Language Misconceived: Arguing for Applied Cognitive Sociolinguistics

  • 13 85 9
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up

Language Misconceived: Arguing for Applied Cognitive Sociolinguistics

Karol Janicki LANGUAGE MISCONCEIVED Arguing for Applied Cognitive Sociolinguistics LANGUAGE MISCONCEIVED Arguing for

1,310 152 2MB

Pages 213 Page size 432 x 648 pts Year 2007

Report DMCA / Copyright


Recommend Papers

File loading please wait...
Citation preview

Karol Janicki

LANGUAGE MISCONCEIVED Arguing for Applied Cognitive Sociolinguistics

LANGUAGE MISCONCEIVED Arguing for Applied Cognitive Sociolinguistics

LANGUAGE MISCONCEIVED Arguing for Applied Cognitive Sociolinguistics

Karol Janicki University of Bergen, Norway



Copyright Ó 2006 by Lawrence Erlbaum Associates, Inc. All rights reserved. No part of this book may be reproduced in any form, by photostat, microform, retrieval system, or any other means, without the prior written permission of the publisher. Lawrence Erlbaum Associates, Inc., Publishers 10 Industrial Avenue Mahwah, New Jersey 07430

Cover design by Kathryn Houghtaling Lacey

Library of Congress Cataloging-in-Publication Data Janicki, Karol. Language misconceived : arguing for applied cognitive sociolinguistics / Karol Janicki. p. cm. Includes bibliographical references and indexes. ISBN 0-8058-5680-3 (cloth : alk. paper) ISBN 0-8058-5682-X (pbk. : alk. paper) 1. Linguistics—Philosophy. 2. Sociolinguistics. 3. Essentialism (Philosophy) I. Title. P121.J277 2005 410¢.1—dc22

2005052262 CIP

Books published by Lawrence Erlbaum Associates are printed on acid-free paper, and their bindings are chosen for strength and durability. Printed in the United States of America 10 9 8 7 6 5 4 3 2 1


Preface I



1 Meaning and Definitions Misconceived


Introduction 3 Explicit (Overt) Essentialism 5 Covert Essentialism 6 Criticism of Essentialism 7

2 Word Meaning, Definitions, Linguists, and Philosophical Commitments


Introduction 11 Endnote 37

3 Integrational Linguistics and Nonessentialism


Introduction 38 Roy Harris’ Integrational Linguistics 39 Conclusions 47

4 Variations on Meaning in Cognitive Linguistics


On the Fuzziness of Concepts 49 More on the Imprecision of Concepts 55




Slipping Into Essentialism? 59 On Flexibility and Creativity 63 Treating Language “Seriously” 65 On Unconventional Metaphor 66 On Misunderstanding 72 Conclusion 81 Endnotes 82



5 Language in Politics


The Clinton–Lewinsky Case 89 The Florida Vote Conflict 96 Abortion, Euthanasia, and the Stem-Cell Research Conflicts 98 Political Correctness 108 Miscellaneous 120 Conclusions: From Essentialism to Conflict and War 125 Endnotes 130

6 Language in Law


Misconceptions About Defendants 131 Misconceptions About Language in the Legal Context 132 Instructions on Interrogation 134 Judges’ Ideologies and Language 138 Conclusions 140

7 Language in Academia


Introduction 141 Is Julia Kristeva a Scientist? 142 Conclusions 150 Endnote 151

8 Language in Education


The Standard Variety 152 Ebonics 157 On New Words—Education Toward Nonessentialism 158 Conclusions 169 Endnotes 170



9 Viewing and Studying Language a Nonessentialist Way








Author Index


Subject Index



The main goal of this book is to show how some linguists and many laypersons alike misconceive language and what the consequences are of such misconceiving. In the former case, the consequences concern mainly the domain of linguistics; they include spinning the linguistic wheels (e.g., returning to the same questions again and again) or delayed research progress at best. In the latter case, the consequences include a number of disagreeable phenomena, such as intolerance, dogmatism, conceit, and conflict. In Part I of the book, I address a number of related misconceptions about language. In my opinion, misconceiving language (i.e., having an unrewarding view of the mechanics of how it works) can be traced to one’s basic philosophical beliefs, which we all have, regardless of whether we are aware of it. In Part I, I discuss the philosophical view called essentialism and its negative impact on the linguist’s work and the layperson’s life. I later propose and promote a nonessentialist view of language, which may be seen as liberating and rewarding in that it easily allows us to understand the many linguistic phenomena around us. Essentialism fails to let us understand these phenomena. The misconceptions discussed in Part I can be seen, largely, as a result of essentialist thinking. In other words, essentialism may be thought of as the common denominator of most of the faulty thinking about language. In chapter 1, some fundamental misconceptions and questions pertaining to meanings and definitions of words are discussed. The view promoted here is that meanings of words are fuzzy and imprecise, and precise definiix



