impossibility of carrying out deictic acts in written texts can lead to different interpretations in sentences that look very much alike on the surface. His examples deal with sentences like a) Today, there are 17 people in room 237. b) In 1801 Tewkesbury’s total population was 18,708. It would be difficult to imagine how people in a) would not agree in face of the evidence. It is very much the shared reality of their primordial speech situation. But for b), there could not be a comparable shared reality, not even among the people back in 1801. We would expect there to be documentation in which the town borders are defined, and a register, or several, kept by authorised institutions, in which those denizens would be listed who count, based on some definition, as residents. No summons of everyone to the commons could ever create a situation similar to that in a). Some people would lie sick in bed, some would have left for London; non-residents from the surrounding villages would have come for their shopping. Any group assembled on the commons would not be fully co-extensive with the population of Tewkesbury, in the sense that the 17 people, and only they, in a) were indeed occupying room 237. Without writing, sentence b) would not be possible.
152
Meaning, Discourse and Society
The profane and the arcane origins of writing The advent of literacy changed society and the way people understand themselves as being part of society. Today, mainly two theories of the origin of writing are discussed. The theory of the profane origin of writing is backed up by the Sumerian cuneiform writing system. Its original purpose may have been to take stock of possessions such as herds, or the contents of larders, for instance honey pots and grain bushels. The priest caste seems to have been in charge of fiscal affairs. The written records were permanently available, and their advantage was that they could be consulted in lieu of the real thing, making bookkeeping much easier. But more importantly, texts (if we can call these early lists texts) were now seen to contain knowledge, a kind of knowledge not hidden away in people’s first-person experiences which could be accessed only in the form of not-veryreliable testimony, but archived knowledge. It was knowledge that could be made available in the absence of the scribes and the people who had given them their instructions. It was knowledge dislocated from the reality to which it referred, knowledge about past or distant realities. It was authoritative knowledge, because only those in charge could command the arcane art of writing. The other theory about the origin of literacy is that writing was introduced for religious or divinatory purposes. Surrounded by inexplicable events, people looked for illumination. They ascribed things they could not understand to spirits, demons and divine beings. Such deities were considered much mightier than humans; they could make thunder and lightning, rain or drought. They could make demands on us. Deities also know more than humans. They see beyond the here and now. Our desire to profit from their might and to tap their knowledge gave rise to a profession of mediators between us and them, priests or shamans and diviners. Before the invention of writing, shamans or priests (the professionals responsible for supernatural affairs, namely those realms that could not be inspected in primordial speech situations) had nothing but their rhetorical skills to argue that they could influence reality, or the course of things. More circumstantial evidence was needed to convert the sceptics. At first they turned to the non-symbolic concepts, as Claude Lévi-Strauss calls them, sights in the shared environment of a group which leave an impression on people’s memories, particularly those concrete objects in the environment that could function as symptoms linked with recurrent events, such as a particular tree with early leaves in the spring, or an intermittent source. Once the priests became recognised as experts in their exegesis, these concepts morphed from mere symptoms to symbols. If the arrival of the first swallows used to be associated with the arrival of summer, this event was now seen as a mark. Thus gradually symptoms morphed into signs, while the relationship between the object and its meaning became increasingly arbitrary. From then on it took experts, shamans or diviners, to tell what they meant. If someone wanted
Differences between oral and literate societies
153
to know if his pregnant partner was bearing a girl or a boy, the shaman might cut open a frog and look at its entrails. He might say: ‘If you offer the gods a large deer as a sacrifice, it will be a boy.’ If it turned out to be a girl, the deer had just not been big enough. Material evidence in the form of signs provided a chance to convince people that shamanist intervention went beyond their shared reality. If they could demonstrate a referential link between two material objects, then they could claim that this link was in fact causal. Such a referential link is a link between a material sign and another object, action, process or state. The layout of animal bones or intestines could be connected to almost anything. In the diviner’s interpretation, it could imply the sex of an unborn child or the day when it would begin to rain. It must have been in the interests of shamans and diviners to replace symptoms (which could be interpreted by anyone) with signs, only meaningful to the initiated. Accidental marks fitted the bill. Harald Haarmann insists that the Vinca script found in the Balkans and dating back to 5000 BC served such religious purposes (Haarmann 1990: 70ff.). To me, it seems plausible that such marks, probably unavoidable in the process of glazing, preceded the direct intervention of shamans, who added their own marks on purpose. All these marks were presented to the wide-eyed audience as signs marked by the deity himself or herself, while the priests or diviners claimed for themselves the monopoly of their interpretation. Seemingly being messages mailed to a group of people by the immortals themselves, they had to be true, presaging what would happen in the future, and explaining what had been missed by experience. This story compares well with what is known about the origin, or one of the origins, of Chinese characters. Here, too, we find them on objects used for divination, the shells of tortoises that were exposed to strong heat and thus cracked. On the site of Yin Xu near Anyang in the province of Henan, more than one hundred thousand such pieces engraved with characters of some sort have been found (Feng 1994). However, these characters in the shells look surprisingly similar to the cracks caused by the heat. Writing may have come about as a systematisation and extension of the natural cracks, a way to turn a more intuitive divinatory talent into a trade that could be taught and learned. While the Yin Xu artefacts date from 1700 to 1023 BC, it is also argued that the custom of incising tortoise shells can be dated to a much more distant past. In his paper ‘Questions on the origins of writing raised by the Silk Road’, Jao Tsung-i claims that ‘[f]inds such as the still-disputed bone implements of the Lung-shan culture in Shensi and the newly discovered tortoise plastrons from Chia-hu, Honan prove that the characters used for divination with tortoise shells and animal bones since Shang and Chou times actually have remote origins dating back seven or eight thousand years’, even though so far no evidence has been found to link them to the Yin Xu findings (Tsung-i 1991: 3).
154
Meaning, Discourse and Society
Even today religions insist that their teachings must be true because they are God’s word in written form. According to one of the two accounts of the Covenant or the Ten Commandments, in the Old Testament (Deuteronomy 9), the original tablets were inscribed by God’s own hand (‘written with the finger of God’), brought down from Mount Sinai by Moses and then smashed by him in view of the idolatry committed by the Jews, to be later replaced by a second set of tablets inscribed by Moses according to God’s dictate. Their materiality was essential, for even this second set was too holy to be read by anyone but Moses. They had to be locked away in the ark, with the priests being their sole interpreters. In God’s word, we are told, we still find the authentic relationship between the sign and the thing it stands for; and we can rest assured that those sentences are true where such links obtain. In these early beginnings of writing, the link between the sign and the thing is not subject to negotiation in a speech situation, but vouchsafed by God himself and revealed by his priests. The important thing is that in arcane writing the referential link is indistinguishable from a causal link. What is written is part of the discourse-external reality. It is the dislocation of a text from a speech situation that explains its causal force in predominantly oral societies. In spite of what Ferdinand de Saussure and a multitude of Western linguists of the nineteenth and twentieth centuries want us to believe, writing can be looked at as something more than just transcribed speech. What is it exactly that links the content to the writing, be it no more than an unintelligible sign on a clay pot? How can this content achieve permanency? Why will it not fade in the same way as the content of our ancestors’ chats in front of their dwellings, around their fireplaces, which has long since been forgotten? How unbreakable for them was the link they saw between a divine sign and the thing it stands for? Would they believe that if the sign was material, so must the thing it stands for be ‘real’ too? Would the thing referred to have to be as real as the shared reality to which the sign belongs? It is this aspect of factuality and truth which has made people believe that written language is more reliable than speech. Indeed, this might be the reason why Protestant and Catholic ministers alike do not rely on their memories, but insist on reading out the written text whenever they perform a sacred rite such as baptism, wedding or the transubstantiation of bread and wine. Does a memorised text not have the same power as a written text? The phrase for it stands written is used, to this day, as a warrantor of truth, as these examples, found on Google, show: For it stands written: Worship the Lord your God, and serve Him only. (wwwdaily. presbycan.ca/devotions/1997/97-02-25.html) For it stands written in the law: the sun should not set on one that has been put to death (www.mithraism.erudition.net/append/passage3.htm) For it stands written in the Kitvei HaKodesh, HINENI YISSAD B’TZION EVEN PINNAT YIKRAT (www.beittikvahsynagogue.org/obh/engtexts/1skefa.htm)
Differences between oral and literate societies
155
For it is written, in Surah Al-Maa’idah 5:116, And behold, Allah will say–O Jesus Son of Mary, did you say to men, worship me and my mother as gods besides (www.theonion. com/onion3513/christ_islam.html)
Thomas Love Peacock, in his novel Headlong Hall, satirises this attitude: ‘Nothing can be more obvious than that all animals were created solely for the use of man.’–‘Even the tiger that devours him?’–‘Certainly,’ said Dr. Gaster.–‘How do you prove it?’–‘It requires no proof,’ said Dr Gaster: ‘it is a point of doctrine. It is written, therefore it is so.’ (Peacock 1816: 29)
While we may have to accept, as true believers, that what is written must also be true, we may still have doubts about its meaning. Even if we are assured that the written word has a causal force, we need an interpreter to tell us what it will cause. As Jean-Jacques Rousseau sees it, the power of the clergy rests on their monopoly of interpretation. But this monopoly requires that no one else can understand what is written: In all these three revelations, the sacred books are written in languages unknown to the people who believe in them. The Jews no longer understand Hebrew; the Christians neither Greek nor Hebrew; the Turks and Persians understand no Arabic, and even the modern Arabs themselves speak not the language of Mahomet. Is not this a very simple manner of instructing mankind, by talking to them always in a language which they do not comprehend? (Rousseau 1909–14: Part 4, § 193)
In writing, the traditional reference to a shared reality, commonly seen as a discourse-external reality, has been replaced by something else, namely by meaning. A written text is not a mirror of what we jointly experience, nor is it the expression of a speaker’s intentions. The content of a text is its meaning. This meaning is the ways in which this text is like, or different from, all other texts. Together, the members of a discourse community will negotiate the meaning of a word, a phrase, a predication we find in a written text, for instance in a poem. This meaning is co-extensive with the knowledge we have about the discourse object (a concrete or abstract object, a process, a state, etc.) for which the lexical item or the predication stands. We cannot share any knowledge that has not become part of the discourse, that does not exist as a symbolic representation. Written text stays. Speech, unless recorded, is a transient phenomenon. Of course, oral societies cherish knowledge just as literate societies do; and in the course of generations they developed a culture of memorisation. Talented young people were trained to memorise huge amounts of knowledge, and methods were developed to package such knowledge in a way that made it easy to remember, using metre and rhyme, for instance. The knowledge that can be preserved in oral communities tends to be of the kind that can be enumerated and listed. While Goody seems to believe that such lists mark the beginning of a literate culture (Goody 1977: 74–111), I agree more with Walter Ong, who accepts lists, for instance genealogies, as part of oral culture (Ong 1982: 46ff.).
156
Meaning, Discourse and Society
Such lists, often lists of geographical names, of animals and plants, support a community’s identity, and also the belief that one’s community is a stable, rather than dynamic, organism. For the members of such a community, the names stand for concrete entities and are not seen as words in the linguistic sense. They provide referential links between the collective mind, present in the memorised lists, and the discourse-external reality. Probably they should not be seen as conventionalised signs whose meaning can be negotiated by the discourse community. Just as primordial speech situations still play an important part in our literate societies, so is our understanding of writing to some extent still infused by the arcane origin of it. Once we forget that written texts are contingent symbolic content in need of interpretation, we tend to accord them an irrefutable, absolute authority. They must be true because they are not the contingent result of negotiation. We largely accept whatever the Encyclopaedia Britannica has to tell us, but we have strong qualms about Wikipedia entries. In Chapter 13 I will have more to say about the relationship between knowledge and perceived truth.
Some effects of literacy Law One of the fields of social life in which the effects of literacy are particularly strong is our collective sense of justice. In an oral society, there cannot be a legal system based on an ever-growing multitude of laws. As Walter Ong knows, ‘[t]he law itself in oral cultures is enshrined in formulaic sayings, proverbs, which are not mere jurisprudential decorations but themselves constitute the law’ (Ong 1982: 35). The Proverbs of Solomon point to a past when judges were making decisions based on a limited number of adages that were passed on from one generation to the next, embedded in contexts that facilitated memorisation. In his book The Logic of Writing and the Organization of Society, the anthropologist Jack Goody explores the limits of a legal system in an oral society. In his view, ‘there is no effective distinction between “law” and “custom” as sources of judicial decisions, although certain rules may be seen as justiciable and others not’ (Goody 1986: 135f.). When a society turns literate, documents such as contracts and testaments replace witness statements. The question of who owns what becomes removed from the shared reality and will be decided on written, documentary evidence, deemed to be true until revealed as a fake. The wealth of the Catholic Church in Italy to this day is owed to a large extent to the forged Roman Imperial edict called the Donation of Constantine, probably faked in the eighth century, a useful reminder of the almost ineradicable belief in the truth of the written word. In England, the common law of the traditional oral society was, in modern times, gradually replaced by written laws, and controversial claims were increasingly decided on the basis of documentary evidence.
Differences between oral and literate societies
157
We can assume that literacy was successful because it tended to work in favour of the ruling class able to afford scribes and literate lawyers. What we see happening now all over the developing world, the expropriation of smallhold land owners due to the absence of written deeds, has been shown as one of the detrimental consequences of the sudden imposition of literacy by the colonial powers. A similar redistribution of land titles happened, for much the same reason, when Oliver Cromwell’s army confirmed British rule over Ireland. In the absence of written documentation, eleven million acres in the counties of Leinster, Munster and Ulster were confiscated as crown property and subsequently redistributed to Protestant supporters. More repossessions followed. While in 1603 90 per cent of the land belonged to Catholic owners, this relationship had been reversed by 1778. Now 95 per cent of all Irish land belonged to Protestants (Edwards 1973: 166, et passim). Unsurprisingly, though, the majority of scholars tend to see literacy as progress. For Georg Elwert, there can be no doubt that ‘the rule of law and growing literacy rates are correlated. The rule of laws requires the written text in order to create realms of foreseeability … Organization of family life … as well as economic activity profit from foreseeability. One knows what is forbidden and what is possible’ (Elwert 2001: 63). However, not in every case does the arrival of literacy eradicate former common-law principles. Harry Hallowes squatted on prime land in London for over twenty years without being challenged, and was eventually granted title to the land, as there was no written(!) deed contradicting this claim (Guardian, 24 May 2007). These squatters’ rights, which can, under certain conditions, even render written documentation worthless, if backed up by oral witness statements, date back to preliterate times when property claims were still negotiated on the basis of a shared reality. Interestingly, all court proceedings seem to require orality up to this day. Just as it would be unimaginable to hold a religious service in writing, the verdict on a criminal must arise from proceedings which have to be held in a kind of ritualised primordial speech situation, requiring collaboration among the participants. Perhaps this is why a growing number of legal acts, such as the imposition of fees, the levy of fines, the service of notices of all kinds including antisocial behaviour orders (ASBOs) are now being imposed in writing, thus denying the need for collaboration between the involved parties. What is negotiated in legal proceedings in oral societies is a shared reality, in the form of claims by the plaintiff, and counter-claims by the defendant, in the presence of the whole community or their representatives whose task it is to watch over the equitability and fairness of the proceedings and to pass sentence. Claims refer either to a shared reality, or alleged shared realities, for which witness statements are needed. Without the (consensual or non-consensual) negotiation of a shared reality arising from the proceedings, there can be no conviction. In legal proceedings of literate societies, what is discussed is not so
158
Meaning, Discourse and Society
much a shared reality as the meaning of written texts, namely of the documents introduced into a case. Meaning replaces reference. History In his book From Memory to Written Record: England 1066–1307, Michael Clanchy sums up the difference between the awareness of history in oral and in written societies: ‘Remembered truth was also flexible and up to date, because no ancient custom could be proved to be older than the memory of the oldest living wise man … Written records, on the other hand, do not die peacefully, as they … can be resurrected to inform, impress and mystify future generations’ (Clanchy 1979: 233). Again the keyword is shared reality. What an old man remembers is believed if it is confirmed by the memories of other men. Of course, there were always some members of a society specially trained to memorise long lists of importance for group identity, e.g. lists of former rulers and past events. As Goody has shown, these lists were not as stable as the audiences believed; apparently they were updated whenever the political situation demanded it. While memorisation and recitation of lists seems to be aimed at establishing identity and continuity in oral communities, these memorised lists were subject to constant modification, conscious or not, reflecting changed environments. Their recitation was more a ritual than an assertion of shared knowledge. The awareness of history in our modern understanding, as a more or less accurate account of a chain of subsequent and potentially linked events, is linked to literacy. History in oral communities hardly exceeds the space of two or three generations. Oral societies are concerned with sameness, not with progress. They believe in a foundational event setting up their society in days long past, and perhaps a few more events of a mythological nature strengthening their identity, but (with the exception of the genealogies conferring legitimacy to their rulers) they try to keep the list of events worth remembering short. On the other hand, they are more keen on circular, recurrent chains of event types, such as the four seasons, or longer units covering seven or twelve years. They may also believe in gradual decay, and in the eventual return of a Golden Age. This perspective does not entirely die off with the coming of literacy. Our modern view of history as progress, owed largely to Hegel and his student Karl Marx, presupposes a comparison with earlier times requiring written documentation. But it does not eradicate a deeper belief in underlying laws that make history move in circles. Indeed, the French Revolution was aimed at what was believed to be the re-establishment of past conditions, of revolving the wheel of history again to its foundational condition. While the stories old wise men have to tell (perhaps even including stories they have learned from previous generations when they were still young) might still be understood as referring to a shared reality of past times, there
Differences between oral and literate societies
159
is another set of narratives, normally recited by especially trained people (professional bards), namely foundational myths and comparable religious legends, which serve principally to create and maintain a sense of group identity, but are not expected to refer literally to a discourse-external reality. The stories of the old wise men might still be subject to some kind of negotiation, by comparing different variants referring to the same event. Myths, however, are not delivered for discussion, but as a top-down ritual. Timeless, as they are, they do not invoke a shared reality. Rather, they should be seen as forerunners of those arcane narratives that we find, in a literary society, under the heading ‘For it stands written’. Schools and knowledge As Jack Goody has pointed out, ‘schools are essentially instruments of written culture’ (Goody 1987: 235). This is certainly the case. Literacy and numeracy are, after all, the basic skills that define almost all schools, with the exception of those in which pupils are taught to memorise texts. Even there, for example in the madrasas, the teachers are expected to be literate and to interpret the texts they have memorised. The types of things that are customarily taught in schools are hardly the kind of practical skills that can be learned by sharing a reality with the person who exercises them. To become a beekeeper, one does not have to study books. An apprenticeship with an established apiarist will be more useful. A mason would teach his apprentices his art and skill by discussing the shared reality of bricks, mortar, and visible, common-sensical standards of stability, asking for thicker walls the higher the buildings were designed to grow. Romanesque churches were built without written instructions, relying instead on the masons’ experience. Gothic cathedrals, on the other hand, already presupposed plans drawn up by architects; they made use of abstract knowledge existing and transmitted in writing. A mason then had to be trained to read and interpret an architect’s plan, a symbolic representation of the future building. This is what one learns in schools. But for abstract knowledge, like grammar, rhetoric and mathematics, for historical and geographical facts, and for the cornerstones of the natural sciences, we need books. Even though the study of physics and chemistry includes the awareness of a shared reality, in the form of experiments carried out in range of the pupils’ perception, an understanding of the underlying processes presupposes a theoretical knowledge that is too abstract to be extracted from observation alone. Schools are socially constructed institutions. Their activities are normally regulated by written documents. The Vedic tradition in India, however, seems to present a stunning exception to this rule. Many scholars have suggested that it was a pronouncedly oral culture, in which students learned to memorise and to interpret texts, mostly narratives but also Panini’s grammar, from their teacher’s
160
Meaning, Discourse and Society
recitals, and that even the teachers refused to put down their knowledge in writing. In his book Education in Ancient India, Hartmut Scharfe insists: ‘The outstanding feature of the oldest Indian education and Indian culture in general, especially in the centuries BC, is its orality’ (Scharfe 2002: 8). It is not that writing was unknown. The Indian scholars had learned about the writing systems of neighbouring societies, and it seems to have been a conscious decision not to use it for their own studies. Scharfe quotes Kumarila: ‘That knowledge of the truth is worthless which has been acquired from the Veda, if the Veda has not been rightly comprehended or learned from writing’ (p. 9). Among the reasons given for the renunciation of writing, Scharfe mentions ‘[t]he fear that the sacred knowledge could fall into the hands of members of the lower castes, or that the student may not be fit for the sacred and supposedly powerful knowledge’. Furthermore, the goal of reciting ‘the Vedic mantras with correct accents and intonations’ meant that the students had to be exposed to oral renderings (p. 8). I still find it hard to believe that Panini’s grammar, certainly not a small or easy text, was transmitted exclusively through memorisation over half a dozen centuries, in spite of being broken up in a quite endless list of sayings. Goody, too, is sceptical as to an exclusively oral tradition and assumes that there had to be literacy at least initially: ‘[T]he Brahmins seem to have continued over a long period to reproduce orally knowledge that derived from literate sources’ (Goody 1987: 115). Perhaps we can compare the early Indian situation with the madrasas or with apprenticeships of young bards. It is certainly atypical for oral societies in the sense that the content of the teaching does not refer to a ‘shared reality’, but to texts. A possible explanation is that the early Indian education system was focusing less on scientific ‘explanations’ (in Wilhelm Dilthey’s sense) of objects and their relationships with each other, than on learning to ‘understand’ (i.e. interpret) texts collaboratively (Dilthey 1883). Property Just as many species are known to defend their territories and their offspring, it is not surprising that the concept of property can be found in all human society, including oral societies. Even those societies subscribing to the idea of collective property, and rejecting the idea of marriage and inheritance, would distinguish between ‘us’ and ‘them’ and feel offended if outsiders attempted to take hold of what they own. In oral societies, however, property seems to be linked much more directly to a shared reality than in literate societies. Over the centuries, two main theories have been forwarded to explain how people can obtain property. Today they co-exist peacefully, and those defending property rights have become resourceful in their alternative use of them. The older approach is known as occupatio rei nullius, the occupation of some concrete thing, mostly land, that was not claimed by others. Jean-Jacques Rousseau tells
Differences between oral and literate societies
161
us how he thinks this stratagem was invented (Discourse of the Origin of Inequality): The first man who, having enclosed a piece of land, thought of saying ‘This is mine’ and found people simple enough to believe him, was the true founder of civil society. (Rousseau 1754)
The other theory, developed by John Locke in his Second Treatise, is that property results from work. Land, unless it is worked on, soil, unless it is tilled, remains worthless. Thus the native Americans, unlike the settlers, have no right to their land: ‘As much land as a man tills, plants, improves, cultivates, and can use the product of, so much is his property’ (Locke 1764: Chapter V, Section 32). It was a theory fit to support the Whigs’ claim to participation in the state’s offices on a par with the landed Tories, whose wealth was said to date back to times when it had still been possible to pick up things, including land, that did not have an owner. The weak point in Locke’s theory was the question of the ownership of the workman. For in the end it was the person to whom the workman had contracted himself, and in whose name he carried out his work, who was the rightful owner of the product: ‘Thus the grass my horse has bit; the turfs my servant has cut; and the ore I have digged in any place, where I have a right to them in common with others, become my property, without the assignation or consent of any body’ (Locke 1764: Chapter V, Section 28). But those who took the native Americans’ land under their plough placed little confidence in Locke’s property guarantees. They wanted to have additional proof of their ownership: written documents, deeds, certifying that the land was rightfully theirs, whether they decided to till it or not. This happened not only in the colonies; it also applied in Britain between 1750 and 1850, during the enclosure movement, when private bills of parliament divided up huge amounts of common land among the landlords. People who might have thought they had common law granting them, and their cattle, access to the commons, soon found that a written private act of parliament assigning the land to interested parties proved to be more powerful. The ‘common right’, because not written down, was declared mere custom, and custom was considered ‘illegal’, for instance by Sir John Clapham in his book An Economic History of Modern Britain (1929, quoted in J. M. Neeson 1993: 55f.). The complex stratification of our modern society is based on writing. In most instances, it is written documentation that backs up a property claim, and sometimes these written claims seem to contradict our perceived and shared reality. The houses we live in, the pensions we expect, the bank accounts in our name, the trains we use, are not ours; they are controlled by housing societies, by our employers, by private equity or by banks. This extent of our expropriation is largely invisible, as the titles, the signifiants, on which they rest refer less
162
Meaning, Discourse and Society
to a discourse-external reality than to a multitude of cascaded signifiants, small print cross-referenced with small print elsewhere, and from there to ever more documents which are often extremely cumbersome to access, and so almost ad infinitum, creating a network of texts whose opacity is designed to imbue fear, uncertainty and doubt. Ever since the invention of writing, the three foundations of property, occupatio rei nullius, work and trade, have been skilfully played against each other to the disadvantage of the weak. Writing, or the absence of it, has nearly always played the decisive role. Because the Australian aborigines had not set up land registers, the British conquerors presumed that they did not own the land on which they lived. Sometimes, however, written documents gave claims which those in power were not keen to accept. In such cases, strangely enough, oral traditions were use to abrogate written treaties, for instance with the argument that they had not been ratified by parliament in an ancient oral ritual of voting. This is what the Maori found out, when they wanted to reap their part of the fruits from the Treaty of Waitangi. The British conquerors argued that, due to the lack of ratification, this contract between the representatives of the Maori and of the Crown ‘confers no rights cognizable in a court of law’ (Brookfield 1989: 10). Today the World Bank forces developing countries to auction off agricultural land which is not backed by written titles to international corporations, even though this land has been tilled by smallholders for generations (Chossudovsky 1997). In the nineteenth century, it had become customary among European powers to legitimise their conquests by producing a treaty. This was a document signed by an African chieftain to confer the possession of the land he ruled (or was said to rule) to a civilised nation. Thus Zimbabwe became the property of Cecil Rhodes, and eventually of the Crown, by such a piece of paper. It was ‘signed’ by someone who did not know that there are such things as writing and signatures, and in whose culture there was no concept of land-ownership by absentee landlords. ‘The very word treaty is a euphemism, for many chiefs had no idea what they were signing. They were being asked to mark their X’s to documents in a foreign language and legalese. The idea of signing over one’s land to someone on the other side of the ocean was inconceivable’ (Hochschild 1998: 71–2). As Alan Samson writes, Rhodes’ ‘dodgiest-of-all-dealings involved those with the Ndebele leader, King Lobengula. Invoking a Treaty of Friendship, Rhodes’ agents coerced the king into accepting a mining concession that would allow white men to dig “a big hole”. In the process, Lobengula signed away the rights to his kingdom’ (Samson 2002). Once documents came to be accepted in lieu of witness statements, theft and related crimes were used by those in power to redistribute property on a large scale. Documents could be forged easily, once they became disconnected from a shared reality. Pierre-Joseph Proudhon knew already in 1840: ‘La propriété, c’est le vol!’ Indeed, the redefinition of property as something represented by a
Differences between oral and literate societies
163
document gave rise to the new concept of intellectual property rights, something inconceivable before the invention of the printing press. These rights presuppose writing. Only written texts can be copyrighted, and while Kentucky Fried Chicken can interdict pub owners from using in writing the phrase family feast which they have trademarked, they can hardly charge us for using it in spoken language. Writing, unlike spoken language, can be owned, not necessarily by the authors who conceived the text, but, as in the case of papers in academic journals, by their publishers. More important are patents, i.e. texts that provide ownership to knowledge, so that this knowledge may not be exploited by other people. Whoever successfully rephrases in writing knowledge concerning a production process of a living or a dead thing that already exists in the discourse can monopolise the process or claim ownership to the thing or plant or animal. Again it is the meaning of the text that replaces its reference to the outside world. The more opaque the meaning of the patent text, the more difficult it will be to demonstrate that the process or thing described is not new but has already been part of a shared reality.
Literacy, meaning and reflection While the effects of literacy on the distribution of the common wealth have not always been beneficial for most members of society, it cannot be denied that literacy had a profound effect on societies’ modalities of thinking. The discourse of oral societies deals largely with concrete things. There is little venture into abstraction. People are just not interested in reflections on how things are connected on a deeper level, nor are they inclined to consider hypothetical situations outside of their shared experience. For them, such thoughts would make no sense. As it would be impossible to preserve the fruits of such reflections for future use, every effort spent on them would prove ill spent. IQ tests, as we know them, would be totally inapplicable. Even simple syllogisms seem to be beyond comprehension. The Kashgars, an ethnic minority in Xinjiang, formerly belonging to the Soviet Union but now a part of China, were still overwhelmingly illiterate (though they knew about writing) when, in the 1930s, Aleksandr Lurija tested their understanding of syllogisms. In one of many such instances, he confronted Abdurakhm, a peasant aged 37, with this syllogism (108, slightly shortened): ‘Cotton can grow only where it is hot and dry. In England it is cold and damp. Can cotton grow there?’ ‘I don’t know.’ ‘Think about it.’ ‘I’ve only been in the Kashgar country; I don’t know beyond that …’ ‘But on the basis of what I said to you, can cotton grow there?’
164
Meaning, Discourse and Society
‘If the land is good, cotton will grow there, but if it is damp and poor, it won’t.’ The syllogism is repeated. ‘If it’s cold there, it won’t grow; if the soil is loose and good, it will.’ While literacy seems to be responsible for the ability to think in and about hypothetical and counterfactual cases and thus to develop new solutions to old problems, an even greater achievement is the capability to recognise and reflect on symbolic content. It is the step from reacting to a speaker’s intentions to consciously interpreting what has been said. Literacy makes people conscious of the fact that they can discuss what a text (segment) means, disregarding the author’s possible intentions, but also conscious of the fact that between writer and reader there is no reality shared in a primordial speech situation. It takes literacy to make people aware that the meaning of what has been said or written can be described without involving a discourse-external reality. Yuri Lotman has pointed out that literacy has a profound effect on the way people deal with symbols. In oral societies, symptoms/symbols are isolated material things or events, such as lightning (which can be interpreted as telling us what is going to happen), or a tree as the materialisation of a spirit. If a symbol comes to be seen as a recurrent part of the culture, it will normally be dealt with by some ritual as a mnemonic device (for instance dancing around a boulder at solstice), though the specific meaning of the symbolic act will not be spelt out. But some kind of consensus has to be reached whenever an object of a shared reality is declared a symbol. What is missing is a contextualisation of the symbol within the discourse. In an oral society, the context of a symbol is other objects in its environment. ‘The material objects which represented the mnemonic sacred symbols are found not in verbal texts but in ritual ones. And in relation to that text they preserve a certain freedom: their material existence continues outside the ritual, or they may be included in many different rituals, which gives them multiple significance.’ In a literate society, ‘[t]he verbal text (and in particular the written one) is based on syntactic bonds … These signs … acquire their meaning only through the syntagmatic chain where they form chains of signs … We can see the difference between [literate and oral signs] if we compare a sentence (a chain of linguistic symbols) with an ornament (a chain of magic-mnemonic and ritual symbols)’ (Lotman 2001: 249f.). The largely illiterate Kashgar peasants only had recourse to their shared reality with its objects and events, not to the syntagmatically conjoined lexical items of a text embedded in a discourse and thus linked to a universe of other texts. They did not have at their disposal recurrent paraphrases for the lexical items or discourse objects these items stand for. While they were engaging in a discourse, they were not aware of the discourse as a concept of their shared
Differences between oral and literate societies
165
reality, just as we in a literate society are not normally aware of a discourse object ‘gesturing’ as something comparable to the discourse object ‘language’. It is the materiality of writing that makes language visible. Once we can ‘see’ that we can do things with language, we become aware that negotiating content, for instance explaining things to other people, is something we often do. As soon as people learn that the meaning of a symbol can be found in the written discourse (‘For it stands written …’) they are also in a position to realise that symbols, for instance the lexical item cotton, come with a meaning even if the discourse object is not part of their shared reality but somewhere far away. Meaning, they find, is something not inherent in the objects of a shared reality, but rather in the objects of the discourse, removed from a claim to reality. A syllogism is a riddle whose solution is not to be found in the ‘real’ environment we share, but only in the discourse. The conscious act of negotiating what has been said is thus at the origin of all reflection. Reflection comes about as the result of the collaborative practice of discussing meaning, of interpreting a piece of language. Reflection presupposes that one is aware of what one is doing. While people in primordial speech situations interpret what they hear, they are not aware that they are interpreting. Yet once utterances begin to be recorded in writing and can be read in the absence of their author, their meaning can be discussed by their addressees. This is the kind of interpretation that people no longer carry out routinely and quasi-automatically, but one that they are aware of. People begin to reflect on their intellectual collaboration. They begin to observe themselves. For Blumer, it is the observer who ascribes interpretation to what he or she sees the objects of observation doing. He would probably accept Durkheim’s stance that we need sociology because ‘social life must be explained not by the conception of it formed by those who participate in it, but by the profound causes which escape their consciousness’ (Durkheim 1897 [1982]: 171). Reflection, however, becomes pervasive (and sometimes even an obsession) in literate societies. Literate societies become their own observers. As I see it, societies in which we find reflection are perfectly capable of making sense of their ‘social life’, as plurivocal and diverse as these interpretations may be. External observers, for instance anthropologists, may or may not agree with (some of) them. But their interpretations are, in principle, as good as those of the people they observe. I would like to stress once again that the reflection I am talking about is a social, collaborative activity. Once people have mastered it, they may carry it out in monadic seclusion. But this is not how they learn it. Society and the individual person presuppose each other. This is, according to Michael Holquist, how Mikhail Bakhtin sees the discourse: ‘Indeed, it was language that enabled the society which in turn permitted the existence of individuals: consciousness of meaning is social in its origin … thought remains [even] in its abstract form sublimated conversation. Thus reflective consciousness implies a social situation which has been its precondition’ (Holquist 2002: 56, his emphasis).
12
Empirical linguistics deals only with recorded language
I have devoted so many pages to accentuating the difference between traditional oral and more recent literate ways of communication, because we have to bear in mind the differences between them when we want to find out what the study of language can and can’t do. In our society we find, side by side, primordial speech situations and metalinguistic reflections of arguably great depth. But while it is impossible to reduce a primordial speech situation to a written text, this can easily be done with oral discussions that do not involve extensions, such as a shared reality in visible reach of all participants. Hansard, for instance, is, in most cases, considered commensurate with, and as authentic as, the oral parliamentary proceedings of which it is a (sometimes quite heavily postedited) transcript. Yet transcripts of pub chats normally leave a lot to be desired. They will almost always contain bits and pieces we cannot, with the best of intentions, make sense of. The problem with the primordial speech situation is that it is impossible in principle to separate what is said from the setting in which it takes place, and one cannot distinguish the linguistic from the non-linguistic symbolic interaction that takes place in it. Not even the most detailed annotation of these non-linguistic factors can replace the immediacy of being part of the situation itself. Primordial speech situations are those in which there is no clear dividing line between the symbolic content of the texts spoken, and the less symbolic extensions of the discourse: the shared environments with their props, the behaviour, gestures, facial expressions and audible peculiarities of those taking part in them. Whenever linguists want to discuss with their peers a piece of speech, they have to rely on a transcript or a recording. It is, first of all, a transcript of the texts spoken. It seems that no annotation system can do justice to the non-linguistic extensions of the discourse. Once a primordial speech situation has taken place, it is irretrievably lost and never fully recoverable in recordings or transcripts. This does not mean that transcripts are useless. But if they are discussed by linguists, then it must be as transcripts of spoken language, not as an equivalent of a primordial speech situation. For linguists working on spoken language, however, there is no alternative to working with transcripts. Empirical sciences need empirical data. Any academic 166
Empirical linguistics deals only with recorded language
167
discipline, including linguistics, presupposes a community of peers. They have to discuss the data they are confronted with, with the aim of interpreting it. Without a community of linguists, and without their discourse, there would be no discipline of linguistics. Yet it is impossible to record the ‘shared reality’ in full that is an integral part of a primordial speech situation. Everything outside of the text itself is uncertain. The only reality accessible for investigation is the recorded text embedded in a discourse. To the extent that linguistics views itself as an empirical science, it has to deal with the data available to it, which comprises only the signifiants, never the signifiées insofar as they are thought to populate a reality outside of the discourse. Linguists therefore have to accept that any exhaustive interpretation of a primordial speech situation is outside their remit. Still, many linguists, importantly Ferdinand de Saussure, insist that spoken language is the primordial form of language. Historically they are right, of course. There was spoken language before writing. For this reason many linguists have claimed that it is, first of all, spoken language which linguistics should study. Roy Harris, however, has argued, in my eyes very convincingly, that ‘Saussurean linguistics is essentially the linguistics of “typographic man”’ (Harris 1987: 51). By treating originally spoken texts (and accompanying data) as symbolic representations and not as someone’s intentions embedded in a shared reality, we have already moved from speech to written (or otherwise recorded) language. As I have shown in the previous chapter, it is beyond a discipline of linguistics to investigate how texts are meant by their speakers, or how they are understood by their hearers. I can never fully express what a particular text, for instance my favourite poem, means to me. By reading it, I may have, somehow, converted it into an irreducible, untranslatable first-person experience, and no textual testimony will ever give full justice to this experience. That does not mean it is not possible, in a dialogic interaction, to negotiate its meaning. How I personally understand a text is actually quite irrelevant. The role I play within a discourse community depends on my observable symbolic behaviour, and not on the ineffabilities of my mental state. Dan Sperber and Deirdre Wilson’s relevance theory (see above, Chapter 6) claims to capture the speaker’s intentions, and therefore to be able to deal with primordial speech situations. But the only evidence they can provide is in the form of statements of intention, and they cannot avoid the infinite regress, namely that all statements of intentions again come with their own intentions. As I see it, relevance theory is not an approach that allows us to tackle primordial speech situations. Roy Harris’s programme of ‘integrationist linguistics’ aims to understand language as a creative process, part of human interaction based on the cotemporality of the world outside of a speaker, a world that includes their memories, their knowledge of the world and powers of reasoning, their interaction with their audience (including their non-linguistic behaviour), and the
168
Meaning, Discourse and Society
situational setting in which the interaction takes place. Such an approach might help to explain the ‘all-pervasive – and perhaps only authentic – characteristic of the individual’s involvement in language’ and ‘would provide an account of linguistic experience’ (Harris 1981: 164ff.). But this programme has to fail in view of the unavailability of the primordial speech situation for analysis. It would be impossible to exclude the speakers and their intentions from such an analysis. In the case of written language, matters are radically different. There we, the readers and the linguistic observers, are confronted with the meaning of the text, not with the speaker’s intentions. Any interpretation of a text is disconnected from its speaker. The author’s intentions are irrelevant to the meaning of a text. This is the stance of contemporary hermeneutics, and I will present it, in more detail, below in Chapter 15. The two modes (speaker’s intentions v. text meaning) are, it seems to me, incompatible, and they signify the key difference between the primordial speech situation and written language. When someone tells me there are many apples on the tree over there, the Gricean maxims of quality (truthfulness), quantity (informative content) and relation (relevance) suggest that the speaker has a reason for passing on this information, perhaps because she wants me to pick and have one of them. When, on the other hand, I find in a written text a sentence telling me that there are many apples on the tree, my interpretative task is to find out which role this sentence has in the wider text or in the discourse in which it occurs, and to reconstruct its meaning from that analysis. If I find my sentence in this context: ‘There are many apples on the tree – Il y a beaucoup de pommes sur l’arbre. There is, there are ou comment traduire “il y a”’ (www.anglaisfacile. com), I know that I am not offered a free apple but shown how il y a is rendered in English. If, on the other hand, I come across this sentence: ‘“There are many apples on the tree,” an old saying supposedly runs, “but only one Big Apple”’ (www.straightdope.com) then I will be left in the dark until I find, in a similar discourse, a text that might explain this sentence, for instance: ‘In 1924 Fitzgerald wrote that he first heard the term from stable hands in New Orleans who referred to New York racetracks as “The Big Apple”, or the ultimate goal of every horse trainer and jockey in the racing world. “Apple” referred to the prize being awarded for each race’ (www.classbrain.com). However, it could also be a quotation from John Burroughs’ book The Breath of Life, published in 1915: ‘The earth is not alone, it is not like a single apple on a tree; there are many apples on the tree, and there are many trees in the orchard’, lending itself to yet another interpretation (www.infomotions.com/etexts/Gutenberg). Three conclusions can be drawn. One is that language studies should refrain from endeavouring to describe people’s linguistic experiences. We just do not have access to people’s minds. The second is that, as long as it is impossible to record all the other data that are inexorably part of symbolic interactions, including memories, knowledge of the world and reasoning powers, any
Empirical linguistics deals only with recorded language
169
integrationist approach, as attractive as it seems, will have to be severely limited. The third is that spoken language can only be dealt with by the kind of linguistics we have today, i.e. a linguistics that is limited to the investigation of the elements that we posit for the discourse, and their interrelationships. Our quest for meaning will never leave the prison of la parole écrite, of the discourse available in recorded form. The world outside will always be hidden from us. Even the ‘shared reality’ is open to our (linguistic) inquiry only to the extent that it is contained in the negotiations of the discourse community. We will never find out if there are ‘really’ trees outside of the discourse, and if on these trees there are ‘really’ apples. As linguists, all we can find out is whether the participants of a primordial speech situation agree on their existence. But that is enough. If Charles tells Julianne that she is to pick an apple from a tree in his little orchard, and if she tells him later how delicious it tasted, and if we learn about this dialogue from a text, then we have acquired another smidgeon of the kind that constitutes the largest part of what we call knowledge, namely knowledge extracted from the discourse, not based on personal experience. What personal experience might add to this knowledge is hardly relevant. We believe that the apple Julianne has tasted will be similar to the fruit that other people have taught us to call apples. Neither Charles nor the investigative linguist will ever find out what the word delicious means in Julianne’s private language, and indeed it hardly matters to us, or even to her. It is part of her first-person experience, and as such it is as intractable for her as it is for us. Julianne has no way of ascertaining that what she calls delicious today is really the same as what she called delicious a year ago. All she knows is that it is an adjective that can be applied to apples, and that people prefer delicious apples to those that are not. Google lists 150,000 occurrences of ‘delicious apples’. We do not need first-person experiences of the reality constructed in discourses. If we are able to negotiate competently the meaning of texts or text segments we encounter with other members of the discourse community, then we have done all that people will expect us to do. What is contained in our memories, in our first-person experience of the world, in our powers of reasoning and in the pragmatics of language use, had to get there somehow, and the most plausible assumption is that it got there by being voiced by members of the discourse community, that it has become our share of the discourse as the collective mind of which we are a part. As an illustration, let us take the key example for dealing with the pragmatics component of linguistics, as we find it right at the beginning of Jenny Thomas’s introduction to pragmatics, Meaning and Interaction (Thomas 1995: 1): It’s hot in here.