tions of words about which there would be a total agreement cannot be reached. Many important consequences of this view are also mentioned. In chapter 2, I discuss the importance of philosophy for linguists and what kinds of consequences may follow from adopting one or another view of a word’s meaning. I refer to what seems to be a general lack of awareness of the significance of philosophy, especially on the part of the beginning linguist. This section of the book is intended to help the fledgling linguist (and some experienced colleagues, perhaps) avoid unnecessary problems in their work. It is my contention that such problems, which delay research progress, may be due to limited knowledge about or misconceptions of philosophical issues related to language. In chapter 3, I take up integrational linguistics as formulated by Roy Harris. I refer to the characteristics of integrational linguistics that are relevant to my discussion of nonessentialism, and I offer some critical remarks on a selection of Harris’ formulations. Chapter 4 takes up cognitive linguistics. Cognitive linguistics may be viewed as an exponent of the general view of language and linguistics that is promoted in this book. Some of the fundamental tenets of cognitive linguistics, as well as some issues that connect cognitive linguistics and nonessentialist thinking in general, are discussed. In chapter 4, the phenomenon of misunderstanding is elaborated on in some detail. The claim is made here that misunderstanding is an integral part of language and cannot be avoided. This claim springs from cognitive linguistics and nonessentialist philosophy. In Part II of the book, I enlarge the discussion in Part I into four domains standing outside linguistics—namely, politics, law, academia, and education. In this section of the volume, I attempt to show how the four domains are still largely dominated by essentialist thinking; I suggest what can be done to change that state of affairs. As for politics, in chapter 5, I discuss in some detail, among other things, the political correctness debate, the debates on abortion and euthanasia, as well as the recent Clinton–Lewinsky scandal, the Florida vote debacle, and the stem-cell research dilemma. As for law, in chapter 6, I draw on other scholars’ research (mainly Roger Shuy’s) and show how essentialism still permeates the legal profession. In chapter 7, I discuss an example of how academics argue about scientific methods and the value of colleagues’ work. This example illustrates my belief that academics generally (not only linguists, as shown in chap. 2) exhibit an essentialist frame of mind. As for education, in chapter 8, I address the question of the standard dialect and language variation; I also discuss Ebonics to show that the educational system and its current participants are deeply rooted in essentialism. I also make the claim that languages lack lexical resources (i.e., words) necessary to successfully handle the intricacies of



the universe. I discuss the proposal that the coining of new words should be encouraged to handle these intricacies. Building on the nonessentialist reasoning formulated in Part I, I suggest what can be done to avoid the trouble that essentialism seems to create in all the four domains. In Part III of the book (chap. 9), I summarize the differences between the essentialist and nonessentialist views and clarify the reasons for embracing the nonessentialist position for handling language in general and for doing sociolinguistics in particular. The core of chapter 9 is written in the form of a set of instructions, or pieces of advice, which I offer to the student of language who finds the ideas developed in this book attractive. These instructions are supposed to help the student avoid essentialist pitfalls and make practical decisions consonant with the nonessentialist agenda. This book encourages linguists to do applied linguistics. Applied linguistics is obviously not a new endeavor, and the excellent work of linguists such as Deborah Tannen, Roger Shuy, and Peter Trudgill, along with many others, demonstrates this fact. Nevertheless, I feel that the general awareness on the part of the lay language user of the significance of linguistics for everyday life is still very low. This same awareness may not be high on the part of many a beginning linguist. One of the main aims of this book then is to make young linguists realize how they can pursue linguistics in a highly rewarding way. Another aim is to help them realize how important linguistics is for handling everyday life. Still another aim is to encourage them to promulgate this message. This book has been written in the spirit of qualitative sociolinguistics (see Johnstone, 2000). Although many examples are given, no data are provided that would allow for statistical analyses or large-scale generalizations. I am tempted to say, however, that the questions discussed and the evidence provided are powerful enough to make the reader consider or reconsider his or her basic views about language. How significant the claims made in this book are in statistical terms remains to be disclosed. This book is intended primarily for graduate and PhD students of linguistics, especially those with some interest in applying linguistic knowledge to other fields such as politics, law, and education. I hope that this book can encourage young linguists to make philosophical decisions leading to what I see as a very profitable and promising stance within linguistics. It can help them ponder how useful and helpful a linguist can be for people outside of linguistics. The book may also be recommended to the practicing and experienced linguists as well as students and researchers in communication, sociology, psychology, and education. It may encourage some of them to approach and investigate language in ways that they had never considered before.



ACKNOWLEDGMENTS I would like to thank Stephen Warner for his editorial work on this book. I would also like to express my gratitude to the anonymous reviewers for all their valuable comments and suggestions for improvements on earlier versions.