Conventional linguistic wisdom has it that if the speaker of this utterance is a person of authority in a room with windows shut in which the air has become hot and sticky, and if the audience consists of people in awe of the speaker, then
170
Meaning, Discourse and Society
the speaker probably intends with this utterance that one of the hearers open the window. This looks like a prototypical primordial speech situation. The intentions of the speaker can be gleaned by a perusal of the segment of reality in reach of this group of people (hot air, windows closed, no air-conditioning), by their memories of situations comparable to the one given, and by appropriate inferences on the part of the hearers. Often when we talk about pragmatics, we talk about situations like these in which apparently the speaker’s intentions are not clearly expressed but hidden behind a rather inconspicuous remark. To reveal the ‘true’ meaning of this sentence, it would take a Sperber and Wilson relevance approach or an integrationist approach as suggested by Roy Harris. Or would it? Instead of carrying out an integrationist analysis of the utterance and the situation in which it occurs, we can try to query the discourse for the meaning of this utterance. For the speaker did not come up with this utterance on a tabula rasa, a blank slate. He or she was making a contribution to a discourse that had been going on for thousands of years, in plenty of languages all over the globe. What he or she said can easily be interpreted in terms of this background. The utterance has been made countless times before. On 27 November 2006, Google finds 58,800 occurrences of ‘It’s hot in here’. If we add to our query the words window and open, we still find 10,800 hits. That tells us that in 20 per cent of all cases when someone says ‘it’s hot in here’, there is also talk about opening a window. There is even a web page (http://openthewindow.blogspot.com/) which is labelled ‘open the window … because it’s hot in here’. Typical are text segments like this one, ‘It’s hot in here. It’s unbearably stuffy. Please open the window!’, found on this web page: www.jewishworldreview.com/ hillel/goldberg_2004_03_25.php3. The example shows that once we restrict the investigation of meaning to la parole écrite as the only means of symbolic representation, there is indeed an alternative to a pragmatic or integrationalist approach. The (written) discourse can tell us all we need to know about the possible implications of a sentence such as ‘It’s hot in here’. The discourse opens an escape route from the realisation that it is, at least for the linguist, impossible to determine all the relevant conditions, such as communicational setting, memory, analogy, comparability, the ways in which we reason and draw inferences. For it is the discourse that has taught us the relevance of all these conditions. I do not wish to imply that it is impossible to intuitively open the window when someone complains about the temperature of a room, without ever having encountered this link in the discourse before. But that would be beginner’s luck. It is the discourse that tells us what a discourse community accepts as rational behaviour, not so much in abstract terms but in an infinity of examples. It is these examples that we find in the discourse which inform our conceptualisation of rational behaviour, not the other way around. In the discourse, and nowhere else, linguists as well as all other people can find out what texts, or segments of texts mean.
13
Meaning, knowledge and the construction of reality
Truth and meaning It is my contention that discourse analysts can easily do without the discourseexternal reality. The external reality and the reality constructed in the discourse are only related to each other indirectly, through primordial speech situations, largely outside of our investigative reach. It is true that if we read the sentence ‘There are many apples on the tree’, we can never be sure this sentence refers to a shared reality in which there were actually ‘real’ apples hanging on a ‘real’ tree. But to me it seems more important that the discourse tells us everything people have said about apples being on trees. Our ‘shared reality’ is not the reality out there. What we see there is not just stuff or matter; it is, to a large extent, an assembly of conceptualisations of concrete and abstract objects, properties, states, processes and actions that owe their existence to foregoing negotiations forming the residue of our memories. Without people discussing them, there would be neither apples nor pears. It is this realisation that makes me so suspicious of what is commonly called truth. If there is no link between the non-symbolic reality outside of the discourse, and the symbolic content of the discourse, how could truth be an issue? What we are concerned with is not whether Julianne’s apple was truly delicious, but what it means if she calls it delicious. For me, the choice between truth and meaning is easy. I will never find out if there are really angels out there. What counts for me is what is ‘known’ about them, that is what people have said about them. How do we know that angels have wings? Is it because we have seen them in paintings, stained-glass windows and sculptures with their wings? But how did the first artists who painted angels know that they had to paint them with wings? They were led to assume that they must have wings, just as they knew other beings able to fly are equipped with them. It is in Daniel 9, 21 where we find the textual source telling us that angels indeed can fly: While I was still speaking in prayer, the man Gabriel, whom I saw in the vision at first, approached me in swift flight about the time of the evening offering.
All conceptualisations of the visible (or audible or tangible) world are textual. We could not tell an apple from a pear if the difference between the two concepts 171
172
Meaning, Discourse and Society
had not, at some stage, been pointed out to us. We would not know what is depicted on a painting if we had not learnt about these objects in the contributions to our discourse. There are cultures lacking the concept of winged angels, and there are languages like Italian which force us to distinguish ble from azurro. For people whose native language is English, ble and azurro are both blue. But there is no way to find out if colours are more than discourse constructs. Colour adjectives are useful for demonstrating that our ‘shared reality’ is not the reality outside of the discourse. Would there still be colours if no one was interested in them? Apparently there is no straightforward answer to this seemingly simple question. For Alan Dix, professor at Lancaster University, pink, for instance, is not a ‘real’ colour: One way to define colour would be via wavelength of light. A pure light of a given wavelength (like the strong sodium yellow) is a colour. This would include all the colours of the spectrum, rainbow or standard colour wheels … However, this would then not include ‘unsaturated’ colours like pink. Pink is pure red mixed with white – there is no ‘pink’ wavelength. This ‘spectrum’ definition of colour is called ‘hue’ and is perhaps the closest we can come to a clear idea of colour at a physical level. (www.hcibook.com/alan)
I do not know if it makes sense to say that there are colours in the reality outside of the discourse. For me, it does not matter. Colours are something talked about and discussed in discourses, even by people who have not actually seen them. By and large, we experience only what is already conceptualised in the discourse. Blind people may well have mental concepts of colours, not innately, but owed to the discourse they have been exposed to. Our whole reality is the reality constructed by the discourse. A non-mediated glimpse of a discourseexternal reality leaves us clueless. We find it almost impossible to make sense of an image that has not been negotiated as shared reality, as for instance with the puzzle picture shown in Figure 1. However, we are all primed to believe that the ‘shared reality’ is really there outside of the discourse. We are convinced that our first-person experiences must be authentic and must mirror reality as it really is. Yet to see a chair as a chair, to categorise an apple as an apple, to find evidence for globalisation or to distinguish a public uprising from a riot would not be possible if these objects had not first been negotiated within the discourse. Our ‘shared reality’, and everything we find in it, has been collaboratively constructed. We can only see angels if they have already been introduced into the discourse. Whether there is a discourse-external reality in which we would find winged angels is of no importance. One reason why we believe that what we see is the authentic reality, is that the chairs, apple trees and angels that populate it have been around in the discourse for a long time before we had a chance to encounter them ourselves. When we finally meet up with them in a primordial speech situation, we are being
Meaning, knowledge and the construction of reality
173
Figure 1. A puzzle picture
introduced to them not as objects of a ‘shared reality’ but of reality itself. Indeed we tend to accept many objects as given that we have never actually encountered in primordial speech situations. We accept other people’s assertions as sufficient reason to believe in their ‘reality’, particularly if attested by people we trust, such as friends, experts or priests. Few of us have actually ever seen a platypus, and even fewer have doubts about their existence. But when the first specimen was presented to an incredulous audience, many naturalists believed it to be fake. The conviction that there is a reality out there with real objects including platypuses seems to be pervasive. John Searle has absolutely no doubt that there are ‘mind-independent phenomena’ such as ‘hydrogen atoms, tectonic plates, viruses, trees and galaxies’ (Searle 1998: 13–14). But no one has ever seen an atom with their naked eyes. Ernest Rutherford’s atoms make a plausible narrative, because they are backed up by similar stories about analogous solar systems. Other physicists, however, like Jim Al-Khalili, believe that the quantum account of subatomic reality provided by the famous Copenhagen Interpretation should demand more attention. Its disadvantage is that ‘[i]t is difficult to translate what is essentially advanced mathematics into words that make sense’ (Al-Khalili 2007). I am not convinced that mathematics is better equipped to represent the reality out there than language. The mathematicians’ formulae are subject to negotiation as much as our use of language. Only if these formulae were accurate representations of reality, could they be called mind independent. But I am not so sure if we really can argue that they were discovered rather than invented.
174
Meaning, Discourse and Society
Are the trees we can see with our naked eyes more mind independent? It is not reality that makes us call what grows there trees, shrubs or bushes. Where do we draw the borderline between them, and how ‘objective’ can it be? Searle’s stance is similar to that of the early Hilary Putnam, when he still referred to himself as a ‘metaphysical realist’ and when he maintained that the word ‘water’ means what water really is (Chapter 2). But Putnam later realised that such a position leads to an impasse. In his book Representation and Reality, first published in 1991, he sees it off with these words: I have described metaphysical realism as a bundle of intimately associated philosophical ideas about truth: the idea that truth is a matter of Correspondence, and that it exhibits Independence (of what humans do or could find out), Bivalence, and Uniqueness (there cannot be more than one complete and true description of reality) … What I used to find seductive about metaphysical realism is the idea that the way to solve philosophical problems is to construct a better scientific picture of the world. (Putnam 1998: 108, emphasis in original)
He calls the new paradigm he subsequently develops ‘internal realism’: The internal realist suggestion is quite different. The suggestion … is that what is (by commonsense standards) the same situation can be described in many different ways, depending on how we use the words … The suggestion I am making, in short, is that a statement is true of a situation just in case it would be correct to use the words of which the statement consists in that way in describing the situation. (p. 114f., emphasis in the original)
In Putnam’s new theory, it is the discourse community that decides matters of correctness: Meanings are not objects in a museum, to which words somehow get attached; to say that two words have ‘the same meaning’ … is just to say that it is good interpretive praxis to equate their meanings … But sophisticated interpretative praxis presupposes a sophisticated understanding of the way words are used by the community whose words one is interpreting. (p. 119)
This is implicitly a rejection of the idea that the objects and properties existing out there are mind independent. It is up to the language community to rule if there is just one property ‘blue’, as in English, or two distinct properties ‘azurro’ and ‘ble’, as in Italian. The old paradigm of metaphysical realism had suggested that: there is a single system … which contains all the objects that anyone could refer to … But, from my ‘internal realist’ perspective at least, there is no such totality as All The Objects There Are, inside or outside science … What looked like an innocent formulation of the problem – ‘Here are the objects to be referred to. Here are the speakers using words. How can we describe the relation between the speakers and the objects?’ – becomes far from innocent … From an internal realist position, the very problem is nonsensical. (p. 120)
Meaning, knowledge and the construction of reality
175
With his ‘internal realism’ Putnam has offended large parts of the analytic philosophy community. This is how Thomas Nagel, one of its American luminaries, describes it: A recent example of subjectivism, usually presented as a way to transcend the outmoded subjective–objective distinction, is the view known as ‘internal realism’, according to which our apparently objective world picture should be understood as essentially a creative product of our language and point of view, and the truth of our beliefs should be understood as their survival in an ideal development of that point of view. (Nagel 1997: 87)
For Nagel, as for Descartes, reason is objective and universal. What reason makes us believe is more than just a point of view; it tells us if a proposition is true or not. This is why, according to Nagel, internal realism has got it wrong: Internal realism fails its own test of rational acceptability … Even if we concluded, as some physicists do about the quantum theory, that the best systematic account of what we observe cannot be given a realistic interpretation, that would still be a belief about how the world is, period – not a belief that it would be correct to qualify with an ‘internalist’ reading. Reason is used to arrive at it, and in the reasoning is not merely a development of our point of view, but objective thought about how things are. (p. 88)
As Nagel notes gleefully, Putman has since partially retracted his new position, quoting him as saying: ‘Whether I am still, to some extent an ‘internal realist’ is, I guess, as unclear as how much I was including under that unhappy label’ (p. 89). Outside of the arcane circles of analytic philosophy, however, the idea that what we call truth cannot be seen independently of the conventions of a discourse community is quickly gaining ground. Again we have to admit that the idea is not new at all. It was already the position of Charles Peirce, the founder of American pragmatism, who maintained that it is the community that defines what is true and what is real: The opinion which is fated to be ultimately agreed to by all who investigate, is what we mean by the truth, and the object represented in this opinion is the real. (Peirce 1878: 5.406–7) The real, then, is that which, sooner or later, information and reasoning would finally result in, and which is therefore independent of the vagaries of me and you. Thus, the very origin of the conception of reality shows that this conception essentially involves the notion of a COMMUNITY, without definite limits, and capable of an indefinite increase of knowledge. (Peirce 1878: 5.311)
This is the ‘shared reality’ of the primordial speech situations of ‘all who investigate’, of a ‘COMMUNITY’.
176
Meaning, Discourse and Society
Meaning and knowledge Our modern society is often called a knowledge society. But what is knowledge? Where do we find it? What does it mean to say that we know something? Obviously there is a tradition relating knowledge to truth. When I say ‘Peter knows that he has seen an angel’, both Peter and I must be convinced that there are ‘really’ angels. Knowledge is at the centre of modern epistemology founded by René Descartes. For him, the only knowledge (scientia) was true knowledge. Propositions can be true or false. Sensory data can be misleading and have to be mistrusted. Therefore we have to be sceptical of propositions that rely on our perception. However, there are a-priori truths, such as the propositions of Euclidean geometry. It is reason that ultimately has to decide if a proposition is true. Reason is a faculty of the mind. Therefore, by applying reason, individual minds can recognise, at least for those propositions not involving sensory data, whether they are indeed true. According to Richard Rorty, Descartes’ insistence on the monadic mind as the sole engine of reason has had disastrous consequences for Western philosophy. In his view, there are three properties that define the Cartesian mind: first, there is a ‘glassy’ substance containing representations that, accurately or not, mirror nature; second, some of these representations are a priori and therefore can be immediately known; and third, only mental contents can be known with certainty. Just as it is seen in metaphysical realism, knowledge is only a representation of reality if it is true. This is the view that Rorty rejects. For Rorty, following Peirce, knowledge always has a linguistic and a social component. Not the monadic mind, but the discourse community, is the judge of what counts as knowledge. Once we shed the notion of knowledge as ‘the mirror of Nature’, we can see it ‘as a matter of conversation and of social practice’ (Rorty 1979: 170f.). His critic Alvin Goldman, a rather traditional (social) epistemologist, is somewhat unhappy to throw overboard the Cartesian concept of knowledge as situated only in the monadic mind. He insists: ‘Rorty is wrong in saying knowledge is a “matter of what society lets us say.” … [I]f a proposition is true and a person has an appropriately formed belief in it, then he does know, no matter what his peers think’ (Goldman 1981: 427, emphasis in the original). His view is supported by Philip Kitcher, another social epistemologist who is keen on reconciling a positivistic belief in the facticity of science with a social perspective. For him there can be no doubt: in the end it can only be the monadic mind that knows that it knows: ‘Individuals are the primary subjects of knowledge. To ascribe knowledge to a community is to make an assertion about the epistemic states of members of the community’ (Kitcher 1994: 113). I am content to side with Rorty’s knowledge as social practice and with Putnam’s internal realism. Knowledge in Goldman’s understanding has to be
Meaning, knowledge and the construction of reality
177
private: it is a proposition that is known by one person to be true, even if all other members of the discourse community might disagree. The consequence is that knowledge, as ‘true’ knowledge, could not be exchanged or shared. There would be no point in talking about it. My suggestion is therefore to distinguish between private and public knowledge. Private knowledge is knowledge no one else knows about. It is the knowledge based on my first-person experiences, which I take to be authentic. I may always remain convinced that there is balllightning because I have an ‘authentic’ experience of a ball lightning event. However, once I give testimony of this knowledge, it becomes subject to negotiation by the discourse community. Once I have made my knowledge public, it ceases, for those who share it, to be proven ‘true’, except for the speaker. I have to admit, though, that shedding the notion that knowledge must be true would contravene common linguistic usage. As I said above, the way we use the verb know indicates that the person using it has to believe that what is said is true. On the part of the addressees, however, there is no such requirement. For them, it is a piece of information, just like a piece of news they read in a paper. To use the verb know is to appeal to the hearer to accept a statement as true. It should be noted, though, that the artificial intelligence community has a tradition of equating information with knowledge, and is not interested whether it is ‘true’ or not. What used to be called in early days ‘information retrieval’, has later resurfaced as the more trendy ‘knowledge extraction’. Once we forego the claim to truth, we can also return to Putnam’s internal realism and equate, as he does, meaning and knowledge, thus throwing up the issue of reference. If, while still a metaphysical realist, he was concerned with reference that relates an expression to the real world out there, the reborn internal realist Putnam now claims that the very problem of reference is ‘nonsensical’. For him, too, it makes no sense to demand that knowledge has to be true. If we can never be sure that what we call knowledge refers to a discourseexternal reality, then it has to be the discourse that decides over knowledge. In that case, the knowledge expressed in a text (segment) becomes identical with the meaning of this text (segment). What distinguishes knowledge and meaning is but a difference in perspective. From a linguistic point of view, the meaning of a sentence (discounting the role of its grammatical structure) is contained in the meaning of its lexical items. From an epistemological perspective, a sentence mentions objects, properties, states, processes and actions (all of which I will henceforth refer to as ‘objects’) about all of which knowledge is available. It needs to be said once more that all these objects are objects of the discourse, not of the reality out there. They are represented by lexical items. My claim is that the knowledge we can have of a discourse object is co-extensive with the meaning of the lexical item that stands for it.
178
Meaning, Discourse and Society
Traditionally, word meaning and world meaning, or lexical meaning and encyclopaedic meaning, are kept apart. Lexicographers used to be advised, as Umberto Eco reminds us in the chapter ‘The platypus between Dictionary and encyclopaedia’ of his book Kant and the Platypus (1999), that the proper way to deal with words is by first pointing to the genus proprium (the hypernym, in modern linguistic terminology) and then by explicating the differentia specifica (what distinguishes it from its co-hyponyms). Thus the revised entry for pig in the Oxford English Dictionary reads: ‘An omnivorous, domesticated, even-toed ungulate derived from the wild boar Sus scrofa, with a stout body, sparse bristly hair, and a broad flat snout for rooting in the soil, kept as a source of bacon, ham, pork, etc.’ (online edition, draft revision June 2007). The genus proprium is ungulate, meaning ‘hoofed quadruped’. The pig differs from other hoofed quadrupeds in that it is, among other things, domesticated and even-toed. Traditionally, the differentiae specificae were seen to refer to essential, unalterable properties. A pig is a domesticated swine; a swine that is not domesticated would not be called a pig. Therefore we can say that ‘domesticated’ turns a swine into a pig, that it is a property essential for using the word pig correctly. But the pig is still a swine, domesticated or not. This is why in an encyclopaedia such as the Columbia Electronic Encyclopedia, not concerned with words but with things, there is, instead of an entry of its own, no entry for pig, only a reference from pig to swine, as the concept of ‘pig’ does not differ essentially but only accidentally from the concept of ‘swine’, namely by referring to the exemplars that happen to be domesticated. A swine, we read, is the ‘name for any of the cloven-hoofed mammals of the family Suidae, native to the Old World. A swine has a rather long, mobile snout, a heavy, relatively short-legged body, a thick, bristly hide, and a small tail. The name swine is applied mainly to domestic animals, which are also known as hogs. Sometimes these are called pigs, a term which in the United States is more correctly reserved for the young animals’. The difference, I hope, becomes quite clear. A dictionary entry tells us what we have to know about the lexical item pig, to distinguish it from other lexical items. The encyclopaedia entry presents knowledge concerning the discourse object. The link from pig to swine tells us that both lexical items stand for the same object. But there the clarity vanishes. We find lexical information in the encyclopaedia (‘Sometimes these are called pigs, a term which in the United States is more correctly reserved for the young animals’), and we find encyclopaedic knowledge in the dictionary definitions (‘kept as a source of bacon, ham, pork’). In practice, it seems, there is no clear borderline between lexical and encyclopaedic knowledge. The difference makes sense only as long as we distinguish between the discourse and a discourse-external reality. Such a distinction, as I have shown, makes little sense. All our knowledge is firmly situated inside
Meaning, knowledge and the construction of reality
179
the discourse. Otherwise it could not be public and could not be shared. All we find in the world unmediated by the discourse is shapeless stuff. It takes innumerable primordial speech situations to turn this stuff into distinguishable objects (including people, properties, states, acts and processes) of a shared reality. But there is no clear-cut line between such objects and objects which have no place in this shared reality, and only can be distinguished by the different contexts in which they occur. William of Ockham was convinced that in the lingua mentis we only find concepts referring to distinct entities, unlike in natural languages in which there are many synonymous expressions. For Ockham, these objects, actions, processes, etc. are distinct entities of the real world outside the discourse (and outside the individual mind). But it makes little sense to ask ourselves if the verbs flout, jeer, scoff, gibe, and barrack refer to the same activity or to five different activities. The discourse will tell us to what extent we find similarities and differences between the contexts in which these verbs occur. Observing people involved in such an activity will not tell us which verb we should use to describe it. Both knowledge and meaning are discourse internal. When someone points to an object in our visual range and tells me that this is a tree, this tree becomes an object of our shared reality, and an extension of the discourse. ‘All the pointings to, demonstrations of, and descriptions of brute reality are inevitably semiotically mediated and communicated,’ as Derek Edwards, Malcolm Ashmore and Jonathan Potter say in their noteworthy contribution to the social construction of reality ‘Death and Furniture: The rhetoric, politics and theology of bottom line arguments against relativism’ (Edwards, Ashmore and Potter 1995: 27). There are no things per se. ‘Objectivists therefore need a device for introducing reality, and having it stand as a per se refutation of relativism. They need to render the acts of construction, of categorization, and of rhetoric invisible’ (p. 37). They do not have to work hard. It goes against our intuition to reject as a figment of the discourse all that we experience as authentic reality, the stuff which constitutes our intentionality, our aboutness, and thus our identity. It feels like the bottom has dropped out of this world. Why should we take a claim seriously that tells us that the things we encounter are just objects of our discourse and not of the world out there, and that the knowledge we have of them is no more than an arbitrary arrangement? For me, to accept that the only reality we ever deal with is a reality we have collaboratively constructed in our discourse empowers us to reconstruct it whenever we think we would benefit from such a reconstruction. It delivers us from the reality that we are fed by those in charge. It is, in Immanuel Kant’s famous words, ‘man’s emergence from his self-imposed immaturity’. But while Kant insists that it is up to the monadic mind to substitute one’s own understanding for the reality dictated by authority, we have to accept the fact that the construction of reality is a collaborative, not an individual act.
180
Meaning, Discourse and Society
The collaborative construction of reality It seems to be impossible to escape the notion that it is we, the members of a discourse community, who negotiate and determine the way in which we cut up into more or less discrete entities the spatial, temporal and motional stuff of the world out there, for our own purposes. All these negotiations involve categorisations that refer to a socially constructed reality, one in which we can jeer or flout or gibe at the ways in which people flaunt their property. It is up to us whether we distinguish swine and pigs. Pigs and swine do not seem to care. For me, the meaning of a lexical item and the knowledge of the discourse object represented by this lexical item are co-extensive, are the same. The word pig in a discourse means everything that has ever been said about pigs in that discourse. The same is true for the word swine. Everything that has been said about pigs and swine is also everything we can know so far about pigs and swine. Of course, much of what has been said is lost. What pig and swine mean for us and what we know about them today is what is still accessible, what has been recorded, what has been written down, and what has not been lost in the meantime. I call the concepts of all concrete and abstract things, all properties, all states, actions and processes that are talked about in a discourse, the objects of this discourse, or discourse objects. In what sense these are mental concepts, concepts of an individual mind, is, as I have shown in the previous chapter, something we cannot know. What we can see, however, is that they are discourse objects, i.e. concepts of the collective mind constituted by a discourse. What a concept is in a philosophical context is something I will not discuss here. For my purposes, there is no need to distinguish concepts from discourse objects. In an everyday discourse of primordial speech situations, we can expect a word like chair to have been applied, by acts of explicit or implicit ostentation, to concrete chairs in the environments of these speech situations. The word chair will also be used to talk about hypothetical chairs, chairs that some, but not all group members have seen, or possible chairs a member wants to buy. There may even be a discussion on the nature of chairs and what distinguishes them from, say, barstools. All occurrences of the word chair (and anaphora referring to chair) will contribute to the meaning of the word, and to the knowledge about the discourse object it represents. In the world outside there are no chairs; there is only matter, stuff. When I wake up with the urgent desire to buy a new chair, then my concept of a chair is based on everything I have heard or read about them, accompanying deictic acts included. It is the same if I wake up with the desire to do justice, except that there are no accompanying deictic acts. I will not have heard or read everything that has been said about chairs or justice in a discourse. Other members of my discourse community will have heard some of the very same utterances that I
Meaning, knowledge and the construction of reality
181
have heard, but also others. Our knowledge of chairs and justice will partially overlap. When we engage in a conversation about chairs or justice, we will permute and reformulate what we have heard, perhaps with a common interest in reaching an agreement on the nature of chairs or justice. In any case, our negotiations, as permutations and reformulations, will add new paraphrases. Once we have concluded our conversation, the meaning of the word chair, or justice, will have been amended, modified, by what we have said. We could also say that something has been added to the knowledge about chairs or justice in the discourse in which our conversation took place. The meanings of lexical items, or the knowledge about the discourse objects they represent, are collaboratively constructed in the contributions made by the members of a discourse community. This knowledge is not a mirror of the reality out there; rather, what we perceive as the reality out there is just a mirror of the knowledge gathered in the discourse. The only reality directly available to us is the reality collectively constructed in the discourse. In the paragraph above I insisted that the construction of meaning is not so much a solitary as a collaborative act. This is a view shared by many, and is by no means new. It is at the forefront of contemporary social psychology. Gavriel Salomon asked the question in the introduction to the book Distributed Cognitions: Psychological and Educational Consideration: If social and perhaps other ‘external’ processes are to be taken as integral parts of the cognitive process, maybe the whole concept of cognitions ought to be re-examined. Are they perhaps distributed rather than located in the head? (Salomon 1993: xiv)
In their book Collaborative Cognition, David Bearison and Brice Dorval stress the pivotal role played by Jean Piaget and Lev Vygotsky. They also share the view concerning children’s cognitive development that ‘knowledge is not necessarily something that individuals possess or that involves inside the head but rather something that individuals do together such that their social processes become intrinsic to their mental operations’ (Berison and Dorval 2002: 1). But while they mention fleetingly ‘the emergence of collaborative cognition’ (p. 19), the focus of their study is still primarily to gain insight in to the workings of the (individual) mind to the extent it is ‘socially contexted’ (p. 1). I do not doubt that individuals possess intentionality. They can make themselves conscious of what they know and of the fact that they are the authors of their contributions to the discourse. But whatever there is in their minds (if it indeed makes sense to hypostatise minds) is so intricately connected to what is outside that we should perhaps not focus so much on it. François Cooren, too, asks us to rethink the concept of the monadic mind: Where is the mind situated? At first sight, this question appears to be easily disqualified if we refer to that well-established line of thought that goes from Descartes (1960) to Chomsky (1990) and Fodor (1990) through Locke (1959) and Hume (1927). For these
182
Meaning, Discourse and Society
earlier thinkers, and most of today’s cognitivists, the answer appears obvious: ‘In the head, of course!’ The concept of distributed cognition calls into question this ‘obvious fact’ by showing that cognition can also be interpreted as shared and distributed among participants. (Cooren 2004: 520)
This is also Jerome Bruner’s view: [A] ‘person’s’ knowledge is not just in one’s own head, in ‘person solo,’ but in the notes one has put into accessible notebooks, in the books with underlined passages on one’s shelves, in the handbooks one has learned how to consult, in the information sources one has hitched up to the computer, in the friends one can call up to get a reference or a ‘steer’, and so on almost endlessly. (Bruner 1990: 106)
Whoever talks about knowledge as the result of some collaborative symbolic action must have given up Descartes’ concept of knowledge as episteme, as something that we know, through our individual power of reason, to be true. The only knowledge worth talking about is public knowledge. Private knowledge residing in the minds of individuals becomes effective only if it is entered into the discourse. Whenever decisions have to be taken collectively, knowledge will be assembled and made public in a collaborative act. It will be discussed, involving the recombination, permutation and reformulation of knowledge pieces. Such a discussion will generate a previously unavailable level of awareness going well beyond the sum of the original knowledge bits. Today, researchers like Edwin Hutchins are exploring the ways in which people with distributed skills and knowledge organise themselves to take decisions collectively, when it comes to navigating the Pacific Ocean, flying large passenger planes or structuring boardroom meetings. In his book Cognition in the Wild (Hutchins 1995) he demonstrates that in such situations we see the emergence of genuinely new content. The construction of knowledge is a collaborative activity which can be observed and described not by modelling individual minds but by investigating what is happening between people, by analysing the discourse. One of the foundational texts of modern social constructivism is Peter Berger and Thomas Luckmann’s book The Social Construction of Reality: A Treatise in the Sociology of Knowledge (Berger and Luckmann 1966). This book hijacks the label ‘sociology of knowledge’ furnished, decades ago, by Max Scheler and especially by Karl Mannheim, for what was then a truly sociological theory aimed at an understanding of knowledge as being socially constructed. The classical sociology of knowledge emerged in the 1920s as an analysis of the relationship obtaining between social processes and structures and intellectual life, including its modes of knowing. It investigates the contingencies of intellectual structures, or forms of knowledge, in different social and historical settings. Mannheim’s focus was on ‘ideology’, and he did not extend his notion of knowledge being socially constructed to the natural
Meaning, knowledge and the construction of reality
183
sciences, which he thought well covered by Cartesian epistemology. For Berger and Luckmann, however, the construction of reality has to be investigated, not in Mannheim’s symbolic transactions of a society, but in the experiences and mental activities of individuals. This focus on the individual mind may account for the immense popularity their book still enjoys. We read: The reality of everyday life is organized around the ‘here’ of my body and the ‘now’ of my present. This ‘here’ and ‘now’ is the focus of my attention to the reality of everyday life … The reality of everyday life presents itself to me … as a world I share with others … I also know, of course, that the others have a perspective on this world that is not identical with mine. My ‘here’ is their ‘there’. (Berger and Luckmann 1966: 36 f.)
Thus Berger and Luckmann import Mannheim’s project into Husserl’s phenomenological programme, the result being a new amalgam, focusing again on a monadic (as opposed to collaborative) interpretation or ‘understanding understanding’, as Ronald Hitzler, Jo Reichertz and Norbert Schröer call it in their introductory article ‘Das Arbeitsfeld einer hermeneutischen Wissenssoziologie’ [‘The agenda of a hermeneutical sociology of knowledge’] (Hitzler, Reichertz and Schröer 1999: 10). Indeed, one has to agree with Jonathan Potter, who comments that Berger and Luckmann’s approach ‘tends to obscure the interactional and rhetorical nature of fact construction, while reifying a mental world which itself [is] a major element in factual discourse’ (Potter 1996: 13). Throughout their book, Berger and Luckmann resist the temptation to explore the perspective of the discourse as the collective mind. Husserl’s phenomenological approach to hermeneutics is, it seems, rather closely connected with the cognitive agenda and its impasse, and this is certainly the main reason why Hans-Georg Gadamer was led to reject it when he developed his version of hermeneutics, which is largely unconcerned with the author’s mind. To us, having grown up in a world in which we extol the individual, and the uniqueness of their artistic and intellectual creations, a perspective that focuses solely on the discourse shared by a community seems counterintuitive. We almost intuitively believe that it is only the individual mind which generates symbolic content. The idea that cognition can only be explored as a social phenomenon is hardly plausible. While everyone agrees that there is an interrelationship between the discourse and the minds of the members of a discourse community, most researchers like Vygotsky, Piaget and even Bruner are first of all concerned with the impact of symbolic interaction on the individual mind, and not with the self-referentiality of the interaction. My perspective is different. I can only study the emergence of knowledge or meaning in the discourse. The question is, however, whether we should look at the discourse as a process, as something taking place without agency, or whether we describe it as an activity, as something being effected by intentionality. How we describe it depends on our perspective. We can see the
184
Meaning, Discourse and Society
discourse as a result of the actions people take, and equally we can see people’s behaviour, including their symbolic behaviour, as being caused by the discourse. How we decide between these two options depends on our goal. In my case, my goal is to explain how meaning or knowledge is constructed. The only way for people to let us know how they experience their reality, and how we are to see their actions and interactions, is for them to tell us. The only way we can find out the realities people construct for themselves is by listening to them. The inaccessibility of people’s minds may be a cogent reason by itself to focus on the discourse. Yet I believe it is equally important to realise that we construct our personhood (and that of the people around us) in a never-ending dialogue with others. Without such a dialogue, the monadic mind would indeed be a slate void of any symbolic content, void of intentionality. We create our own self-awareness by interacting with others. Without communication, we are reduced to the status of automata reacting to stimuli. Thus Daniel Defoe found himself forced to supply Robinson with his Friday, when the diary device (in itself a quasi one-person dialogue, written for us rather than for himself) had run dry. Robinson is only able to construct his self as a reality interpreting his experiences and activities because he, in his sudden isolation, carries on with his contributions to the discourse in the way he has done before, as if nothing had happened. The reality he constructed on his island was, in many respects, the mirror image of the reality he had brought with him from his past. Meaning or knowledge is the essential ingredient of a discourse. Meaning alone enables us to interpret what is going on in our reality, a reality that has been and is continually constructed by our never-ending negotiations of what is happening around us. Even if we had access to the mental representations of individuals, we would find there little more than very partial, distorted and incomplete copies, bare reflections, of the symbolic representations we find in the discourse. The social constructionism developed by social psychologist Kenneth Gergen and his community, which focuses on the discourse as our collective mind, instead of on the individual mind, seems to me an important achievement. From Gergen’s perspective, the traditional concept of epistemic truth, the knowledge known to be true, has become obsolete. ‘[T]he very concepts of truth and falsity as traditionally conceived are themselves so flawed as to be irrelevant’ (Gergen 2001: 429). This does not mean, though, that ‘“anything goes”, because little goes except within a tradition of social collaboration. In this sense, what constructionism does invite is a reflexive posture toward the often blinding force of tradition’ (Gergen 2001: 423). This is a position that we find expressed in a very similar way in Hans-Georg Gadamer’s attitude towards what he calls ‘prejudices’ in his Truth and Method, to which I will return below. It takes a
Meaning, knowledge and the construction of reality
185
lot of negotiation and reflection to convince a discourse community to throw cherished traditions or prejudices overboard. Social constructionism is neither solipsistic nor relativistic. Gergen’s ‘realism is essentially situated … [T]here is a reliance on local conventions of sense making (“what we mean together by these words here and now”)’ (Gergen 2001: 424). Once we leave behind the primordial speech situation with its ‘shared reality’, once we begin discussing more abstract ideas, our propositional statements become, by necessity, more ‘free-floating’. There we have no possibility for ostensive definitions. We may point to an object and say ‘That is a chair’, but we cannot point to a cognition or an attitude … We may point to a given facial expression, but we do not say ‘that is fear’; rather we say it is an ‘expression’ of an event that is essentially ‘off stage’ … In this vein I have tried to test the limits of the possibility that the rules of reference are so loose that virtually any proposition about the mind can be supported by virtually any behavioral data. (Gergen 2001: 421)
However, the key reason for choosing this article (out of the fast-growing mountain of publications in this area) to endorse my case for social constructionism, is its reasoned rejection of our Western tradition of argumentation. This is, I believe, one of the more serious (and so far hardly acknowledged) consequences of adopting the view that our reality is a contingent construct of the discourse. Once we accept that truth is no more than the social acceptability of a proposition, we begin to see why the logic of an argument is never enough to convince us. A healthy dialogue is not one in which one view is accepted as truth, but one in which many voices clamour for their often incompatible positions. A univocal dialogue only repeating what has been said is bound to be sterile and stagnant. A plurivocal discourse, on the other hand, can give rise to new permutations, recombinations and reformulations, and can therefore make new ideas possible. In a discourse in which all people shared the credo of social constructionism, this paradigm would soon be dead. The progress of a discourse community depends not on the inescapable ‘logical’ force of arguments, not on an algorithmic process in which a new argument is based on its predecessors, but on the unpredictability of the results to which clashes of opinion lead. New paradigms do not replace old ones because no new arguments proving the old paradigm right can be found, but because a new narrative, a narrative we can view as the result of random mutations of the old one, enters the discourse as a competing voice, and that, for reasons largely lost on us, this new narrative is so successful that it leaves traces in subsequent texts. Gergen tells us that he can ‘no longer find the tradition of argumentation a viable one’, and he asks ‘[a]re there other forms of interchange that might enable us to build forward from common visions …? I am in sympathy with many other scholars confronting the limits of abstract theorizing, and believe that major efforts must now be devoted to enriching the forms of social practice’ (Gergen 2001: 431).
186
Meaning, Discourse and Society
I hope I have succeeded in showing that there is already an established tradition of viewing reality as something that emerges from our negotiations within the discourse, in other words as a social construct. This social reality hinges on the meanings that we have collaboratively assigned and are continually assigning to the lexical items we use to refer to the objects we have been constructing and continue to construct. How does this happen? At some point, permutations and recombinations of things previously said will generate something that is subsequently accepted as a new idea. The example I choose is the lexical item global warming, standing for the discourse object ‘global warming’, closely related to the lexical item climate change and the discourse object ‘climate change’. Both concepts, global warming and climate change, are relatively recent. One of the first occurrences of global warming is found in the New York Times of 3 August 1982: Until there is indubitable proof of a global warming caused by CO2, the greenhouse effect must remain a hypothesis. A few months later, on 21 October 1983, we find, in the same paper, this excerpt from a scientific report: … The dominant sources of uncertainty stem from our inability to predict future … developments that will determine the global demand for energy and the attractiveness of fossil fuels … If we project increases in all these gases, climate changes can be expected significantly earlier than if we consider CO2 alone. From climate model simulations of increased CO2 we conclude with considerable confidence that there would be global mean temperature increase. With much less confidence we infer other more specific regional climate changes … Changes in temperature and rainfall may be amplified as changes in the annual discharge of rivers … If a global warming of about 3 or 4 degrees centigrade were to occur … it is likely that there would be a global sea-level rise … Such a warming might also bring about changes in Arctic ice cover, with perhaps … associated changes in high-latitude weather and climate … It is important to confirm the various predictions of climate changes … Rapid climate change will take its place among the numerous other changes that will influence the course of society, and these other changes may largely determine whether the climatic impacts of greenhouse gases are a serious problem … As a human experience, climate change is far from novel … Man-made emissions of greenhouse gases promise to impose a warming of unusual dimensions on a global climate that is already unusually warm … Viewed in terms
Meaning, knowledge and the construction of reality
187
of energy, global pollution and worldwide environmental damage, the; ‘CO2 problem’ appears intractable … Another two days later we read in the same paper a far less frightening account of global warming claiming to be a digest of this report: The greenhouse effect is the prediction that the earth will get hotter as carbon dioxide gas released from the combustion of coal builds up in the atmosphere and traps the sun’s heat like the glass of a greenhouse … Surely the seas may rise and the crop-growing belt may be pushed northward. But … crops could be genetically engineered to grow in changed climates. Indeed, all the predictable effects of a global warming can be coped with. The majority of people reading these articles might never have come across the newly coined phrase global warming before. The greenhouse effect, however, was so familiar that it needed no explanation. It remains unclear from our first citation whether global warming is offered as a synonym for greenhouse effect, whether the two discourse objects are interchangeable, or whether their relationship is more ambiguous. Obviously global warming was not yet a fixed phrase in 1983, i.e. a generally accepted lexical item; rather, it was one of many possible kinds of warmings. Therefore we find it preceded by the indirect article a, something not at all common now. Indeed, in the second text we find two occurrences of warming not preceded by global. We also find in this text five occurrences of global not linked to warming. In addition to two singular occurrences of climate change there are also two plural occurrences of climate changes, which, too, would sound rather strange today. Besides climate change we also find temperature increase; but change dominates, with altogether eleven occurrences. As the lexical item greenhouse effect (and the discourse object ‘greenhouse’) was common fare in 1983, it is somewhat strange to see it here explained and defined, something one normally does only with less familiar lexical items. The explanation given may be read as a shortcut, avoiding all the negative details we find in the report abstracted in this article. Again its relationship to global warming is left in the dark. Is ‘global warming’ something that we can cope with, while the ‘greenhouse effect’ was seen to inspire fear? The second text shows how new discourse objects are the contingent results of permutations and recombinations of elements that existed previously. In a greenhouse effect discourse in which climate, change, temperature, increase, global and warming are keywords, the aleatory generation of the item global warming was an event more or less bound to happen. A closer analysis might suggest why it replaced the greenhouse effect. For the year 2006, we find only 36 articles in the Guardian featuring the latter, while there are 1,026 articles talking about ‘global warming’. The evidence tells us that subsequent texts
188
Meaning, Discourse and Society
began using this new phrase, and the more it became the common label, the less it was felt to be an ad hoc adjective-cum-noun phrase. It did not take long before it became a lexical item in its own right. This was when it began to be used without an indefinite article. Initially it acquired its meaning from the context in which it was placed, as in the second text quoted above. Today, having come across so many occurrences embedded in contexts repeating themselves, we no longer need to be told what it means. Instead we can use it to explain the meaning of other lexical items. These are three random quotations taken from the Guardian: We have been slow to take up the challenge presented by global warming. (19 November 2007) Global warming will send Asia’s social and economic progress into reverse unless immediate action is taken to tackle climate change, according to a report today. (19 November 2007) All sources of carbon pollution – from flights to inefficient light bulbs – must become more expensive if the world is to tackle global warming, an influential panel of scientists and government officials will say today. (17 November 2007) Unlike the early texts, these three quotes do not tell us what ‘global warming’ or ‘climate change’ is. The two discourse objects, or the two lexical items, are used interchangeably. What we learn is that they constitute a challenge, and that this challenge must be tackled. Because we have learned the technical meaning of these items, we no longer need to be told why. What we learn now is that global warming, or climate change, is something we must fight. The meaning of a lexical item is everything that has been said about it in the discourse under investigation. To some extent, it is up to us, as the researchers, to define the discourses we investigate. Not all the texts that constitute a discourse will say the same things. Normally, discourses are plurivocal. There are, even in the liberal Guardian, articles sceptical about global warming as a ‘real’ phenomenon, while most texts take it for granted. Different articles will, in any case, have different things to say about global warming. But everything said, contradictory as it might be, adds to the meaning of the lexical item. Some explanations come up repeatedly, and therefore they may carry more weight than something said just once and never repeated. Some context patterns may be prominent at one time, and then fade out, to be replaced by new ones. New things will be said in tomorrow’s papers, adding new aspects to the existing meaning. As long as a discourse continues, meaning is always provisional. Other discourses (for instance the discourse of American newspapers) will have different things to say about global warming.