Meaning and Definitions Misconceived

INTRODUCTION For many of the people for whom language study is a meaningful concept, meaning is the heart of such study. Any user of language needs to handle meaning all the time; language totally devoid of meaning (blurred as it often is) is hardly imaginable. This is perhaps why meaning has a central role in many linguistic studies and why it is at the core of the lay person’s intuitions about language. If we agree that meaning is central to language, in that language users convey meanings of various types all the time, it is no wonder that our (mis)conceptions about meaning will be of utmost significance in how we handle the universe around us. Language permeates all areas of our lives, so what we believe about language and meaning will also affect almost our entire life. People are normally unaware of what they believe about how linguistic meaning works. They do have beliefs about it, however. They act on these beliefs without being aware of what these beliefs are or how they act. If these beliefs are misconceptions, the consequences may be dire. As is shown in Part II of this book, these consequences may include, for instance, intolerance, conceitedness, and, if worse comes to worst, war. Word meanings are associated with definitions because definitions are usually understood to give the necessary information about the meanings of words. No wonder then that misconceptions about meaning and definitions usually constitute a package. In what follows, the two are also discussed as aspects of what might be thought of as basically one phenomenon. 3



Many lay people, many a beginning linguist, and even some experienced linguists tend to believe that words have, or should have, one meaning and that this meaning can be defined in some absolute sense; it is believed that this one ultimate meaning is something people agree or should agree about. Seen from this view, anyone who does not agree may be seen to be either a deliberate maverick causing problems or a person lacking knowledge. This position with respect to meaning is a misconception. More important, it is a very significant misconception with both theoretical and practical consequences. The common belief in the existence of one clearly definable word meaning goes back thousands of years. In the philosophical tradition, it can be traced back to Plato and Aristotle. This tradition is often discussed under the heading of essentialism. Essentialism refers to a set of beliefs about the world that have a direct reflection in assumptions about language. Extensive treatments of essentialism have been offered by, for example, Popper (1945, 1957, 1972), Hallet (1991), Bealer (1987), La Porte (1997), Escobar (1999), Teller (1975), and Fine (1994). Aristotle (1983) claimed that when we try to define a term (a word), we in fact try to define the name of the essence of the thing for which the term stands. Our definition is, in turn, the description of that essence. In other words, Aristotle thought (and millions of his followers still do) that words refer to concepts and that these concepts (which are somewhere beyond language users; they “float around” in the universe, as it were) have some essence (some essential characteristics) that allows us to distinguish one concept from another. This essence is what allows us to tell, for example, the concept of lamp from that of chair. Obviously we use different words in different languages to refer to these concepts. Aristotle claimed that the main goal of all inquiry should be compiling a kind of encyclopedia that would contain the intuitive definitions that we can give of all essences. From this perspective, an academic should collect the names of all essences of concepts (i.e., words) and provide the definitions for these names (words). According to Popper (1945), Aristotle considered that kind of encyclopedia-compiling activity the progress of knowledge. As such, an encyclopedia would be gradually expanding, as would our knowledge be gradually accumulating. In this view, research progress follows from our ability to define words properly. According to Aristotle, the procedure is very simple. Let us consider the sentence, “A lamb is a young sheep.” In it we point to a certain thing by saying “lamb” and then describe it as “a young sheep.” The term (word) lamb points to the essence, and a young sheep describes that essence. Aristotle believed that through this kind of exercise, we determine or explain the meaning of the term, lamb. In this way, the definition of lamb that we have— that is, a young sheep—answers two related questions. One is “What is it?”



(i.e., What is a lamb?) and the other question is “What does it mean?” (i.e., What does lamb mean?). The two questions may be perceived as asking “What is the essence of ‘lamb’?” (the former question) and as asking “What does ‘lamb’ mean?” (the latter question). For the last 2,000 years since Aristotle, essentialism has flourished. It is still rampant. It has taken a variety of forms, the most important of which is, as mentioned earlier, the common belief that words have or should have one meaning. Essentialism may surface as an explicit intellectual commitment, a salient practice, a covert realization, or a combination of the three.

EXPLICIT (OVERT) ESSENTIALISM Essentialism as an explicit intellectual commitment can be seen, for example, in religiously motivated discourse, in which the belief is conveyed that one must call things by their names, that good and evil, as well as many other concepts with a heavy axiological load (those with much value assigned to them), are clearly distinguishable and that it is possible to define them clearly. Morality is another concept believed, in this view, to be definable in an undebatable way. Explicit essentialism hardly allows compromises on the meaning of words. Words, it is believed, do not mean what they mean to the millions of their users in a particular language. Words really mean what they mean to a select few philosophers, thinkers, ideologists, or leaders. The tacit or explicit argument here is that the unsophisticated, ignorant outsider does not know the right, real, ultimate, and correct meaning of a particular word; this is known only to the enlightened insider. It is crucial to realize that the explicit essentialist commitment relates directly to, or rather follows from, the belief in the existence of objective truth. People who explicitly show their commitment to essentialism not only tend to believe in the existence of objective truth; they also believe that this objective truth can be known, if only to the select few. The practice of essentialism most often materializes when people pose explicit “What-is” questions. Following Popper (1972), the term what-is questions is used in this book to refer to questions such as “What is love?,” “What is science?,” “What is freedom?,” “What is justice?,” and so on. More important, the term what-is questions implies that these questions are meant to be answered in an unambiguous way. Sometimes such questions are asked without the speaker striving for one unambiguous answer. In such cases, the speaker may not believe in the existence of one such answer, but will still formulate the question the essentialist way; for example, “What is freedom?,” meaning “What do we call freedom?,” or “What do you call freedom?,” and so on. We may say that the question stylistically reminds us of essentialism, but philosophically it may not or is not such.



Telling an essentialist what-is question, which implies a belief in the existence of an unambiguous answer, from a what-is question that is such only stylistically is not an easy task. A misjudgment may lead, unfortunately, to taking the latter for the former. This, in turn, may lead to an unjust attribution of essentialism to a person. It must be stressed that one should not declare a what-is question to be essentialist on the basis of the question alone. One needs to have access to an extensive context in which the question was asked. Although we will never have 100% certainty as to whether a what-is question is philosophically essentialist, we can often say with a fair amount of certainty whether somebody reasons in an essentialist way (and thus takes what-is questions seriously) on the basis of a wider context. That wider context is what the person writes within the same paragraph, the same chapter, or the same book; what the person says on various occasions; how the person behaves; and so on. Some formulations are so obviously essentialist that not much context is necessary to draw conclusions about the person’s beliefs. For example, consider the sentence, “What is man? Philosophers have debated this question for millennia, reaching no very serious conclusion” (The Economist, 1996, p. 13). There is little doubt that the author of this passage strives after some ultimate definition of man or that she or he refers to others who do.

COVERT ESSENTIALISM Essentialism may also take a covert form that may be more difficult to detect. It is hidden. Hallet (1991) mentioned this kind of essentialism: . . . essentialist thinking has been far commoner than the formulation of explicit, general essentialist doctrines. . . . These doctrines are but surface waves; the sands beneath are drenched with essentialistic practice—with piecemeal essentialistic reasoning. More extensive, more diffuse, and less obvious, hence less exposed to effective criticism, this underlying essentialism may long survive the surface variety. (p. 28).

One of Hallet’s examples of hidden (covert) essentialism is the following: Observe this argument of Waismann’s: “What I have to say is simply this. Philosophic arguments are not deductive; therefore they are not rigorous; and therefore they don’t prove anything. Yet, they have force.” Here two explicit premises beget an explicit conclusion with the aid of two tacit premises. Whereas the spoken premises are not essentialistic, the unspoken ones are. The first “therefore” suggests that only deductive arguments are or can be “rigorous”; the second “therefore” suggests that only rigorous arguments do



or can “prove” anything. Such is the essence of rigor; such is the essence of proof. The essentialism looks multiple. (Hallet, 1991, p. 70)

Another common realization of essentialism that might be called covert essentialism is to be found in statements such as, “We do not know as yet what life is,” where “as yet” is a crucial part of the sentence. “As yet” implies that in principle we may know what life is; it is just that we may not have the sufficient knowledge yet, or that we may not have used the appropriate methods, or that we may not have worked long enough on the topic. Any of these reasons may have led, the reasoning is, to the fact that it is difficult now to arrive at the correct definition of “life.” One day, however, the reasoning continues, we will know what “life” really is, what it means.

CRITICISM OF ESSENTIALISM Essentialism is usually understood to stand for the belief in the existence of one correct meaning of words and in the power of definitions that give us ultimate information about the essence of the things to which the words defined refer. It is my contention that the essentialist position on meaning and definitions understood in this way is highly debatable; it may in fact be conceived of as a misconception. There are at least two reasons that that position seems to be wrong. One is that it is highly doubtful that intuitive definitions, which were extolled by Aristotle, let us gain access to the essence of things designated by the words defined (Popper, 1945). The other reason is that defining words always requires further defining of the words in the definition. We are thus led to an infinite regress of definitions (Popper, 1945, 1959). Consider the following example:

Step 1: What is fudge? Fudge is a soft crumbly or chewy sweet made from sugar, butter, and milk or cream. Step 2: What is a sweet? A sweet is a small shaped piece of confectionary made with sugar. Step 3: What is a piece? A piece is a portion of an object or of material, produced by cutting, tearing, or breaking the whole. Step 4: What is a portion? A portion is a part of a whole; an amount, section, or piece of something.



“What is exactly a part?,” “what is chewy?,” and “what is a section?” are some of the possible further what-is questions that might be asked as further steps and that would certainly lead to asking further such questions. This procedure leads to an infinite regress of definitions, wherein more and more what-is questions are asked about more and more defining terms. If one proceeds in this way, one does not achieve any ultimate precision or ultimate meaning of fudge. In fact, the meanings to be explained get even more complicated. The previous example showed how one gets into the infinite regress of definitions mode virtually any time ultimate definitions are sought. Popper (1945) showed very convincingly that we do not achieve much by proposing definitions. In fact we make things worse because we lose much by introducing an unnecessary amount of verbosity. With reference to definitions, it is useful to remember that anyone adhering to the Aristotelian method of definition reads definitions “from left to right.” If one reads definitions from left to right, one believes that in the sentence, “A fad is an intense and widely shared enthusiasm for something, especially one that is short-lived and without basis in the object’s qualities,” “an intense and widely shared enthusiasm for something, especially one that is short-lived and without basis in the object’s qualities” defines what fad is. One can easily see the danger of definitional infinite regress as the questions that immediately come to the fore are: “what is short-lived?,” “what is intense?,” “what is basis?,” and so on. These questions will, in turn, elicit answers that will bring forth an array of further questions, and the regression will continue. Popper teaches that as this procedure hardly leads anywhere, we should by all means discard this way of handling definitions (Popper, 1945). Definitions should be viewed from right to left—that is, with reference to the prior example, “A fad is an intense and widely shared enthusiasm for something, especially one that is short-lived and without basis in the object’s qualities,” we should understand it or preferably paraphrase it as “An intense and widely shared enthusiasm for something, especially one that is short-lived and without basis in the object’s qualities is called a fad.” This right-to-left view of definition gives us an entirely different perspective on the issue. Truly, such definitions do not explain anything, but essentialist, left-to-right definitions do not either. What the right-to-left view of definitions does is introduce one short label ( fad) for a much longer set of labels (“an intense and widely shared enthusiasm for something, especially one that is short-lived and without basis in the object’s qualities”). Such definitions cut a long story short. The earlier argument should not lead the reader to the conclusion that definitions are entirely useless. The reasoning presented before proposes to show only that definitions should not be taken too seriously; that is, they should never be seen as final and always as debatable. If they are taken seri-