Meaning, knowledge and the construction of reality
189
Meaning, as I use the term, is only in the discourse, not in the head. Normally no one has come across everything that has ever been said about a lexical item. No two people will have come across the same accidental selection of texts belonging to a discourse. No one will remember correctly all they have learned. The meaning of a lexical item is not the sum of the mental representations of all members of a discourse community. Meaning, here, has nothing to do with people and their individual intentionalities. Meaning is what we find said concerning a lexical item, in the texts that make up a discourse. Meaning and knowledge are co-extensive. If we look once more at the New York Times text of 21 October 1983, we find that it explains the discourse object ‘global warming’, while it does not explicitly set out to define the meaning of the lexical item global warming. It provides us with knowledge about ‘global warming’, for instance by relating it to sea-levels, CO2 emissions, energy and pollution. It tells us what is changing, and that these changes have a global impact. It presents us with the full global warming scenario. If we add this information to the information extracted from all the other global warming articles in Englishlanguage newspapers, we have everything we need for an entry on ‘global warming’ in an encyclopaedia. As many people still believe, it is not at all clear whether such a thing as global warming exists outside of the discourse, in the world out there. In our Western world, science seems almost to have a monopoly on questions of reality, a monopoly it has to share only with economics and theology. This is why I will discuss in the next chapter the question of whether the language of science can assert a privileged access to reality.
14
The language of the scientific experimental report
Both science and engineering are about new knowledge. While engineering is about inventing a new gadget or a new process, science is about discovering something that was already there but was not known about before. What is invented and what is discovered are part of the real world out there (or so it seems). Some doubts might not be amiss, however. New inventions are said to serve a purpose or to have a function. A plough is there to plough. But who decides what should be called ploughing and what not? A new pill is certainly real, but how real is the depression which it is supposed to cure? Artefacts and inventions presuppose a discourse. The discourse makes us aware that there is something we are missing. It gives us the idea that something could be done about it. Even before these artefacts and these inventions are around, we turn these objects of the world out there into discourse objects. We talk about our ideas, explain them in relationship to other discourse objects, and discuss what we like and do not like about them. As discourse objects, they become part of our knowledge. One of the reasons why even chimpanzees are pretty poor at engineering is that they cannot discuss their needs and the possible options to meet them. Science is about facts. According to the Cobuild Dictionary (3rd edition 2001), facts are ‘pieces of information that can be discovered’. But information is already a representation of something, not the thing itself. According to the Random House Dictionary (1975), a fact is ‘the quality of existing or being real’. We are told that the first sense of real is ‘true, not merely ostensible or nominal’. But what is true? Truth is, again according to the Random House Dictionary (leaving out technical details etc.) ‘1. true or actual state of the matter. 2. conformity with fact or reality. 3. a verified or indisputable fact, proposition, principle. 4. state or character of being true.’ Reality is ‘1. the state or quality of being real. 2. resemblance to what is real. 3. a real thing or a fact.’ This is not only circular, but also full of loopholes. Are logarithms part of reality? Is the second law of thermodynamics real? Did it exist before it was discovered? Is this law the thing itself, or is it a textual representation of what is the case in the world outside the discourse? 190
The language of the scientific experimental report
191
Google lists (8 December 2007) 4,350 hits for ‘science is about facts’. One of them, taken from Theory and Practice of Nursing: An Integrated Approach to Caring Practice by Lynn Basford and Oliver Slevin (2003: 272) reads: Put in simple terms, science is about facts, the truth of which is established by agreed upon evidence in regard to observed phenomena.
Do scientific facts have to be endorsed by a consensus? Take the instance of global warming that I discussed in the previous chapter as an example of how new discourse objects are introduced into the discourse. Is it a scientific reality or not? ‘The debate is over. The science is in. It’s now time for action’, we are told by Californian governor Arnold Schwarzenegger (http://carboncoalitionoz.blogspot.com/2006/07/science-is-real-us-government.html). Not everyone agrees, though. Robert Carter, professor at James Cook University and member of the Institute for Public Affairs, funded by Woodside Petroleum, Esso Australia (a subsidiary of ExxonMobil), and over a dozen other companies in the energy industry, disagrees. Testifying before the US Senate Committee on Environment and Public Works, he said on 16 December 2006: The Governor is deluding himself, because the science of climate change has never been more uncertain. Furthermore, science is about facts, experiments and testing hypotheses, not consensus; and science is never ‘settled’ … As Margaret Thatcher famously observed [‘The Downing Street Years’, 1993: 167]: ‘Consensus is the process of abandoning all beliefs, principles, values and policies in search of something in which no one believes, but to which no one objects; the process of avoiding the very issues that have to be solved, merely because you cannot (otherwise) get agreement on the way ahead’. (http:// community.tri-cityherald.com/?q=node/436/15808)
In a lecture on ‘The myth of dangerous human-caused climate change’ (Carter 2007a) Carter also quotes the bestselling author Michael Crichton as his witness against scientific consensus: Let’s be clear: the work of science has nothing whatever to do with consensus. Consensus is the business of politics. Science, on the contrary, requires only one investigator who happens to be right, which means that he or she has results that are verifiable by reference to the real world. In science consensus is irrelevant. What is relevant is reproducible results. The greatest scientists in history are great precisely because they broke with the consensus. There is no such thing as consensus science. If it’s consensus, it isn’t science. If it’s science, it isn’t consensus. Period. (Crichton: 2003)
One could argue that Carter and Crichton are primarily opposed to a consensus they do not share. It seems that in less antagonistic contexts Carter is not in principle opposed to the usefulness of an agreement between scientists: ‘Reclassifications such as this should not be imposed by international bodies,
192
Meaning, Discourse and Society
but arrived at by consensus within the appropriate [scientific] user groups’ (Carter 2007b: 192). In spite of certain reservations concerning possible motives, I agree with Carter that the insistence on consensual views would be detrimental to any discourse community. Only plurivocality engenders innovation through recombination and variation of discourse elements that keep it from turning into a loop producing nothing but replicas of things said before. It is due to the plurivocality of the discourse that our new ploughs outperform those of the stone age. Plurivocality does not mean that anything goes. Scientific research is only successful if it is accepted by one’s peer group. It is accepted if subsequent contributions to the peer group’s discourse refer to a scientific publication by discussing it. The climate change discourse is a good example. Members of this discourse community agree and disagree in many ways. Yet no one has denied Robert Carter his voice. More important is Carter’s insistence that ‘science is about facts, experiments and testing hypotheses’ and Michael Crichton’s claim that results must be ‘verifiable by reference to the real world’. Both apparently agree that scientific knowledge mirrors a reality out there, and therefore is something other than meaning which is subject to collaborative interpretation. Scientific knowledge is contained in reports about experiments, carried out by a single researcher (or a small group of researchers) and is submitted to the peer group of the discipline in question. In his article ‘What written knowledge does’ Charles Bazerman looks at the conditions which have to be met so that the claims made are accepted as objectively and truthfully referring to reality (Bazerman 1981: 362f.). A successful scientific report has to: 1. use the accepted terminology of the discipline, thus professing that terms refer to language-independent concepts; 2. use ‘explicit citation and implicit knowledge [indicating] an article’s relationship to the previous literature on the subject’; 3. be aware of ‘the knowledge and the attitudes’ of the audience; and 4. construct authorship, i.e. ‘present a persona … which makes the reader aware of the author as an individual statement-maker’. [Today we frequently find multiple authorship, representing all relevant team members named individually.] This is not so different from other written texts. They all give us the illusion that they refer to a reality outside of the discourse; they all have to be interpreted on the basis that they are reactions to what has been said before; they all presuppose a language-independent shared reality and insist that the text is a piece of intellectual property. As far as these claims go, the scientific report cannot demonstrate that it refers to a discourse-external reality. Instead of verifying its truthfulness, we, the members of the peer community, only have to interpret it, which may or may not lead to a consensus among us. Nor does
The language of the scientific experimental report
193
terminology do the trick of transubstantiating meaning into reference to an external reality. The claim on which terminology is based, namely that ‘a tightness of fit between name and object indicate the quality of tie between text and the world’ (Bazerman 1981: 362) rings hollow, as long as such a link can only be textually (re)presented. So where does reality ‘really’ come in? The report analysed by Bazerman ‘describes a geometric model, elaborated in qualitative and quantitative terms, that is claimed to correspond to the structure of a substance (deoxyribose nucleic acid) found in nature. This act of geometric naming depends on the substance being discrete and robust and its structure being consistent through repeated observations’ (Bazerman 1981: 364). The ordinary objective of experimental reports is to establish a regular relationship between two states of affairs. In our case one of these states is what is professed to be a real-world object, a robust and discrete substance, and the other is what is claimed to be its language-independent definition by a particular geometrical model. This case is similar to that of an object in the shared reality of a primordial speech situation, let’s say something to sit on, and its subsequent definition as a chair or a barstool. One important difference is that the chair/ barstool definition is only provisional and anyone can, whenever they feel like it, attempt to change it, while the substance definition is implicitly proposed to become standardised, so that, once accepted by the peer community, it can only be altered if the experts agree that the ‘qualitative and quantitative terms’ on which it is based are faulty. The second difference is that the interpretation, i.e. the geometric model, has to be ‘elaborated in qualitative and quantitative terms’. There must be something that can be counted and/or measured, using a method that will always reproduce the same results for the same data. Unless the model requires otherwise, it must provide the same results not only here and now but tomorrow, as well as on the peak of the Mount Everest or at the bottom of the sea. The result must have a general validity, suggesting a regular, necessary and not just contingent correlation between substance and model. Thirdly, counting and measuring has to be objective. Not only the original researcher, but each member of the peer group would be expected to arrive at the same results. Again there is a parallel with the case discussed above in which the participants of a primordial speech situation agree on the number of persons in an environment such as a room. Measuring and counting have to be objectified. Both objective measurement and objective counting are based on non-human devices. It is not enough for the researcher to say: ‘It feels very radioactive in here’. There must be a Geiger-counter to verify it. Without instruments, tools, gadgets there would not be experimental science. It is to these devices that we owe our trust in objectivity. Yet it is easy enough to turn this line of argumentation on its head. Our scientific knowledge differs from other kinds of knowledge in that it depends on the existence of devices. They measure time, latitude and longitude,
194
Meaning, Discourse and Society
temperature, height and depth, velocity, weight, and all the other attributes we, the lay people, and they, the scientists, assign to whatever there may be found outside of the discourse. But what is cause; what is effect? Could we not just as well say that our concepts of time, space, temperature and so on are the result of things (normally artefacts) that we learn to interpret as measuring devices? The first sundial was a stick, and someone had the idea that what the shadow showed could be called time. After Philo and Hero of Alexandria came up with a device showing that certain substances expand and contract in different situations, someone had the idea to say that when a substance expands the environment has a higher temperature, and vice versa. Would we talk about time or temperature if they could not be measured? Are they real just because they can be? The discourse object ‘time’ exists because we can ‘read’ what a stick, or a clock, or whatever else, measures, but this does not make it real. It stays a convention. The discrete and robust substance referred to by Bazerman is related to a geometric model defined in qualitative and quantitative terms that have been discursively constructed. Does it make sense to assume, as we are told, that this substance was already there before ‘the historical, human act of isolating and identifying the substance’ (Bazerman 1981: 365)? Whatever we count and measure is, like time, contingent; it is a social construct. As long as we do not make it an object of our shared reality, it is nothing but stuff. Of course there is a reality out there. But whenever we try to turn some of this stuff out there into our shared reality, we have already constructed hundreds and thousands of discourse objects, of categories such as chairs, radioactivity, temperature, the colour pink, genes, intelligence and so on. These categories determine how we deal with the stuff of which the reality out there consists. We call these categories scientific once the scientific community has established a consensus about a category, while agreeing on an appropriate method for counting or measuring what this category is about, and once it has invented a gadget that does the counting for us. These activities are closely interlinked. Whenever a new category has been introduced in such a way, it will be called a scientific language-independent concept referring to a discourse-external reality. New findings in the natural sciences are commonly disseminated in the form of contributions to scientific journals and as presentations to scientific conferences. Such texts are highly standardised, even though the standards may be different from one discipline to the next. But on the whole they follow the prescripts that Bazerman described in his article discussed above. In the introduction, the problem the author(s) set out to solve is identified. We then find a report on the status quo. This is followed by the presentation of a methodology for demonstrating that the innovation or new idea actually delivers a scientific result. The main part is devoted to describing in detail whatever is being
The language of the scientific experimental report
195
identified, counted or measured by the application of the methodology, and explaining how reproducible this result is. As conclusion, we are told to what extent the problem has been solved and what still remains to be done. The acceptance of the result by the scientific community hinges largely on the main part, on the analysis of the phenomenon under research. This is the report type analysed by Bazerman. It describes how the scientists in their laboratory go about conducting their experiment. It has to be described in such detail that it is possible for their peers to carry out the very same experiment in their laboratories, with the promise that they would obtain the same results. What the journal article does in effect is to create a virtual primordial speech situation. It takes place in a reproducible environment. The standardised apparatuses the researchers/authors apply correspond to the senses we normally employ. We are not really there in the original situation, but the report gives us the blueprint to reconstruct it. It details the data and the apparatuses which are applied to the data. This blueprint corresponds almost isomorphically to the act of pointing that takes place in the speech situation. Thus the work of scientists constructs a reality that can be shared by other members of their peer community. But just as the props in the immediate environment of a primordial speech situation are only an extension of the discourse, and not distinct from it, the lab, with all the data to work on, all the methodologies determining how the work is carried out, and all the equipment found in this environment, are no more than an extension of the scientific discourse. The knowledge generated is discourseinternal knowledge. It is not a mirror of the world outside the discourse. The stuff of which the discourse-external reality consists is neutral in relationship to the question of whether there is global warming or not. For us, it will never be anything but stuff. What scientists can tell us about global warming is their interpretation of the reality which they construct and share in their discourses. Obviously there is more than one interpretation of climate change, and consequently more than one shared reality. But not all contributions to this discourse are equally successful. Over time, we will find that some interpretations will leave fewer traces in subsequent texts than others, and will increasingly be seen as irrelevant. Why one scientist, or group of scientists, uses a certain methodology for analysing a certain set of data and comes up with the apposite findings, is something we can never be sure about. People’s intentions are not open to scrutiny, scientific or otherwise. But the discourse as a collective mind can be investigated. There we can find that certain pieces of knowledge are repeated, rejected or modified. We can assume that texts, to the extent we can ascribe to them an intentionality borrowed from their authors, want to be successful and leave as many (positive) traces in subsequent texts as possible. It is this tendency which will over time determine what counts as the scientific consensus on climate change. As Foucault has shown, this is the normal development for discursive formations.
196
Meaning, Discourse and Society
The idea that there is no other reality than the one constructed in the discourse is beginning to be seriously considered outside of the rather narrow confines of the social constructionists and the recently revived sociology of knowledge, as we find it, for instance, in Nico Stehr’s and Volker Meja’s Society and Knowledge (2005), a collection of theoretical and empirical studies. But it has not yet really caught on in the philosophy of science, which is moving only slowly away from the Cartesian understanding that the reality outside of the discourse is accessible to us, and that it is up to each scientist, using the power of reason afforded by their own mind, to be the judge of whether a statement dealing with this reality out there is true or not. It is this focus on truth that makes it so hard to reconcile this paradigm with that of the more radical social epistemologists, a small (but growing) minority among the philosophers of science. One of the reasons why the traditional model remains so popular is that society in general, and academics within it are still accustomed to attributing scientific achievements to individual scholars. Even if it has become the standard for scientific research to be carried out by project groups, and for scientific reports to give large lists of names as their authors, instruments such as the Nobel prize and the way science is reported and presented in the media, still affirm our belief in solitary and rather monadic ingenuity. ‘One reason to think of individuals as the locus of rationality is to preserve the practice of making judgements of epistemic responsibility and assigning praise and blame’, says Miriam Solomon (1994: 341). The focus on truth, rather than on meaning, can also be seen as a consequence of the prevalence of all kinds of realism in much of contemporary AngloAmerican philosophy. Common sense tells us that there is a reality out there with which I can deal. Antirealism, on the other hand, ‘thrives on supposing that there is an enormous gulf between the realist’s claims and everyday ideas and judgments, on accusing realists of importing unnecessary metaphysics’, maintains Philip Kitcher (2001: 24). He insists that people’s behaviour reveals a direct link between their mind and the reality out there: ‘The correspondence we recognize between [a person’s] mental states and the world is disclosed through her behaviour, not created by our observation and explanation of it’ (Kitcher 2001: 24). His example for demonstrating that reality exists even if we cannot observe it, is people’s ability to use the London Underground system. The traveller’s behaviour reveals that she is certain to reach her destination even if she cannot see it at her starting point. For Kitcher, this proves that it makes sense to assume that objects exist independently of us and that we have direct access to them: ‘Although the objects [the Underground traveller] confronts and with which she coordinates her behaviour exist independently of her, her access to them is perfectly direct: while waiting on the platform at the station, she may have indirect access to the approaching train by observing the monitor that
The language of the scientific experimental report
197
announces its arrival, but, when it actually comes, there is no plausible sense of ‘direct’ in which her access to it is less than direct’ (p. 25). But he neglects that what has turned the stuff out there for her into platforms, trains, monitors and codes of conduct are many negotiations that have taken place in primordial speech situations. That a platform is a structure where one has to go and wait for the train to come and stop beside it so that one can step right into a carriage through doors opening automatically, and that we find such platforms at all Underground stations, are things we have to be made aware of. My belief that there were always trains running on the tracks even if I could not see them was shattered one particular 25 December when traditionally all public transport services are shut down. Before people were kind enough to pass on this piece of knowledge, my behaviour was rather erratic and revealed the absence of any direct access to reality. To behave adequately in a cultural environment in which we have grown up, or which is familiar to us, is no great miracle. It is when we leave our everyday lifeworld that we begin to feel out of touch with reality. When we are confronted with the mysterious ancient structures of geometric designs (drawings?) found in the Palpa and Nasca region of Peru, we have no clue what this stuff means. Some seem to show figures. But others just do not have any meaning for us. There is no primordial speech situation that could turn them for us into a shared reality. The point of departure from the classical Cartesian concept of truth is, for the relatively new field of social epistemology, that there must be a social consensus for what we take to be real. ‘Proponents of [this] anti-classical approach have little or no use for concepts like truth and justification. In addressing the social dimensions of knowledge they understand “knowledge” as simply what is believed, or what beliefs are “institutionalized” in this or that community, culture, or context’, as Alvin Goldman describes their stance in his entry ‘Social epistemology’ in the Stanford Encyclopedia of Philosophy (2006). In her book The Fate of Knowledge, Helen Longino makes it clear that this approach requires the primacy of the discourse over the discourse-external reality: ‘Critical discursive interactions are social processes of knowledge production. They determine what gets to remain in the public pool of information that counts as knowledge’ (Longino 2002: 129). But Longino does not discard the notion of scientific objectivity altogether. She advocates an epistemology that comprises both the realist and the constructionist approach: ‘A satisfactory epistemology should be open to theoretical plurality or theoretical unity being the final result of inquiry’ (p. 95). Social epistemology accepts the social determination of what counts as scientific practice, the need to see scientific findings within the contexts of their production, and the impossibility of resolving our differences. But that is only one face of the coin. The other face is still embossed by the Cartesian agenda: the scientific agenda ‘should not foreclose metaphysical questions’
198
Meaning, Discourse and Society
(Longino 2002: 96). This ambivalence fits in nicely with Goldman’s conclusion in his entry in the Stanford Encyclopaedia to the effect that social epistemology does not give up the pursuit of truth; it only accepts that this pursuit is hampered by the constraints and imperfections of social reality: Whereas Descartes thought that truth should be pursued only by the proper conduct of ‘reason’, specifically, the doxastic agent’s own reason, social epistemology acknowledges what everyone except a radical skeptic will admit, namely, that quests for truth are commonly influenced, for better or for worse, by institutional arrangements that massively affect what doxastic agents hear (or fail to hear) from others. To maximize prospects for successful pursuits of truth, this variable cannot sensibly be neglected. (Goldman 2006)
As long as the philosophy of science continues to adhere to the traditional concept of truth, it has to presuppose not only that there is a reality out there, but also that this reality is structured in ways that exist independently of the observers and their discourse; that there are apples and pears as distinct metaphysical categories, regardless of observers and their discourses. For me, the existence of such a reality is a futile question. The reality out there, structured or not, could only become relevant for us if we could all agree on a mechanism that makes it accessible for us. In the absence of such a consensus, I suggest that we focus on the reality to which we have full access, namely the reality that people have constructed for us in the discourse. This is the only reality on which collaborative action on our side can have an impact. Truth ceases to be a parameter for this reality. For the reality of the discourse is a symbolic reality. What is said in the discourse has a meaning. In the next two chapters I will discuss how we can seize on this meaning.
15
Diachronicity, intertextuality and hermeneutics
Hermeneutics is the art of interpreting texts. Its name is derived from the god Hermes. In Cratylus, Socrates describes the connection between Hermes and interpretation: I should imagine the name Hermes has to do with speech, and signifies that he is the interpreter (ermeneus), or messenger, or thief, or liar, or bargainer: all that sort of thing has a great deal to do with language. (Translation: Benjamin Jowett; taken from the Project Gutenberg online version)
Truth is not an issue closely associated with Hermes’ character. He is more concerned with the meaning attached to a message. A text, a text segment, a phrase or a simple lexical item, means. When we ask what it means, we expect to be given a paraphrase of the text, an interpretation of it. If I ask ten people, I will probably be told ten different paraphrases. If I want to know what a Chinese sentence means, I have to ask for a translation. Ten different translators will give me ten different translations. Translations are in this sense like interpretations. There will never be the one and only, the perfect translation. Translations and interpretations have to be negotiated, and it does not matter if a consensus is reached or not. There is no mental mechanism that generates the one and only translation in a computational process with a predefined outcome. Translations and interpretations are the contingent results of collaborative efforts of discourse communities, an English Literature class or a community of bilingual speakers. Modern hermeneutics is about interpretation. It is interested in the meaning of texts, text segments, phrases or simple lexical items. It is not interested in the relationship between the text and reality. The hermeneutic act never leaves the discourse. Contemporary hermeneutics, for instance the hermeneutics of HansGeorg Gadamer and Paul Ricoeur, thus shares with Jacques Derrida the view that Saussure’s signified, the concept, the pre-linguistic idea, will always be outside of our reach. There is no metaphysical given. Meaning is only in the discourse. The meaning of a text segment can be explored by comparing it to similar segments in other texts of the discourse. No text starts from nowhere. Every text is a reaction to previous texts. Many texts have subsequently been referred to, or interpreted. It is these intertextual links that we can take as clues for the meaning. 199
200
Meaning, Discourse and Society
Hermeneutics, insofar as it is concerned with language, is only concerned with writing, not with speech. Wilhelm Dilthey states explicitly that it is about the ‘interpretation of the written testimonies of human life’ (Mueller-Vollmer 1986: 27). This is hardly surprising. We can only explore the meaning of recorded (usually written) texts. Unrecorded spoken utterances are lost. With the absence of the author, the content of the text itself occupies centre stage. Paul Ricoeur, who was crucial in wedding hermeneutics to post-structuralism, agrees that this is what sets writing apart from speech: Inscription becomes synonymous with the semantic autonomy of the text, which results from the disconnection of the mental intention of the author from the verbal meaning of the text, of what the author meant and what the text means. What the text means now matters more than what the author meant when he wrote it. (Ricoeur 1976: 29f.)
For Ricoeur, this is exactly the reason why our concept of language presupposes the primacy of writing over speech. If we cannot map what is written onto the author’s intentions or some discourse-external reality, we have to accept that meaning is only in the discourse. This is not the place to sketch the history of hermeneutics. Even more than analytic philosophy, hermeneutics is characterised by a plenitude of irreconcilable and incompatible attitudes. Kurt Mueller-Vollmer’s hermeneutics reader (1986) and Lawrence Schmidt’s Understanding Hermeneutics (2006) are excellent introductions to this fascinating continental alternative to AngloAmerican mainstream philosophy during the second half of the twentieth century. Within hermeneutics, there has always been contention as to whether understanding means having access to the author’s life process, or mental history. It is true that Schleiermacher defines the task of interpretation as ‘die Rede zuerst ebensogut und dann besser zu verstehen als ihr Urheber [to understand the utterance first as well as, and then better than, its author]’ (Frank 1977: 94). Fritz Hermanns points in this context to Kant who may have inspired Schleiermacher. Kant is quoted as having said that it was not uncommon ‘sowohl im gemeinen Gespräche, als in Schriften den Verfasser […] besser zu verstehen, als er sich selbst verstand [both in casual talk, as in his writings, to understand the author […] better than he had understood himself]’ (quoted in Hermanns 2003: 139). Mueller-Vollmer warns us against interpreting this citation from our modern, psychologically informed perspective. In his view, ‘the traditional label of the psychologism of Schleiermacher’s position can no longer be maintained. For Schleiermacher, the author is “something fluid and dynamic … an act rather than a substance” (MuellerVollmer 1986: 11). Be this as it may, August Boeckh, the eminent philologist of the German Romantic era, states more succinctly what the interpreter’s task is concerning the ‘subjective conditions of the thing communicated’:
Diachronicity, intertextuality and hermeneutics
201
The author composes according to the laws of grammar and style, but is as a rule unconscious of them. The interpreter, however, cannot fully explain without consciousness of these laws. The man who understands must reflect on the work; the author brings it into being, and reflects upon his work only when he becomes as it were an expositor of it … The interpreter consequently understands the author better than the author understands himself. (Mueller-Vollmer 1986: 136, 138)
There can be little doubt that up to the twentieth century, secular hermeneutics was always interested as much in the author as in the text. For hermeneutics, the revelation of the author’s intentions was always seen as a crucial part of the enterprise. This was to be achieved through empathy, or, in the words of Wilhelm Dilthey, the activity of sich hineinversetzen (‘placing oneself in the situation of another person’). It should be noted that English empathy and French empathie trace their origins back to this tradition. They are calques from German Einfühlung (Hermanns 2003: 147). Dilthey was open to empathy. He was also the first to distinguish the text and our interpretation of it ‘from the mental (psychisch) processes in the poet or his readers’ (Dilthey 2002: 107). The early preoccupation of hermeneutics with the human mind is not accidental. Once hermeneutics had liberated itself from its theological origins, it co-evolved with the modern Western conception of the person as an autonomous individual, subject only to reason, and fully responsible for his actions. The autonomy of the individual led consequently to the elevation of authorship. Hermeneutics is an art. The fact that Gadamer agreed to change the title of his magnum opus from Fundamentals of a Philosophical Hermeneutics to Truth and Method, following his publisher’s suggestion, shows his deep sense of irony. For him, hermeneutics, as he defined it, had neither a method nor was it in pursuit of the kind of truth his colleagues in analytic philosophy were after. Gadamer sees interpretation, the core of hermeneutics, as an act, determined by the contingent intentionality of the people who carry it out. They are the ones who decide which tracks are relevant or otherwise. There is no manual of instructions that will take me from my first reading of a text to the one and only valid interpretation. I will show in Chapter 16 why there is no alternative to the subjectivity of the interpretation. But within this freedom that the hermeneutic act gives us, there are a few principles on which the peer community can agree. The two instruments a hermeneutic analysis can use are intertextuality and paraphrase. Both instruments make use of the assumption that the discourse is self-referential. Texts often refer to previous texts, by discussing text segments, more complex phrases and single lexical items found there. They create something new by placing them in more or less cognate textual environments. The more synonymous the new environment is with the previous ones, the more allegiance the new text claims to the discourse to which these existing texts belong. The more the new environment deviates, the more originality it claims. A suitable
202
Meaning, Discourse and Society
example is the role the phrase stand shoulder to shoulder played in British politics after the attack on the World Trade Center. On the same day, on 11 September 2001, Tony Blair read a statement distributed on BBC News that contained these two sentences: This is not a battle between the United States of America and terrorism, but between the free and democratic world and terrorism. We, therefore, here in Britain stand shoulder to shoulder with our American friends in this hour of tragedy, and we, like them, will not rest until this evil is driven from our world. A few days later, on 14 September 2001, this phrase was reused in four contributions to the House of Commons debate on ‘International terrorism and attacks in the USA’. In the days after 11 September 2001, speakers of all major parties varied the phrase standing shoulder to shoulder. According to Hansard, these are relevant contributions, in temporal order: Iain Duncan Smith (Conservatives, leader): That is why we are united in the House in our determination not only to extend our genuine and heartfelt sympathy to the United States but to defend civilised values against those who seek to bring them down by violence. I have absolutely no hesitation in giving the Prime Minister my party’s full support for his immediate pledge to stand shoulder to shoulder with our strongest friends and allies. Michael Ancram (Conservatives): The bond between this country and America is strong – strong because of the values and interests that we share and because of the personal bonds of friendship and the family ties between our two countries, but strong, most of all, for the times that we have stood shoulder to shoulder against evil. Charles Kennedy (Liberal Democrats, leader): An American writer once observed that the terrorist attempts to wash an impure world clean with the blood of innocent victims. The impurity here is the dreadful deed of the terrorist. On that, this House stands shoulder to shoulder in full support of our American cousins. Dennis Skinner (Labour): Does the Prime Minister agree that there is a world of difference between standing shoulder to shoulder with the American people in their fight for justice and hanging on to the coat tails of an American President whose first act when firefighters stood 10 ft tall in the rubble of the World Trade Center was to scurry off to his bunker? New contributions to the discourse do not repeat verbatim what has been said. Rather, they bring in more or less subtle changes. In cases in which a
Diachronicity, intertextuality and hermeneutics
203
subsequent text agrees with the reference text, we often find a movement towards hyperbole. Thus in Blair’s contribution the partners in the shoulderto-shoulder exercise are our American friends. For Iain Duncan Smith, they are our strongest friends and allies. Ancram wants to stand shoulder to shoulder with America. For Kennedy the partners are our American cousins, a variation on Ancram’s family ties. Skinner’s text is disdainful of Blair’s statement; while not rejecting the phrase outright, it distinguishes between the meaning it ascribes to Blair’s statement (shoulder to shoulder paraphrased as hanging on to the coat tails of an American president) and a better meaning which should replace it: shoulder to shoulder with the American people. The common enemies in Blair’s statement are terrorism and evil. What needs to be defended is for Blair the free and democratic world; for Iain Duncan Smith it is civilised values; for Ancram it is shared values, interests and personal bonds; for Kennedy it is innocent victims and for Skinner it is the fight for justice. The enemy is described as terrorism or terrorists (Blair and Kennedy), those who seek to bring civilised values down by violence (Smith), and evil (Blair and Ancram). On the same day, 14 September, the House of Lords was also debating ‘International terrorism’, and there, too, we find four citations for shoulder to shoulder: Baroness Symons of Vernham Dean: As the Prime Minister stated, Britain stands shoulder to shoulder with the United States, in total solidarity with our closest ally. We have deep ties of history and friendship and share a common language, but above all we share common values. We are part of the same community. Lord Davies of Coity: Of course a response must come from the whole of the democratic world and we must unequivocally stand shoulder to shoulder with the Americans in their hour of great sorrow and great need. Yes, now is the time to be cool and calculating but also committed and determined to rid the world of these indescribable people who have taken ‘man’s inhumanity to man’ to new depths of evil. Lord Selkirk of Douglas: On the first occasion, [the Scottish Parliament] expressed condolences and yesterday, in the Prime Minister’s words, the desire to stand shoulder to shoulder with the United States after what were acts of war against the US. The Lord Bishop of London: The council is right to brand this as an attack on the whole civilised world. As bishop of a city which has suffered in the past from terrorist attacks, it is good to stand shoulder to shoulder with British Muslim leaders in an unequivocal condemnation of these crimes.
204
Meaning, Discourse and Society
Alongside repeats of what we have already encountered, there are some new expressions here concerning the partners with whom we stand shoulder to shoulder (e.g. closest ally, British Muslim leaders), for the reasons to engage in a fight (e.g. common history, common language, unequivocal condemnation), and for the enemy who needs to be fought (e.g. these indescribable people, terrorist attacks, crimes). They tell us that standing shoulder to shoulder is something one does in times of a deadly threat and against an enemy who is outside of the civilised world. It is only the bishop of London who comes up with an innovative twist, worthy of an experienced practitioner of the religious craft. He uses the battlefield metaphor to put a claim on the leaders of the British Muslims, recruiting them as allies of the civilised world, thus driving a wedge between good and evil Islam followers. However, he does not make it clear whether he will be standing shoulder to shoulder with the Muslim leaders, or if he expects them to stand shoulder to shoulder with him. Paraphrases are the other instrument for discovering meaning. The discourse, in reflecting on what is said and how it is to be interpreted, abounds in paraphrases, particularly of key expressions, whether positive ‘banner words’ or negative ‘stigma words’, to use Fritz Hermanns’ terminology (Hermanns 1994: 81). What I call paraphrases are text segments that attempt to explain, affirm, modify, reject or merely elaborate on the discourse object corresponding to the expression in question. Paraphrases, as I see them, can distinguish between the expression and the discourse object expressed. But often they do not distinguish. To illustrate this discourse object ‘paraphrase’, I will give a few paraphrastic citations on the expression civilised world that we encountered in the excerpts above. The most common and straightforward form of paraphrase uses the copula is. These examples were found in texts on the internet from the time after 11 September 2001: The civilized world is at war with terrorists. The civilised world is fighting an ongoing ‘war on terror’ against evil, exotic forces that want to destroy freedom and progress everywhere. Looking around the world today, it is not hard to see that the civilised world is under siege from Islamic terrorism. Of course, which countries constitute the civilised world is a matter for discussion. The civilized world is rallying to America’s side. The uncivilized world is rallying to America’s side. One is tempted to think that what George Bush Jr. and Tony Blair like to call the ‘civilized world’ is not so civilized after all. The last two citations reject the standard post-September 11 paraphrase, associating the civilised world closely with its enemy, terrorists or Islamists. Both
Diachronicity, intertextuality and hermeneutics
205
are comments on the expression, the lexical item, but go deeper and also aim to redefine the discourse object for which it stands: the so-called civilised world is perhaps not so civilised. The first of them has become very successful: Google lists 85,000 occurrences of it. The second one comes, unsurprisingly, from a Muslim country that has to struggle hard to be counted among the elect few and is found in a leader in The Jakarta Post (Tamimi 2001). Not all paraphrases are so straightforward. Many citations contain paraphrastic content that can be extracted from the contexts in which the expression is found. Here are a few examples taken from the Guardian during the days after the attack on the World Trade Center: So: we have a group of people in Afghanistan who are the sworn enemies of everything the civilised world stands for, who have killed once on a vast scale and will kill again unless stopped. They can’t be negotiated with. They refuse to yield to justice. Terrorists could not be ignored or appeased; they must be destroyed since they threatened everybody. ‘This conflict is a fight to save the civilised world,’ Mr Bush declared. Asked whether any military strikes would in effect be an American operation in which Britain was obliged take part, Mr Blair said: ‘This has been an attack on the whole of the civilised world.’ ‘Attacks on the US are attacks on the basic democratic values in which we all believe so passionately and on the civilised world,’ he said. These were attacks on the basic democratic values in which we all believe so passionately and on the civilised world. There is no reason to expect that the paraphrastic contents we find in a special discourse do not contradict each other. But even if we may hope that Guardian journalists are given leeway to express a wide range of opinions, contradictory voices in this quality newspaper are mostly found among the letters to the editor. Here are some examples, two of which again take up Tony Blair’s shoulder to shoulder statement: George Bush did not stand shoulder to shoulder with the civilised world when he turned his back on the Kyoto protocol. Was he with us or against us? Like many other readers, I have yet to talk to anyone who wishes to stand ‘shoulder to shoulder’ in this coalition of the so-called civilised world. Backing oil sheiks, military strongmen, and absolute monarchs, in an exercise that inflames many of their subjects, is not the best way to defend ‘the civilised world’.
206
Meaning, Discourse and Society
One might feel tempted to follow up the lead given in the first regular Guardian citation: to Google what the civilised world stands for. Discounting duplicates, there are only twenty-four occurrences of the phrase ‘the civilized world stands for’, and while none of them actually answers our question, they show how similar content tends to recur in different guises: ‘… Taliban sponsors, whom he called the ‘sworn enemies of everything the civilized world stands for.’ The latter are the ruthless and implacable enemies of everything the civilized world stands for. ‘We have a group of people in Afghanistan who are the sworn enemies of everything that the civilized world stands for’ It is that these terrorists hate all that the civilized world stands for, and they are willing to lose their lives for the cause. ‘When radicals say they loathe everything the civilized world stands for, believe it. Terrorism is a serious challenge to the values systems that the civilized world stands for, and a threat to what humanity strives to achieve.’ The civilised world is what is hated and loathed by its sworn, ruthless and implacable enemies: the Taliban, terrorists and radicals. One way to find contexts with a more specific paraphrastic content is to extend the query expression. In order to learn more about the nature of the civilised world I have added the word values from the last citation to our expression. We find 1,130 Google hits for ‘values of the civilized world’. The overwhelming majority of the first 100 citations deal with the enemy activities against these values, and there are only three instances in which some of the values are actually listed: The culprits of these incidents who are the executers of gloomy acts of terrorism beyond the borders, their enmity with the understandings the high values of the civilized world like peace, security, human rights, democracy and social welfare are in contrary with the values we believe in, therefore they are always trying to disorder the regional and international peace. Official website of the Ministry of Foreign Affairs of the Islamic Republic of Afghanistan, 31 July 2008 … the chance to adopt the ‘universal’ values of the ‘civilized’ world, such as the values of democracy, market economy, and protection of human rights. Tsygankov and Tsygankov 2004
Diachronicity, intertextuality and hermeneutics
207
They apparently live in an enclave that has abandoned the shared values of the civilized world, one where the fundamental values of respect for physical integrity of others, compassion, and elementary social solidarity have been suspended. Waldmann 2007 It may not be a coincidence that these authors come from outside the AngloAmerican sphere and are therefore not representative of the tendency to leave the concrete nature of the civilised world unspelt. The very first citation of the phrase ‘civilised world’ that is accessible in the Guardian archive throws an interesting light on its rhetorical function. It is the English translation of the speech made by Hitler’s successor, Karl Dönitz, in May 1945: German men and women, our Fuhrer, Adolf Hitler, has fallen. His life was one single service for Germany. His action in fighting against the Bolshevist springtide was waged beyond that, for Europe and the entire civilised world.
Paraphrases and other contexts containing paraphrastic material tell us what can be known about the meaning of a lexical item, be it a single word or a phrase. If we search the web, as a sample of the discourse at large, we obtain a fairly comprehensive representation, with often contradictory material, covering a wide ideological space. If we, on the other hand, investigate the special discourse of a newspaper like the Guardian, the results will by and large appear more streamlined. At the same time, even in a rather limited discourse, a certain amount of plurivocality is vital. Innovation presupposes a discourse that allows the recombination, permutation and variation of what has been said before; it makes deviation possible. It is one of the functions that the letters to the editor (which belong to two discourses, that of the Guardian but also that of the Guardian consumers) fulfil. It is this absence of almost any plurivocality that made most newspapers in the Eastern Bloc so unreadable. Intertextual links and paraphrases are the key material that provides textual evidence for an interpretation and thus makes it negotiable. Other members of the peer community may adduce other texts, and extract additional paraphrastic content, and they may reject some of my choices. But it will be a negotiation of the meaning of a text that is based on the reality of the discourse, not on inaccessible first-person experiences, or on claims about a discourse-external reality. Hermeneutics is not concerned with truth but with meaning. Meaning is what the entirety of the paraphrastic content available for a given lexical item, a phrase, a text segment or a full text within a given discourse (which may well be the discourse at large or a sample thereof) tells us. Meaning thus is co-extensive
208
Meaning, Discourse and Society
with the knowledge of the discourse object for which an expression stands. There is no distinction between lexical and encyclopaedic knowledge. Meaning can be discussed by the members of a discourse community and, indeed, is discussed whenever disagreement arises. Meaning is what can be communicated verbally. Meaning can be found neither in the minds of speakers and hearers nor in the reference of a text segment to some perceived or metaphysical reality. Meaning is what is exchanged and shared in contributions to a discourse. Meaning is social, not mental. Meaning is in the discourse. It is the collaborative act of interpretation that makes meaning available to a discourse community. We can discuss what a given instance of language use means, and agree or disagree about it, and this negotiation is an integral part of the discourse. This is how children learn to speak, how we learn foreign languages, how the newspaper editorials explain to us what ‘standing shoulder to shoulder’ means in the postSeptember 11 world. Obviously there is a certain affinity between Niklas Luhmann’s concept of society, as part of his systems theory, and this (my) version of hermeneutics. Systems theory is increasingly being invoked in the social sciences to describe social (i.e. symbolic) interaction, including communication. It shows how the medium of communication is essential for determining the relationships between the parts, elements and variables of the system, and under which conditions of social subsystems emerge. One of the guiding principles of systems theory is self-reference. ‘Über Mitteilung bezieht das System sich auf sich selbst. Die Mitteilung aktiviert die Möglichkeit, rekursiv weitere Kommunikationen auf das System zu beziehen [by relaying messages, the system refers to itself. The message opens up the opportunity to relate subsequent communications to the system]’ (Luhmann 1998: 98). Equally relevant is his insistence that people’s intentionalities are outside of society proper. From the point of view of systems theory, the discourse (‘communication’) itself is an autopoietic system that generates new text after new text, the speakers being no more than aleatoric production mechanisms. The ensuing (and continually evolving) discourse thus creates its own intentionality, only loosely coupled with (and, as Paul Ricoeur puts it, ‘dislocated from’) the minds of the contributors. In my version of hermeneutics I adopt the strict separation of communication and the individual mind, as suggested by systems theory. The way a person understands a text can be described neither as a mechanical process nor as a conscious act. Understanding is experiencing, and experience is non-symbolic. Experience has no meaning. Only my (verbal) testimony of my experience becomes symbolic and thus means. As testimony, however, it is no longer reducible to experience. Making sense of the discourse, or a lexical item, phrase or text (segment) of it, takes place not as someone’s monadic experience but as an event happening within the discourse, a collaborative activity. Any interpretation arrived at becomes part of the discourse, as the collective mind of a discourse community.
Diachronicity, intertextuality and hermeneutics
209
In computational linguistics, automatic procedures have been developed to summarise texts. While such summaries (for instance of papers in academic journals) are certainly useful for researchers who want to find out what might be relevant for them, they will hardly tell us if a paper provides a breakthrough. That is a conclusion based on interpretation. Even after a lengthy discussion among a peer community there may not be agreement. For Luhmann’s systems theory, intentionality as such plays no role in what he calls ‘communication’. For hermeneutics, it is essential. How can the two approaches be reconciled with each other? We can look at the discourse from two perspectives. In Luhmann’s sense, discourse, as the emanation of society, is an object of investigation. Indeed, he does not discuss the issue of the vantage point of the researcher. He took over the concept of second-order observation from Heinz von Foerster, one of the pioneers of cybernetics, and a supporter of radical constructivism (cf. Von Foerster 1993). Second-order observation seeks to avoid ‘blind spots’ which necessarily come with first-order observation, by instituting a second observer who observes the first observer’s act of observing. This reminds me of the never-ending succession of Searle’s homunculi that we need for making sense of mental representations (see above, Part 1). Luhmann, as I see it, re-introduces the discourse observer through the back door, so-to-speak, after having evicted people as such from his concept of society. But it is the autopoietic discourse itself, not the intentionalities of the members of a discourse community, that by observing itself generates interpretations. There is no second observational vantage point outside a discourse. Even if we distinguish between the discourse of those taking part in an interpretation and the discourse containing the lexical item, phrase or text (segment) that is being interpreted, this separation is cancelled out in the act of interpretation. Any interpretation from the outside has to take into account how the discourse element is interpreted by the discourse to which it belongs. As Gadamer puts it, the two horizons of (what he calls) understanding have to be merged in the interpretive act. Thus on the one hand we have the discourse as an object. It is the representation of the collective mind, of the symbolic content shareable by the members of a discourse community. From this perspective, the discourse can be observed and interpreted. But there is also another perspective. Discourse is symbolic interaction, something we do together. Collaboration means that each of the parties involved is performing their part of the action. Each participant experiences and exercises their intentionality. But this individual intentionality is not symbolic. The individual contribution to the discourse is an act that becomes symbolic only in the moment it enters the discourse, in the moment it becomes accessible to the other participants. The collective intentionality of a discourse community is that community’s discourse. Unlike individual intentionality, it is symbolic and can therefore be observed. What we say is important and has an impact on society. But the mental processes or acts leading to the utterance, as
210
Meaning, Discourse and Society
important as they may seem to us, will be without consequence. The discourse is both a collaborative activity and an object. It is the act that creates it, but it is only as an object that we can observe it. This is why the primordial speech situation can be made an object of investigation only once it has been recorded and thus turned into accessible discourse. Any discourse has, of necessity, a diachronic dimension. We never start at point zero. Whenever we make an utterance, whenever we contribute something to the discourse, we react to something that has been said before. That is even true for fiction. While authors and readers agree that they are dealing with an invention, any fictional narrative will, at least implicitly, refer to other fictional or factual texts. Each novel, unique as it may be, carries with it the tradition of novel writing. Whatever is said (or written) has to be read against the background of what has been said or written before. Each signifyer of a new text is connected by an inexhaustible richness of links to previous texts, just as in Derridaean grammatology signifiers, signifiants are seen in hermeneutics as never referring to discourse-external signifiées but only ever to previous signifiants. Derrida, though, was not the first to understand this. Michel Foucault expressed the same thought about a decade earlier in his 1964 essay ‘Nietzsche, Freud, Marx’: There is nothing absolutely primary to interpret, because at bottom everything is already interpretation [tout est déjà interprétation]. Each sign is in itself not the thing that presents itself to interpretation, but the interpretation of other signs. (Foucault 1990: 64)
This is the point originally made so impressively by Mikhail Bakhtin or V. N. Volosinov (to the extent that it is useful to distinguish between their texts): But the monologic utterance is, after all, already an abstraction, though, to be sure, an abstraction of a ‘natural’ kind. Any monologic utterance, the written monument included, is an inseverable element of verbal communication. Any utterance – the finished, written utterance not excepted – makes response to something and is calculated to be responded to in turn. (Volosinov 1986: 72)
Michael Holquist, a key exegete of Bakhtin, formulates it in these words: ‘An utterance … is always an answer to another utterance that precedes it, and is therefore always conditioned by, and in turn qualifies, the prior utterance to a greater or lesser degree’ (Holquist 2002: 60). For the interpretive act this means that we can only make sense of a piece of text if we know to which texts it is connected. It was Julia Kristeva who introduced the term ‘intertextuality’ in her exegesis of Bakhtin’s terms dialogue and ambivalence: ‘[A]ny text is constructed as a mosaic of quotations; any text is the absorption and transformation of another. The notion of intertextuality replaces that of intersubjectivity’ (Kristeva 1986: 39, Kristeva’s emphasis). What Kristeva is describing here is the mental process that, in the perception of intertextual links, engenders a communication between the author and the reader.
Diachronicity, intertextuality and hermeneutics
211
Since the introduction of the new term, it has been used in different ways. For me, intertextuality cannot count as a mental phenomenon. It does not matter if we feel ourselves reminded of certain other texts when we encounter a phrase or larger text segment. Intertextuality is a feature of the discourse. The discovery of intertextual links is part of the task of interpretation. Whether the similarity or even identity between a phrase or text segment in a text under investigation and other texts can count as an intertextual link is something to be negotiated. It is only relevant if we can construct an impact, direct or indirect, of one on the other. Not all previous occurrences of shoulder to shoulder, or of civilised world, are relevant for the meaning of these phrases in the post-September 11 discourse. Intertextuality is, like the concept of the discourse, negotiable; it is not an objective parameter. Kristeva has done little to elucidate her take on intertextuality, or to clarify her own interpretation of Bakhtin’s dialogism. I agree with Graham Allen who sums Kristeva’s stance thus up: ‘[T]he Bakhtinian notion of the dialogic has been rephrased within Kristeva’s semiotic attention text, textuality and their relation to ideological structures. While Bakhtin’s work centres on actual human subjects employing language in specific social situations, Kristeva’s way of expressing these points seems to evade human subjects in favour of the more abstract terms text and textuality’ (Allen 2000: 36). Intertextuality fills hermeneutics, the art of interpretation, with life. It gives those of us who discuss the meaning of a word, a phrase or a text segment something to do. While the traditional linguist, and also the corpus linguist of the last forty years, was mostly concerned with the meaning of recurrent lexical items, seen as more or less extended units of meaning, in general, intertextuality allows us to discuss the occurrence of these entities as unique events, as something happening in a given text that stands in a specific relationship to other texts of a discourse. Traditional linguists, lexicographers in particular, would be keen to develop a methodology leading them to a dictionary entry for the idiom shoulder to shoulder that holds for most occurrences in the general Englishlanguage discourse of, say, the last one hundred years. Corpus linguistics has been developing ever finer blueprints of such a methodology based on finetuned statistics applied to real language data for detecting what links all these occurrences together. Traditional linguistics, including corpus linguistics, is still in search of the systematicity of a language, including its lexis. Hermeneutics, on the other hand, is interested in the meaning of a given occurrence of a lexical item or text segment in a given text. Intertextuality is a perspective that allows us to compare this occurrence to other occurrences in previous texts to which the text in question may be seen as a reaction. It allows us to investigate what our occurrence has in common with them and where it departs from them. It shows us how a given text segment is a recombination, a permutation or a reformulation of what has been said before.
212
Meaning, Discourse and Society
In the nineteenth century, it was the philologists who analysed the text corpora of antiquity (and also of the Middle Ages) from this kind of intertextual perspective. Their glossaries and dictionaries carefully go through citation after citation, telling us how the same word was used differently at different times and by different authors. They were the ones who turned the theoretical approaches of Friedrich Schleiermacher, Wilhelm von Humboldt and Philip August Boeckh into lexicographic practice. They discussed how to translate a word like the Greek proseuchomai, found in the first sentence of Plato’s Republic, into English, as pray, say a prayer, make my prayers, offer up my prayers, pay my devotions, pay my devoirs, or worship, all of which were used in the various translations I have collected over the years. They were aware that in Plato’s time people did not pray as they do today in a Christian framework, and they analysed carefully the paraphrases they found, to the extent that they related to Plato’s text. What the hermeneutic theory of the age of philology did not spell out was how we have to read the texts of times past in the context of our own interpretation of the discourse objects we encounter. This is the key question that Hans-Georg Gadamer discusses in his Truth and Method. It is, then, in principle not possible, he says, to reconstruct the meaning of a word like proseuchomai as it is used in Plato’s text. We should not hope ‘to bridge the temporal gap and reconstruct the original situation of the text, but to discover what the text has to say to us’, as Lawrence Schmidt tells us (2006: 104). Gadamer makes us aware that we cannot ever expect to hear a Bach sonata as his contemporaries heard it. We may re-create the original instruments and pay painstaking attention to the details of musical performances in the early eighteenth century as much as we like, but we cannot avoid hearing his music against the background of a tradition including Mozart, Schubert and Shostakovich. What sounds Baroque and quite impersonal to us used to infuriate Bach’s critics, because they saw it as an expression of individual sentiments that have no place in music. Whenever I listen to a Bach composition, I will bring to it the horizon of my contemporary bias. It is this history of effect (Wirkungsgeschichte), the effects a past text has on our present discourse, that is the focus of Gadamer’s interest, and indeed it is this history of effect that will determine any new interpretation. We cannot escape the fact that our present collective way of understanding has to be seen as the result of a coming-to-grips with the past. Even before we begin the act of interpretation of a past text, we already have a preliminary, if vague, knowledge, a certain anticipation of the meaning of it. With his pronounced sense of irony, Gadamer insists on calling our expectations prejudices, giving this word a surprisingly positive twist. Prejudices are what we have learned from the voices of authority to which we have been exposed. We should not be too disdainful of authority, he tells us. After all, they have been successful in their impact on our discourse.
Diachronicity, intertextuality and hermeneutics
213
For Gadamer, the task of interpretation is the fusion of the horizon to which the text belongs (in the case of the Republic, Plato’s Athens) and the horizon of our present-day discourse. While the text’s horizon is fixed, our own horizon is far from stable: In fact the horizon of the present is continually in the process of being formed because we are continually having to test all our prejudices. An important part of this testing occurs in encountering the past and in understanding the tradition from which we come. Hence the horizon of the present cannot be formed without the past. There is no more an isolated horizon of the present in itself than there are historical horizons which have to be acquired. Rather, understanding is always the fusion of these horizons supposedly existing by themselves. (Gadamer 1989: 305, Gadamer’s emphasis)
Gadamer sees interpretation as a dialogue, as a collaborative act taking place in the discourse, involving the discourses of the texts under consideration, but also the discourse community carrying out the interpretation: ‘Understanding is to be thought of less as a subjective act than participating in an event of tradition, a process of transmission in which past and present are constantly mediated’ (1989: 291, Gadamer’s emphasis). Later he points out the role of language, here to be understood as la parole, the discourse: ‘[T]he fusion of horizons that takes place in understanding is actually the achievement of language’ (p. 370). A few sentences further down we read: ‘To reach an understanding in a dialogue is not merely a matter of putting oneself forward and successfully asserting one’s own point of view, but being transformed into a communion in which we do not remain what we were’ (p. 371). For Gadamer, the even more important step than reading a text in its original horizon is the discussion of this reading in the light of our present horizon. As a trained classicist, he is, of course, more concerned with the texts of antiquity than with the post-September 11 discourse. Fusing the horizons of Plato’s Republic and our current understanding of it, means, for instance, making ourselves aware of how the concept of praying that characterises Christianity today has emerged over the last two millennia. We may also discuss what praying means today in other religions: in Judaism, in Islam, etc. Interpreting a text therefore means for Gadamer not only a new reading of it, but also an increase in awareness of what it means for us, a reading that takes into account our current discourse. In this sense, reading the Republic can lead to a reconsideration of the discourse object ‘praying’: ‘Part of real understanding, however, is that we regain the concepts of a historical past in such a way that they also include our own comprehension of them. Above I called this “the fusion of horizons”’ (Gadamer 1989: 367). The discourse hermeneutics deals with is the recorded, written discourse. Only in a literate society do we find a sense of history, because texts, dissociated from their speakers, cease to be expressions of their intentions and become an
214
Meaning, Discourse and Society
objects of interpretation: ‘The full hermeneutical significance of the fact that tradition is essentially verbal becomes clear in the case of a written tradition. The detachability of language from speaking derives from the fact that it can be written’ (Gadamer 1989: 391). This view is shared by the other eminent figure in twentieth-century hermeneutics, Paul Ricoeur: Inscription becomes synonymous with the semantic autonomy of the text, which results from the disconnection of the mental intention of the author from the verbal meaning of the text, of what the author meant and what the text means. What the text means now matters more than what the author meant when he wrote it. (Ricoeur 1976: 29)
As John B. Thompson interprets Ricoeur’s approach to hermeneutics, the exclusion of the author’s presumed intentions from the interpretation of a text does not imply the reader’s suspension of ‘any judgement concerning the referential dimension of the text, treating the latter as a wholly worldless and self-enclosed entity’. Rather, the reader is asked to ‘seek to unfold the nonostensive references of the text’ (Thompson 1981: 14f.). What Thompson means are references, not to some discourse-external reality, but first of all to all the other texts with which the text under consideration is intertextually linked: ‘The emancipation of the text from the oral situation entails a veritable upheaval in the relations between language and the various subjectivities concerned … [W]e shall have to go still further, but this time beginning from the upheaval which the referential relation of language to the world undergoes when the text takes the place of speech’ (Ricoeur 1981: 147). It is the achievement of hermeneutics to have recognised that the reality we deal with is not the reality out there. The world we find ourselves in is a discourse construct. We are born into it. We encounter this world continually in the texts we consume. It is a world full of prejudices delivered to us in the various voices of authority. We, the text consumers, are the victims of the content forced upon us. As long as we remain passive consumers, we have no role in the construction of this reality. At the origin of interpretation lies a question. We must, first of all, ask ourselves what lies behind what is said. Once we start questioning the construction of the reality that we are supposed to accept, we have taken the first step towards constructing the reality as it seems fit to us. Each interpretation of the status quo changes the status quo. As long as we only experience individually the reality we find ourselves exposed to, nothing will change. The way we experience ourselves is nonsymbolic. But the reality of the discourse is symbolic, and therefore only the collaborative act of interpretation can have an impact on it.
16
Meaning and the interpretation of a haiku
In this chapter, I want to show how we can work together to negotiate the meaning of a text. What I present is not a methodology, not a manual for interpretation that delivers a result quasi-automatically by going through the process step by step. What I want to show is that while interpretation is an activity with an unpredictable outcome, for those who collaborate there is one reality on which they can base their negotiations: the reality of the text embedded in the reality of the discourse. This is the reality available to all those concerned in the interpretation. This is the common basis. I have purposefully chosen a text that is rather removed from our ordinary life-world, this very unassuming haiku: abrasive heron blares, blue jittery songbirds stampede bouncily
This is a haiku of little if any poetic value. But that is not the point. What I am interested in is what it means. As with any new contribution to the discourse, the only way to make sense of it is to look at it as a reaction to what has been said before. If we were to take it as a text that has not emerged out of the discourse we could never assign an interpretation to it. We would still see all those words looking so familiar, but we could not be sure what they mean. It would be a text constructed on Humpty Dumpty principles. If we, the readers, could not read it as being constructed out of the elements that the discourse consists of, it would be meaningless. But taken as a text added to a discourse already in existence, we can hope that the meaning of the lexical items we find, for instance the nouns heron and songbirds, qualities expressed by adjectives such as abrasive or jittery, or verbs like blare or stampede, initially mean what they have meant in the discourse so far. Indeed, we can only make sense of our haiku if we embed it in the reality which has been collaboratively constructed so far by the discourse community. The heron we meet in this haiku has to be understood in the context of all that has been said about herons before, and the same is the case with the other words. This new text takes our socially constructed reality as a starting point. But it also introduces something new. In our interpretive endeavour, we have to approach it 215
216
Meaning, Discourse and Society
as a text that, through varying, rehashing, recombining and reformulating what has been said before, creates something that has never been said in exactly the same way before. Thus it may well be that abrasive herons or stampeding songbirds are new additions to the discourse at large, or that blares was never found as closely to stampede as in this haiku. The reality of our discourse will have changed if some future contributions to it also contain these novel constellations. If we do not find any traces of this haiku in any future texts, then it will have had no impact on the reality we keep reconstructing by our contributions to the discourse. This is surely the fate of most of the innovation brought into the discourse. Will this haiku be successful in adding a new morsel of knowledge to the reality of the discourse, namely that songbirds can stampede? If a phrase such as songbirds stampede bouncily has never occurred before, it is nevertheless possible to come up with an interpretation of it that can be subjected to communal negotiation. Its meaning is far from arbitrary. The discourse is full of stories telling us what is public knowledge about the discourse object ‘songbird’, and likewise we have learned what it means to stampede and who or what it is that is said to stampede. Even if no object of the discourse has ever been said to stampede bouncily, we have been told time and again what bouncing means and what kinds of things are said to bounce. This knowledge, contained in the discourse, has to be brought to the table and made visible for those with whom we want to collaborate in our interpretive endeavour. It is this activity of interpretation based on discourse evidence that I want to illustrate in this chapter. By interpretation I mean the assignment of a meaning to a text in a given situation, as a contribution to the discourse of a specific discourse community, for instance teacher and students in a class. The haiku itself is already part of the discourse at large, this limitless repository of everything that has been publicly communicated. As an object of interpretation it now enters a special discourse, defined as the sum of all contributions made by a teacher and students in a class interpreting this haiku. Once offered by a teacher or student, an interpretation is up for negotiation by the students. They may agree with it, repudiate it or modify it. The haiku and my interpretation of it will make an impact on this discourse to the extent that the students engage in the interpretive act and refer, in their contributions, to the haiku and to the reading that I, their teacher, give it. At the end of the class, my reading will probably have provoked other, competing readings, collectively making up the interpretation of our haiku. All interpretation is bound to be provisional. In a class next year, the haiku might be interpreted in a new way, perhaps informed by the previous interpretations. None of the readings offered, none of the interpretive acts, will ever exhaust the meaning of this haiku. Its meaning will grow over time. Meaning is the result of adding, over time, one interpretation on top of the other. The discourse has of necessity a diachronic dimension. A hermeneutic interpretation
Meaning and the interpretation of a haiku
217
of a text is aware that the meaning of a word like heron next year will differ from this year’s meaning, for many new things will be said about herons that have not yet been said, whether in poems or in the news or in literary criticism or in other genres. Meaning is never stable. When we, contributors to the discourse at large and members of sundry special discourse communities, including a discourse community concerned with the interpretation of texts (like staff and students of English Literature departments), have to interpret a text, perhaps a poem, we are expected to be aware of its previous interpretations. Our heron haiku, however, has never before been interpreted. There are several aspects that will have to be considered in this situation. What makes this concatenation of words a text? Has it been carved out of a larger ensemble of words? Should we consider it as a single sentence? How sure can we be that it is a haiku? Could it be that it is a translation from another language, perhaps Japanese? Is there something we have to know about haikus that would help us to interpret this text? While all these questions deserve answers, I want to abstract from them, in order to focus strictly on the meaning of a textual sequence that I take to be a haiku. For the interpretation which I will offer to the community, I will regard it as a text consisting of a single sentence. By categorising this text as a haiku, I make certain assumptions guiding my interpretation. I spend more effort on it than if it were the headline of a newspaper article. For one has to ponder over a haiku and its deeper sense. A haiku website (www.haiku.insouthsea.co.uk/form. htm#form) tells us that ‘[a] haiku is the smallest language construct that can generate enough complexity to create tension and resonance between its parts and take on symbolic power’. Thus, when we take a text to be a haiku, this is what an interpretation will try to bring to light. How the haiku affects me as a first-person experience is irrelevant. What I think about it can only become a matter of collective awareness if I tell other people. Whatever I say will have to bear in mind the nature, conventions and expectations of the discourse community to which I contribute my interpretation. What I say is, to a considerable extent, determined by my anticipation of possible reactions. It is the initiation of a collaborative act. Others will venture ideas about the meaning of the haiku, its inherent tension, vibrations, and symbolism. Once we bring this collaborative act to a temporary end (at the end of the class, for instance) our joint interpretation will hopefully have gone far beyond my initial comments. There is no need for a common denominator. The readings offered do not have to be mutually reconcilable. An abrasive heron can stand for an evil person, for a boss who does not accept our ways, a dark threat to all that is loveable and for much more. The songbirds may be jittery because they are frightened by the heron, or both the heron and the songbirds are frightened by something not mentioned. A collaborative act of interpretation may bring about agreement as well as any number of incompatible readings.
218
Meaning, Discourse and Society
When people try to make sense of a text collaboratively, they form (part of) a discourse community. Discourse communities do not exist as such, though. They are constructs of a discourse. People will perhaps accept that the students taking part in the interpretation together with their teacher form a discourse community. They can be said to be part of a wider community, that of all students of English Literature, at a given university or at all universities in the country. The discourse community to which I contribute my interpretation of the heron haiku consists of the readers of this book. Each of them will belong to their own discourse communities, determined by the people they interact with, the activities they carry out and the temporal and spatial constraints they are subject to. Stanley Fish talks in this connection about ‘interpretive communities’, and he shows that different communities will follow their specific conventions and come up with different interpretations (Fish 1980: 345ff.). In a similar way, Michel Foucault insists that discursive formations are identifiable by the rules that determine what can be said and what not (Foucault 1972: 117, et passim). Among my friends in the pub, there is probably not very much interest in a haiku about an abrasive heron, and I would have to deal with it (if I insisted on pushing the topic) in a rather jocular fashion. I would say things about the heron haiku there that I would not write down in an academic text, and vice versa. What one academic journal may accept as an appropriate contribution to its discourse is perhaps bound to be rejected by another. Interpretations may always be provisional and depend on the discourse which brings them about, but they are not altogether arbitrary. The readings offered have to be argued for. If I say something without a shred of evidence, people will reject it and it will leave no traces in the subsequent discourse. But what is the evidence in the case of a haiku that has never been discussed before? What is the meaning of a new text? As I have argued in Part I of this book, I will disregard whatever might be said about the author’s intentions. A text can have one or many authors, or no author at all, as the books in the example of Jorge Luis Borges’ Library of Babel. If I happen to know the author and ask her what her intentions were, what she tells me would be no more than one of any number of possible interpretations. Her ideas of what the haiku means are only as good as anyone else’s. Introspection as such does not provide shareable evidence. In my interpretation, I will focus on the lexical material that the haiku consists of. I could, of course, also analyse in painstaking detail the syllabic, syntactic, rhythmic or linear structure. But life is short, so I will come back to these aspects only where they help me with my interpretation. The lexical material covers more than the single words and the meanings that can be ascribed to them in isolation. All words are embedded in their context, and it is this context that determines their meaning. What words in isolation mean is often rather ambiguous, vague and fuzzy. When I discuss what words like abrasive, blare, jittery and bouncily mean, I will be guided by the context in which they occur in this
Meaning and the interpretation of a haiku
219
text, i.e. the words to the right and to the left. Equally as important as the single words themselves are the combinations found in the haiku, contiguous or not, e.g. abrasive heron, heron + blare, blue songbirds, blare + stampede, stampede bouncily. Syntax may guide us to disregard some potential combinations. From a syntactic perspective, there is nothing that links heron to bouncily. The haiku tells us that the songbirds may do things bouncily, not the heron. In accordance with the terminology of corpus linguistics, I will call combinations worth looking at collocations. Meaning is only in the discourse. But which discourse? One possible discourse might be the discourse of all haikus, or at least all haikus accessible in English. But that would seem too restrictive. When we interpret poetic texts, we do this in the context of the discourse at large, perhaps constrained by a few parameters like language, period, modality (e.g. writing) and accessibility. What makes a text poetic is the tension between its language and the language of the discourse at large. Baffled by the rather enigmatic text of our haiku, I will choose the largest possible corpus to which I, and my readers, have access. This is the web, as made accessible by Google. It is a segment of the written discourse at large, with an unknown composition and size, and with a preponderance of fairly recent texts. I will only look at English texts. When we look for the meaning of a word or a collocation, we are looking for possible ways to paraphrase them. Such a paraphrase can be something like a dictionary definition. But dictionaries are highly specialised texts, and there are many rules and conventions concerning these definitions. Thus a lexicographic definition must normally be substitutable for all occurrences of the word in the discourse. But many paraphrases that we find in the discourse cannot be generalised. Indeed the ones which help us most with our interpretation are those that we would never expect in a dictionary. There are contexts in which it makes sense to say that ‘[a] heron is a huge target. Hard to miss’; but it is not what a dictionary would tell us. Paraphrases like these abound in the discourse. They will add meaning to the lexical item heron and provide knowledge on the discourse object ‘heron’ beyond the anorak definition found in dictionaries or encyclopaedias. Google gives us 73 hits for the phrase ‘a heron is a’ of which it lists 55, implying that the other 18 are duplicates of the listed ones. A phrase like ‘a heron is a …’ is the most prevalent paraphrase type we find in the discourse. It tells us what a heron can mean. Here are a few examples: A heron is a large wading bird in the family Ardeidae. A Heron is a bird that hangs out around the water; by the beach. A heron is a beautiful creature. A heron is a huge target. Hard to miss. An image is no more a representation of an existing reality than a heron is a representation of the landscape through which it is flying.
220
Meaning, Discourse and Society
A heron is a lonely and solitary creature often found sobbing in ladies washrooms lamenting its gothic heritage. To some, the appearance of a heron is a good omen, but to others, it warns of fire and drought. A Heron is a bad omen, heralding parsimony and evil. Citations like these tell us what is said about the lexical item heron, co-referential with the discourse object ‘heron’. They show us that the discourse is selfreferential. Discourse objects, herons for instance, are constructed and negotiated within the discourse. We can make sense of them without referring to a discourse-external reality, without ever having seen a heron. This is what I mean when I say that meaning is in the discourse. For this approach to meaning, I do not have to take refuge in some inaccessible cognitive concept corresponding to the lexical item heron or to the dubious reality of my sensory perception. The meaning of the lexical item heron is the paraphrastic content of the contexts in which we find it. Sometimes, as in the cases above, these contexts have the form of rather straightforward paraphrases, explaining a lexical item/discourse object in terms of other lexical items/discourse objects: something is something. But paraphrastic content can be found everywhere in the context of a given lexical item. It is also present in contexts in which the lexical item is used rather than talked about. In the following citations, taken from the British National Corpus (BNC), a hundred-million-word public-domain corpus representing British English of the early 1990s, I highlight the paraphrastic content using italics: a tall stooped grey figure, like a heron statuesque heron stared balefully into shallow pools a heron … would hunch on the garage roof as I left for work, and I could well imagine the thoughts of piscicide running through its reptilian brain. This long neck and dignified strut make the heron’s posture an enduring image. A heron scraiked below the steep slabs of stone Forget shrill heron’s stark cry watching a heron gliding around like a poker with a crick in its neck a heron rose on gaunt wings, its legs dangling The old heron … flaps his melancholy wings. That most attractive bird, the heron can be observed more easily Meaning is paraphrase, explicit paraphrases and paraphrastic content taken together. From this perspective the meaning of the lexical item is not someone’s mental concept corresponding to it, and not something existing outside of the discourse. The meaning achieved according to the discourse perspective is neither stable nor finite. It is impossible to compile all the citations featuring
Meaning and the interpretation of a haiku
221
heron in the discourse at large. Even if we focus on special discourses to which we have full access (like the set of all electronic editions of the Times for the year 2007), we still have to decide in a rather arbitrary way how much context we should look at. There could easily be a full-length article on herons in the paper. But herons could also just be briefly mentioned in other contexts, for example in this letter to the editor of the Guardian, published on 9 October 2006: ‘I used to negotiate with a college principal who filled my mind with wonderful images by telling me I had brought a red heron into the argument.’ Any decision as to what counts as relevant can be disputed. Even for those only interested in word meaning (as opposed to world meaning or encyclopaedic meaning) there is no operational answer. My own approach does not distinguish between the two aspects. For me, the meaning of the lexical item heron is the same as all the public knowledge (true or false) we find in the discourse at large for the corresponding lexical item heron. But what is relevant for the interpretation of a given text can only be a subjective decision. Meaning and knowledge are not absolutes. When we want to be told what a lexical item means, we normally do this with a question in mind to which we are looking for an answer. There is something we want to know about this expression, or about the discourse object for which it stands. What we are looking for always depends on the nature of our enquiry. If we want to find out how a heron is distinguished from a cormorant, we search for paraphrastic content different from the kind of meaning/knowledge that might help us to interpret a haiku featuring a heron embodying ‘symbolic power’. If we are interested in haikus in which a heron creates a ‘tension’ with songbirds, we find other relevant citations than those in haikus relating herons to fisheries. As Gadamer tells us, no algorithmic method can ever tell us what is relevant and what is not for an act of interpretation. Embarking on our act of interpretation we have only our prejudices to guide us. But contrasting the evidence of the discourse with them will lead our interpretation into unpredictable waters. Where the hermeneutical circle will eventually take us cannot be predicted. The collaborative act of interpretation cannot be reduced to the intentionalities of each participant. The symbolic interactions between the participants engender a collective intentionality which will feed back in different ways to each participant. The outcome is, in principle, open. The quest for meaning and knowledge which I have sketched here is about the interpretation of a particular text or a text segment, a piece of the discourse. I want to make sense of the haiku in question; I am interested neither in the lexical item heron in general, nor in the ‘heron’ as a discourse object in general. On the one hand, language is the discourse, i.e. this entirety of all that has ever been said or written by members of a discourse community or the discourse community at large. On the other hand, there is the language system, part of which is its lexicon, the entirety of the lexical items. Over the last century, linguistics has almost
222
Meaning, Discourse and Society
exclusively occupied itself with language as a system defined by a set of rules (e.g. for syntax) and a set of elements (e.g. lexical items, morphemes, phonemes). In this way, linguistics wanted to compare itself with the systems we postulate to make up the natural sciences, such as chemistry or physics. My goal here is not to show what heron means in general. Linguistic theories have been competing with each other to provide foolproof methodologies that would render, if scrupulously followed, the ‘true’ meaning of lexical items. This is also the case for corpus linguistics, which has been, since the 1960s, an alternative to these top-down theories: a system based on the actual evidence of real language data, i.e. corpora. Such a bottom-up system will discover no inviolable rules but rather regularities and trends. In the natural sciences, meteorology could be seen as a comparable system: a system allowing predictions with variable force. Corpus linguistics as a science (as opposed to the humanities) needs a strict methodology, so that the same results will come up every time these methods are applied on the same data. As in the case of meteorology, it does not posit laws or rules but works with the notion of statistical significance, claiming that this significance somehow translates into semantic relevance. This methodology has enabled corpus linguistics to provide incontrovertible evidence, based on real language data, of the ways the word heron is commonly used, much superior to previous approaches such as universal grammar or cognitive linguistics, which used introspection and top-down reasoning to tell us about the meaning of a word. Corpus linguistics argues that the meaning of a lexical item can be determined according to the words with which it tends to immediately co-occur in a statistically significant way. Such words are called collocates. Of course, there has to be a sufficiently large corpus. I use here the Bank of English of August 2008, a fairly balanced corpus of c. 600 million running words. For any given lexical item under investigation, statistical operations will deliver a profile which defines its meaning. What these statistics show is how unlikely, and therefore how significant, is the co-occurrence of each word in a profile list with the node word heron, where co-occurrence is defined as occurring within a narrow window of five words left and five words right of the node word. Assuming that statistical significance, translates into semantic relevance, it should then be possible, at least in theory, to construct a paraphrase, using some key collocates, that could serve as a dictionary definition of heron. The table shows three main columns. The first column presents collocates of heron in their order of significance, according to the statistical measurement ‘t-score’. It privileges the more frequent collocates. In the second column we find collocates ranked according to their significance as far as the measurement ‘mutual information’ is concerned. The third column gives collocates in the order of their raw frequency. Collocates occurring in all three lists are underlined. Those occurring only in the first and second lists are given in italics. Collocates occurring in the first and the third lists have an asterisk.
Meaning and the interpretation of a haiku
223
Table 1. Heron: Bank of English statistics t-score patrick scott gil island s rc flight blue and* liz night* great* ronson artist kingfisher chick was* st crowned gerald painter taylor james international by* pippa nolan green abstract mike fish city * lyne purple stakes painting white grey ives 001 1981 tower 2* a
Mutual information 79 66 51 42 149 28 28 28 276 22 27 26 17 14 13 13 93 15 11 11 11 12 13 14 64 9 9 11 9 10 10 14 122 8 8 8 8 12 8 7 7 7 7 17 218 9
8.851986 8.064827 7.136570 6.358066 5.778779 5.283019 5.193544 5.102982 4.984057 4.661392 4.482823 4.333431 4.120879 3.648895 3.597146 3.595990 3.512002 3.463732 3.305039 3.285298 3.284370 3.282829 3.278711 3.065846 3.027831 2.997651 2.982007 2.973864 2.968457 2.964786 2.913201 2.884250 2.847704 2.825549 2.785454 2.751572 2.728219 2.717118 2.710228 2.638379 2.625273 2.556060 2.552501 2.547713 2.522361 2.517748
cyclerama aorg triandis 6730 allyn merganser 0749 lanyon grebe ronson bishopsgate metronome gil pippa 19990322 cormorant lyne herons rc dipper snipe sisterhood 729 crested osprey kingfisher chick ives paragon crowned burghley kempton patrick denton heron nocturnal beak turtles mystique breasted nolan albans liz scott 001 002
6 3 4 4 5 3 4 3 5 17 4 5 51 9 4 4 8 3 28 3 3 3 3 4 3 13 13 7 3 11 3 5 79 3 4 3 3 4 3 3 9 3 22 66 7 3
Raw frequency 15.523910 15.331246 13.576183 13.102205 12.395493 11.938590 11.679830 11.658454 11.497974 10.856020 10.690538 10.567891 10.522654 10.319519 10.094710 9.990863 9.941615 9.302739 9.285664 9.145754 9.026837 8.946905 8.942592 8.808995 8.753161 8.745627 8.559589 8.488213 8.190983 8.161973 7.997359 7.949833 7.940231 7.906339 7.749760 7.692048 7.542567 7.511884 7.497575 7.438675 7.382086 7.356036 7.337060 7.100897 7.014135 6.977266
the and* a of s to in
* was* is patrick for on scott by* at with as gil that his island from it but has one this an blue flight rc he night* or had all great* who said liz be are our which up
373 276 218 171 149 126 125 122 93 81 79 77 76 66 64 60 55 53 51 50 43 42 36 32 32 29 29 28 28 28 28 28 27 27 26 26 26 26 24 23 22 21 21 19 19 18
224
Meaning, Discourse and Society
Table 1. (cont.) t-score corporation lake baby park
Mutual information 7 7 8 9
2.506733 2.504475 2.498620 2.475114
stevie suzuki headingley cooks
4 3 3 4
Raw frequency 6.951845 6.927396 6.803897 6.751075
when will 2* ronson
17 17 17 17
Once we deselect proper names, such as Patrick, Ronson, Liz, Scott, etc., there are few collocates which make it on all three lists. On the whole, the t-score list has the highest number of relevant collocates, such as night, flight, kingfisher, blue, crowned, fish, grey, and lake. To recognise these as relevant we have to have some knowledge about herons, for instance that they can fly, that there are different varieties (crowned, blue, grey herons), that they like fish, are found near lakes, and so on. The mutual information list does not have much to offer. It adds cormorant to the kingfisher of the first list, both being comparable birds. This is how heron is described in the Cobuild English Dictionary (2001, 3rd edition), the first dictionary strictly based on corpus evidence: A heron is a large bird which has long legs and a long beak, and which eats fish.
What is described here is the prototypical discourse object ‘heron’. The definition shows how much even corpus linguists have to rely on knowledge that is not readily available through the application of computational tools to real language data. Without human interpretation, without extracting paraphrastic content from the discourse, it would be impossible to generate dictionary definitions. In addition, there are certain conventions for describing birds in dictionaries. That they have a long beak seems to be more important than that they prefer to hunt at night. Collocation profiles are hard facts. Every time we apply the same statistical procedure to the same corpus, we will come up with exactly the same profile. In this sense, corpus linguistics can claim scientific status, more so than the models provided by the various schools of cognitive linguistics. But the task I have set myself in this chapter is different from this traditional remit of (corpus) linguistics. I am not interested in a general dictionary definition of heron or songbirds that would fit each occurrence of these items throughout the discourse at large (or in a sufficiently comprehensive corpus). What I am looking for are instances in the discourse that can help to throw light on the meaning of these two lexical items as unique occurrences in our haiku. That a ‘heron is a large wading bird in the family Ardeidae’ may not be helpful at all for this task. Their ‘dignified strut’ may well be more to the point. Would my partners in the act of interpretation be inclined to consider it as such? It depends
Meaning and the interpretation of a haiku
225
on the interpretive community, in all its unpredictability. The same community may tomorrow come up with a different interpretation. Different discourse communities will disagree over their interpretations. Nothing symbolic is ever fixed. Meaning and knowledge are always provisional. Interpretation is never science; it belongs to the realm of the Geisteswissenschaften. Understanding, as Wilhelm Dilthey defines the remit of the humanities, always involves taking decisions. Any interpretation has to be selective. The discourse at large is quite boundless. Even the discourse made accessible to us by Google, a minuscule fraction of the discourse at large, is so big that it cannot be exploited as a whole. There are over 21 million occurrences of heron in it. Taken together, they embody the meaning of heron in the web discourse. Each of them would tell us something about the discourse object ‘heron’ and about the word heron. But while a computer can give us the collocates of heron ranked according to their statistical significance, such a profile is not equivalent to an interpretation of the meaning of the word in general, and certainly does not help us to deal with the heron in our haiku. For this, we need the collaborative act of interpretation, involving the collective intentionality of the ‘interpretive community’, as Stanley Fish calls it. What I have selected in the short lists above, guided by my ‘prejudices’ (in the sense that Gadamer uses this word), are the explicit paraphrases and the paraphrastic content from Google and the BNC, and I believe some of them might aid us in the interpretation of the haiku. This is, the supposed interpretive community agrees, what we have learnt so far about the lexical item heron and the discourse object ‘heron’ from these citations: A heron can be found close to the water, by the beach. It displays a tall stooped, statuesque, posture, featuring a long neck and a dignified strut. Its wings are gaunt and perhaps even melancholy. When it flies, it may look like a poker with a crick in its neck. The heron’s cry can be called shrill and stark, a kind of skraiking. It lives by piscicide, and humans may disagree with the baleful intentions ascribed to its mind. Because it presents a huge target, it can easily be shot dead by those who do not like it. Others may find it a beautiful creature and a most attractive bird. Whether herons are seen as a good omen, or a bad omen, warning of fire or drought or heralding parsimony and evil, is an open question. In some cultures, a heron is ridiculed as a king among a hundred crows. More context would be needed to understand why it is said that a heron is a lonely and solitary creature often found in ladies’ washrooms lamenting its gothic heritage.
As long as I tell the other members of the discourse community where my evidence comes from, and as long as they can obtain access to the full data from which I gleaned my evidence, my partners can participate competently in the collaborative act of interpretation. They can declare some of my citations irrelevant, adduce other quotes, and they can dispute the sense I make of them. It is not possible to achieve more than this in a hermeneutic enterprise. Any new interpretive act is free to select other citations from the chosen discourse, and might therefore produce different results.