ously (i.e., if they are viewed as potentially final), we get caught in the infinite regress of definitions. We do not gain much. Rather, we get overwhelmed by a sea of words, which mean less and less, the more of them we use in the subsequent definitions that were meant to make things more precise. They usually do not make things more precise. Let us stress again that left-to-right definitions, which are believed to be giving us the ultimate information about the essence of things that the defined words stand for, do not give us that information. They confuse us. They may leave us frustrated. Taking left-to-right definitions seriously leads us into a wild-goose chase. We seem to be getting closer and closer to the ultimate meaning (essence), but, in fact, we never get there. Many of us, striving at the ultimate meaning, use a lot of intellectual energy in vain, while others never stop looking for an ultimately good and satisfying definition. Many people believe that left-to-right definitions make things simpler and more precise. The fact seems to be that these definitions make things more complicated and less precise. The more steps in the defining process we go through, the more confused we become. That is, any definition entails the necessity of further definitions, which entails a continuously growing number of terms to be further defined. As opposed to left-to-right definitions, definitions approached the rightto-left way may be useful, and not only because they make a long story shorter. They may make things clearer in a discussion. This is, in fact, what happens in an operational or a working definition of a term. Consider the term learner, for example. The English word learner may mean very many different things. A toddler is a learner as she or he learns how to walk and speak. A 20-year-old student of economics is a learner as well. She or he is learning the principles of economics. Any person (independent of age) learning to drive is a learner as well. Obviously, dozens of different examples of the sort of people who might be referred to in English as learners could be given. When we are involved in the field of language teaching, for example, we have to limit, to some extent, the label learner to refer only to that kind of learner whom we are interested in in the context of language teaching. We might then adopt a more limited perspective of learner as “foreign language learner” and define him or her as “a person who is acquiring a language other than his/her native one.” But, it must be remembered that “a person who is acquiring a language other than his/her native one” is not meant to tell us what a “foreign language learner” really is (that would be reading our definition from the left to the right). All that happens here is that, instead of using 14 words (“a person who is acquiring a language other than his or her native one”), we use 3 words (“foreign language learner”). That is how we cut a long story short. Having cut the long story short, we obviously do not know exactly what a “foreign language learner” is (because we do not know, e.g., what exactly “a language” is). However, we



have managed to make our concept of foreign language learner a bit clearer in the sense of “more limited,” and we have to accept the amount of ambiguity that the words in the definition still exhibit. We might certainly further clarify our definition of foreign language learner, but no ultimate clarity (or precision) is attainable. Under any circumstances, “a person who is acquiring a language other than his or her native one” should be treated as a “working definition” of foreign language learner. This definition, imprecise as it obviously is, should nevertheless be helpful in studying the educational phenomenon of interest to us in a particular situation or in solving the educational problem that we are addressing. Essentialism is a widely held view of meaning and definitions and concerns many lay people and some academics. This also concerns linguistics, which indicates that some linguists also misconceive the way meaning and definitions, arguably the two most fundamental tenets of language, work. In chapter 2, we expand the notion of essentialism as it applies to the discipline of linguistics.



Word Meaning, Definitions, Linguists, and Philosophical Commitments

INTRODUCTION Many people are interested in philosophical questions, but they are also intimidated by philosophy as a discipline, as a body of knowledge reflected in philosophical treatises, in books, and in articles. Many people are also interested in religious questions, and no doubt religious questions translate into philosophical questions to a significant degree. Other people, however, avoid philosophy at all costs. One reason for this may be that philosophy is often viewed by the average person as an elitist discipline that is incomprehensible or difficult to understand. Another reason may be that people find philosophy simply irrelevant to leading an ordinary life. Unfortunately, many linguists also shun philosophy. There are probably several reasons for that. One is that they feel (like many students or the general public) that philosophy is not interesting to them. This attitude can be understood, but it should be discouraged. Another reason is that some linguists think that philosophy has nothing or very little to do with linguistics. This stance is highly inadequate. In fact, philosophy has a lot to do with linguistics; there is a subdiscipline called the philosophy of language, which may be considered an interdisciplinary enterprise. Philosophical questions often arise, sometimes unexpectedly, when linguistic questions are discussed. Consider, for example, a discussion about aspect and tense. If you are deeply involved in what you do, you will quickly find yourself engrossed in fundamental philosophical questions. You will soon reach the point of having to talk about time as a philosophical category. 11