226
Meaning, Discourse and Society
The heron of the haiku can mean more than just a ‘large wading bird in the family Ardeidae’. The discourse contains citations in which the object ‘heron’ is described as other things, or in which other discourse objects are said to resemble herons. This wider picture is important in the light of what the haiku definition above calls its ‘symbolic power’. The heron can be seen as a sign pointing to a particular aspect of a discourse object, thus saying something new about it, something that has never been said before about a heron. We have learned to expect poetry to be metaphorical in this way. But do such symbols have to be categorised as metaphors? We have all been brought up in the belief that the meaning of lexical items can be summed up in the kind of brief definitions we find in dictionaries. Whatever does not conform to these definitions is often called metaphorical usage. I do not want to engage in a discussion of how helpful the concept of metaphor is. Rather I will show how the word heron is used to characterise or symbolise other discourse objects. Again I turn to Google, and search for citations containing the phrase like a heron. This phrase gives similes rather than metaphors, but I have to choose this option as there is no way to automatically detect metaphors in texts. There are 6,720 occurrences, and the citations below, chosen from the first two hundred of them, might initially be of interest for the interpretation of our haiku: A man who knows the law should not offer even a little water to … the evil man who acts like a heron Like a heron, I pose above its currents, scratch the Rio de Amapari’s belly with my plate. your soul leapt like a heron sailing from the salt, island grass. into another heaven I loved her like a heron, my cursive neck penning letter s to f to l, looping my head into the water When she was in a high wind her light body was blown against trees and banks like a heron’s. He began pacing the room, waiting, like a heron in a pond, waiting for his fish. When the Big Boy took wing up-court, he looked like a heron in full flight. My grown heart rose like a heron, cackled like a goose, my aged heart Lay under the back-bent sky, was a dark stone in the sky’s belt. It is perhaps a bit surprising that herons seem to give the discourse objects to which they are conjoined in a paraphrase a touch of attractiveness and sometimes even glamour. It can be male or female. The haiku’s heron, on the other hand, does not seem particularly attractive, but there might well be something imposing about him (or her), or perhaps just some delusions of grandeur. He or she is abrasive, and discourse objects called abrasive do not normally inspire
Meaning and the interpretation of a haiku
227
sympathy, especially if these objects are called birds. Google lists 77 hits. I have selected a few which could throw some light on their nature: So, it was with some disappointment that I realized the woman living across the street from me is a fanatic for this loud, abrasive bird. Containing the threat posed by this clever yet abrasive bird to our native species and habitat will need more regional cooperation to succeed. ‘Oh,’ replied the bashful fellow, ‘I noticed that my toffees were being carted off by a rather large, abrasive bird, and so I followed him today. This isn’t the rather abrasive bird who gets pissed on lager and tries to pick a fight with the boy she’ll be shagging later, this is the pretty, quiet one in the corner with the unusual scars, who you swear has been stalking you. Again the last citation obviously signifies a female person. There is more to know about abrasive people (and animals including birds). We can tap this knowledge by looking for adjectives conjoined with abrasive by the connector and. Google lists 413,000 hits for abrasive and. Here are some (taken from the first one hundred) that seem more to the point: abrasive and unpleasant blunt, abrasive and scrupulously honest abrasive and land-hungry abrasive and aggressive so abrasive and vicious-tongued an abrasive and persistent campaigner abrasive and morose (well, she’s a girl, ain’t she) he is an obnoxious soul, abrasive and paranoid With the exception of the second citation, abrasive seems to denote rather unpleasant people, be they male or female. Strangely, there is not a single citation among the first one hundred of and abrasive (altogether 86,500 Google hits) that does not tell us about some aspect of the grinding and abrasive technology, a topic that does not enlighten us concerning the heron haiku. In the BNC we find, however, the phrase glib and abrasive. As this might also apply to our heron, I checked this phrase on Google, finding ten hits, among them: Downey’s Stark is glib and abrasive, and he’s ready with a quip when somebody insults him back. (6 occurrences in slight variations) It is easy enough to dismiss Rushdie’s manifesto as a glib and abrasive put-down of South Asian vernacular literature
228
Meaning, Discourse and Society
It is not my intention to be rude and I will try to improve my crits in future to be more helpful and less glib and abrasive In text – in game and on the forums – he seemed glib and abrasive (although I dare say I had a little too much rivalry going on he removes all possibility of nuance from the conversation, in much the same way that glib and abrasive conservatives do Obviously, a glib and abrasive person is someone who likes to make a rash pronouncement but will not listen to others. Is our heron like that? The next question I want to consider is how blares resonates with abrasive. Google lists altogether c. 53,000 hits for blare/blares/blaring/blared (in the following: blare*) in connection with abrasive, among them 20,800 for blares + abrasive. (As an aside, it may be noteworthy that Google suggests looking also for Blair + abrasive [107,000 hits], though I am not sure this helps with our interpretation.) Interestingly, Google also lists a quite endless number of occurrences in which abrasive is connected to music (as blare is, too). Here are some citations: Blending electronic synthesisers, guitars and early samplers, these bands created an aggressive and abrasive music fusing elements of rock with experimental electronic music. This is the type of music that parents don’t mind because it’s edited and screened and is not too harsh and abrasive for themselves to listen to also. The most immediate audible characteristic of my music is its noisiness. Abrasive, loud, fast. This music is rarely aggressive or abrasive. While the sheer number of these citations indicates that there must be many who enjoy this kind of music, most of the texts we find are rather sceptical. Both abrasive and blare* seem to share an often quite negative connotation. Indeed the connection between the two words is well documented in the web discourse. Among the first twenty hits for blares + abrasive we find these citations: From the next room, the television blares, as inane and abrasive in German as it is in English. strikes me as the archetypal Off the Beat track, with its blaring intro, abrasive energy and seeming absence of catchy melodies. Surprisingly there’s a ‘brass band’ instrumental section with horns blaring over the abrasive guitars but it’s very short. The deafening, abrasive sound of ambulance sirens blares through California’s San Fernando Valley. Embracing something that is blaring, abrasive, or inaccessible doesn’t necessarily endow it with any outward meaning.
Meaning and the interpretation of a haiku
229
They are the spokespeople, the loud, abrasive megaphones through which the siren song of this cult blares. According to Google, herons are not said to blare. There is just one citation among the first fifty Google occurrences of bird + blared (of altogether 45,700 occurrences) in which a bird is blaring: Relief flooded Riley as the bird blared overhead. When humans are said to blare, blare is normally used as a transitive verb, like in ‘He blared his horn’. Among the first fifty Google occurrences (of 5,300 altogether) for he blared, there are only these two citations for the intransitive blared (but quite a lot for blared in connection with direct speech): He intoned, he moaned, he pleaded – he blared, he crashed, he thundered. Though he blared like a trumpet, a note of uncertainty could be detected in his voice. ‘Shake it up, baby,’ he blared. Normally it is things like music, horns, sirens, trumpets, loudspeakers and headlines that we find in connection with the intransitive use of blare. The situation for ‘she blared’ is very similar. So far, the discourse according to Google does not contain a single heron that is blaring or abrasive. The first part of our haiku thus adds something new to the discourse. What sense we can make of this new unit of meaning, this new discourse object ‘abrasive heron that blares’, has to be based on those contextualised occurrences of these lexical items which have compatible paraphrastic content that is not self-contradictory. The citations given above are of this kind. This is the interpretation of the first part of our haiku that I want to offer to the interpretive community for discussion: An abrasive heron blares. Something, someone, a bird or a person with heron-like features, is taking centre-stage. He or she does not endear himself/herself to us. There may be something statuesque and lonely about our heron with a tall, stooped, grey figure, a long neck and dignified strut and gaunt, melancholy wings, but we must not forget that herons are baleful. A heron in general can be a good or a bad omen, a beautiful creature and a most attractive bird, but also rather ridiculous when he or she glides around like a poker with a crick in its neck. They make shrill unpleasant noises described as scraiking. Our heron is abrasive, a trait associated with loudness, unpleasantness, aggression, moroseness, persistence and perhaps even paranoia. He or she blares, making sounds we would
230
Meaning, Discourse and Society
normally prefer not to hear. Our heron thus is as off-putting as the music characterised as abrasive (at least for the non-initiated). The sound of our heron, male or female, reminds us of ambulance sirens, and while it is unpleasantly loud it can also be seen as meaningless. This makes it very unlikely that our heron is attractive or beautiful. Perhaps it is a heron that acts like an evil man. An interpretation is a collaborative act. It begins with my reading of a text (segment). But I do not read it for myself. I read it for my audience. I want to see my reading of this haiku discussed by the other members of this interpretive community. I base my reading on the segment of the discourse at large made accessible by Google. The discourse is the only reality that is shareable and directly accessible to us. There I search for citations that enlighten me on the meaning of the words and phrases that make up my text (segment). I present my findings, the citations that I have selected from the web, citations providing the shared raw material for our interpretation of the text (segment). The result of this collaborative act can be a single reading we all more or less agree on, or a number of different readings that we accept as being compatible with the discourse evidence. This act of interpretation takes place within a discourse of a given interpretive community. Other contributions may select other citations from the web. All citations can be read in different ways. New contributions to this interpretive discourse will take up what has been said before, will accept, modify or reject it. The outcome of it all is unpredictable, though perhaps not entirely arbitrary. Each discourse community has its own conventions. Even if it is always possible to flout one or the other of them, they provide the necessary continuity and are therefore relatively stable. Contributions that violate too many conventions will not be accepted, and therefore not referred to, by subsequent contributions. They will have no impact on the interpretation. I will now proceed in my reading of the heron haiku. This is the second half of it: … blue jittery songbirds stampede bouncily
Again, the standard dictionary definition of songbird will not do justice to this unique occurrence of the word. The context in which we find it embedded distinguishes our songbirds from all other occurrences. This is the Cobuild Dictionary definition: ‘A songbird is a bird that produces musical sounds which are like singing’. This is very general, and it is unclear if it bears any relevance for our songbirds. So what can we find in the discourse that throws some light on them as we encounter them in the discourse? What do we learn from investigating the context words in the discourse at large? Is there anything that might apply to the songbirds in our haiku? The first step is always to look for straightforward paraphrases of the type ‘a songbird is …’. Google lists c. 550
Meaning and the interpretation of a haiku
231
occurrences. Few of them actually paraphrase the discourse object ‘songbird’, and to the extent they do they are overwhelmingly concerned with the prototypical songbird. Here is a selection of paraphrases taken from the first fifty hits: A songbird is a bird belonging to the suborder Passeri of Passeriformes. A Songbird is a bird that chirps in soothing tones that sound like musical notes. A songbird is small. A songbird is concerned with food, too, but also with enemies and potential mates. As the songbirds in our haiku occur in the plural, it might be useful to also investigate the phrase ‘songbirds are’. Google lists 49,000 occurrences. This is a selection of paraphrases taken from the first one hundred of them. Technically songbirds are members of a particular group, the suborder Oscines Songbirds are small and very adaptable. It is common knowledge now that British songbirds are in trouble Many songbirds are killed by house cats. Tens of thousands of wild British songbirds are being trapped illegally Songbirds are perching birds Songbirds are always farting. In Indonesia, songbirds are prized pets. Songbirds are our national heritage. All our recent Songbirds are invited to perform in what could be a very exciting Showcase. ‘Songbirds are not dumb; They don’t buy a crumb of bread, It’s said.’ ‘Summer’s gone and no songbirds are singing.’ Some of these paraphrases could be relevant for our haiku. We certainly learn that our songbirds are highly appreciated, by cats, as pets, as delicacies perhaps, and as part of our national heritage. Do they stampede because they notice a cat, or because they fear to be trapped, or because the summer is over? I will add three citations with paraphrastic content (shown here in italics) from the 15 occurrences of songbirds in the BNC: Her heart was like a songbird in the morning and she was as lovely as the lilies of the field. I stayed at home and listened to the Top Forty top volume while I shouted out Nora’s big speech from The Doll’s House: ‘I was simply your little songbird, your doll.’ We must have words with your songbird.
232
Meaning, Discourse and Society
This small selection of citations does the full meaning of the lexical item songbird little justice, but it allows us to come up with this new paraphrase: Songbirds are appreciated, but also endangered. They chirp in soothing terms that sound like musical notes. They are small and very adaptable. They are preoccupied with food, enemies and potential mates. British songbirds are a national heritage and they are in trouble. Many are killed by house cats, and many are being trapped illegally. Elsewhere they are kept as pets, put into cages and hung in trees. Some people say that songbirds are also farting. Still, they are everybody’s favourites. Lovely girls’ hearts can be compared to them. Also some women may call themselves their partner’s songbird. Humans called songbirds show up in TV shows, and one can have words with them. They feature occasionally in lyrics.
Of course, much more has been said about songbirds. The web offers only a minute fraction of the post-World-War-II English language discourse at large. My selection of citations is supposed to help with the interpretation of our haiku, and was not intended as a ‘representative’ sample of the web discourse. But I did not make them up. I did not use my introspection. My paraphrase only contains things that have been said in the discourse. It certainly constitutes evidence, perhaps not to the liking of everyone in the discourse community to which I present my interpretation. Other members are welcome to present additional evidence. The next item I will consider is jittery. As a first approach, I will present jittery in connection with other adjectives as found by Google (jittery and, and jittery). We can assume that these adjectives will be partly synonymous with jittery, as is often the case with binomials. There are 142,000 occurrences of jittery and, and 87,000 occurrences of and jittery. My citations are taken from the first fifty listings of each: many importers are still jittery and unsure of what to expect I took 50 mg of diet ephedra i feel jittery and not good. i am SO jittery and hyper and in a good mood I get very jittery and shaky Six months ago, I was feeling jittery and unable to get any sleep. Can decongestants make you feel irritable and jittery Are you nervous and Jittery? well, i’m antsy and jittery–just thought you should know. I’m jumpy and jittery all of a sudden. I’m so jumpy and jittery and can’t sit still for a minute Someone who is jittery is often also jumpy, shaky, antsy and irritable. If one feels jittery, one can be in a good, even exhilarated mood, but it can also indicate that one isn’t feeling well or that one is unsure. While Google does not feature jittery songbirds, it lists seventy-five jittery birds. Here are a few citations that might help to make sense of the haiku:
Meaning and the interpretation of a haiku
233
The wheat-green sea of the Hebrides smashes in a grey cave among dewed, jittery birds. They are very quirky, jittery birds. They skittered and jerked around like two jittery birds. Like a pair of some kind of weird, jittery birds, Moody thought. I can imagine that having a resident hawk could result in a lot of jittery birds! She exclaims again and again how amazing are two rambunctious squirrels chasing jittery birds across high branches Jittery birds can be quirky, we learn; they skitter and jerk around. People are said to behave like jittery birds. And we learn that birds feeling threatened become jittery. Indeed, jittery behaviour, whether of people of birds, is often linked to intentional content. If people are confronted with a situation that forces them to take action, they become jittery. Google lists 52,000 occurrences of jittery about. Here are some of them, taken from the first fifty items: Air Carriers Jittery About Fuel Prices. Green activists jittery about Dubai’s man-made isles. Strike talk leaves customers jittery about United’s future Danish consumers jittery about the future Why Economists Are Jittery about the Stock Market. When people (or birds) collectively become jittery about something, this may be an indication that they are in a negotiation about a course of action. This could be relevant for our haiku. Can songbirds be described as bouncy? After all, birds, songbirds at least, can fly. It would be our expectation that birds, when threatened, would fly off, rather than bounce. But Google lists quite a number of cases (2,010) of bouncy birds. There even seems to be a pop band by the name of Bouncy Birds. For the haiku, these citations could be telling: Bouncy Birds. Rockhopper penguins do just what their name says. House sparrows are little, bouncy birds that hop around on the ground in small flocks while foraging As these little bouncy birds flock through our locale, many frequent backyard feeders, enriching our day with their energetic antics. When your bright, glowing days offer a welcome to the most colorful and beautiful flowers naturally grown to offer a habitat for bouncy birds Penguins cannot fly. They bounce, though, we are told. But there also seem to be bouncy sparrows, and birds that frequent bird feeders (like songbirds), and birds
234
Meaning, Discourse and Society
who tend to nest among beautiful flowers. Among the first fifty citations for bouncing birds, I have selected these: I named ’em bouncing birds for the way they play the air currents. Although I guess it was worth it to see that one pic above, not to mention the bouncing birds actually made me laugh. A noisy flock of goldfinches arrived in Gomps by mid-April; the song of these bright yellow bouncing birds is a sign that spring is on the way. He set up one of those bouncing birds to click on a button but later the bird ending up hitting a wrong button and setting off the nuclear plant. Bouncing birds, including goldfinches (which are a kind of songbird) really do occur in the discourse, and even can be seen on pictures. Even in the act of flying they can be said to be bouncing. There are, unexpectedly, also two citations for bouncy songbird(s): This song showcases what a bouncy songbird Miss Bush truly is and in no uncertain terms, and any person with the gift of hearing is not safe from her Ha Ha! I hear you bouncy songbirds! In both cases, the songbirds are obviously human (and probably in the music business). We cannot be quite sure how much fun it is to encounter them. While jittery often seems to indicate a feeling of nervousness and excitability, the citations above make bouncing birds appear as being very much in their element. So do jittery + bouncy/bouncing go together? Quite well, it seems, as these citations, selected from the first fifty of 24,000 listed by Google, show: The road feel was solid but not jittery or bouncy. Overall, the ride is smooth but is jittery and bouncy when taking on rough surfaces The jittery, bouncy drums and chopped keys on ‘The Bones Of My Pets’ show that even when not rocking with the band A bouncy, jittery arrangement provides the contradiction in a track peppered with random sonic tidbits His whole jittery Miyavi bouncy persona was gone. He was calm, and kinda serious and polite. She was all bouncy and jittery like a little kid who has been too long in a car ride and needs to use the restroom Also, I feel very jittery and bouncy. This feeling typically subsides within a short period of time, but it is not a pleasant feeling.
Meaning and the interpretation of a haiku
235
There are three context types in which jittery and bouncy seem to co-occur. One has to do with transport. Rides or flights or roads can be jittery and bouncy, and it is not something that is appreciated. In music, there seems to be a certain ambivalence. When the adjectives are applied to people, they indicate a state of behaviour that is felt as unpleasant by those who exhibit it, and by those who are confronted with it. Songbirds who are jittery and bouncy are not calm; they are excited and on the brink of doing something. So far, stampeding songbirds have not been part of the discourse to which Google gives us access. If birds are said to stampede at all, then the talk is frequently about birds which cannot or do not normally fly, as in these two citations (out of six): and sets fire to the warbird stables, creating a considerable diversion, and is almost run over by the stampeding birds. birds can leave their nest exposed to predators and heat, and ‘stampeding’ birds can crush eggs. However, there is textual evidence of birds stampeding before or while they begin to fly: They swung open the pen doors and the birds stampeded out onto the runway looking very excited to get underway … I poured on the coals to begin my takeoff – desperately trying to stay ahead of the birds who were airborne and flying right beside me. Not far from his goal he came to a steep precipice and the birds stampeded, not down the cliff but into the air. Suddenly, the birds stampeded. They spread their wings, flew over the precipice and raced away. Why do birds, other animals or people stampede? Here are a few citations taken from the BNC, which lists fifteen occurrences of stampeded: He’d used an Armalite semi-auto on the creatures as they stampeded from crag to crag. Nobody was hurt, but the congregation panicked and stampeded for the exit. Then, as if to break the spell, the sheep dogs began to bark frenziedly and the nearby sheep stampeded as if they sensed impending danger. The Government are being stampeded by a bunch of loony, rabid, right-wing fanatics But if Gloucester chose to act against Hastings, rather than being stampeded into it by his discovery of a conspiracy, why did he
236
Meaning, Discourse and Society
choose to show his hand before his troops reached London and before he had control of the other possible claimants to the throne? But while England and, for that matter, Scotland, are being stampeded into adopting them before the ink on the bill is even dry, there is a definite lack of urgency west of Offa’s Dyke. Animals and people stampede when they act instinctively, when they forsake reason. According to the discourse, this happens normally when they feel threatened, when they panic, when they sense danger. Something must have happened that stampeded our songbirds into spreading their wings and racing away. As to the second part of our haiku, there is still one question unanswered. What does the discourse have to say about blue songbirds? Google lists 277 occurrences of blue songbirds. In an astonishingly high number, they are equipped with highly desirable attributes. Here are seven citations, taken from the first forty occurrences: Keep building those nesting boxes so we can continue to increase the population of these beautifully colored blue songbirds in our state. Barn swallows are pretty blue songbirds with forked tails. Oil Painting on Canvas depicting the small blue songbirds perched among the lilac flowers. mountains unfolding layer after layer on the horizon, a pasture full of bright blue songbirds. Sure the fairy princesses and blue songbirds are kind of cute, but let’s face it, they’re getting overplayed. I had a bunch of guinea-corn to cook, but this week two pairs of indigo buntings (brilliantly beautiful azure-blue songbirds) ate ALL of it Overhead a cluster of tiny blue songbirds twittered to each other in the boughs of the Rynacca trees. Their cheerful calls brought a smile to Leia’s face. Blue songbirds are pretty, bright, brilliantly beautiful, beautifully coloured, have cheerful calls, come with fairy princesses and are the subjects of paintings. There can be little doubt: while the discourse object of an abrasive heron hardly evokes any feelings of sympathy, it is part of the meaning of blue songbirds that people tend to find them attractive. Google lists an astonishing number of 600,000 occurrences for heron + songbirds. A look at the first fifty occurrences gives us a clue as to how they are seen to be connected: Endangered there are all diving birds, herons, songbirds, and terns, as well as most ducks, geese, falcons, rails, and wading birds.
Meaning and the interpretation of a haiku
237
The refuge is also used by peregrine falcon, bald eagle, northern harrier, Caspian tern, great blue heron, songbirds, and a variety of waterfowl. The riparian trees provide nesting habitat for a variety of raptors, herons, and songbirds. We’ll search for wildlife visiting the lake at twilight such as osprey, herons, beavers, and songbirds. A whippoorwill joined the chorus of croaking and barking herons and chirping frogs, with songbirds tittering and tweeting occasionally. Herons and songbirds share the same type of habitat. Other birds and other animals are to be found in these places, as well. There is, however, no symbiotic relationship between them. Do they get into each other’s way? Is that why the heron is blaring? Is he or she frightening our blue songbirds? We are told that herons are likely to commit the crime of piscicide, and thus we would probably be told if they also murdered songbirds. The silence of the discourse thus indicates that they are safe from herons. As we have learned above, a ‘haiku is the smallest language construct that can generate enough complexity to create tension and resonance between its parts’. Part of our interpretive task is to find out what kind of tension exists between the two parts of this haiku. The key question is how the two verbs, blare and stampede, are connected. Above we have seen that a perceived danger, or the ensuing panic and fright cause animals and people to stampede. The Google discourse tells us that blaring can indeed be interconnected with panic and fright. There are 66,000 occurrences for panic + blare, and 44,000 occurrences for fright + blare. Here are some relevant citations, taken from the first few occurrences: Wash’s panicked voice blared over their coms. ‘Simon, you need to get back here. Something’s wrong with Zoe.’ That afternoon, our family was riding in the car when a siren blared. I panicked and asked, ‘What’s that?’ Welcome to the latest health scare on the Internet, where e-mail or ‘consumer alerts’ can suddenly spark panic by blaring about ‘DEADLY POISONS.’ As I hear the Sirens Blaring I sit frightened by the feeling growing within me Frightened by the sirens blaring outside my window And on Halloween night, a steady stream of frightened and tormented wailing blared from a set of speakers on the front porch. But the way I frightened the folks and the way they ran and hollowed and blared the tin trumpet … made about as sweet a row as you can imagine.
238
Meaning, Discourse and Society
It seems that in a large number of citations the blaring causes panicky behaviour. But the first, fifth and sixth example the fright or panic happens first, and causes voices to blare. As I see it, our haiku does not provide an easy answer to the question of whether the songbirds stampeded because they were frightened by the heron’s blaring, croaking voice, or whether there was an incident causing panic and fright to the heron and to our songbirds alike. Several readings are possible. The heron, in his (or her) statuesque grandeur, admirable but also repulsive, produced an abrasive, ‘skraiking’ sound that frightened the songbirds and made them stampede into flight. This would be unreasonable behaviour, as the reality constructed in the discourse does not show that herons pose a threat to songbirds. Sharing the same habitat, songbirds, we might assume, are genetically adapted to pay no undue attention to herons. But another reading is equally possible. The heron’s blaring can be processed by the songbirds as an indication of an imminent danger. For them, it may have the function a klaxon has for us. Not knowing what troubles the heron, making a quick escape may well be the best choice of action. For me, the preferred reading is that the haiku shows us two different ways of reacting to a peril. What the peril is need not be specified. It is sufficiently indicated by the heron’s blaring and by the songbirds’ stampeding. What is contrasted are two modes of behaviour. The heron’s bearing is defined by a pretence of grandeur, majesty and posturing, by strutting along in a presumption of dignity, and by a belief in one’s invincibility. Our heron has an abrasive temperament, and this quality is often linked to negative qualities such as being aggressive, morose, and obnoxious. The heron, it seems, is the bad guy in this haiku, clearly in stark contrast with our blue songbirds. Solitary, and equipped with a reptilian brain, it is complacent and self-righteous in its pursuit of shameless piscicide. It does not interact with its environment. It feels invincible. Thus, perhaps unsurprisingly, the heron only pays lip service to the peril by blaring out indignantly without really feeling threatened. The songbirds, on the other hand, do not pretend to know better in the face of danger. They find little value in the ramblings of solitary minds. They conduct themselves as an egalitarian community. What we call jittery may be their way of interacting, of dealing collectively with their perceptions, and of negotiating what to do next. What looks like a mindless stampede to our Western minds, fixated as they are on hierarchical societies with strong leaders, can also be seen as a collaborative act of escape from a more or less clearly perceived danger. So while the songbirds are up and away, our heron presents herself or himself unwittingly as ‘a huge target, hard to miss’, as we have read above. This haiku can be seen to convey a moral message. Herons and songbirds have been likened time and again to people, and people to all kinds of birds. What we make of it is, however, hardly an act of interpretation, because it is not
Meaning and the interpretation of a haiku
239
based on concrete discourse evidence. Interpretation, however, cannot be strictly separated from the construction of new realities, discourse realities waiting to be interpreted in their turn. The whole discourse can be seen in its diachronic dimension as unceasingly adding layers of interpretation upon existing layers of interpretation. The haiku has added some new touches to the reality as it had been constructed by the discourse. For apparently the first time, there is now discourse evidence that herons can be abrasive, that there is a discourse object ‘jittery songbirds’ and that songbirds can be said to be stampeding. This is new potential knowledge, and for social constructionists it does not matter whether realists of various denominations would call it true or false. What matters is whether subsequent contributions to the discourse will refer to it. Only then will it have made an impact on the discourse, be it the discourse at large, the discourse as it is represented by Google, or the discourse of the small community that aims to collaboratively interpret this haiku. My reading of our haiku will be initially determined by my prejudices, and so would everyone else’s. I was bound to select corpus citations as I saw fit. But the reality of the citations led me in ways I did not foresee. Gadamer would call it a dialogue between the interpreter and their evidence. So far, the interpretive community is only imaginary. But I have selected and presented my case in such a way that I hope to find some readers to agree with my reading. This interpretation can only be provisional. Readers will want to pursue different alleys, select different citations from the discourse and rephrase them as they see fit. Any new contribution, like my interpretation, will only have made an impact on the discourse if subsequent contributions refer to it. Otherwise it will have been a useless, solitary exercise. The discourse is, of necessity, the result of collaboration. There will never be a final interpretation. There is an unlimited terrain for making sense of a text. As long as readings of it can be justified by discourse evidence, by the reality of citations taken from the discourse, which is the only reality available to us all, they count. As I see it, it is not necessary that the ‘interpretive community’ evoked by Stanley Fish agrees on just one reading. Several readings may be endorsed, and together they make up the interpretation as it has been negotiated by this community. Other readings may be rejected, based on a lack of support from discourse evidence. This may lead to their dismissal and a subsequent lack of references to them, but it could also result in a split of the community. In this chapter, I wanted to show that it is never possible to grasp fully the meaning of a word, a phrase, a text segment of a full text. The meaning of any unique discourse occurrence results from its intertextual links with all other discourse occurrences. But it is we, the members of the interpretive community, who have to decide what constitutes such a link. And it is we who have to select
240
Meaning, Discourse and Society
the links to follow up. While our starting point is the prejudices, the foreknowledge and the fore-judgements we have learned from authority, our dialogue with the evidence of the discourse will liberate us from these preconceived ideas and take us to new shores. So will the interaction between the members of the interpretive community. The ensuing collective intentionality will be more than the agglomeration of the individual intentionalities. All discourse is self-referential. Whenever we say something, we do no more or less than interpret what has been said before. New interpretations recombine, permute and rephrase what has been there all along. But this is how innovation comes about. Only by interpreting the world created by the discourse can we change this world. We will never know at the beginning where this collaborative act will take us. I readily admit that I could have presented my reading of this haiku without wasting ten thousand words on a tedious, unexciting and wearisome process of explication. But this exercise was less about making sense of it than an illustration of the hermeneutic act itself, as I see it. I wanted to show how unnecessary it really is for our sense-making efforts to seek the answer in our individual minds. My point is that we have little to gain from undertaking the vain endeavour of accessing them. Indeed, cognitive linguists have never been able to demonstrate the existence of any mental concept, any cognitive representation, that could not also be extracted from the discourse. We do not need to look into our heads to interpret texts. We also do not need to look into the head of the author. In this whole chapter I have never referred to him or her. What she may have thought, which intentions she may have made, is entirely irrelevant. Most readers will have guessed this by now anyway. Our haiku does not have an author. It is the random output of a computer program generating haikus, namely ‘the Genuine Haiku Generator’ (www.everypoet.com/haiku/default.htm).
Conclusion
Authorship, intentionality and mental states: can the quest for meaning dispense with the investigation of the solitary mind? In this book, I have presented my ideas on meaning, discourse and society. Meaning, I have said, is to be found only in the discourse. If it is also in people’s heads, we would not have access to it there. Our quest for meaning would not be helped by an attempt to construe a monadic mind endowed with a mechanism for processing mental states. My contention is two-fold: first, there is no such object as the ‘mind’ whose workings we might observe directly. If there are mental representations of what is said, we do not have direct access to them. We have to rely on what people tell us about them. What we then get, are people’s contributions to a discourse, thus part of the discourse, and not the mental representations themselves. Second, as long as mental representations are confined to the mind and not expressed in any kind of behaviour, including symbolic behaviour, they are meaningless. What someone feels inside but studiously refrains from expressing in any form cannot have an effect on society. It is irrelevant. This leads us, I believe, to the unavoidable conclusion that whatever is said about a person’s intentions cannot be based on knowledge of that person’s mental states. We have no way of knowing what it is that made a person express herself in a certain way. We cannot know if her behaviour was caused by physiological processes, by carrying out a conscious intention or by sheer randomness. It is her behaviour that matters, not her mental state. If we want to make sense of a contribution to a discourse, or to the discourse at large, we have to abstract from the intentionality of its author and her intentions. We need to find out what the contribution means. Whether it was said by someone in full control of their mind, or whether it is but the output of a computer program is not for us to know. We investigate the meaning of any contribution under the assumption that it has a meaning. This meaning is not just another word for the intentions of the speaker. Meaning, I have argued throughout this book, is not a mental representation of what is said; meaning is everything said in the discourse that impacts on a text or a text segment and helps to interpret it. As I have shown in the preceding chapter, it is even possible to make sense of a haiku not conceived by a conscious mind but generated by a computer program. Whether the poem has an author or not is quite irrelevant. It has no effect on its meaning. 241
242
Meaning, Discourse and Society
Making sense of what is said, interpreting contributions to the discourse, however, is an act. There is a long-standing discourse on the nature or essence of what it takes to look at something happening as an act. An act, we have learnt, presupposes agency, and someone who exercises it. An act involves making decisions. Computers, or brains, do not make decisions. They process input by following rules or instructions. The results they come up with are predictable. Acts, on the other hand, are, from the outside, indistinguishable from random events. We cannot tell in advance whether the flapping of a butterfly’s wings will cause a hurricane. From the point of view of intentionality, decisions are of a different essence. When I make a decision, I have viewed the pertinent evidence, reflected on what I want to see as the outcome and considered my options before I resolve the issue. I could have decided differently. I am aware of what I am doing. I exercise agency. But just as we cannot foretell whether there will be, in a given situation, a hurricane, or a tsunami, we cannot predict how a person will decide confronted with a situation. Event and act are both contingent. If we only look at what has been said, at the contributions to the discourse, there is not much to be gained by attributing it to the intentionality of the speaker. A decision that can go either way defies causal explanation. When I claim that texts mean regardless of what their speakers think, I am contributing to a discourse that sees discourse as an autopoietic, self-referential system. When I claim that interpreting a text is an activity, I seem to be participating in a different discourse, a discourse in which concepts such as consciousness and intentionality are taken for granted, in which acting means carrying out a preconceived plan. Hermeneutics, the art, or the craft, of interpretation, has always presupposed intentionality on the part of those who carry out the interpretation. This is why computers may be able to summarise texts, but cannot interpret them. But now I say the texts that are interpreted (which are interpretations in their own right) do not presuppose intentional authors. What happens to the intentionality of those who offered these interpretations? For if we accept that the discourse is selfreferential, that it does not refer to a discourse-external reality, then everything said is a comment on what has been said before. Thus there seems to be a contradiction that invalidates my whole argument. If for the interpretation of an interpretation its author’s intentions are irrelevant, then why should we assume that the new interpretation must be understood not as a process but as an act involving intentionality? Would it not be better to redefine interpretation as a process? In the philosophy of mind, a mental state is viewed as the correlation between a physiological state, for instance the ‘firing’ of a cluster of neurons or the activation of synaptic connections, and the behaviour a person exhibits, including her symbolic behaviour, i.e. her participation in a discourse. Frequently mentioned as mental states we find things such as beliefs and desires, emotions, memory, recognition and intention. What I am interested in here are intentions. Are we
Conclusion
243
in charge of them or do they come about by neural processes over which we have no control? As I have argued in the first part of this book, there is no right or wrong answer to this question. There are only different ways of talking about these issues. If someone is convicted of a criminal act, that person is seen as capable of making a decision about whether or not to carry out the crime. But the perpetrator could also be acquitted on the grounds of diminished responsibility, implying that she or he was not in a position to make a conscious decision. While the standard assumption is that adult people, or, in the case of Scotland, even children at the age of eight, are equipped with a sense of right and wrong and are to be held responsible for their acts, explicit arguments are needed to exempt someone from the default assumption that they are responsible. There are two major discourses concerning the workings of the mind, and they take opposite stances. On the one hand we find the mainstream discourse of the cognitive sciences. For them the mind is a computational engine, embedded in the hardware of the brain. It is driven by its routines, whether hard-wired or acquired, and even if their outcome is (still) unpredictable, it is not controlled by intentionality, by our conscious planning or by the decisions a person takes. The mind carries out processes, but it is neither a conscious agent, nor is there a conscious agent, a homunculus, in charge of it. Successive theories and models of the mind have been developed, and while there are (still) some unresolved issues, it is claimed that every new model emulates the mind better than the previous one. This proves, we are told, that the cognitive sciences are on the right track. This discourse has been exceedingly successful. It pledges that it is within our power to understand the mind, and that we can use this knowledge for unparalleled technological advancement. Intentionality can be disregarded, as a supervenient phenomenon, or as an illusion. As I have already pointed out in Chapter 7, for Daniel Dennett, the mind is a complex system giving us the impression that that there is a ‘whole wonderful person’. But in reality, he says, there is no such thing. The whole wonderful person is an illusion. What we find instead in our investigation is something else. There are ‘brains – more particularly parts of our brains – [that] engage in processes that are strikingly like guessing, deciding, believing, jumping to conclusions, etc.’. Once we decompose this holistic view into its elements, it is bound to disappear like a mirage at close sight. ‘[W]e can make progress by breaking down the whole wonderful person into subpersons of sorts, agentlike systems that have part of the prowess of a person, and then these homunculi can be broken down further into still simpler, less personlike agents, and so forth – a finite, not infinite regress that bottoms out when we reach agents so stupid that they can be replaced by a machine’ (Dennett 2007: 86ff., emphasis in original). It is this approach that has laid the foundation of artificial intelligence. Expert systems, surpassing the hormonal limitations of reason to which we humans are subject, will show the way towards a world in which unbiased
244
Meaning, Discourse and Society
rationality obliterates the need for human decision making. This agenda, a programme that defines intelligence as reducible to the processing of data on the basis of rules and probabilities, has been endorsed by legions of expert scientists. It is, we are told, strictly based on proven facts. If doubts were possible, it would not have attracted billions of dollars. In this view, interpretation is what we get once our tools for text summarisation are so refined that they match the intelligence of a ‘whole wonderful person’. This discourse, the discourse of the cognitive sciences, of the philosophy of mind, and increasingly also of the neural sciences, is in opposition to a counterdiscourse, to what is commonly called by these experts the discourse of ‘folk psychology’. In his revised entry to the Stanford Encyclopedia of Philosophy, Ian Ravenscroft defines the ‘external’ concept first of all as: a theory of mind implicit in our everyday talk about mental states. In the everyday traffic of our lives we make remarks linking sensory experiences to mental states; mental states to other mental states; and mental states to behavior. Thus we remark that the smell of freshly baked bread made Sally feel hungry; that Sally wanted to go on a diet because she thought that she was overweight; and that Sally went to the fridge because she desired a piece of chocolate cake. (Ravenscroft 2004)
It is unclear to what extent our ways of making sense of other people’s behaviour are learnt or innate. Ravenscroft complements this view with an ‘internal’ concept of folk psychology, a kind of mechanism that directs our everyday ways of talking about mental states but is largely ‘inaccessible to consciousness’, while the ‘external’ folk psychology is ‘available to conscious reflection’ (Ravenscroft 2004). Within the discourse of folk psychology, we are in the habit of making assumptions about the mental states underlying our own behaviour and that of others. We believe that people often, but not always, know what they are doing and that they have reasons. People can, for instance, be mistaken about their feelings. They can mistake envy for love. But they can become aware of such mistakes. They can reflect on their behaviour. They can be expected to behave responsibly. When they make a contribution to the discourse, they will be in control of what they say. An author of a text cannot disclaim responsibility for it. They must know what they mean. But there may be more to the meaning of a text than its author is aware of. As already said in Chapter 15, the task of the interpreter is, in Friedrich Schleiermacher’s words, to understand the text at first as well and then even better than its author. Since we have no direct knowledge of what was in the author’s mind, we must try to become aware of many things of which he himself may have been unconscious, except insofar as he reflects his own work and becomes his own reader. (Schleiermacher 1986: 83)
The important part of this citation is the demand to understand a text as well as its author. This was assumed in hermeneutic philosophy during the age of Romanticism. Even today, it is not uncommon to claim a knowledge of an author’s intentions, just as we tend to believe we can know why people behave
Conclusion
245
the way they do. We easily talk about their feelings, their desires, their beliefs and attitudes. Google lists 3,320 hits for ‘the author wanted to say’ and another 1,450 for the same sentence in the present tense. We are even admonished to distinguish between our and the author’s understanding of their text: ‘Ensure that you have understood what the author wants to say, and not what you think the author wants to say’, we read on the web page of the National University of Singapore under the heading Reading Journal Articles (www.cdtl.nus.edu.sg/success/sl12.htm). In a book review, David Wootton tells us, ‘Shakespeare’s intention was to engage with the views of his audience, rather than to pioneer new ways of thinking’ (Wootton 2004), one of 301 occurrences for the phrase ‘shakespeares intention was to’. Are we really innately primed to talk about ourselves and about others as if we had knowledge about their mental states? Is there some kind of genetically inherited mechanism that drives us to explain behaviour, whether our own or that of other people, in terms of their thoughts, feelings or intentions? Or is this way of talking specific to certain discourse communities, for instance to the educated upper middle class society we find in the Western world from about the middle of the eighteenth century? It is interesting to see that our modern concept of empathy seems to have arisen with the Romantic movement. The word empathy is the English rendition of German Einfühlung, which was only coined in the second half of the nineteenth century, either in 1858 by the philosopher Rudolf Lotze, or, as some claim, in 1873 by Robert Vischer in his theory of aesthetic sensibility. But the verb einfühlen (‘empathise’) and other expressions largely synonymous with it seem to have been around much longer. It was only when people began to define themselves by their internal, mental states and not by their social relationships, that a need for empathy came about. Then it was no longer possible to explain people’s behaviour by the norms applying to a specific situation and to people of a specific social status. In Defoe’s Moll Flanders, none of the characters ever feels an urge to empathise. The feelings people expressed in these pre-Romantic times were determined by the rules governing social life. If someone violated the conventions, it was perhaps seen as an act of rebellion. But it was not discussed whether such behaviour was the expression of a character’s ‘true’ feelings. It was only the doctrine of the autonomous person, of the uniqueness of people’s interior life, that engendered the ubiquitous discourse on mental states we find today in the Western world. In his paper ‘Why folk psychology is not universal’, David Ohreen presents the case of Samoan culture which ‘doesn’t seem to take much interest in the mind the way we do … According to Samoan epistemology, people are not supposed to guess what another person could be thinking because there is a strong sentiment that they cannot know what others are thinking’ (Ohreen 2007). Alessandro Duranti also doubts whether the Samoans would, on their own and without missionary intervention, talk about intentions. According to him, they do not even have a word for this concept (Duranti 2006).
246
Meaning, Discourse and Society
Thus a third discourse seems to be possible. We can deal with texts and their meanings not as psychological but as social phenomena. From this perspective, they would neither be brought about by mental processes, by the interactions of a multitude of stupid mechanisms, nor would they be the outcome of a person’s intentionality, planned and executed by a conscious mind. Rather texts would be embedded in discourses, and their meanings could be inferred by their adherence to and deviation from the conventions underlying the discourse to which they belong. All we are concerned with are the texts once they are entered into the discourse. In such a discourse the question of whether a text is the result of mental processes or the enactment of a deliberate stratagem would be outside our remit. But what about interpretation? As I have said above, interpreting contributions to the discourse is an activity. As such, it presupposes agency and intentionality. But what appears to be a contradiction is easily resolved once we accept that agency and intentionality do not have to be confined to individual minds, to autonomous persons. We can talk about interpretation as a collaborative activity, as an act unfolding in a virtual or a real interpretive community. Once the relevant textual and intertextual evidence is presented (as I have shown in the preceding chapter) and shared, each member is called upon to react to it. What any contributor will come up with is rarely something new. Over time, interpretive communities have given rise to a discourse of interpretation, a vast repertoire containing almost everything that could be said when one is asked to comment on a text. In each interpretive situation, a participant will pick one of the options offered by this long-standing discourse of interpretation. The other members of the interpretive community may take for granted that this input will reflect the intentionality of the speaker and the conscious decisions she has made. But what matters is that they accept her contribution as meaningful and as being in order, in conformance with the conventions of what can be said in interpretive communities. What makes an utterance acceptable does not depend on the intentionality of a solitary speaker, on having made the right decisions, but on social norms, the rules governing the language game of interpretation. The next speaker will choose another option, as a reaction to the evidence and to the preceding contribution(s), and so on. Most of what will be said in this situation will thus have been rubber-stamped by repeated successful use in previous comparable situations. What engenders the novelty of each new interpretation is the recombination, the permutation and reformulation of what has been said before. Whether a contribution is based on conscious decisions or the result of a mental (or a real) computer is not for the participants to know, and it does not matter. They react to an utterance as meaningful if it fits their bill of expectations. What is important in such an interpretive discourse is the argumentation that links a new contribution to preceding ones, is what is happening in the open,
Conclusion
247
between the members of the interpretive community. Interpretation is not a solitary act. What a member of the interpretive community thinks about the text is irrelevant. Only what she says counts. The intentionalities of the contributors, the decisions they have made, the intentions they may have had, will not visibly affect the success of their utterances. Success, measured by the number and weight of traces a contribution leaves in subsequent texts, can only be measured ex post, after the contribution has been made. What I call here an interpretation is the collaborative act of an interpretive community in dealing with a text or a set of texts. While unfortunately there are many interpretive situations in which the teacher gives her reading of a text, usually of a poem, which is then customarily accepted by the pupils without further ado, this is not what should happen in a democratically organised community where each member is accorded the same rights. There, at the end of an interpretive act carried out in collaboration, there is a set of contributions, each reacting to the text and to whatever has been said before, which in their diversity, in their plurality, make up the interpretation. What I am trying to say is that what appears to be a contradiction may be resolved on a higher level. Texts can be viewed as contributions to the discourse. What they mean is not how they are represented in the mind of a given reader. What they mean is what they refer to in the discourse and how they are referred to in the discourse. It is the discourse that creates society, by constructing personhood, by turning objects of the environment into people, and by negotiating the meaning conventionally assigned to people’s interactions, whether verbal or otherwise. As Niklas Luhmann has shown (see Chapter 9), it is possible and may be useful to keep the people themselves, their intentionalities and their mental states, outside of such a concept of society, with their symbolic behaviour being the only link between them and society. This concept of discourse does not preclude texts from talking about intentionality, about mental states, about people making decisions and carrying out acts. The discourse is self-referential; texts do not talk about ‘real’ mental states but only recombine, permute and reformulate what has been previously said about these discourse objects. When we are participating in an interpretive situation, as members of an interpretive community, we assign to each of us discursively, whether in an explicit or an implicit way, the status of agents, of ‘whole wonderful persons’, equipped with mental states and intentionality. Thus we can look at an interpretive situation as we can look at a play, an opera, a film or a TV drama. What makes it more interesting to watch such interactions of people than, for instance, the growth of grass, is that we can recognise and understand the situations enacted, and that they show human behaviour in the way it has been presented and explained to us before, that we experience it as relevant for us, and that it engages us as quasi-participants in the unfolding drama. We do not watch it as experts in the communication of ant colonies would watch their ants, for whom the ants are just props. In the
248
Meaning, Discourse and Society
performance of a drama, we see behaviour of the kind we constantly encounter in our own lives, behaviour that determines what is going to happen with ourselves and the people around us. In order to cope, we can never desist from interpreting behaviour (including our own), and from making sense of it. From morning to evening we never cease to contemplate the meaning of behaviour we encounter, of facial expressions, of what people say. We categorise it by using the pigeonholes supplied to us by the discourse. What a gesture, a smile, a tear in the eye, blushing cheeks, and a phrase like ‘I love you’ mean is what has been decreed about them in the discourse. But the discourse is plurivocal and the conventions it exacts are always negotiable; we can choose between the many options offered. It is this personal involvement, this necessity of making choices, that keeps us engaged. This is what made Aeschylus’s plays catch the attention of Athenian audiences, people who had never considered the existence of mental states, the authenticity of feelings, or the idea that people are the makers of their destinies. What I suggest is that we widen our concept of intentionality. It often makes sense to view intentionality as involving a person’s consciousness of herself, an awareness of what someone’s behaviour, a situation she encounters or a text she reads, is about, of her capability of making a plan and carrying it out. It is this perspective that made the Enlightenment so uniquely successful, that has laid the foundation of our modern world. This is what we could call embodied intentionality. But this understanding of individual intentionality has almost erased another intentionality, an intentionality that emerges in situations of social interaction. Intentionality thus becomes the defining force of a community; it is a community’s constant interpretation of the situation in which it finds itself, an intentionality which gives up a crude cause–effect perspective in favour of a Darwinian vision of evolution and adaptation. Any contribution to its discourse, unpredictable as it is in the way it reacts to previous contributions, takes the awareness of the group members a step further, to ever new frontiers. This intentionality is the result of collaboration. We could call it collective or symbolic intentionality. It is the essence of each play, each film and each TV drama. What catches our interest there is the meaning which the characters assign to each other’s behaviour and which we may endorse or reject. What makes us call the happenings of a play actions, and not processes, is that the characters themselves take a vital part in the assignment of meaning. This is what makes a play different from what is happening in an ant colony. What makes an act an act is the self-referentiality of the situation. In this sense, we can talk about acts without invoking the intentionality of a solitary mind. Plays do not offer us any insight into the inner lives of their characters, in the way stories or novels sometimes do. All we can ever see are incomplete holograms of the kind of personhood as it is defined by our modern Western discourse. But this absence of inner lives does not diminish the agency of its characters.
Conclusion
249
An interpretative situation is like a play, with the members of an interpretive community as its characters. It is a play without a script. The participants have to improvise. What turns this situation into an act is not the assumed personhood of the players. It is their symbolic, their verbal behaviour, a behaviour to which meaning has been assigned in the past, which has been interpreted in various ways. We have all been pupils once. Time and again a teacher has tried to discuss the meaning of a poem with us. We have learned the conventions of interpretive situations. We know what we are expected to say and what not. But unless we have a dominant teacher the outcome of the situation is open. Jointly we develop our interpretation, each new contribution being understood as a reaction to previous ones. This is how the participants experience themselves as being involved in a collaborative act. They do not need access to the other participants’ inner lives to make sense of their contributions. The emerging collective intentionality of the interpretive situation is the result of the selfreferentiality of its discourse. The Athenian audiences perceived Aeschylus’s tragedies as action, not as process. They did not waste empathy on the characters. They were not interested in Orestes’ inner life. What they experienced as tragic were not his guilty feelings about the murders he had committed, but the apparent inevitability with which the whole chain of events emerged, each symbolic act provoking the reaction required by the conventions. It is not Orestes’ intentionality they are concerned about; it is the intentionality of the collective mind of all those involved in the play. It is not the intentionality of the one who happens to say something but the interaction of all the members of an interpretive community that determines the result of an interpretive situation. How real are the realities we experience? At the beginning of this book, I introduced the idea that it is the discourse that constructs the reality in which we experience ourselves. Without a discourse, I claimed, our reality is nothing but stuff. This way of looking at things informed the second part of my book. It is no doubt a clear contradiction of our intuitions. It seems to be natural to us to take the reality in which we find ourselves, the reality constructed through unceasing negotiations within the discourse, as undistorted data of our perceptory apparatus. Such an intuition seems to be appropriate for a period of human development that preceded linguistic communication. In those times people could not ask others for advice. Each person had to act on their own intuitions and drives, without collective attempts to make sense of the sensory input. With the advent of the discourse, of the ability to share and exchange symbolic content and to negotiate the meaning of the reality out there, things changed. Surreptitiously, our interpretations began to control and structure what we perceived, while we still believe that we see the world as it ‘really’ is. We seem unable to distinguish between sensory input via
250
Meaning, Discourse and Society
our perceptory apparatus and the input we obtain from participating in the discourse. Sceptical as many philosophers may have been about the link between our ideas and the world out there, few went as far as Bishop Berkeley who conceived his solipsistic version of idealism some three hundred years ago. For Berkeley, there was no reality out there; there were only our perceptions, our ideas. My stance, the stance that the reality which we are able to communicate is a reality constructed by the discourse, does not deny the existence of a world outside of discourses. It only claims that once we accept discourse input on a par with other sensory input, the discursive categorisations we are exposed to tend to take precedence over our intuitions and drives. This is how the stuff out there becomes sorted through the mediation of the discourse. Both positions, though, mine and that of Bishop Berkeley, are relativist. Samuel Johnson would find as much fault with my version of social constructionism as he does with Bishop Berkeley: After we came out of the church, we stood talking for some time together of Bishop Berkeley’s ingenious sophistry to prove the nonexistence of matter, and that every thing in the universe is merely ideal. I observed, that though we are satisfied his doctrine is not true, it is impossible to refute it. I never shall forget the alacrity with which Johnson answered, striking his foot with mighty force against a large stone, till he rebounded from it – ‘I refute it thus.’ (James Boswell: Life of Samuel Johnson book 3)
Aren’t the stones we kick at real? Would kicking a stone hurt otherwise? Would anyone in their sound mind declare that the stone we just kicked at is a construct of the discourse? It is not surprising that the authors of the famous or infamous article ‘Death and furniture’ found themselves at the centre of this never-dying controversy between relativists and realists (Edwards, Ashmore and Potter 1995). Common sense, it seems, has no patience with relativist views. Kicking a large stone hurts. This shows unequivocally not only that stones actually exist, but also that we are able to experience the world out there directly. Or does it? Mountain climbing involves, as far as I can see, a lot of stone kicking. But no one ever mentions that it hurts. It seems that for mountain climbers the reality of rocks and stones is quite different from that of the great lexicographer. When, in the Sunday Times of 27 August 2006, Sir Ranulph Fiennes calls the Eiger ‘his kind of therapy’, the author of the article in question, Stephen Venables, extols the sublime pleasure of climbing the Eiger north wall: ‘a fantastically varied journey through a stupendous vertical landscape’, in which one could ‘be fighting for your life … pounded by a barrage of falling water and rocks’. But apparently those rocks do not hurt. Mountain climbers obviously do not feel pain when they hit or are hit by rocks. It is other things that can agonise: ‘[B]oots hurt that old skin graft on my right foot’. As we can learn from David Morris, it is to a large extent determined by the culture in which we live whether or not a pertinent physiological state is experienced as pain (Morris 1993). Such a view is still more acceptable than to deny the ontological reality of rocks. Not even an exceptionally well-presented
Conclusion
251
version of relativism such as we find in ‘Death and furniture’ could persuade the majority of my students to doubt their manifest existence. Perhaps it might help to engage in a thought experiment. Let us assume, for the moment, that we find ourselves miraculously transported to a planet in some far-away galaxy, perhaps even outside of our own universe. Happily we experience ourselves having no problem breathing. Of course, we do not know how similar the air is to the air we have on Earth, and we cannot even be sure if it is ‘really’ air, or perhaps some kind of thin, dispersed liquid, a kind of vapour. Unlike the air we are familiar with it seems neither humid nor dry. Also what looks like water does not give us the sensation of wetness. Will it harm us if we drink it? When we look up, is it the sky we see? It does not look like the sky we know. Is it something else? Can we be sure that we are ‘really’ looking upwards? Does gravity as we know it apply to this world? We are, it seems, treading on something, but on what? Apparently it is not metal, not soil, not stone and not asphalt. What would we find underneath? It does not seem to be uniformly hard, or soft, or elastic. Is it resonating with our steps? Can we be sure we are not sinking in? Around us are all kinds of shapes, not reminiscent of anything we know from Earth. Are some of them alive, whatever life means here? Are they similar to trees, or to polyps, or to mammoths? Are they close or far away? How solid are they? Are some of them perhaps permeable, or just holograms without any corporeal existence? Are all these shapes, including the ground on which we walk, of the same substance, intrinsically connected and inseparable, or can we distinguish spatially separated objects of different natures? How rigid and durable is their shape? It seems as if some of these objects ‘really’ move around; but perhaps it would be better to compare the movement we discern to water waves. At times the ‘air’ seems to become concrete and impenetrable like glass, and sometimes what looks like a rock offers no resistance. Hungry as we are, what should we bite into? If something feeds us today, how sure can we be that it will feed us also tomorrow? But there is no today or tomorrow. How cyclic are the changing states we observe? We have no way of finding out if such cycles are of equal length. None of our colour adjectives have any use for us; and neither can we decide how bright it is. We cannot say if it is light or dark, and there is nothing that looks red or green or blue. If there are odours, they do not smell like anything we know. We find ourselves surrounded by stuff of which we can make no sense. There is a reality, but it has no meaning. For chimpanzees or feral children it would not matter so much whether they are here in our world or there in this other reality. In both places they would try to follow their instincts and drives and satisfy their needs. They would look for something to eat, for companions with whom to associate, for a place where they could feel safe and so on. Of course, life for them would be easier on Earth than on this far-away planet, at least initially. On Earth they could follow their intuitions which might guide them as to what looks and smells edible. They
252
Meaning, Discourse and Society
would try to stay away from boggy ground, and they could rely on an innate or at least internalised clock. In that other world, they would have to learn everything from scratch by trial and error. Of course, they would not undertake a conscious effort to learn. They do not know what learning is. Yet just as they do not pause to reflect on their instinctive or acquired knowledge in their accustomed environment, they would not be proud to succeed in surviving in their new world. They survive by repeating routines that have been effective. They do not worry about realities. Humans, as long as they are members of discourse communities, are different. On that far-away planet, they would feel as if they were half-awake in a dream that did not make sense. They would desperately try to apply the categorisations they have learned from participating in the discourse. They have learnt to distinguish gases from liquids and solid stuff. They can tell up from down. They know what is called hard and soft. They have negotiated what can be called an object in its own right and what is just part of something else. They have also learnt to accept that there is a never-ending succession of days and nights of equal length. They think it is only logical that visible objects can be sorted according to colour and smelly objects according to their odour. Making sense of one’s reality means putting things into pigeonholes. For humans, these categories have become part of their reality. Even if they are acquired, they seem inescapable. But they are contingent. We could envisage discourse communities that have constructed their realities by adhering to rather different categorisations. It is possible to imagine a society in which people do not pay attention to colour, and in which they only accept things as existing which they find in their immediate presence. When something disappears in the jungle or behind the next bending of the river, it ceases to exist. It is imaginable that in their discourse quantity is negotiated without any recourse to numbers. What I have sketched here are, it is argued, vital characteristics of a small tribe, the Pirahãs, who have recently become the object of a heated discussion, not only among linguists and anthropologists, but also among a wider public, by virtue of an article by John Colapinto: ‘The interpreter: Has an Amazonian tribe upended our understanding of language?’ (Colapinto 2007). Over the last five years, there has been a major controversy over the issue of whether the language of the Pirahãs has to be seen as unique among human languages, or whether it is just one of many ways of talking about things like colour and quantification and what is in their immediate presence (the ‘immediacy of experience’ principle, as Dan Everett, the key linguistic expert on the Pirahã language, calls it). The issue of colour has always separated realists from relativists. There is a strong tradition in nativist language theories in the line of Noam Chomsky to assume that we are innately endowed with the faculty to classify colour, even though the particular language in which we grow up may not provide the full realisation of this potential. The foundational text is Brent Berlin’s and Paul
Conclusion
253
Kay’s book, first published in 1969, Basic Color Terms: Their Universality and Evolution (Berlin and Kay 1969). They assume that our faculty to perceive colours and to label them is innate, that it is defined by biological, not cultural, parameters and that it operates largely in a mechanical way. This doctrine appeared to be proven to many mainstream colour experts, in experiments which are seen as flawed by others. It is perhaps conspicuous that the eleven supposedly ‘basic colour terms’ proposed by Berlin and Kay happen to correspond to the eleven colour words listed in Edward Thorndike’s The Teacher’s Wordbook, first published in 1921. This is the contention of B. A. C. Saunders and J. van Brakel in their discussion of the vast amount of literature on colour perception and colour discourse. Adopting a more relativist stance that favours nurture over nature, they throw doubt on much of the allegedly empirical research in this area (Saunders and van Brakel 1997). A key problem in the experiments carried out by researchers adhering to the paradigm of realism, such as those of Berlin and Kay, is, as Saunders and van Brakel see it, an ambiguity about the object of study. ‘The referent could be a set of colour chips (colour in the world), a set of neurons in the brain or a functionally defined term in a language-of-thought (colour in the head), words in different languages labelling basic colours (colour semantics), or the experience or sensation associated with basic colour categories (phenomenal colour)’ (p. 169). Are the colours in the world out there also the colours we perceive, and are the colours we perceive also the colours that we talk about in our discourse? Opinions are predictably divided on how successful Saunders and van Brakel were in their endeavour to refute the kind of realism which takes the categorisations we find in our Western languages for real. Saunders and van Brakel insist that the evidence concerning colour perception and the way it is neurophysiologically processed is far from clear. Colour is a fuzzy concept, they claim, and they suggest that hue, brightness and saturation are intrinsically and intricately linked. The Western conceptualisation of colour is only one of many possibilities and far from universal. How the perceived data are interpreted depends on the discourse in which one grows up: ‘People in the Euro-American world are trained to distinguish hue. Cross-cultural research reveals the distinction to be contingent – hue, we must conclude, not being naturally salient’ (p. 196). Even in ancient Europe hues were not as much talked about as today. Homeric Greek seems to be widely unconcerned with colour and more focused on the texture of a surface and the kind of lustre reflected by it. Without being embedded in a discourse that constructs colours in relationship to objects of a shared reality, people seem to be at a loss to put into a pigeonhole what they see. This is why it would be so hard for us on that hypothetical planet in a far-away universe to assign colours to what we see. This is also why people who have never discussed the colour of things in their environment cannot provide the answers for which their interrogators are looking. Scientific research
254
Meaning, Discourse and Society
used to be a Western domain; it is not surprising that it is based on the Western discourse. We often take it for granted that our way of analysing things is the only scientific approach there is and is therefore more advanced than what we find in other societies. Paul Kay, for instance, believes that languages move from backwardness ‘toward the precise and explicit speech of the analytic philosopher, the scientist and the bureaucrat’ (Kay 1977: 30). The problem with the realist position is that it tends to assume that our reality is the only one there is. The reality of colour perception is called into question whenever the uniqueness of the Pirahã language is discussed. According to Dan Everett, who lived among the Pirahãs for many years, ‘[t]here are no color terms in Pirahã’ (Everett 2005: 14). Instead when pressed they come up not with terms, but apparently rather with phrases coined on the spot such as bii’-sai (‘blood’ for ‘red’) or xahoasai (‘immature’ for ‘green/blue’) or kobiai (‘see though it’ for ‘white/ clean’). Different speakers will offer different such phrases in different situations. Paul Kay, in his peer comment attached to Everett’s article, admits flaws in the experiments that Steve Sheldon, a colour perception expert in line with Berlin’s and Kay’s theory, carried out with some twenty-five Pirahãs, aimed at eliciting their colour judgements from a palette of colours. But Kay does not accept that the inability to extract a set of recurrent expressions referring to colours could be due to the Pirahã discourse; he would expect to find the existence of ‘true color terms’, once these experiments are conducted without flaws. Sheldon himself, who so far has left the existence of Pirahã colour terms in doubt, is a key researcher in the ongoing project World Color Survey, closely based on Berlin’s and Kay’s Basic Color Terms. I have no doubt that Pirahãs can if necessary distinguish a green snake from a brown one, saying perhaps that the one looks immature and the other one rotten. But this does not mean that talking about colours is part of their habitual discourse. As I see it, colours are not part of their shared reality, in spite of their being neurophysiologically equipped to perceive them as well as we can. Where I disagree with Kay is when he makes the connection that ‘color terms … are directly traceable to complex peripheral neural structures – in the retina’ (p. 53). How can we be sure if there are languages without colour terms as we know them? In much of the scientific approach to the categorisation of colour in the anthropological literature we still find a bias assuming that our categories are universal. But if they are symbolic, if they are the contingent results of negotiations between people, they cannot be innate. No doubt the Pirahãs are as good as we are at distinguishing in their discourse the objects of their environment according to their needs, and apparently they can do that in most cases without reference to colour. In Western culture, at least in modern times, colour has become a dominant discourse topic. It does not have to be so. Fifty years of black-and-white films tell us that even we can make sense of our environment, whether ‘real’ or fictitious, without reference to colour. Perhaps we could learn
Conclusion
255
from the Pirahãs which other ways, ways of which we are not aware, are available to collectively cope with one’s environment. To talk about colours is only one option out of many. Even when we in the West have to distinguish between finer shades of colours, colour terms are not necessarily helpful. Hardly anyone can remember the exact shade of the colour of an object. If my wife wants to find a lipstick matching her new red dress, she will take the dress along to the store. Yet even then it might not be easy to decide on the closest match, as different surface textures make such a comparison very difficult. What looks like a perfect match in the artificial light inside the store may well appear incongruous in bright sunshine. Our colour terms are conspicuously vague. My Roget’s Thesaurus lists for the entry red these adjectives: blood-coloured; maroon, wine, ruby, vermillion, crimson, cardinal, scarlet, cherry, rosy, rose; auburn, pink, coral, flame. I have seen cherries of all hues, and I wonder if anyone can distinguish burgundy from bordeaux (two hues of red not mentioned by my thesaurus) as two different colours, considering that trends in wine-making change rather often. There may be colours outside of the discourse; but our colour words do not so much refer to a discourse-external reality as to the reality constructed in our discourse. Crimson occurs in the 100-million-word BNC 377 times, and these are the more frequent nouns it precedes: ball (3), blooms/blossoms (4), carpet (3), flower (11), lips (3), mask (3), rope-light (6), silk (5), skirts (3), velvet (8), wings (3). On the other hand, dresses tend to be maroon (6 out of 136) but are, at least in the BNC, never called crimson. However, there are no maroon skirts, and there is only one maroon flower. We know which colour attributes go with which discourse object. It is much more difficult to assign a colour word to a thing not talked about. How little colour words have to do with the reality out there is quite obvious when we look at western children’s limited creativity with colouring books. The sky and water are always painted blue, lips and apples are always red, chocolate is brown and grass is green. We have all been trained to memorise the prototypical colours of the kind of objects of which we are given visual images. However, colour training was not always and is not everywhere such a central part of growing up. The role it plays for us today may well have been initiated by the fact that painting epitomised Renaissance arts and that ever since it has played a prominent part in the liberal arts education of the elite. The obsession with coloured images has put its stamp on the Baroque churches, advertising the superiority of Catholicism in the age of the counter-reformation, thus informing also the minds of the less privileged classes. Over the centuries, colours have played an ever-growing part in the design of our human-made environment. Dyes in clothing have distinguished people’s origins, professions and status. They have told the side to which a soldier belonged. Colour is an important aspect in discriminating last year’s fashion from this year’s. In designing the outsides and
256
Meaning, Discourse and Society
insides of buildings, colours have been replacing or been added on to ornamental relief. The aristocracy had to paint the walls of their mansions’ state rooms in glaring colours. Gradually more and more people began to take account of colour. Up to the middle of the last century, advertisements were still mostly black and white. Today we find colour everywhere. It is not the reality out there, it is the discourse that focuses our attention. It would be worth analysing the kind of stories that have been told young children. Two hundred years ago, Grimm’s fairly tales hardly mentioned colours. Today, I would assume, colours play a more dominant role. The colours we see on objects are the colours given to these objects in the discourse, not necessarily the ‘real’ colours, if there were such a thing. As the impressionists knew, our colour assignments are less based on unbiased observation than on our expectations informed by the discourse. Hardly anyone remembers in which colours Edouard Manet painted the sky in his Monet painting in his floating studio. Unless we have been told that it is remarkable because it is not blue, we would not notice that this sky is all but blue. Our basic colour words, red, brown, blue, green and yellow have become so pervasive in our discourse not due to the reality out there, but due to the role they play in our discourse. There is no basic blue for Italians; they have to tell azurro from blu. Did the astronauts on the moon ever use colour terms for the objects they saw in their lunar environment? Can we be sure of the colours of objects we find at the bottom of the oceans? It is the absence of an established discourse dealing with the world we find on the hypothetical planet of a far-away galaxy that would make it so hard for us to assign colours to the objects we find there, or indeed to assure ourselves of the existence of these objects themselves. Before we have turned the stuff surrounding us there into a reality we share, discuss and negotiate, it remains without meaning for us. In the shared realities of our Western societies, colours as objects of the discourse are ever present. We find it hard to imagine that children growing up in a culture without colour pencils, without colouring books, without illustrated children’s books, without glossy magazines full of colour pictures, children growing up in a world in which everything appears more or less grey, might not be very interested in discerning different colours. We should not find it so hard to believe that the Pirahãs can so easily live their lives without talking about colours. Just as we are neurophysiologically equipped to see what we can interpret as colours, we can also discern odours. Interestingly, in our Western world we normally do not find a discourse on odours that can match the one on colours. There are no lexical items that could be compared to the three, or six, or nine basic colour terms we have designated as such in our Western languages. We use only very few of the adjectives preceding odour or smell just for talking about smell, among them acrid (also used for flavour), musty (though derived from must – ‘fermented juice’), pungent, rank and perhaps stale. But these words are not as basic as adjectives like red or green, which we believe to be fairly uniform as categories. There is something like a prototypical red or green.
Conclusion
257
But is there also a prototypical smell we would call pungent? Among the things called pungent we find, according to the BNC, the smell of love, Muscat, sundried tomatoes, badger aroma, aroma of apples, sesame oil, cod’s roe, female odour, garlic, cigarettes, cloves, urine and much else. It would be hard to come up with a common denominator for these sensations. Many of the adjectives we find with smell or odour indicate objects which are the source of the scent or can be associated with a smelly thing: fishy, foul, fresh, herbal, milky, putrid, fetid, rotting, or smoky. This is not so dissimilar from the ways the Pirahã talk about colours. It might be worthwhile to ponder the possibility that there are ethnicities that have a way of talking about odours as we are talking about colours. In our Western societies we tend to classify odours normally as pleasant (pleasing, nice, comforting) or as objectionable (horrible, bad, awful, unpleasant). These are more intuitive or even instinctive categories than those we have for colours. Even though culture can play a major role in determining whether something smells good (the Swedes are said to love the smell of rotten herring), the way we experience odours is perhaps less mediated by language than is the case for colours. Chimpanzees or feral children can distinguish between pleasant and unpleasant smells, but they may not have abstract concepts of a small set of basic colours. Their reality is not a reality constructed in a discourse. It consists of spurious memories of fuzzy glimpses of unmediated experiences, prompted by chance encounters with memory-causing prompts in their environment. It is, I believe, as unstructured as our encounter with the world on a planet in a far-away universe. It is a world in which instincts and drives help us to survive, but a world without meaning. Unlike them, we may be conscious that our categories do not fit with what we find there. But when it comes to odours, we are not much better off than them. We cannot explain which scent we like and which not. Finding one perfume nicer than another one is overwhelmingly a matter of immediate experience, not something that lends itself to discussion. Sometimes, however, we find it necessary to classify odours in a systematic way. Then efforts will have to be made to develop an expert discourse, for instance for experts talking about perfumes or wine or tobacco or coffee. Much has been said about the adjectives that are supposed to describe the taste and aroma of wine. For me, such tasting notes have not been too helpful. There are also many specialised vocabularies for other objects which appeal to consumers though their flavour/odour profiles. A Google search for odour vocabulary coffee delivers impressive 73,000 hits, the first being the web page of the International Coffee Organization. These are the adjectives proposed there to describe the aroma, not the taste, of this drink: animal-like, ashy, burnt/smoky, chemical/ medicinal, chocolate-like, caramel, cereal/malty/toast-like, earthy, floral, fruity/ citrus, grassy/green/herbal, nutty, rancid/rotten, rubber-like, spicy, tobacco, winey, woody. The adjectives come with their definitions. Earthy is defined as ‘[t]he characteristic odour of fresh earth, wet soil or humus. Sometimes associated
258
Meaning, Discourse and Society
with moulds and reminiscent of raw potato flavour, considered as an undesirable flavour when perceived in coffee’. Such a vocabulary will make sense for a discourse community of coffee aroma experts. They will construct their reality through primordial speech situations, as I have discussed them in Chapter 10. Sitting around a table and comparing their impressions of sample after sample, trying to use the prescribed vocabulary in the prescribed way, they will create for themselves a world of coffee aromas which they can communicate among themselves. It is a collaborative act, in which the definitions are interpreted and applied to the samples they share, with the aim of creating some kind of consensus. Yet their reality will not be my reality. Like other normal people, I call the coffee I drink pleasant or unpleasant, aromatic or stale. Odours, but not colours, are full of associative potential. We can easily recall long-forgotten memories. They invoke vague and fuzzy images of experiences that we have not made, via interpretation and reflection, part of our conscious history. Often we cannot place these memories into a strict lattice of place and time. The more we try to make sense of them the more elusive they become. Perhaps odours are as close as we can hope to approximate the immediacy of an experience that is not distorted by the categorisations we find in the discourse. Because the classification of odours is rarely part of the discourse, we experience them to a large extent in isolation, as solitary minds. In this sense, they are unlike most of our other experiences, which, even when they seem to be authentic, owe more to discursive prefigurations than to the immediacy of our apperceptions. The social dimension of experience and intentionality How important is it really for us to know that our experiences are authentic? One of the goals of this book has been to show that there is an alternative to our trust in the individual, solitary mind that has dominated Western society since the advent of modern times. Not for a second would I doubt that this venerable tradition, the essence of our modernity, has given rise to fantastic developments. In the eighteenth and nineteenth centuries, the celebration of genius has set free a creativity that has no equal in other parts of the world. More than other societies, Western society has extolled the notion that each person is unique; it has promoted the concept of intentionality, has put a premium on individual responsibility, and has encouraged the individuality, difference and autonomy of each person. This would not have been possible without the reification of the human mind and the credo that each of us is unmistakable, and that all we are and all we do has to be original and authentic. But there is also the perspective that alone we are nothing. Growing up without human carers who talk with us, we remain speechless feral children. In solitary confinement, deprived of the discourse, we tend to go mad. Robinson Crusoe needed his Friday to pull him out of his depression. Our instincts and our
Conclusion
259
drives may be strong enough to ensure survival. But no one would build a cathedral, write a book or spend an evening in a pub if there were not the expectation of some sort of interaction. After all, some things, like clapping, require more than one hand. Individuality, originality and autonomy are only thinkable if there are others with whom we can engage and who we expect to take notice and to let us know. We may not be born as a blank slate, as Steven Pinker has reminded us (Pinker 2002), but there is no assignment of meaning that comes in the form of genes. Content is symbolic; it can only emerge in a discourse. Genes can only have had an impact on non-symbolic phenotypes, observable properties that can be detected, counted, and measured by approved devices. Everything that counts as symbolic content cannot be part of our genetic equipment; it is but the result of joint negotiation. All our ideas about herons, about minds, about intentionality or about the pleasures of a pub visit are ideas we have picked up in our verbal interactions with other people. We do not know what is inside people’s minds. Minds, intentions and even what counts as pleasure (for instance the smell of rotten herring which the Swedes find so attractive) are objects that we have jointly constructed in our discourse. They are not objects of a discourse-external world. As we grow up, we become immersed in the discourses of our carers and our peers. Through interacting with them, we gradually internalise the reality as it is presented to us. To a large extent we accept it unquestioningly. This is how symbolic representations are transformed into first-person experience that is felt to be ‘embodied’ and ‘authentic’. Some first-person experiences may indeed be ineffable. For the Pirahãs, colours may well be such a phenomenon. For people in Western societies, this is often the case where the quality of an odour is concerned. But much of what we believe to be a unique experience is a modulated repeat of experiences that we were told about. What we register as the true and authentic feeling we wake up with on a given morning is a reflection of and a reaction to what we have learnt about other people’s feelings when they get up in the morning. They tell us what they feel, and this is what we will feel as soon as a comparable situation arrives. What we call feeling includes giving what we feel a name. The name must have been made available to us, and applying it to a feeling involved reflection and thus consciousness. Giving feeling a name, interpreting what we feel, is stepping outside the authenticity of what we feel. Babies have authentic feelings, because they cannot contemplate what they feel. Naming a feeling is applying a category that is not embodied but discursively constructed. As soon as one is aware of one’s feeling, this experience is no longer true or authentic. It is mediated. Giving an experience a name is a conscious act. As an act, it involves intentionality. Intentionality, though, is a very tricky discourse object. As I have argued above, embodied intentionality, the intentionality of having an authentic
260
Meaning, Discourse and Society
feeling, excludes consciousness and reflection. It cannot be observed. This, however, is so counterintuitive that we must probe deeper. Firstly, there is raw experience, a kind of ‘feel’ I have, perhaps a familiar smell that puts me into a certain mood. No conscious reflection is needed. Another example would be my early-morning routine. Many years ago there was probably a time when I had to remind myself what to do next once I had got up. But over the years getting up has become a quite automatic kind of procedure that no longer involves any conscious planning. This procedure would be felt as raw experience. It would not make me switch on my reflections. But if I ask myself or someone asks me how I feel, I would have to find words for the quality of an experience that was, as ‘raw’ experience, ineffable. Suddenly I have to reflect on what I am feeling while I am getting up. In this situation, I will find a label for my feeling, and I will intuitively take it for an authentic expression of what I feel. But I will have fooled myself. Looking at it from an outside perspective, what I do is scan the applicable content memorised from the discourse in which I have participated. There I can find a number of options for how I could label my ‘raw’ feeling. The discourse (Google lists 1,270 citations for ‘in the morning i always feel’) offers us a wide choice: in the morning, I always feel sleepy and have no strength to eat. in the morning i always feel the most anxiety I’m not hungry in the morning, I always feel like I’m force feeding myself. in the morning i always feel hungry in the morning, I always feel groggy and unproductive. in the morning, I always feel terrible In the morning I always feel like I’m walking on broken glass. in the morning I always feel OK. In the morning I always feel better. in the morning, I always feel like I can conquer the day! This is a long, but by no means infinite list. In Google, there is not a single hit for ‘in the morning i always feel guilty’ and there is probably no one who has ever described their early morning feeling in these words. Yet if there had been a bestselling novel Feeling Guilty in the Morning, things might be different. Because this book still remains to be written, we are not induced to use this label for our feeling in the early hours. The idea that our experiences are, as soon as they are labelled, no longer ‘authentic’ is not at all new; it is very much how George Herbert Mead saw it: The individual experiences himself as such, not directly, but only indirectly, from the particular standpoints of other individual members of the same social group, or from the generalized standpoint of the social group as a whole to which he belongs. For he enters
Conclusion
261
his own experience as self or as individual, not directly or immediately, not by becoming a subject to himself, but only in so far as he first becomes an object to himself just as other individuals are objects to him or in his experience; and he becomes an object to himself only by taking the attitudes of other individuals toward himself within a social environment or context of experience and behaviour in which both he and they are involved. (Mead 1934: 138)
As soon as this communal experience becomes appropriated and begins to define one’s identity, it begins to feel ‘authentic’. It feels ‘natural’, ‘embodied’, as George Lakoff would call it, but it is no longer a ‘raw’ feeling of which we are unconscious (Lakoff and Johnson 1999). Yet it seems nearly impossible to draw a line between ‘raw’ feeling and a feeling we have labelled. Awareness is a matter of degree. Our intuition tells us that our feeling, even if labelled, can still be authentic. This is why it feels direct and authentic, and not a symbolic representation. Such an appropriated feeling is neither a mere automatic process nor a fully fledged conscious act. Thus the feeling I have when I get up is for my intuition, but not for those who have asked me, an ‘embodied’, not a symbolic intentionality. My intuition, however, is unreliable. What really happens is this: waking up feeling rotten, I will look for ways to label this experience. I may come to identify my feeling as cheerless or dispirited. In my memory, I will seek for possible causes I have learnt about for harbouring such a feeling. Am I perhaps clinically depressed? Did I lose a close family member? Is it grief that I feel? Or did I perhaps have a drink too many last night? If I opt for this last explanation for my state of feeling rotten in the morning, I may decide to take an AlkaSeltzer. This means taking a decision, making a choice, planning a course of action and carrying it out. But all this may not involve ‘real’ consciousness. I will have enacted my intentions without formulating them for myself or others. This sort of reflective intentionality, somewhere between embodied and symbolic intentionality, too, cannot be accessed from the outside. Yet when my wife happens to ask me how I feel before I have hastened to take my Alka-Seltzer, things will be different. Now there is symbolic content, content that is articulated. This changes things. The reflective intentionality is turned into a symbolic intentionality, an intentionality no longer confined to inaccessible reflections, but realised within a discourse, as something that can be observed and analysed. As long as all my symbolic behaviour indicated that I was taking an Alka-Seltzer to offset my hangover, perhaps day after day, it would not matter if I thought I was clinically depressed. Only if I stopped drinking alcohol and still claimed hangovers, would there be observable discrepancies in my behaviour. As long as I behave consistently as if I were not clinically depressed, my belief to the contrary would not matter a bit. I call this kind of intentionality exhibited by symbolic behaviour, symbolic intentionality. It is an intentionality that is accessible, that can be interpreted by other people, and therefore the only one that matters.