Space might be the next topic to be discussed. You will soon realize that you should have a view on how time relates to space. It is interesting to watch discussions at linguistics conferences. The speaker and a discussant (or several discussants from the audience) often argue fiercely about a point. At a certain moment, however, their argument collapses. It either becomes totally incomprehensible to both the involved discussants and the audience, or the two parties involved get confused in the discussion and are unable to reach any conclusion whatsoever. There may be several reasons for such an outcome. One may be that the interlocutors differ philosophically. They differ as to the basic assumptions about the world, about what linguistics is all about, and so on. They differ, but they do not know it, or they do not know where they differ. In other words, they differ philosophically, and they are totally unaware of that fact, and they are not able to locate their differences with any accuracy. They may attempt to discuss these differences, but such attempts usually only complicate matters further and leave everybody with a sense of frustration and failure. Very likely, each feuding party leaves the stage with the feeling that the other party does not understand the point being made. Unlike what many linguists seem to believe, philosophy is very important for linguists. In fact it is very important not only for linguists, but also for any academic. Philosophy is important for linguists because once one has made certain philosophical assumptions, a number of conclusions naturally follow. These conclusions may materialize in the form of theoretical statements or in the form of very practical decisions that linguists make all the time in their daily work. For example, if you make the philosophical assumption that language is a human faculty only, several theoretical conclusions follow from that assumption. One is that no matter how relevant to your considerations some animal communication system may seem to be, you cannot conceive of that animal communication system as comparable to (treatable on a par with) human language. As far as practical activities are concerned, if you conceptualize language as behavior only (i.e., if you make the philosophical assumption that only language as behavior can be attributed some existence, and the rest is a chimera, or a product of human fantasy), you are obliged to treat as linguistic data (or facts) only those kinds of linguistic activities that you will label as behavior. Collecting as data what people know or what they think they know would be a practical activity inconsistent with your philosophical assumptions. Another point concerns ontological commitments—that is, our commitments as to whether something exists or whether something has an existence different from that of something else. Differences in ontological commitments, which we are often unaware of, may contribute significantly to our not being able to communicate successfully with others in the academic community. In a way, the prior example on human versus animal



communication may also be treated as an example of an ontological commitment. A more salient example is this: Let us assume that one claims that there are more concepts than words; that is, the claim is that we have concepts at our disposal (some sort of mental entities perhaps) to which individual words do not correspond. This is quite an obvious case of an ontological commitment. Perhaps a still more obvious case is to assume that the body exists and the mind does not, or the other way round. There may be all kinds of important theoretical and practical consequences of such commitments. If, as a person formulating an argument, you forget having made that commitment and diverge from it, you may get confused. Given that your listener or reader has missed your having made (explicitly or implicitly) that commitment, that listener or reader simultaneously misses all or some of the implications of it. We often seem to make commitments of this sort in an implicit way, which makes it hard for others to detect them or to follow in a discussion. The fact that we do not make such commitments explicitly all the time is not surprising. Most discussion time, or most of our conference talks time, or most of our writing space would be taken by making our commitments explicit. However, it is still my contention that, to a reasonable degree and to the extent possible under given practical circumstances, philosophical commitments should be made explicit. Some linguists do not realize the importance of the philosophical commitments that they, in fact, always make (as speakers, writers, or listeners and readers) regardless of whether they are aware of them. We may suspect that they often are not. Some support for making this claim comes from professional conferences where we can witness some linguists getting significantly confused when it comes to discussing fundamental philosophical issues. Discussions usually end when they should not. One can rarely hear a clear, “I think we disagree here philosophically,” which could be taken as a valuable, pleasant, and inoffensive way of concluding a heated debate. Linguists often leave the discussion forum thinking that the other party is not competent enough to understand the point that is being made. The fact seems to be, however, that linguists often do not understand each other because they make different philosophical assumptions and different philosophical commitments, and, in the worst case, no party in a discussion is aware of these assumptions and commitments. They work in different philosophical worlds, and thus they talk at cross-purposes. Making philosophical commitments is very important for a researcher. One such possible commitment is that to essentialism. Another one is to nonessentialism. In other words, we may consider being an essentialist or being a nonessentialist to be examples of philosophical commitments. These two commitments, examples only as they are, are instances of commitments with serious academic and nonacademic consequences. In a later section of this chapter, we discuss the consequences for the academic field