262
Meaning, Discourse and Society
As I see it, there is no obvious way to distinguish between experiential and reflective intentionality; they are difficult to tell apart, because they are internal to the individual mind and because neither of them can be examined. We also cannot observe how exactly they feed into the symbolic variant of intentionality. But our symbolic intentionality is open to inspection by the discourse community, as it takes the form of contributions to the discourse. In this book I have claimed that we are not in a position to observe how an individual, solitary mind memorises and processes symbolic content. We can only study a person’s non-symbolic properties, their phenotypes (for instance the colour of their eyes) and now also their brain scans, for what they are worth. But neither a person’s observable properties nor whatever their brain scan shows has any meaning. As attractive as it would be to make the symbolic content of a person’s mind visible, we must accept that meaning does not exist unless it is put into language and communicated. For cognition to be meaningful it has to be expressed in language. Therefore I have suggested in this book an alternative perspective on cognition. The cognitive sciences can become empirical as soon as they give up on individual minds and start dealing with the collective mind. Making sense, assigning meaning to what has been said, interpreting the reality of the discourse in which we find ourselves, is a collaborative undertaking. Of course, I do not deny that I am an individual person, endowed with raw experience and reflected intentionality. I experience myself in my environment, I have dispositions, desires, fears, I want to do things and have expectations, I can decide to have a glass of wine, I make myself work on my laptop. To myself, it does not matter how much of this is driven by the discourse. But if I were, like Robinson Crusoe, on an island, and if neither a Friday nor a boat ever came to bring me into contact with society, all these things would remain entirely inconsequential. What matters more than our hidden experiential and reflective intentionalities is the symbolic, the collective intentionality expressed in the discourse. Only when we begin to see ourselves as the distributed elements of a collective mind, do we become aware of our full potential. Only in symbolic interaction can we make sense of ourselves. Why the media discourse advertises individual agency In this book, I have argued for a perspective that leaves out the people who contribute to the discourse. Instead I have suggested that we should focus only on texts and their links with other texts. This is against our intuitions. Since we invented ourselves as autonomous individuals, we have learned to empathise with other people. Because I tend to see myself as being fully in charge of myself, I also see other people as autonomous agents. Just as we cannot forego the illusion that we are dealing with the ‘real’ world outside and not with a
Conclusion
263
vicarious reality that is but a discourse construct, our relatives, friends and neighbours appear to us as real as we appear to ourselves. But can we really trust our intuitions? All we know about other people, close as they may be to us, is our perception of them. We interpret their looks, their eye colour, hair colour, height, figure, ‘natural’ (involuntary) body movements, agility, etc. against the background of generalisations we have derived from the discourse. But even more important than their looks is their interpersonal behaviour, for instance their facial expressions, particularly inasmuch as they express symbolic content. Taking looks and behaviour together, we can indeed know a lot about a person. But their experiential and also their reflective intentionality remains hidden from us. All the evidence about a person (apart from their looks) that observers can access and share is their symbolic behaviour, is what they choose to tell us. The discourse is the only reality we can be sure of. I matter to other people not for my first-person experiences or my private thoughts, but for my (symbolic) behaviour, for what I contribute to causes common to the group of which I am part. What I know of the other people are not their experiences or thoughts, but the way they interact with me. Yet our trust in the power of empathy is strong enough to make us believe we can actually read their thoughts and know what they feel. This is not how most nonWesterners tend to see it. As I have shown above, we are told that it is taboo for Samoans to make assumptions about the mental states of other people. Thoughts are private. Apparently in Islam thoughts, however they may violate the spirit of the religion, will have no effect on a person’s eternal fate. It is only one’s acts, one’s behaviour that counts. This was also true of Christianity before the invention of the modern human as an autonomous person with an internalised conscience. These are good reasons to disregard the monadic mind and to focus instead on someone’s contributions to the discourse. Only by collaborating with others in the interpretation of what has been said can the members of a discourse community make sense of themselves and bring about change if their world constructed by the discourse has lost its appeal. Acting together, the people can reinterpret their reality and thus make a change. But they can only make a change if they raise their voices. Alone, a person has no chance of reconstructing the discursive reality. Only if we collaborate can we put ourselves in charge of our reality. The public discourse to which we are exposed, the newspapers, magazines, radio and television all want to make us believe that what they tell us is an accurate mirror image of the reality out there. There is a reason for this. Once we are made to believe that there is a ‘real’ world out there, we have foregone any leverage. Google lists 22 million occurrences for ‘have to accept the facts’. This is what we are constantly told, and if our politicians fear we do not trust their selfless motives, they bring in the experts who claim to have proven their
264
Meaning, Discourse and Society
factuality. These are the data, this is the evidence; you have to go along with it, is their mantra. Here are our unquestionably scientific methods to identify, count and measure what the case is. These are the results. You can talk about them as much as you like. But you cannot change reality. Once we accept the facts, we have rendered ourselves powerless. We have put ourselves in the charge of specialists and experts. Take the economy. What we call the economy has been constructed as a discourse object that is as elusive as the concept of language as a system. It is presented to us as a highly complex mechanism that only experts can understand, and fooling around with it will be the ruin of us all. Asking the big corporations to pay their fair share of taxes, introducing measures to reduce the gap between the rich and the poor, having air passengers pay VAT on their tickets as we do on our local buses spells disaster, we are told. But what do our experts know about this alleged system? So far, economists have not come up with a single undisputed regularity, and the success rate of their predictions correlates nicely with randomness. We can also never be sure how much of the growth they attribute to the economy is really due to an increase in the number of refrigerators produced rather than to the creativity of the auditing business. The unpredictability of the credit crunch is an obvious example. We are told that specialists are there to know and tell us the truth. But in reality their main job is to keep us from asking nasty questions. The reality they present to us is not the reality out there; it is a reality that suits those who sit in the boardrooms of the big corporations. There is no necessity to accept the ‘facts’ that have been constructed for our obliging consumption. Unless we renounce media consumption altogether, we will be affected by the public discourse, however much we try to exercise our critical minds. What makes it even more persuasive is that, unlike the media in the former socialist countries, it is not strictly monovocal. Many of my students are convinced that our national British newspapers cover the whole range of the political spectrum from left to right. Indeed it is true that those who have the time to scan through half a dozen papers on a pretty regular basis will find dissenting opinions. The system is just about multi-faceted enough to make the illusion of factual information appear real. Yes, it was reported some years ago that the Taliban had eradicated practically all opium production in Afghanistan. Today more opium is produced than ever before. But many British discourse consumers believe what they are told, namely that fighting the Taliban’s opium production is one of the British army’s main targets in that country. Those of us who want to compare the different realities have to invest more time than most of us can afford. The easiest way to disempower people in a democratic society is to persuade them to accept a reality constructed for them as the only reality. To make them believe that what they are told are incontrovertible facts, scientifically established by the best experts in the field, is the best way to disenfranchise them. If a whole
Conclusion
265
nation lets itself be persuaded that there is such a scientific fact as race and that some races are poison for other races, then apartheid and the holocaust will be the consequence. A large majority of Germans obediently accepted a reality that had been constructed by racist ideologies as scientific and factual. They did not insist on having their own say. In Nazi Germany, the public discourse was controlled by the close collaboration between media tycoons and an undemocratic totalitarian regime. It was not impossible, though, to take part in alternative discourses. There were a few Germans who chose exile over silence. In the Western world today, there is hardly more than a handful of corporate discourse merchants, joined all too often by common interests. They exercise control over all symbolic content, whether in the media, in the book market, in the market of academic journals and wherever the construction of reality takes place. Ever tighter capillaries control the negotiation and dissemination of new ideas. Alternative ways of looking at things have to pass through filter after filter before they are accepted. Our Western public discourse is on the brink of becoming as petrified as that of the Socialist Bloc of Eastern Europe many years ago. The central theme of this Western public discourse is the elevation and glorification of individual agency. Our economic leaders deserve their bonuses and our gratitude because it is to their genius that we owe our prosperity. If we, the normal people, do not do as well as we would like it must be our own fault. TV series offer millionaires as role models, and we are invited to observe how their profit-oriented policies of hiring and firing keep the British economy growing. If we were to practise our agency as determinedly as they do, we are told, we would become like them. Failure is a sign that we are not up to it. Only when things do not go as planned does this adoration of personal agency become briefly dented. Thus we are told that the global meltdown of our financial institutions cannot be blamed on our economic leaders. It was a natural catastrophe, a tsunami, for which no one bears responsibility. The capitalist economy is a system based on its inherent laws and regularities. It is executed by ‘rational actors in the marketplace’. It has its ups and downs, its winners and its losers. There is no reason to blame our system managers. The docile consumers of the public discourse won’t mind accepting this new reality. This is what we read about the collapse of the Lehman Brothers in a comment in the London Times of 17 September 2008, titled ‘Crisis and capitalism’: The world is in the grip of a banking crisis. It would be easy to blame it all on the bankers, to watch their gloomy desk-clearing with barely concealed Schadenfreude and to look to government and regulators to rebuild a wrecked financial system. But it would be wrong. It may seem an inopportune moment to say that bankers might be the solution, rather than the cause, of the current turbulence. Yet it needs to be said … There is no shortage of gravediggers happy to declare the market has failed us. In fact, the opposite is true: it is working … Rumours of Lehman’s impending downfall allowed traders to move money into safer havens. Predators will happily take over any assets with residual value. Profitable parts of
266
Meaning, Discourse and Society
the business will find a new home and the weaker parts closed down. This is painful and worrying but the opposite of a disaster. It might be brutal and unforgiving but this is how capitalism works. The market ensures that those who make mistakes are accountable for them. What critics are too hasty to see as capitalism in crisis is, in fact, capitalism in action … To do nothing is often more difficult than it sounds. But in this case it would be wise. ‘Capital is money, capital is commodities’ said Marx. ‘By virtue of it being value, it has acquired the occult ability to add value to itself. It brings forth living offspring, or, at the least, lays golden eggs.’ The source of the solution will be the same as the problem – rational actors in the marketplace.
According to the media, our economic leaders are not to blame for the economic crisis. Instead it is the people’s fault if they find themselves in hardship. On the other hand, we find that over the last twenty years, in much of the public discourse, it is the small people who are increasingly constructed and presented as individual agents of their economic fortunes. The keyword is personal responsibility. Here are a few citations (out of 2,305), representing the global Anglophone newspaper discourse of one month, taken from Google News on 7 October 2008: But Maryland’s success in welfare reform should provide a model for Annapolis in dealing with other policy areas. As the results of the study show, policies that encourage individual choice and personal responsibility empower citizens and strengthen the economy. (Baltimore Sun, United States, 8 Sep 2008: ‘The welfare reform model’) So on the one hand, folks like this always tell others – especially the poor and people of color – to take ‘personal responsibility’ for their lives, and not to blame outside factors (like racism, or the economic system) for their problems. But on the other hand, these same persons then demonstrate that their own ability to blame others for their personal setbacks, or the nation’s problems, knows no rival. (Trinicenter.com, Trinidad and Tobago, 30 September 2008: ‘Racism as reflex’) At this point I should say that I have never had a problem with the moral justification of the market. In the first place, genuinely free and open markets enable individuals to promote their own, self-chosen goals (which do not have to be selfish) as opposed to being exploited in order to promote other people’s goals. So markets promote liberty, autonomy, personal responsibility and self-determination. They have been central to the emancipation of humanity from oppression, both in its old forms (like feudalism) and in its modern forms (like totalitarianism) (Scoop.co.nz, New Zealand, 1 October 2008: ‘Moral markets’) For me, the most important word is responsibility. Personal responsibility. Professional responsibility. Civic responsibility. Corporate responsibility. Our responsibility to our family, to our neighbourhood, our country. Our responsibility to behave in a decent and civilised way. To help others. That is what this Party is all about. Every big decision; every big judgment I make: I ask myself some simple questions. Does this encourage responsibility and discourage irresponsibility? Does this make us a more or less responsible society? Social responsibility, not state control. Because we know that we will only be a strong society if we are a responsible society. (Guardian, UK, 1 October 2008: ‘David Cameron’s speech in full’)
Conclusion
267
[Our task] begins with personal responsibility, something in short supply these days … Taking responsibility for one’s actions is, arguably, the single best character trait to have. Taking responsibility means we don’t drive after drinking alcohol, we don’t buy things we can’t afford, and we pay our bills … Personal responsibility does not allow for laziness or procrastination in this or any issue. It makes some pretty big demands on us, with the best of intentions. Taking responsibility for our actions means we don’t blame our problems and hang ups on a lack of money or time, stress, our boss, our kids, our parents, our job, or a hundred other factors … But it doesn’t end there. Just when we reach that point in life where we have it all together, we’re in the best situation to then offer to take some responsibility for something that needs our help – a person, a cause, a charity, or an event that could use a helping hand. (Daily Gleaner, Canada, 6 October 2008: ‘The challenge of living responsibly’)
The death knell to the welfare state sounds sweeter if it glorifies our individual agency. We should not ask for solidarity or act in solidarity with others. The call for personal responsibility discourages all forms of collective action. We should learn from successful entrepreneurs, these mavericks who take things into their own hands. They should be our role models. They are the ones who turned Britain into the global nation we find her to be today. Meanwhile, other societies with different conceptualisations of agency are advancing rapidly. Neither Japan nor China nor India were ever as enthusiastic as the west about a world-view that elevated the individual monadic mind and the spirit of competition at the expense of collaboration. These societies have looked with disdain upon the western concept of individual autonomy. Their philosophy is that things should be done jointly. They cherish reliability more than asocial idiosyncratic behaviour. Innovation in these societies is the result of patient negotiation in never-ending discussions. In the reality that they have constructed for themselves, they prefer an engagement with arguments proffered by other participants to a competition between different voices. It is true that this collaborative approach has in the past been curtailed in its efficiency by delegating individual responsibility to an unchallenged hierarchy. Confucian disciples had to take care not to contradict the voice of their teachers. Successful civil servants practised obsequiousness. It is an approach that did not stand a chance when it was confronted with the western lust for conquest in the nineteenth century. Today, however, top hierarchies in the western world fashion themselves increasingly on the eastern model of collaborative problem solution. The sinologist and philosopher François Jullien has produced a host of publications in which he extols the superiority of the Chinese approach to problem resolution and has given innumerable key addresses at management seminars. It does not matter so much that much of what he says has been common knowledge among sinologists all along, and that his interpretations of the philosophical texts he quotes leave much to be desired, at least according to his main critic, Jean François Billeter of the University of Geneva (Billeter 2006). What matters more is the
268
Meaning, Discourse and Society
fact, as François Cooren (see above, Chapter 13) has shown, that this classical Chinese approach is rapidly becoming the style of western boardroom meetings. The higher up in the corporate hierarchy, the more important becomes collaboration. The reality that we are presented with in our media, however, is different. Our celebrity culture broadcasts individuality. Success, we are told, is a matter of individual agency. At closer sight, though, our celebrities are little more than puppets on the strings controlled by our content merchants. The reality they design for our consumption is the opposite of the reality they exercise among themselves. We normal people are not part of these discourses and the realities they construct. We have no good way of knowing how collaborative they are. It could well be that important decisions are not made by lonely Rockefellers or Vanderbilts, but collectively in boardrooms and even more in backrooms, or in exclusive holiday resorts, or on shiny yachts, involving powerful people about whom we normally do not hear much. Far away from media limelight they negotiate a reality which often affects us all, but it is one we hear little about. Only rarely, and often more by accident, are we permitted a glance into this hidden world, for instance when Peter Mandelson, David Osborne, Nathaniel Rothschild, Elisabeth Murdoch and the Russian oligarch Oleg Deripaska met on the beautiful island of Corfu in October 2008. What was planned there we were not told, and we would not have learned about this meeting had not David Osborne divulged it because he was unhappy about how he was treated. We can look at such negotiations as evidence for successful problem management. We can also see them as confirmation of a pervasive cronyism and sleaziness of our elites that is normally carefully concealed from us. What is carefully obscured is that it is the collectivity of people we find in boardrooms, not the individual entrepreneurship in a free market, that determines our reality. There is no corporate CEO taking lonely decisions. The boardroom acquires and exercises an agency of its own. Personal responsibility apparently does not rank prominently in this discourse. For all we know, personal responsibility does not rank very highly. Collective agency seems to overrule the agency of its individual members. In the public discourse, however, we are told a different story. Teachers, politicians, managers and the content merchants keep impelling us to believe in ourselves and go it alone. No one encourages young working-class kids to collaborate for their own interests and to place their trust in their own collectivity. The only kind of teamwork we are persuaded to engage in is less in our own interest than it is a way of increasing the overall productivity of our work unit. Neither the schools we have attended nor the mainstream media or our moral authorities invite us to work and fight jointly for our interests, for instance to exercise solidarity in associations like the trade
Conclusion
269
unions. They know that as long as we compete for our interests individually, we will remain solitary voiceless consumers of a monopolised discourse. However, it seems worth taking a moment to ponder why the public discourse in our streamlined, highly controlled and carefully calibrated post-free-market society is so much more successful than the socialist system in Eastern Europe. The standard answer we are given is that our society, unlike socialism, offers choice. Indeed in Britain we have the choice between four or five national ‘quality’ newspapers, three nation-wide parliamentary parties, four supermarket chains, four or five TV suppliers, etc. As Naomi Klein has shown in great detail (Klein 2001), the content merchants themselves offer us respite from the kind of sameness proverbial in socialist countries by letting us choose between different brands. It does not matter if Pepsi and Coke taste the same and may well be owned by the same hedge fund, if there are people who exercise their individuality by refusing to take a sip from the other brand. Branding allows us to define ourselves as unique persons. Choosing between brands gives us the feeling of personal agency. We can collaborate to change our reality We are all consumers of a public national discourse, the discourse of newspapers, magazines, radio and television, in which we have no voice. Apart from letters to the editors or to our members of parliament, it is practically impossible for us to contribute to this discourse. What we, the text consumers, negotiate among ourselves has less and less relevance. We behave as docile sheep which are happily eating the grass they are fed. We have more confidence in this public discourse than in what our family members, colleagues and friends tell us. When we have taken part in an anti-nuclear demonstration we do not trust our own senses as much as its presentation we find in the evening news. Proud as we are to live in a democracy, we have no say in shaping the reality in which we find ourselves. There is no counter-discourse that could stand up to the media discourse. As consumers, we can, at most, boycott what is there. Our participation in public affairs is largely limited to casting a vote. The members of parliament who are there to represent us are not much better off. What they are allowed to say and what not is subject to impenetrable and often unfathomable rules and customs. It is also constrained by the interests of those who control the public discourse. Speaking one’s mind may have unwelcome consequences. In the end, it is again up to the content merchants to decide whether what an MP says will find its way into the media. A healthy discourse is plurivocal, and it minimises the distinction between text producers and text consumers. It is also a discourse that is less competitive and more collaborative. It is a discourse of partners who have equal rights as members of this discourse community. A healthy discourse is anarchic. Everything can be
270
Meaning, Discourse and Society
said. Not everything said will be successful, will survive. It is because we want our contributions to be successful that we mostly say things we hope other people will find telling enough to refer to them. A healthy discourse is a discourse in which participants listen to each other, placing less value on their individual agency than on the collective agency of the group. All contribute to the design of a novel reality by testing out various recombinations, permutations and reformulations of what others have said before. Some new ideas will be picked up by other contributors, while others are never referred to. Those surviving to the end of a boardroom meeting are the successful ones. The less we insist on our individual agency, the more we engage with what has been said, the more chance we have of contributing to the eventual outcome. In an ideal discourse, it should not matter who comes up with the winning phrase. I offer my interpretation of the haiku (see the previous chapter) because I hope others will refer to it. Nothing is to be gained from an interpretive community divided into text producers and text consumers. As consumers of the media discourse we exercise neither an individual nor a collective agency. What would turn us into agents, individually or collectively, would be our power to talk back. A public discourse that we can accept as our discourse has to have us as its partners, and not only as consumers. It is our participation in the discourse that turns our reflective intentionality and our personal agency into a symbolic intentionality, an intentionality that comes out of interaction with others, that puts ourselves collectively in charge of the discourse and the reality it constructs. Such a discourse enables us to give our shared reality the interpretation we see fit. If we are unhappy with an interpretation offered to us, we can replace it with a new one. On 2 October 1989, people in Leipzig, East Germany, took to the streets. They had grown tired of the discourse to which they were exposed, an exceedingly dull, monovocal discourse that never had anything new to say, a discourse whose reality was far removed from the reality that mattered to the people, that they constructed for themselves when they thought themselves free from surveillance. Discontent had grown over many years, fanned also by the media discourse that slopped over from the West. Mikhail Gorbachev’s slogans of ‘perestroika’ and ‘glasnost’ gave them the necessary confidence. Having finally found their collective voice, they started shouting their battle cry ‘Wir sind das Volk’ (‘We are the people’). By hijacking the public discourse, they rejected the reality that their authorities had tried to force on them. They refused to be reduced to mere consumers of the official version. It was, after all, their state. It had to be their discourse. So they sent their old government into retirement, and they forced their rejuvenated one to open the border. On 9 November the Berlin Wall fell. Together, the people of Leipzig, and of other towns in East Germany, had learnt to talk back and were beginning to take control. They created a new discourse and a new reality. They assumed the power of definition. Or so it would seem. In the second half of October, just a fortnight after the first demonstrations, a new, just
Conclusion
271
slightly different version of their battle cry began to appear on the placards: ‘Wir sind ein Volk’ (‘We are one nation’). The people who had come together in solidarity against their leaders, the people who were exercising their collective intentionality in ever more powerful rallies, who had put themselves in charge of the public discourse, were all of a sudden outwitted by a subversive infiltration of their discourse. By stealth and almost overnight the discourse took a new direction. This reformulation of their slogan did not emanate from the protesters or from the East German authorities. It came from West Germany. For the collective actions in the East had already given rise to a few, still rather isolated, initiatives of people living in the West. They too began to call for a new discourse and a new reality, provoked by the news about the peaceful protests in the East. This danger needed to be contained. The tiny variation, the replacement of the definite article by the number one, was enough to turn a call for self-determination of the people into a call for a friendly takeover of the East German state and its assets. It helped to contain the expression of East German discontent and to channel it in a more desirable direction. Was it the boardrooms who decided that something needed to be done? The one-nation discourse has always been a convenient instrument to sweep the divisions between content merchants and discourse consumers under the carpet. Everyone understands that national unity requires both leadership and the acceptance of leadership. By changing the slogan, order was quickly re-established. No one ever found out where these new placards came from. They were the beginning of the end of this revolution from below. The people who had taken to the streets in East German towns, and those who now saw their chance to publish their ideas in the suddenly censure-free media, had all begun to collaborate in constructing a future for themselves, in a collaborative act that looked impressively successful. Their new public discourse made it clear that their rejection of the socialist reality should not be seen as an endorsement of the Western capitalist system (called a ‘free-market economy’ in West German media). For a few months, in between the collapse of the old system and the West German takeover, they tried to develop a joint vision of a Third Way between both systems. Everyone was asked to contribute their ideas. The official newspaper of the reformed but still ruling communist party, the Neues Deutschland, described the situation just one month after the fall of the Berlin Wall, on 11 December 1989, in these words: ‘This third way beyond bureaucratic-centralistic socialism and beyond the tyranny of transnational monopoles, to which our party commits itself, has to be defined more clearly – but in joint, combative discussions of all comrades.’ When this vision began to enchant equally dissatisfied people on the other side of the border, the Western media were quick to steer against it. In a special issue of the liberal taz newspaper, produced on 23 February 1990 for East German consumption, readers were told: ‘All deliberations concerning autonomy [of the East] and a “third way” have vanished as if they had never existed … People are
272
Meaning, Discourse and Society
fed up, also with the third way.’ Only a week later, on 3 March, the West German Rheinischer Merkur formulated the new credo for consumption in the West: ‘Whoever isn’t blind to the real world has become aware that [our] socially regulated market economy already is the often sought ‘third way’, between inhuman capitalism and ineffective (and thus not less inhuman) socialism.’ By the end of the month, the third way was forgotten in East and West. Rather than collaborate in the construction of a societal design to their own advantage, the people in the East proceeded avidly to consume Western-owned media and engaged themselves in political parties funded by Western money. The boardrooms had succeeded in averting the dangerous spectre of people exercising collective agency and a joint intentionality (Teubert 1996). The East German people’s revolution had failed. Perhaps one of the reasons is that once the eastern media had been taken over by West German consortia, there was no space in which a counter-discourse could have established itself. Twenty years later, it looks as if things could be beginning to change. As long as the internet retains its anarchic structure, it offers itself as a forum for disenfranchised people. The web discourse is as plurivocal as one can imagine. It is already leaving traces in the official public discourse. The media can no longer easily ignore popular alternative realities. It is still too early to see if blogging and related web activities will bring an end to the long-standing division between text producers and text consumers. Our authorities are already designing steps to control and censure the web. But it is never too late. By making sense of the realities we find in the discourses to which we are exposed, it is up to us to make new realities happen.
Bibliography
Agnesund, M. (1997) Representing culture-specific knowledge in a multilingual ontology’, Proceedings of the IJCAI-97 Workshop on Ontologies and Multilingual NLP, Nagoya, Japan. Al-Khalili, J. (2007) ‘In a parallel universe, this theory would make sense’, The Guardian (1 December). Allen, G. (2000) Intertextuality, Routledge, London. Ammonius (1998) On Aristotle’s On Interpretation 9, with Boethius, On Aristotle’s On Interpretation 9, transl. D. Blank and N. Kretzmann, Cornell University Press, Ithaca, N.Y. Anon. (2006) ‘The word: common sense’, New Scientist, 15 April. Available: www.newscientist.com/channel/opinion/mg19025471.700-the-word-commonsense.html. Anselm of Canterbury (1980) Complete Philosophical and Theological Treatises, transl. J. Hopkins and H. Richardson, The Arthur J. Banning Press, Minneapolis. Aristotle Peri Hermeneias (‘On Interpretation’). Available: www.greektexts.com/library/Aristotle/On_Interpratation/eng/752.html Armstrong, D. F., W. C. Stokoe and S. E. Wilcox (1994) ‘Signs of the origin of syntax’, Current Anthropology, 35/4: 349–68. Averroes (1998) Averroes’ Middle Commentaries on Aristotle’s Categories and De Interpretatione, transl. C. E. Butterworth, St. Augustine’s Press, South Bend, Indiana. Ayer, A. J. (1956) The Problem of Knowledge, Penguin, Harmondsworth. Bach, K. (2005) ‘The top 10 misconceptions about implicature’ in B. Birner and G. Ward (eds.), Festschrift for Larry Horn, John Benjamins, Amsterdam. Available: http://userwww.sfsu.edu/~kbach/TopTen.pdf. Baddeley, A. D. (1999) Essentials of Human Memory, Psychology Press, Hove. Basford, L. and O. Slevin (2003) Theory and Practice of Nursing: An Integrated Approach to Caring Practice, 2nd revised edition, Nelson Thornes Ltd. Bazerman, C. (1981) ‘What written knowledge does: three examples of academic discourse’, Phil. Soc. Sci. 11: 361–87. Bearison, D. and B. Dorval (2002) Collaborative Cognition, Ablex Publishing, London. Bennett, M. and P. Hacker (2003) Philosophical Foundations of Neuroscience, Blackwell, Oxford. Bennett, M., D. Dennett, P. Hacker and J. Searle (2007) Neuroscience and Philosophy: Brain, Mind and Language, Columbia University Press, New York. Berger, P. and T. Luckmann (1966) The Social Construction of Reality: A Treatise in the Sociology of Knowledge, Allan Lane, New York. 273
274
Bibliography
Berlin, B. and P. Kay (1969) Basic Color Terms: Their Universality and Evolution, University of California Press, Berkeley. Bierwisch, M. (1967) ‘Some semantic universals of German adjectivals’, Foundations of Language 3: 1–36. (1970) ‘Einige semantische Universalien in deutschen Adjektiven’ in Hugo Steger (ed.) Vorschläge für eine strukturale Grammatik des Deutschen, Wissenschaftliche Buchgesellschaft, Darmstadt. Billeter, J.-F. (2006) Contre François Jullien. Editions Allia. Paris. Blumer, H. (1969) Symbolic Interaction: Perspective and Method, University of California Press, Berkeley. Brigham, N. (2000) ‘Machine Translation: Its Past, the Potential and the Problem’. Available: www.nbrigham.org/Brigham_machinetranslation.html. Brookfield, F. M. (1989) ‘The New Zealand Constitution: the search for legitimacy’ in H. Kawharu (ed.) Waitangi: Maori and Pakeha Perspectives of the Treaty of Waitangi, Oxford University Press, Auckland, pp. 1–25. Bruner, J. (1990) Acts of Meaning, Harvard University Press, Cambridge, Mass. Burr, V (1995) An Introduction to Social Constructionism, 1st edition, Routledge, London. Burroughs, J. (1915) The Breath of Life Available: www.infomotions.com/etexts/ Gutenberg). Busse, D. and W. Teubert, (1994) ‘Ist Diskursgeschichte ein sprachwissenschaftliches Objekt? Zur Methodenfrage der historischen Semantik’ in W. Teubert, F. Hermanns and D. Busse (eds.) Begriffsgeschichte und Diskursgeschichte. Methoden und Forschungsergebnisse der historischen Semantik, Westdeutuscher Verlag, Opladen, 10–28. Carter, R. M. (2007a) ‘The myth of dangerous human-caused climate change’, Australasian Institute of Mining & Metallurgy, New Leaders Conference, Brisbane, May 2–3 2007, Conference Proceedings, pp. 61–74. (2007b) ‘Stratigraphy into the 21st Century’, Stratigraphy 4: 187–93. Chalmers, D. (1999) ‘Is there synonymy in Ockham’s mental language?’ in P. V. Spade (ed.) The Cambridge Companion to Ockham, Cambridge University Press, Cambridge, pp. 76–99. Chomsky, N. (1959) ‘A review of B. F. Skinner’s Verbal Behavior’, Language 35: 26–58. (1965) Aspects of a Theory of Syntax, MIT Press, Cambridge, Mass. (1966) Cartesian Linguistics, Harper and Row, New York. (1972) Language and Mind, Harcourt Brace Jovanovich, New York. (1990) ‘On formalisation and formal linguistics’, Natural Language and Linguistics 8: 143–7. (2000a) New Horizons in the Study of Language and the Mind, Cambridge University Press, Cambridge. (2000b) ‘Language and Interpretation’ in N. Chomsky (2000) New Horizons in the Study of Language and the Mind, Cambridge University Press, 46–74. (First published in 1992.) Chomsky, N., M. D. Hauser and W. T. Fitch (2002) ‘The faculty of language: what is it, who has it, and how did it evolve?’, Science 298: 1569–79. Chossudovsky, M. (1997). Exporting Apartheid to Sub-Saharan Africa. Available: www.hartford-hwp.com/archives/37/076.html.
Bibliography
275
Christensen, S. M. and D. R. Turner (eds.) (1993) Folk Psychology and the Philosophy of Mind, Hilldale, Lawrence Erlbaum, New Jersey. Clanchy, M. (1979) From Memory to Written Record: England 1066–1307, Arnold, London. Clapham, J. H. (1926) An Economic History of Modern Britain: The Early Railway Age 1820–1850, Cambridge University Press, Cambridge. Cobuild English Dictionary for Advanced Learners (2001) 3rd edition, Harper Collins Publishers, Glasgow. Colapinto, J. (2007) ‘The interpreter: has an Amazonian tribe upended our understanding of language?’, The New Yorker, 16 April. Available: www.newyorker.com/reporting/2007/04/16/070416fa_fact_colapinto). Columbia Electronic Encyclopedia (2000) 6th edition. Cooren, F. (2004). ‘The communicative achievement of collective minding: analysis of board meeting excerpts’, Management Communication Quarterly 17/4: 517–51. Crichton, M. (2003) ‘Aliens cause global warming’. Speech given at the California Institute of Technology, Pasadena, 17 January. Cronk, G. (2003) Dictionary of Literary Biography [Vol. 270: American Philosophers before 1950], Thomson Gale, Farming Hills, Michigan. Dennett, D. (1969) Content and Consciousness, Humanities Press, New York. (1993) Consciousness Explained, Penguin, Harmondsworth. (1998) ‘Reflections on language and mind’ in P. Carruthers and J. Boucher (eds.), Language and Thought, Cambridge University Press, Cambridge, pp. 284–94. (2007) ‘Philosophy as naive anthropology’ in M. Bennett, M. and P. Hacker, Neuroscience and Philosophy, p. 86ff. Derrida, J. (1974) Of Grammatology, The Johns Hopkins University Press, Baltimore. (1988) Limited Inc., Northwestern University Press, Chicago. Descartes, R. (1968) Discourse on Method and the Meditations, Penguin, Harmondsworth. Dilthey, W. ([1883]1989) Introduction to the Human Sciences, edited by R. Makkreel and F. Rodi, Princeton University Press, Princeton, NJ. (2002) Selected Works, Volume 3: The Foundation of the Historical World in the Human Sciences, Princeton University Press, Princeton, New Jersey. Dudai, Y. (1989) Neurobiology of Learning and Memory, Oxford University Press, Oxford. Dunbar, R. (2004) Grooming, Gossip and the Evolution of Language, Faber and Faber, London. Duranti, A. (2006) ‘The social ontology of intentions’, Discourse Studies 8/1: 31–40. Durkheim, E. ([1897] 1982) ‘The materialist conception of history’ in S. Lukes (ed.) The Rules of Sociological Method, Free Press, New York, pp. 167–74. Eco, U. (1995) The Search for the Perfect Language, Blackwell, Oxford. (1999) Kant and the Platypus, Secker and Warburg, London. Edwards, D., M. Ashmore and J. Potter (1995) ‘Death and furniture: the rhetoric, politics and theology of bottom line arguments against relativism’, History of Human Sciences 8: 25–49. Edwards, R. D. (1973) An Atlas of Irish History, Methuen, London. Elwert, G. (2001) ‘Societal literacy: writing culture and development’ in D. R. Olson and N. Torrance (eds.), The Making of Literate Societies, Blackwell, Oxford, pp. 54–67. Engel, U. (1994) Syntax der deutschen Gegenwartssprache, Erich Schmidt, Berlin.
276
Bibliography
Evans, V., B. Bergen, and J. Zinken (eds.) (2006) The Cognitive Linguistics Reader, Equinox Press, London. Everett, D. (2005) ‘Cultural constraints on grammar and cognition in Pirahã: another look at the design features of human language’, Current Anthropology 46(4): 1–69. (2008) Don’t Sleep, There Are Snakes: Life and Language in the Amazon Jungle, Profile Books, London. Fairclough, N. (1989) Language and Power, Longman, London. Feng, Z. (1994) Die chinesischen Schriftzeichen in Vergangenheit und Gegenwart, WVT Wissenschaftlicher Verlag, Trier. Fish, Stanley (1980) Is there a Text in This Class? The Authority of Interpretive Communities. Harvard University Press, Cambridge, Mass. Fitch, W. T., M. D. Hauser and N. Chomsky (2005) ‘The evolution of the language faculty: Clarifications and implications’, Cognition 97: 179–210. Fodor, J. A. (1975) The Language of Thought, Harvard University Press, Cambridge, Mass. (1990) A Theory of Content and Other Essays, MIT Press, Cambridge, Mass. (1998) Concepts: Where Cognitive Science Went Wrong, Clarendon Press, Oxford. (2005) ‘Give me that juicy bit over there’, London Review of Books, 6 October. (2007) ‘Why pigs don’t have wings’, London Review of Books, 18 October. Fogassi, L. and P. F. Ferrari (2007) ‘Mirror neurons and the evolution of embodied language’, Current Directions in Psychological Science 16/3: 136–41. Foucault, M. (1972) The Archaeology of Knowledge, Tavistock, London. Frank, M. (1977) Das individuelle Allgemeine: Textstrukturierung und interpretation nach Schleiermacher, Suhrkamp, Frankfurt am Main. Frisby, J. P. (1980) Seeing: Illusion, Brain and Mind, Oxford University Press, Oxford. Gadamer, H.-G. (1989) Truth and Method, 2nd edition, transl. J. Weinsheimer and D. G. Marshall, Continuum, New York. Gallese, V. and G. Lakoff (2005) ‘The brain’s concepts: the role of the sensory-motor system in conceptual knowledge’, Cognitive Neuropsychology 22: 455–79. Gergen, K. J. (1994) Realities and Relationships, Harvard University Press, Cambridge, Mass. (2001) ‘The construction in contention’, Theory and Psychology 11: 419–32. Giridharadas, A. (2008) ‘India’s novel use of brain scans in courts is debated’, New York Times, 14 September. Goldman, A. I. (1981) ‘Review of Richard Rorty: Philosophy and the Mirror of Nature’, The Philosophical Review 90/3: 424–9. (2006) ‘Social epistemology’ in the Stanford Encyclopedia of Philosophy. Available: www.plato.stanford.edu; revised version of 18 August. Goody, J. (1977) The Domestication of the Savage Mind, Cambridge University Press, Cambridge. (1986) The Logic of Writing and the Organization of Society, Cambridge University Press, Cambridge. (1987) The Interface Between the Written and the Oral, Cambridge University Press, Cambridge. Greimas (1983 [1966]) Structural Semantics: an Attempt at a Method, transl. D. McDowell, R. Schleider and A. Velie, University of Nebraska Press, Lincoln, Nebr. Grice, H. P. (1989) Studies in the Way of Words, Harvard University Press, Cambridge, Mass.
Bibliography
277
Haarmann, H. (1990) Universalgeschichte der Schrift, Campus Verlag, Frankfurt am Main. Habermas, J. (1971) ‘Zu Gadamers “Wahrheit und Methode” in Theorie-Diskussion Hermeneutik und Ideologiekritik. Suhrkamp, Frankfurt a. M. (1981) Theorie des kommunikativen Handelns, Suhrkamp, Frankfurt am Main (Theory of Communicative Action, Polity, Cambridge, 1884–1887). Hacking, I. (1992) ‘Statistical language, statistical truth and statistical reason: the selfauthentication of a style of scientific reasoning’ in E. McMullin (ed.) Social Dimensions of Science, Notre Dame University Press, Notre Dame, pp. 130–57. Halliday, M. A. K. and M. I. M. Matthiessen (1999) Construing Experience through Meaning, Continuum, London. Harras, G. (2000) ‘Concepts in linguistics – concepts in natural language’ in B. Ganter and G. M. Mineau, (eds.), Conceptual Structures: Logical, Linguistic and Computational Issues, Springer, Heidelberg, pp. 13–26. Harris, R. (1981) The Language Myth, Duckworth, London. (1987) The Language Machine, Cornell University Press, Ithaca, New York. (1990) ‘On redefining Linguistics’ in H. G. Davis and T. J. Taylor (eds.), Redefining Linguistics, Routledge, London, pp. 18–53. Hermanns, F. (1994) Schlüssel-, Schlag- und Fahnenwörter. Zu Begrifflichkeit und Theorie der lexikalischen «politischen Semantik», Univ., Heidelberg (Arbeiten aus dem Sonderforschungs-bereich 245, «Sprache und Situation»), Heidelberg, Mannheim. (2003) ‘Linguistische Hermeneutik’ in A. Linke, H. Ortner, and P. R. PortmannTselikas (eds.), Sprache und mehr: Ansichten einer Linguistik der sprachlichen Praxis, Niemeyer, Tübingen. Hitzler, R., J. Reichertz and N. Schröer (eds.) (1999) Hermeneutische Wissenssoziologie, UVK Verlagsgesellschaft, Konstanz. Hjelmslev, L. (1963 [1943]) Prolegomena to a Theory of Language, The University of Wisconsin Press, Menasha, Wisconsin. Hochschild, A. (1998) King Leopold’s Ghost, Houghton Mifflin, New York. Holquist, M. (2002) Dialogism, 2nd edition, Routledge, London. Hughes, J. A., P. J. Martin and W. W. Sharrow (1995) Understanding Classical Sociology, Sage, London. Hume, D. (1927) Selections, edited by Hendel, Ch. W., Charles Scribners, London. Hutchins, E. (1995) Cognition in the Wild, MIT Press, Cambridge, Mass. Jackendoff, R. S. (1987) Consciousness and the Computational Mind, MIT Press, Cambridge, Mass. (1997) The Architecture of the Language Faculty. MIT Press, Cambridge, Mass. Jackendoff, R. S. and S. Pinker (2005) ‘The nature of the language faculty and its implications for evolution of language (reply to Fitch, Hauser, & Chomsky)’, Cognition 19/2: 211–25. Jerne, N. K. (1985) ‘The Generative Grammar of the Immune System’ Nobel Lecture, 8 December 1984, Basel Institute for Immunology, Basel. Katz, J. J. and J. A. Fodor (1963) ‘The structure of a semantic theory’, Language 39/2: 170–210. Kay, P. (1977) ‘Language evolution and speech style’ in B. G. Blount and M. Sanches (eds.), Variability and Change: Sociocultural Dimensions of Language Change, Academic Press, New York.
278
Bibliography
Keele, R. (2007) ‘Review of Claude Panaccio (2004) Ockham on Concepts, Ashgate, Aldershot’, Journal of the History of Philosophy 45/4: 659. Kenyon, F. (1941) The Myth of the Mind, Thinker’s Library, Watts & Co., London. King, P. (2005) ‘Ockram on the role of concepts’. Available: http://individual.utoronto.ca/pking/presentations/Ockham_on_Concepts.pdf. Kitcher, P. (1994) ‘Contrasting conceptions of social epistemology’ in F. F. Schmitt (ed.) Socializing Epistemology: The Social Dimensions of Knowledge, Rowman and Littlefield, Lanham (Maryland), pp. 111–34. (2001) Science, Truth and Democracy, Oxford University Press, Oxford. Klein, N. (2001) No Logo, Flamingo, London. Kristeva, J. (1986) The Kristeva Reader, T. Moi (ed.), Blackwell, Oxford. Lakoff, G. and M. Johnson (1999) Philosophy in the Flesh, Basic Books, New York. Langacker, R. W. (1997) ‘The contextual basis of cognitive semantics’ in J. Nuyts and E. Pedersen (eds.), Language and Conceptualization, Cambridge University Press, Cambridge, pp. 229–52. (2002) The Cognitive Basis of Grammar, Mouton de Gruyter, Berlin. Laslett, P. (1950) The Physical Basis of Mind, Blackwell, Oxford. Leibniz (1666) ‘Ars Combinatoria’, Doctoral Thesis. Levinson, S. C. (1997) ‘From outer to inner space: linguistic categories and nonlinguistic thinking’ in E. Pederson and J. Nuyts (eds.), With Language in Mind: the Relationship Between Linguistic and Conceptual Representation, Cambridge University Press, Cambridge, pp. 13–45. Lévi-Strauss, C. (1966) The Savage Mind, The University of Chicago Press, Chicago. Lewandowski, T. (1976) Linguistisches Wörterbuch, 2nd edition, Quelle und Meyer, Heidelberg. Lieberman, P. (2007) ‘Review: The Stuff of Thought by Steven Pinker’, New Scientist, 3 October 2007. Locke, J. (1764) Second Treatise of Government. Available: http://oregonstate.edu/instruct/phl302/texts/locke/locke2/locke2nd-a.html. Longino, H. E. (2002) The Fate of Knowledge, Princeton University Press, Princeton. Lotman, Y. M. (2001) Universe of the Mind, I. B. Tauris, London. Luhmann, N. (1998) Die Gesellschaft der Gesellschaft, Suhrkamp, Frankfurt am Main. Mannheim, K. (1936) Ideology and Utopia, Routledge, London. Marx, K. and F. Engels (1888) ‘Theses on Feuerbach’, transl. A. Wood, Marx Engels Werke, 3: 7. McGinn, C. (1991) The Problem of Consciousness, Blackwell, Oxford. (2007) ‘How you think’, New York Review of Books, 27 September. Mead, G. H. (1934) Mind, Self and Society, The University of Chicago Press, Chicago. Melby, A. K and C. T. Warner (1995) The Possibility of Language. A Discussion of the Nature of Language, with Implications for Human and Machine Translation, Benjamins, Amsterdam. Meyer-Oeser, S. (2006) ‘Medieval Semiotics’ in Stanford Encyclopedia of Philosophy. Available: http://plato.stanford.edu; revised version of August 18. Millikan, R. G. (2004) Language, Thought and Other Biological Categories, MIT Press, Cambridge, Mass. (2005) ‘The language–thought partnership: a bird’s eye view’, Language and Communication 21/2: 157–66.