of linguistics. In Part II of the book, we discuss the consequences pertinent to other domains of life. Essentialism in Linguistics (Overt) Linguists, like other academics, differ in their attitude to essentialism. Some endorse it, others do not, and still others take an intermediate position. Although essentialism seems to be an unrewarding approach to language, there still remain many essentialists and semiessentialists in the linguistic community. It might be noted that 60 years ago, Popper (1945) claimed that those sciences that regard definitions the nonessentialist rightto-left way (e.g., physics) are the ones that move forward, whereas those that use the essentialist method do not seem to progress much. Popper’s statement seems to be valid still today. In other words, we might say that, within a particular discipline, those researchers who treat definitions the left-toright way do not achieve much. Their wheels spin around in the multitude of definitions. In contrast, those researchers who take definitions lightly— that is, those who take the right-to-left perspective—move forward. Their primary objective is not to define terms correctly (as is the case with the essentialist left-to-right definition believers), but to do something else—for instance, to solve some problem. In solving a problem, the definition of a term may be of some help, but it is never an important goal. Unfortunately, as Wardhaugh (1997) rightly pointed out: Linguists have found such concepts as sound, syllable, word, and sentence equally difficult to define (in contrast to lay usage, in which they are just assumed to be obvious and uncontroversial). In one sense, linguistics is all about trying to provide adequate definitions for words such as sound, syllable, word, and sentence. (pp. 21–22; italics added)

Examples of essentialism in linguistics are easy to find, for instance, in the subfield of sociolinguistics. Consider the following passages, which clearly express or refer to essentialist what-is questions: Ultimately, sociolinguistics hopes to go beyond comfortably simple theory concerning the nature of communicative competence in the conviction that only an adequate theory of human capacity to acquire and to use a repertoire of interlocking language varieties and their related behaviors will yield an adequate theory of what communicative competence in social man really is. (Fishman, 1970, p. 17; italics added)

From the nonessentialist view, the question, “What is really communicative competence in social man?,” is unanswerable, so it should not be asked at all. Communicative competence may only be treated as an analytical (meth-



odological) category that the investigator is free to define in different ways. In other words, to determine communicative competence is an unattainable goal; we can obviously use the concept in describing human communication, and we can view it as it relates to other concepts (such as speech community or linguistic repertoire). This, however, is significantly different from trying to define communicative competence in some absolute sense. The concept of diglossia has been applied, with varying degrees of conviction, to several types of speech community, and it is now taken for granted that the label should be part of any attempt at typological classification of sociolinguistic situations. This has led in many cases to confusion rather than clarification, and to much tortured debate about the precise meaning of diglossia and what really constitutes a diglossic situation. (Winford, 1985, p. 345; italics added)

From the nonessentialist point of view, we will never know the precise meaning of diglossia or what really constitutes a diglossic situation. In other words, a diglossic situation or diglossia will never be defined unambiguously. We can use these two terms as Ferguson (1959) originally did, or we can use them as others did or do (see e.g., Fasold, 1984), depending on which definition(s) suit our purposes best. We can use Ferguson’s (1959) definition: Diglossia is a relatively stable language situation in which, in addition to the primary dialects of the language (which may include a standard or regional standards), there is a very divergent, highly codified (often grammatically more complex) superimposed variety, the vehicle of a large and respected body of written literature, either of an earlier period or in another speech community, which is learned largely by formal education and is used for most written and formal spoken purposes but is not used by any sector of the community for ordinary conversation. (p. 16)

But we can also apply Fishman’s definition, as reported on by Fasold, “the distribution of more than one language variety to serve different communicational tasks in a society” (Fasold, 1984, p. 40). Neither Ferguson’s original definition nor any of the later ones (including Fishman’s) are appropriate or correct in any absolute sense. Defining a natural language is the main problem faced by sociolinguistics. On the one hand, linguists do have the impression that they know what linguistics is . . . yet on the other hand, the concept of natural language is for the most part a sociolinguistic concept. Indeed, if we were to ask what a so-called national language, a dialect or a patois is, linguistics would be unable to answer the question. (Greimas, 1987, pp. 193–194; italics added)



Greimas may be right in saying that defining a natural language is the main problem faced by sociolinguistics, but he should have said by essentialist sociolinguistics only. Nonessentialist sociolinguists are not interested in the serious pursuit of definitions, including that of “a natural language.” A nonessentialist sociolinguist will adopt a working definition of a natural language if such a definition is helpful for some study. Defining a natural language (like defining a dialect, variety, etc.), disconnected from goals standing outside this definition, is never a nonessentialist’s academic goal. According to Zuengler (1985), “Tanzania’s linguistic composition is complex, for there are today over one hundred vernacular language groups. This is, however, only a rough estimate, since little analysis has been done as to what objectively constitutes a dialect, and what a language” (p. 242; italics added). From the nonessentialist view, distinguishing objectively between a dialect and a language is not possible. This is because neither can be defined objectively. Only if they could be defined objectively (which the essentialist seems to believe is possible) could we make a clear distinction between the two. The nonessentialist will be satisfied with a variety of definitions of both terms. One possible example of an acceptable definition is “A language is a dialect with an army and a navy” (Weinreich; quoted in Romaine, 2000, p. 13). Of course, other definitions would also be acceptable. “ ‘What is German?’ and ‘Who speaks German?’ are excellent potential topics for German language sociolinguistics (cf. Steger, 1980; p. 348)” (Löffler, 1985, p. 59; italics added). “What is German?” is a typical what-is question that, according to the nonessentialist, cannot be answered in only one way. If there is an agreement on this point, “What is German” cannot be treated as an excellent potential topic for German language sociolinguistics. The question, “Who speaks German?,” in turn, can be easily paraphrased as “what is a speaker of German?,” which is another unanswerable what-is question. Neither question has excellent research potential. From the nonessentialist view, anyone who is concerned with such questions is involved in research that can only yield unrewarding results. “We have seen that political discourse analysis first of all should be able to define its proper object of study: What exactly is ‘political discourse’?” (van Dijk, 1997, p. 12; italics added). Here we deal with another example of an explicitly asked what-is question—“What exactly is political discourse?” The word exactly may reinforce the assumption that the question is meant as a philosophically (and not stylistically) essentialist what-is question. By viewing this question as misguided and unnecessary, the nonessentialist will also find the first part of the passage off the mark. The nonessentialist’s reasoning will be that we do not need to know the exact meaning of any notion (because we never can know it) before carrying out some thorough analysis of a phenomenon. Although the nonessentialist has nothing against providing working definitions of terms (such as political discourse) before a study