Bibliography
279
Morris, D. B. (1993) The Culture of Pain. 2nd edition, University of California Press. Berkeley. Mueller-Vollmer, K. (1986) The Hermeneutics Reader, Blackwell, Oxford. Mullins, J. (2005) ‘Whatever happened to machines that think?’, New Scientist, 23 April: 32–7. Nagel, T. (1997) The Last Word, Oxford University Press, Oxford. Neeson, J. M. (1993) Commoners: Common Right, Enclosure and Social Change in England, 1700–1820, Cambridge University Press, Cambridge. Newell, A. and H. A. Simon (1963) ‘GPS: a program that stimulates human thought’ in E. Feigenbaum and J. Feldman (eds.), Computers and Thought, MIT Press, Cambridge, Mass., pp. 279–93. Ohreen, David (2007) ‘Why folk psychology is not universal’. Available: http://facweb.bcc.ctc.edu/wpayne/why_folk_psychology_is_not_unive.htm). Ong, W. (1982) Orality and Literacy, Routledge, London. Onians, J. (2005) Neuroarthistory: from Aristotle and Pliny to Baxandall and Zeki, Yale University Press, London. Origgi, G. and D. Sperber (2000) ‘Evolution, communication and the proper function of language’ in P. Carruthers and A. Chamberlain (eds.), Evolution and the Human Mind: Modularity, Language and Meta-cognition, Cambridge University Press, Cambridge, pp. 140–69. Ormiston, G. L. and A. D. Schrift (eds.) (1990) Transforming the Hermeneutic Context, State University of New York Press, Albany, N.Y. Oxford English Dictionary (2007) Online edition, June. Panaccio, C. (1999) ‘Semantics and mental language’ in P. V. Spade, (ed.), The Cambridge Companion to Ockham, Cambridge University Press, Cambridge, pp. 53–75. Peacock, T. L (1816) Headlong Hall, J. M. Dent & Co, London. Pederson, E. and J. Nuyts (1997) Language and Conceptualization, Cambridge University Press, Cambridge. Peirce, C. (1878) ‘How to make our ideas clear’, Collected Papers of Charles Sanders Peirce 5: 406–7. Pinker, S. (1994) The Language Instinct, Harper Collins, New York. (2002) The Blank Slate, Viking Adult, New York. (2007) The Stuff of Thought, Viking, New York. Plato Cratylus, transl. B. Jowett, Project Gutenberg online version: http://bang.pmc.purdue.edu/victorian/uploads/R00010/Cratylus.pdf. Potter, J. (1996) Representing Reality: Discourse, Rhetoric and Social Construction, Sage, London. Pottier, B. (1978) ‘Entwurf einer modernen Semantik’ in Horst Geckeler (ed.) Strukturelle Bedeutungslehre. Wissenschaftliche Buchgesellschaft, Darmstadt, pp. 68ff. Proudhon, P.-J. (1840) ‘La propriété, c’est le vol!’ (What is property? or, an inquiry into the principle of right and of government). Available: www.Indypublish.com (2002). Putnam, H. (1981) Reason, Truth, and History, Cambridge University Press. (1998) Representation and Reality, MIT Press, Cambridge, Mass. Random House Dictionary of the English Language (1975). Ravenscroft, I. (2004) ‘Folk psychology as a theory’ in Stanford Encyclopedia of Philosophy, http://plato.stanford.edu/entries/folkpsych-theory/).
280
Bibliography
Ricoeur, P. (1976) Interpretation Theory: Discourse and the Surplus of Meaning, The Texas Christian University Press, Fort Worth, Texas. Roehm, D. (2004) ‘Waves and Words: Oscillatory Activity and Language Processing’, PhD Thesis, Marburg. Rorty, R. (1979) Philosophy and the Mirror of Nature, Princeton University Press, Princeton, NJ. (1982) ‘Contemporary philosophy of mind’, Synthese 53: 323–48 Rose, S. (2005) The 21st Century Brain: Explaining, Mending and Manipulating the Mind, Jonathan Cape, London. Rousseau, J.-J. (1754) Discourse on the Origin of Inequality. (1909–14) ‘Profession of Faith of a Savoyard Vicar’, P. F. Collier & Son, New York. Ryle, G. (1973) The Concept of Mind, Penguin, Harmondsworth. Salomon, G. (1993) Distributed Cognitions: Psychological and Educational Consideration, Cambridge University Press, Cambridge. Samson, A. (2002) ‘Rhodes 1930’, University of Otago Magazine, 3: 15. Available: www.otago.ac.nz/news/otagomagazine/otagomagazineissue3. Sampson, G. (2005) The ‘Language Instinct’ Debate, Continuum, London. Saunders, B. A. C. and J. van Brakel (1997) ‘Are there non-trivial constraints on colour categorization?’, Behavioral and Brain Sciences 20 (2): 167–228. Saussure, F. de (1983) Course in General Linguistics, Duckworth, London. Scaruffi, P. (2003) Thinking about Thought: a Primer on the New Science of Mind, Writers Club Press, Lincoln. Scharfe, H. (2002) Education in Ancient India, Brill Academic Publishers, Leiden. Schleiermacher, F. D. E. (1986) ‘General theory and art of interpretation’ in K. MuellerVollmer (ed.) The Hermeneutics Reader, Blackwell, Oxford, pp. 72–97. Schmidt, L. K. (2006) Understanding Hermeneutics, Acumen, Stocksfield. Searle, J. R. (1980) Mind, Brains and Programs, Cambridge University Press, Cambridge. (1990) ‘Is the brain a digital computer?’, Proceedings and Addresses of the American Philosophical Association 64/3: 21–37. Available: http://users.ecs.soton.ac.uk/harnad/Papers/Py104/searle.comp.html. (1992) The Rediscovery of the Mind, MIT Press, Cambridge, Mass. (1995) The Construction of Social Reality, The Free Press, New York. (1998) Mind, Language and Society, Basic Books, New York. (2005) ‘The quest for consciousness: review of Kristof Koch The Quest for Consciousness’, The New York Review of Books, 13 January. Sinclair, J. (1996) ‘The empty lexicon’, International Journal of Corpus Linguistics 1: 99–109. Skinner, B. F. (1957) Verbal Behavior, Appleton-Century-Crofts, New York. (1992) Verbal Behaviour, Copley Publishing Group, Acton, Mass. Smart, J. S. (2006) ‘CLAW 2006 5th International Workshop on Controlled Language Applications’, AMTA 2006, Boston Marriott, Cambridge, Mass. Available: www.geocities.com/controlledlanguage/SMARTCLAW06. Smolensky, P. (1994) ‘Computational models of the mind’ in S. Guttenplan (ed.) A Companion to the Philosophy of Mind, Blackwell, Oxford. Solomon, M. (1994) ‘Social empiricism’, Nous 28/3: 327–43.
Bibliography
281
Solomon, R. (1988) Continental Philosophy Since 1770. Oxford University Press, Oxford. Spade, P. V. (2006) ‘William of Ockham’ in Stanford Encyclopedia of Philosophy. Available: http://plato.stanford.edu; revised version of August 18. Sperber, D. and D. Wilson (1998) ‘The mapping between the mental and the public lexicon’ in P. Carruthers and J. Boucher (eds.), Language and Thought, Cambridge University Press, Cambridge, pp. 184–200. Stehr, N. and V. Meja (2005) Society and Knowledge: Contemporary Perspectives in the Sociology of Knowledge and Science, 2nd edition, Transaction Publishers, London. Stich, S. (1983) From Folk Psychology to Cognitive Science: the Case Against Belief, MIT Press, Cambridge, Mass. Available: http://plato.stanford.edu/entries/mentalrepresentation/. Tamimi, A. (2001) ‘American’s crusade’, The Jakarta Post, 25 September. Teubert, W. (1996) ‘Zum politisch-gesellschaftlichen Diskurs im Postsozialismus’ in R. Reiher et al. (eds.), Von Buschzulage und Ossinachweis, Aufbau Verlag, Berlin, pp. 286–318. (1999) ‘Starting with Trauer: Approaches to multilingual lexical semantics’ in F. Kiefer, F. Kiss and J. Pajzs (eds.), Complex 99 (Proceedings), HAS, Budapest, pp. 153–70. (2007) Corpus Linguistics, Continuum, London. Thatcher, M. (1993) The Downing Street Years, Harper Collins, London. Thomas, J. B. (1995) Meaning and Interaction, Longman, London. Thompson, J. B. (1981) ‘Editor’s introduction’ in Paul Ricoeur, Hermeneutics and the Human Sciences (edited and translated by John B. Thompson), Cambridge University Press, Cambridge. Thorndike, E. (1921) The Teacher’s Wordbook, Columbia University, New York. Toolan, M. (1996) Total Speech: Integrational Linguistic Approach to Languages, Duke University Press, Durham (North Carolina). (1997) ‘What is Critical Discourse Analysis and why are people saying such terrible things about it?’, Language and Literature 6/2: 101. Tsung-i, J. (1991) ‘Questions on the origins of writing raised by the Silk Road’, SinoPlatonic Papers 26: 3. Tsygankov, P. A. and A. P. Tsygankov (2004) ‘Dilemmas and promises of Russian liberalism’, Communist and Post-Communist Studies 37: 53–70. Volosinov, V. (1986) Marxism and the Philosophy of Language, Harvard University Press, Cambridge, Mass. Von Foerster, H. (1993) ‘Für Niklas Luhmann: wie rekursiv ist Kommunikation?’, Teoria Sociologica 2: 61–88. Waldmann, P. (2007) ‘Is there a culture of violence in Colombia?’ International Journal of Conflict and Violence 1: 61–75. Weber, M. (1968) Economy and Society, Bedminster Press, New York. Weigand, H. (1997) ‘A Multilingual Ontology-based Lexicon for News Filtering’. Available: www.uvt.nl/infolab/prj/trevi/trevi.ps. Widdowson, H. G. (1995) ‘Discourse analysis: a critical view’, Language and Literature 4/3: 159.
282
Bibliography
Wootton, D. (2004) ‘Review of Hugh Grady (2002) Shakespeare, Machiavelli and Montaigne: Power and Subjectivity from Richard II to Hamlet’, The Review of English Studies 55/220: 455–7. Wierzbicka, A. (1996) Semantics: Primes and Universals, Oxford University Press, Oxford. Young, J. Z. (1978) Programs for the Brain, Oxford University Press, Oxford. Zeki, S. (2000) ‘Abstraction and idealism’, Nature, 404: 547.
Index
a priori category, 55, see also Kant, Immanuel Aeschylus, 248–249 agent, 88, 100–101, 103, 105, 131, 138, 162, 198, 243, 247, 262, 266, 270 Agnesund, Mattias, 69, 71 Al-Khalili, Jim, 173 Allen, Graham, 211 ambiguity, 16, 67–69, 73, 80–82, 253 ambivalence, 198, 210, 235 American Philosophical Association, 103 Ammonius, 47 analytic philosophy, 11, 24, 26, 51, 89, 141, 175, 200–201 Anglo-American philosophy, 196 anthropology, 33 antirealism, 11, 196 antonym(s), 73 arbitrariness, 39, 126 arbitrary, 14, 17, 35, 118, 126, 144, 152, 179, 216, 218, 221, 230 construct, 83, 107, 128 setting, 77 sign, 4 Aristotle, 14, 46–48, 50 Armstrong, David, 145 artefact, 118, 124, 141, 153, 190, 194 artefactual, 142 artes liberales, 8 artificial intelligence, 13, 39, 41, 67, 95, 243 community, 40–41, 69, 177 project, 71 authority, 5, 121, 156, 169, 179, 212, 214, 240 autonomy, 24, 38, 200–201, 214, 258–259, 266–267, 271 semantic, 214 autopoietic machine, 20 autopoietic process, 138 autopoietic system, 208 Ayer, Alfred J., 38 Bach, Kent, 80 Baddeley, Alan, 98
Bakhtin, Mihail, 134, 138, 144, 165, 210–211 Bank of English, 222 Baroque, 212 Baroque churches, 255 Basford, Lynn, 191 Bazerman, Charles, 192–195 Bearison, David, 25, 181 behaviour verbal behaviour, 33 behaviourism, 13, 33–34, 38–39, 75, 133 antisocial behaviour, 157 asocial idiosyncratic behaviour, 267 automated behaviour, 126 behaviour, 7, 19, 33–34, 42, 85, 90, 118, 122, 124, 127, 129, 132, 134, 140, 151, 166, 184, 196–197, 233, 235, 238, 241–242, 244–245, 248–249, 261, 263 behavioural feature, 133 behavioural output, 42 behaviourist model, 51 human behaviour, 41, 247 interactional behaviour, 113 interpersonal behaviour, 263 linguistic behaviour, 33, 113 meaning of behaviour, 248 modes of behaviour, 238 non-linguistic behaviour, 167 non-symbolic behaviour, 21, 124–125, 144 non-vocal behaviour, 146 object-oriented behaviour, 140 person’s behaviour, 107 rational behaviour, 170 symbolic behaviour, 43, 129, 167, 184, 241–242, 247, 261, 263 verbal behaviour, 33–34 verbal behaviourism, 37 Bennett, Maxwell, 17, 102–106 Berger, Peter, 25–26, 182–183 Berkeley, George, 21, 98, 250 Berlin, Brent, 252–254 Bierwisch, Manfred, 15, 64–65 Billeter, Jean François, 267
283
284
Index
Bloomfield, Leonard, 23 Blumer, Herbert, 33, 133–134, 165 body, 7, 26, 38–39, 42, 68, 82, 95, 98–99, 107, 119, 161, 178, 183, 226, 263 body language, 146 body–mind dualism, 37, see also Descartes, René Boeckh, August, 200, 212 Borges, Jorge Luis, 116, 218 bottom-up system, 222 brain, 17, 40–42, 44, 51–52, 74–77, 85, 87, 88–92, 94–100, 102–108, 242–243, 253 activity, 97–98, 100 artificial, 71 Brain Electrical Oscillations Signature test (BEOS), 97 brain-external stuff, 17, 107 brain scan, 96–97, 100, 102, 262 as digital computer, see artificial intelligence electronic brain, 13, see also artificial intelligence experts, 97 forebrain, 95, 98 human, 40, 125 individual, 25 memory, 98 mind-brain, 58 new brain sciences, 102 person’s brain, 105–107 potential, 98 reptilian, 238 sciences, 98, 107 structure, 101 Brakel, Jaap van, 253 Brecht, Bertolt, 23 brick, 40, 60 basic conceptual, 60 basic semantic, 84 shared reality of, 159 British National Corpus (BNC), 220, 227, 255 Burr, Vivien, 130 Burroughs, John, 168 Busse, Dietrich, 120 Canterbury, Anselm of, 14, 48 Carter, Robert, 191–192 Cartesian agenda, 197 concept, 197 concept of knowledge, 176 epistemology, 26, 183 mind, 176 souls, 41 understanding, 196
cause, 38, 69, 165, 194, 206, 261, 265, 267 cause and effect, 85, 242, 248 mental cause, 38 central phase, 101 century eighteenth, 8, 23, 37, 212, 245, 258 eighth, 156 fourteenth, 49 nineteenth, 5, 8, 9, 162, 212, 245, 258, 267 seventeenth, 37, 58 third, 47 twentieth, 6–7, 9, 11, 24, 33, 44, 51, 68, 73, 125, 139, 163, 200–201, 214 twenty-first, 192 Chalmers, David J., 49–50 chat, 1, 154 pub, 166 China, 163, 267 choice, 6, 171, 207, 238, 248, 260–261, 269 individual, 266 lexical and grammatical, 128 Chomsky, Noam, 4, 13–14, 33–37, 46–47, 53–54, 56–60, 65–66, 93–94, 113, 125, 128, 139, 144, 181, 252 Chomskyan language universals, 36 Chomskyan linguistics, 4, 9, 13 Chomskyan theory of the semantic level, 58 Chomsky’s generative transformational grammar, 42 language faculty, 52 universal grammar, 17, 128, 139 universalist theory of language, 118 Chomsky-sentences, 57 Clapham, Sir John, 161 Cobuild English Dictionary, 4, 190, 224, 230 cognition, 25, 39, 91–92, 95–97, 182–183, 185, 254, 262 animal, 94 collaborative, 138, 181 concept of, 181 human, 65, 82 Cognition in the Wild, 182 Cognitive Linguistics Reader, 42 cognitivism, 33, 99 accounts, 76 acts, 65 agenda, 12, 183 approach, 13 commitment, 42 conception, 88 concepts, 7, 74, 78, 101, 220 development, 181 ethologists, 105 events, 86 intelligence, 41
Index landscape, 76 language, 45 linguistic research, 46 linguistics, 4, 7, 9, 15–17, 38, 42, 44–46, 48, 52, 54, 62, 64, 74–75, 80, 83, 88, 95, 98–99, 101–102, 107, 222, 224 linguistics community, 107 linguists, 127 mechanisms, 41, 78–80, 94 modelling, 40 models, 7, 13 paradigm, 13, 14 perspective, 129 phenomena, 25 process, 92, 181 processing, 86, 87 programme, 97 psychology, 92 representation, 25, 54, 62, 66, 74, 76–78, 86–87, 240 schools, 7 sciences, 13, 17, 33, 38–39, 41–42, 44–45, 51–52, 75, 91, 98, 107, 136, 243–244, 262 semantics, 65, 86 states, 91 theories, 40 turn, 13, 14, 33 Collaborative Cognition, 181 collocate, 56, 222, 224–225 Columbia Electronic Encyclopedia, 178 communication pseudo-communication, 122 communication system, 9, 34, 139 communicative system, 34 communicative situations, 120 community interpretive, 4, 29, 225, 229–230, 239, 240, 246–247, 249, 270 oral, 130, 151, 155, 158 comprehensive grammar, 57 Comte, Auguste, 51 concept, 7, 14–18, 22, 25, 27, 35–36, 38, 40, 45, 47–50, 53–64, 66–74, 76, 78–79, 84, 86, 89–91, 93–95, 98–102, 105–106, 115, 120, 125, 130, 136, 143–144, 147, 150, 152, 160, 162–164, 171–172, 176, 178–182, 184, 186, 192, 194, 197–200, 208–209, 211, 213, 226, 240, 242, 244–245, 247–248, 253, 257–258, 264, 267, 278 basic, 55, 62 conceptual knowledge, 13 conceptual ontology, 15, 68, 71
285 conceptual representation, 15, 48, 69–70, 82–85, 93 conceptual system, 41 mental, 7, 13–16, 25, 37, 47, 50, 53–56, 58–61, 65–67, 73–78, 83–84, 86, 89, 95, 99, 103, 107, 113, 172, 180, 220 synonymic, 50 Concepts Where Cognitive Science Went Wrong, 53, 59 conceptualisation, 14, 16, 25, 86–87, 107, 170–171, 253, 267 Condillac, Etienne Bonnet de, 145 consciousness, 25, 39, 44–45, 50, 52, 76, 86, 90–92, 94–96, 106–108, 119, 127, 129, 136, 148, 165, 201, 242, 244, 248, 259–261 constituents, 35 construct, 7, 11–12, 15, 27, 36, 38, 42, 52, 60, 63–65, 75, 80, 83, 85, 90, 95, 107, 118, 120–121, 128, 132–134, 140, 150, 172, 184–186, 194, 214, 218, 250, 263 language, 217, 237 linguistic, 80 constructivism, 25 Continental, 65 radical, 209 social, 12, 24–26, 85, 182, 184–185, 250 context, 6, 9, 15, 26, 56, 62, 67–68, 73–74, 77–78, 80, 82, 92, 95, 136, 156, 164, 168, 179–180, 188, 191, 197, 200, 205–207, 218–222, 225, 230, 235, 261 context-dependency, 69 context-free, 67 co-occurrence, 222 Cooren, François, 138, 181, 268 corpus linguistics, 73, 211, 219, 222, 224 corpus, corpora, 18, 73, 117, 212, 219–220, 222, 224, 227, 239 Cratylus, 67, 199 critical discourse analysis (CDA), 18, 120–122 Darwin, Charles, 60, 94 Darwinist, 60, see also anti-Darwinist Dennett, Daniel, 17, 40, 91–92, 102–106, 243 dependency grammar, 36 Derrida, Jacques, 28, 48, 131, 199, 210 Descartes, René, 14, 23–26, 37, 39, 96, 175–176, 181–182, 198 Dewey, John, 26, 33, 133 diachronic dimension, 8, 45, 128, 141, 210, 216, 239 Dictionary of Literary Biography, 133 Dilthey, Wilhelm, 160, 200–201, 225
286
Index
discourse, 1–4, 6–12, 14, 18–29, 38, 42–44, 50–51, 72–74, 77, 79–80, 83, 85, 88–90, 95, 99, 103, 107–109, 114–123, 125–126, 128–143, 147–150, 155–156, 163–172, 177–192, 194–202, 204–205, 207–221, 224–226, 228–230, 232, 234–250, 252–266, 269–272 academic, 103 analysis, 120–121 anarchic, 12 autopoietic, 209 boardroom, 138 chair-discourse, 141 community, 1 construct, 27, 121, 172, 214 discourse-external ‘real world’, 85 discourse-external givens, 119 discourse-external laws, 115 discourse-external reality, 11, 18, 20–21, 26, 28, 136, 141–144, 154–156, 159, 162, 164, 171–172, 177–178, 192, 194–195, 197, 207, 214, 220, 242, 255 discourse-external signifiée, 131, 135, 210 discourse-independent, 10 discourse-internal arrangements, 143 discourse-internal content, 148 discourse-internal interpretation, 132 discourse-internal knowledge, 149, 195 discourse-internal reality, 142 factual, 183 Google, 237 healthy, 12, 147, 269–270 history, 97 intellectual, 18 language, 211, 232 mechanism, 137 oral, 23 orders of, 121 plurivocal, 185 post-September 11, 211, 213 public, 263, 265, 269–272 reality, 239 scientific, 9, 26 special, 18–19, 116–119, 207, 216–217, 221 specific discourse community, 52 studies, 76 written, 170, 213, 219 discourse act, 21 discourse community, 1, 7–8, 10, 14, 20, 24–25, 27–28, 43–44, 50–51, 72, 77, 88, 90, 108, 115, 120, 128–132, 134–136, 143, 150, 155, 156, 169–170, 174–177, 180–181, 183, 185, 189, 192, 199, 208, 209, 213, 215–218, 221, 225, 230, 232, 245, 252, 258, 262–263, 269
discourse element, 192, 209 discourse hermeneutics, 213 discourse object, 8, 11, 12, 21, 26, 59, 86, 141–144, 147, 149, 151, 155, 164–165, 177–178, 180–181, 186–191, 194, 204–205, 208, 212–213, 216, 219–221, 224–226, 229, 231, 236, 239, 247, 255, 259, 264 discourse observer, 120, 132 discourse partner, 103 discourse rupture, 119 discourse society, 119 discourse system, 12 discourse topic, 254 Dix, Alan, 172 Dorval, Brice, 25, 181 Dudai, Yadin, 98 Dunbar, Robin, 146 Duranti, Alessandro, 90, 245 Durkheim, Emile, 51, 127, 165 Eco, Umberto, 61, 178 Elwert, Georg, 157 Engel, Ulrich, 36 Enlightenment, 16, 24–25, 27, 37, 51, 248 entity monadic entity, 95, 103, 108 epistemic complex, 120 European romanticism, 24 EUROTRA – European project for automatic translation, 41, 70 Everett, Daniel, 36, 252, 254 evidence, 17, 22, 24, 36, 44, 50, 53–55, 57, 66, 75–76, 80, 82, 84, 86, 97–98, 104, 121, 150–153, 156, 167, 172, 187, 191, 207, 216, 218, 221–222, 224–225, 230, 232, 235, 239, 240, 242, 246, 253, 263, 264, 268 experience first-person, 43, 89, 105, 135–136, 147, 152, 167, 169, 172, 177, 207, 217, 259, 263 expression(s), 16, 21, 25, 36, 44, 54, 60, 67–68, 72, 76–78, 80–81, 83, 87, 90, 99, 101–105, 115, 125, 130, 132, 142–143, 146, 149–151, 155, 166, 177, 179, 185, 204–206, 208, 212–213, 221, 245, 248, 254, 260, 263, 271 Fairclough, Norman, 120, 121 final state, 40, 101 finite regress, 105, see also infinite regress first-order observation, 209 Fitch, Tecumseh, 34, 37, 93–94 FLB, 93–94
Index FLN, 93–94 Fodor, Jerry, 14, 48, 50–51, 53–55, 57–60, 64, 68, 75–76, 84, 91, 93–94, 99, 113 Foerster, Heinz von, 209 formal universals, 34 Foucault, Michel, 18, 118–121, 195, 210, 218 Francesca, Piero della, 9 Frege, Gottlob, 141 French Revolution, 158 Freud, Sigmund, 51, 210 Gadamer, Hans-Georg, 28, 138, 183–184, 199, 201, 209, 212–213, 221, 225, 239 Gallese, Vittorio, 17, 98–103 Geisteswissenschaften, 8, 51, 65, 225 generative semantics, 100 generative transformational grammar, 4, 128 Gergen, Kenneth, 26, 184–185 German idealism, 24–25 Goldman, Alvin, 176, 197–198 Goody, Jack, 150, 155–156, 158–160 Google, 2, 8, 24, 52, 86, 132, 154, 169–170, 191, 205–206, 219, 225–237, 239, 245, 260, 266 Google search, 257 Greimas, Algirdas J., 15, 63, 65 Grice, Herbert Paul, 80–82 Gricean maxims, 81, 168 Haarmann, Harald, 153 Habermas, Jürgen, 27, 122 Hacker, Peter, 17, 102–106 Hacking, Ian, 151 haiku, 29, 215–219, 221, 224–227, 229, 230–233, 236–241, 270 Halliday, Michael, 120 Hallowes, Harry, 157 Harras, Gisela, 78, 81 Harris, Roy, 9, 125, 144, 150, 167–168, 170 Harris, Zelig, 23 Hauser, Marc D., 34, 37, 93–94, 144 Hegel, Georg F. W., 25, 86, 158 Heidegger, Martin, 28 Herder, Johann G., 145 Hermanns, Fritz, 200 hermeneutics, 27, 28, 65, 119, 138, 168, 183, 199–201, 207–211, 213–214, 242 history of effect (“Wirkungsgeschichte”), 212 Hitzler, Ronald, 183 Hjelmslev, Louis, 63 Hochschild, Adam, 162 Holquist, Michael, 165, 210 Humboldt, Wilhelm von, 36, 212 Hume, David, 8, 25, 181 Husserl, Edmund, 183
287 Hutchins, Edwin, 182 hypernym, 68, 73, 178 hyponym, 73 idea, 1, 3–4, 7, 13–15, 20–25, 27–28, 36–38, 40–41, 46–49, 51–55, 58–61, 69, 74–76, 82, 94, 101, 105, 108, 118, 126, 133, 135–136, 145, 150–151, 160, 162, 172, 174–175, 183, 185–186, 190, 194, 196, 199, 217–218, 240–241, 248–250, 259–260, 265, 270–271 Platonic idea, 39 ideal speech situation, 27 idealism, 250 identity, 108, 115, 119–120, 142, 147, 156, 158–159, 179, 211, 261 idiom, 6, 211 image, 20–22, 48–49, 68, 83, 91, 93, 143, 172, 184, 219–221, 255, 258, 263 imagination, 12, 48, 96, 100, 102–103 infinite regress, 50, 54, 167, 243, see also finite regress initial condition, 101 integrationalist linguistics, 124–125, 127, 167 intentionality, 10–12, 17, 20, 23, 42, 44–45, 49, 60, 76, 86–94, 96, 99, 101, 103, 106–109, 114, 122, 129, 130, 132, 135–136, 138, 179, 181, 183–184, 195, 201, 208–209, 221, 225, 240–243, 246–249, 258–259, 261–263, 270–272 individual intentionality, 20, 108, 209, 248 interaction, 1, 13, 19–21, 25, 33, 37, 54, 77, 86–87, 99, 108, 113, 121–122, 126–127, 129, 132–133, 140, 151, 167, 183–184, 197, 240, 246–247, 249, 259, 270 social, 77–78, 125, 133, 248 symbolic, 4, 17, 33, 95, 122, 124–127, 133–134, 136, 139, 144, 150, 166, 168, 183, 208–209, 221, 259, 262 interpretation, 1–4, 8–9, 27–29, 33, 39, 47, 49, 64, 69, 77, 92–93, 97, 114, 117–118, 120, 126, 129, 132–135, 138, 141, 148, 151, 153, 155–156, 165, 167–168, 173, 175, 183, 192–193, 195, 199–201, 207–219, 221, 224–226, 228–230, 232, 238–240, 242, 244, 246–249, 258, 263, 267, 270 reinterpretation, 29, 103 intertextuality, 201, 210–212 Jackendoff, Ray, 39, 41, 62, 75–77, 93 James, William, 26, 33, 133 Johnson, Mark, 98, 261 Johnson, Samuel, 250 Jullien, François, 267
288
Index
Kant, Immanuel, 25, 55, 179, 200 Katz, Jerold, 64 Kay, Paul, 252–254 Keele, Rondo, 50 Kenyon, Frank, 38 King, Peter, 50–51 Kitcher, Philip, 176, 196 Klein, Naomi, 269 knowledge, 4, 7–9, 13–14, 20, 23–27, 40–41, 60, 62, 67–69, 75, 77, 97–99, 104, 106, 108, 116–119, 127, 139, 148–149, 152, 155, 156–160, 163, 167–169, 175–184, 189–190, 192–193, 195–197, 208, 212, 216, 219, 221, 224–225, 227, 231, 239–241, 243–245, 252, 267 Kristeva, Julia, 210–211 Lakoff, George, 17, 98–103, 115, 261 Langacker, Ronald, 16, 84, 86–87 language, 2, 4–19, 23–26, 33–37, 39–78, 80–95, 98, 101–103, 106–107, 113–118, 120–121, 124–132, 134–137, 139–142, 144–146, 150–151, 154–155, 162–163, 165–170, 172–175, 177, 179, 189, 192–193, 199–200, 203–204, 208, 210–211, 213–214, 217, 219, 221–222, 224, 232, 237, 246, 252–254, 256–257, 262, 264 E-languages, 34–35 I-languages, 34–35, 37 Language and Conceptualization, 40 language community, 25, 84 language data, 18, 24, 66, 73, 121, 128, 211, 222 language user, 6, 10, 125 langue, 52, 95, 128, 139 Laslett, Peter, 38 Leibniz, Gotthold W., 61 Levinson, Stephen, 16, 75–76, 82–85 Lévi-Strauss, Claude, 118, 143, 152 Lewandowski, Theodor, 63–65 lexical item, 8, 42, 56–57, 63, 72–73, 76, 78–80, 83, 99, 107–108, 115, 117, 126, 131, 138, 140–144, 151, 155, 164–165, 177–178, 180–181, 186–189, 199, 201, 205, 207–209, 211, 215, 219–222, 224–226, 229, 232, 256 lexical material, 218 lexis, 211 Lieberman, Philip, 55 linguistic(s), 4, 6, 8–9, 22, 25, 33, 51, 63, 72, 78, 95, 107, 115, 118, 120–121, 125, 127–129, 167, 169, 211, 221–222, 224 computational, 209 data, 121 meaning, 82, 99 patterns, 9
theories, 36, 222 theoretical, 4, 6 literacy, 22, 126, 130–131, 144, 147–149, 151–152, 155–159, 163–164 Locke, John, 161, 181 Lodge, David, 137 Longino, Helen, 197 Lotman, Yuri M., 164 Luckmann, Thomas, 25–26, 182–183 Luhmann, Niklas, 20, 129–132, 135, 138–139, 208–209, 247 Lurija, Aleksandr, 163 machine ghost in the, 38 main clause, 35–36 Mann, Thomas, 56 Mannheim, Karl, 26, 182–183, 204 Marceau, Marcel, 125 Marx, Karl, 3, 158, 210, 266 Maturana, Humberto, 129, 131 McGinn, Colin, 45, 55, 136 Mead, George Herbert, 33, 133, 145, 260 meaning, 1–12, 14–17, 19–21, 23–29, 33, 43–51, 53–56, 58, 61, 63–69, 71–82, 86–88, 90–93, 98–100, 102–107, 109, 114, 116, 118, 121, 124–127, 129, 131–135, 138, 140–144, 146, 148–149, 151–152, 155–156, 158, 163–165, 167–171, 174, 177–181, 183–184, 186, 188–189, 192–193, 196–200, 203–204, 207–208, 211–212, 214–222, 224–226, 228, 230, 232, 236, 239, 241, 244, 246–249, 251, 256–257, 259, 262, 274, 280 linguistic meaning, 82, 99 of words, 5, 6, 87 units of, 3, 6, 23, 80, 211, 229 Meja, Volker, 196 Melby, Alan, 15, 70–71 mentalese, 14, 48, 50, 53–55, 60–62, 68, 71, 82, 84, 92, 107 message, 153, 208, 238 Meyer-Oeser, Stephen, 49 Millikan, Ruth, 81–82, 89, 93 mind, 4, 7, 12–17, 19, 22–26, 28, 31, 33–34, 36–44, 46–52, 57–61, 66, 72, 74–75, 77, 82, 84–92, 94–100, 102–105, 107–109, 114–115, 119, 125, 127–131, 133, 135–138, 143, 155, 164, 166, 168, 173–174, 176, 179, 181–185, 196, 201, 208, 217, 221, 225, 228, 238, 241, 243–246, 248, 250, 255, 258–262, 264–265, 269, 273, 280 collective, 8, 11, 20–21, 23, 25–26, 28, 88, 107–109, 133, 135–136, 138, 156, 169, 180, 183–184, 195, 208–209, 249, 262
Index individual, 11, 20, 23–24, 33, 38, 45, 51, 86, 126–127, 130–131, 133, 135–138, 176, 179–180, 182–184, 208, 240, 262–263 monadic, 7, 8, 19, 29, 37, 135, 143, 176, 181, 184, 241, 267 solitary, 238, 258, 262 mind–body dualism, 96, see also Descartes, René moral sense philosophy, 37 morpheme, 101, 222 Mueller-Vollmer, Kurt, 200–201, 244 Muller, Justin, 71 Nagel, Thomas, 175 network of texts, 162 Neues Deutschland, 271 Neumann, John von, 41 Newell, Allen, 91 nominalism, 65 Nuyts, Jan, 40–41, 86 object(s), 3, 4, 7–9, 12, 17–18, 21, 23–24, 27–28, 37, 40, 43–44, 48–49, 51, 57, 59, 65, 72, 85–86, 88, 91, 100–101, 103–104, 108, 115–121, 122, 125, 128, 133–136, 140–144, 147–153, 155, 160, 164–165, 171–175, 177–180, 185–187, 190–191, 193–194, 196, 209–210, 213, 216, 226–227, 241, 247, 251–257, 259, 261 occurrence(s), 2, 19, 28, 43, 52, 68, 74, 86–87, 117, 131–132, 136, 169–170, 180, 186–188, 205–206, 211, 219, 224–233, 235–237, 239, 245 Ockham on Concepts, 50 Ockham, William of, 14, 49–52, 54, 83, 179 Ockham’s razor, 54, 83 Ohreen, David, 245 Ong, Walter, 155–156 Onians, John, 96–97, 99 ontology, 15, 41, 68–69, 71, 90, 245 Orestes, 249 Origgi, Gloria, 81–82 Oxford English Dictionary (OED), 24, 178 Panaccio, Claude, 50 paradigm, 4, 13, 26–27, 33, 36–37, 39, 73, 95, 97, 109, 118, 120, 125, 139, 151, 174, 185, 196, 253 paraphrase, 6, 47, 74, 80, 102–103, 107, 114, 140–141, 147, 164, 181, 199, 201, 204–205, 207, 212, 219–220, 222, 225–226, 230–232
289 parole, 120, 125, 128–129, 134, 139, 150, 169–170, 213 part-whole, 73 Pederson, Eric, 40–41 Peirce, Charles, 26, 33, 133, 143, 175–176 Petit Robert, 86 phenotypes non-symbolic, 259, 262 philosophy of mind, 11, 14, 16, 38–40, 42, 45, 50, 52, 75, 88–89, 96, 136, 242, 244 phoneme, 63, 102, 222 phonological analysis, 63 Piaget, Jean, 25, 139, 181, 183 piece of speech, 166 Pinker, Steven, 14, 48, 54–55, 62, 68, 76, 84, 93, 259 Plato, 22, 24, 67, 74, 212–213, 279 plurivocality, 120, 192, 207 polysemy, 80 postmodernism, 12, 85 post-structuralism, 28, 119, 200, see also Prague school; structuralism Potter, Jonathan, 179, 183, 250 Pottier, Bernard, 15, 63 pragmatism, 26–27, 33, 85, 133, 175 Prague school 63, see also structuralism primordial speech situation, 20–21, 26, 28, 131, 134, 136, 140–152, 156–157, 164–173, 175, 179–180, 185, 193, 195, 197, 210, 258 Proudhon, Pierre-Joseph, 162 purpose condition, 101 Putnam, Hilary, 14, 24, 28, 51, 58, 141, 174–177 quality, 19, 23, 45, 74, 145, 148, 168, 190, 193, 205, 238, 259–260, 269 quantity, 168, 252 Random House Dictionary, 190 Ravenscroft, Ian, 244 realism, 11, 24, 27, 51, 65, 96, 174–177, 185, 196, 253 reality, 1–3, 5, 7–9, 11–12, 18, 20–22, 24, 26–29, 35, 38, 42, 47, 50–51, 57, 64–66, 71–72, 77, 83, 85, 88, 95, 97, 107–109, 114–117, 120–123, 131, 134, 136, 138–200, 207–208, 214–216, 219–220, 230, 238–239, 242–243, 249–260, 262–265, 267–272, 280 recognition, 9, 44, 81, 242 recontextualisation, 132
290
Index
reference, 29, 53, 98, 121, 129, 131, 140–141, 155, 158, 163, 177–178, 182, 185, 191–193, 203, 208, 214, 239, 254 Reichertz, Jo, 183 relativism, 26–27, 65, 85, 96, 179, 250 relevance theory, 16, 76, 81, 167 representation(s), 7, 15–16, 36, 40–42, 48–50, 54, 60, 62, 66, 69–72, 74, 76–78, 82–87, 93, 95, 98–99, 102–104, 106, 145–146, 170, 173, 176, 190, 207, 209, 219, 240 mental, 16, 43–44, 46–47, 61, 72, 74–75, 78, 80, 82, 87, 91, 93, 95, 102, 104, 107, 128, 184, 189, 209, 241, 247 semantic, 15, 58, 84–85 symbolic, 104, 124, 155, 159, 167, 184, 259, 261 re-representation, 106 Ricoeur, Paul, 138, 144, 199–200, 208, 214 Robinson, Daniel, 103 Roehm, Dietmar, 98 Rorty, Richard, 75, 176 Rose, Steven, 98 Rousseau, Jean-Jacques, 145, 155, 160 Russell, Bertrand, 51 Rutherford, Ernest, 173 Ryle, Gilbert, 38 Salomon, Gavriel, 181 Sampson, Geoffrey, 35 Samson, Alan, 162 Saunders, Barbara, 253 Saussure, Ferdinand de, 6, 9, 23, 45–46, 51–52, 63, 95, 118, 128, 139, 150, 154, 167, 199 Scaruffi, Pierro, 40 Scheler, Max, 182 Schleiermacher, Friedrich, 138, 200, 212, 244 Schmidt, Lawrence, 212 Schröer, Norbert, 183 Searle, John, 11–12, 17, 26, 41, 85, 90–92, 95–96, 102–103, 105–106, 173–174, 209 second-order observation, 209 self-consciousness, 133 semantic markers, 64–65 semantic primes, 15, 60–64 séme(s), 15, 63–65 sémeme(s), 15 Shakespeare, William, 9, 16, 23, 68, 114, 117, 245, 282 Simon, Herbert, 91 Slevin, Oliver, 191 Smart, John, 70 Smolensky, Paul, 39 socialism, 269, 271–272
society, 1, 2, 9, 12, 18–22, 27, 40, 43, 51–52, 77, 90–91, 95, 103, 106, 108–109, 113, 115, 119–122, 124, 127–130, 132–136, 138, 150, 152, 156, 158, 160–161, 163, 165–166, 173, 176, 183, 186, 196, 208–209, 238, 241, 245, 247, 252, 254, 256–259, 262, 264, 266–267, 269 literate societies, 22, 134, 144, 147–149, 155–157, 159–160, 164–165, 213 oral societies, 23, 126, 143–144, 147, 149–151, 154–158, 160, 164 Socrates, 67, 199 Solomon, Miriam, 196 Solomon, Robert, 25 Spade, Paul Vincent, 49–50 Sperber, Dan, 16, 53, 76, 78, 81–82, 167, 170 Stanford Encyclopedia of Philosophy, 49, 197–198, 244 starting phase, 101 Stehr, Nico, 196 Stich, Steven, 92 stimuli, 34, 37, 82, 95, 126, 134, 184 Stokoe, William C., 145, 273 structuralism, 63, 65, 118–119, see also poststructuralism; Prague school subject noun phrase, 35–36 symbol, 43, 47–48, 54, 69, 93, 99, 101–104, 143–144, 152, 164–165, 226 symbolism, 217 synonym(s), 14, 50, 61, 73, 86, 187 Systemic Functional Linguistics, 120 Tamimi, Azzam, 205 Tel Quel grouping, 119 Teubert, Wolfgang, 56, 120, 133, 272 The Quest for Consciousness, 95 thesaurus, 9, 255 Thomas, Jenny, 169 Thompson, John B., 214 Toolan, Michael, 121, 126 trait distinctif, 63 transcripts of spoken language, 166 truth, 8, 11, 17, 24–28, 37, 51, 74, 97, 102, 129, 145, 147–148, 151, 154, 156, 158, 160, 171, 174–177, 184–185, 190–191, 196–198, 201, 207, 212, 264 Turing machine, 39 Turing, Alan, 39 Turner, Dale R., 42 universal grammar, 17, 113, 222 Varela, Francisco, 129 verb phrase, 35–36
Index verbal and kinesic behaviour, 90 Vienna Circle, 51 Volosinov, Valentin, 134, 144, 210 Vygotsky, Lev, 25, 139, 181, 183 Weber, Max, 51, 127 Weigand, Hans, 71
291 Weizenbaum, Joseph, 137 Widdowson, Henry, 121 Wierzbicka, Anna, 15, 60–61, 63, 76, 84, 100 Wilcox, Sherman E., 145 Wilson, Deirdre, 16, 53, 76, 78, 81, 167, 170 Zeki, Semir, 104