is carried out, the requirement “define your terms” before doing anything else, especially when understood as “define your terms exactly,” is not part of the nonessentialist agenda. In some cases, the nonessentialist may encourage the researcher to give working definitions of terms before a study using these terms is carried out, but such definitions are never meant to provide exact, precise, or ultimate essentialist meanings of these terms. To the universalist question, What is language?, the current best answer is Chomsky’s: “Language is a set of very specific universal principles which are intrinsic properties of the human mind and part of our species’ genetic endowment (Chomsky, 1986:15ff)” (Downes, 1998, p. 17; italics added). It should be easy now for the reader to guess that the nonessentialist will not ask the what-is question, “What is language,” at all. Nor will she or he treat it as a universalist question. According to Wodak (2001), The ways in which some of CDA research is directly and indirectly related to the research produced in the tradition of critical theory are particularly evident when one considers central concepts with which the various areas work, and social phenomena on which they focus. Examples of these are pertinent in their approaches to questions such as: “What constitutes knowledge? . . .” (p. 11; italics added)

The question that Wodak invokes is a paraphrase of the straightforward “What is knowledge?,” that, being a fairly clear case of an essentialist what-is question should be absolutely rejected by the nonessentialist. The following extracts may be treated as the epitome of essentialist writing or of writing referring to essentialism: “We will be forced to re-examine some basic questions: What is a language?, What is a linguistic fact?” (Labov, 1975, p. 6; italics added) Few problems continue to generate so much endeavor and so much conflict as the problem of style. Even conferences are called to attempt to answer the questions: “What is style?,” how can we study it? . . . Marozean began by admitting that the question of style is as open today as it was two millennia ago. Some progress was made in the nineteenth century when scholars began “traiter de style comme une object de science,” but consequently the question of what style is became more pressing . . . Regardless of the amount of effort, money, and time expended on the problem [of establishing what style is] results curiously fail to materialize. (Gray, 1969, pp. 7–8; italics added)

For a nonessentialist, these are straightforward what-is questions (“What is a language?,” “What is a linguistic fact?,” “What is style?”). The nonessentialist will simply not ask them as she or he believes that they produce no specific answer. To address these questions as goals or elements of im-



portant research requiring one single answer is for the nonessentialist simply an unrewarding way of doing linguistic research. In all the cases quoted earlier, one can assume with a fair amount of certainty that the authors believe or refer to others who believe that the terms of interest to them are in principle definable in some ultimate sense. If these terms have not been defined successfully yet, it is believed, they should and, in principle, can be defined. Hudson (1996) rightly claimed that much sociolinguistics has been essentialist: “What would help the sociolinguist most in his work would be if he could identify some kind of natural speech community with reference to which he could make all his generalizations, and much of sociolinguistics has in fact been carried out on the assumption that this is possible” (p. 28; italics added). In light of our earlier criticism of the essentialist view of definitions, Hudson’s statement can be understood as saying that, unfortunately, much of sociolinguistics to date, rooted in the essentialist belief that some kind of natural speech community can be found, has been misguided. A nonessentialist alternative does not posit the existence of any natural speech community. In the nonessentialist view, no natural speech community exists. In this view, the speech community is an analytical/methodological category, a term, if you like, which can be defined in different ways (as it, in fact, is!) depending on the researcher’s theoretical and/or practical goals. Referring to what might be taken to be a clear token of essentialism, Shuy (1993) rightly claimed that many people believe “Meaning is found primarily in individual words. All people in a conversation understand the same things by their words” (p. 8; see chap. 3 and chap. 4 for more information on this point). In the nonessentialist view, meaning is not found in individual words. From this perspective, meaning is understood to be assigned to words by people. This distinction is of paramount importance in understanding the difference between essentialist and nonessentialist views of language. R. Lakoff (2000) endorsed the view that essentialism is rampant: “The absolutists insist that these (sex, lie, apologize), like all words, have single, decontextualized meanings: Everyone knows what each of these words means, everyone knows what constitutes an instance of each of their referents. Language is fixed. Meaning is certain” (p. 270). For a nonessentialist, words do not have single, decontextualized meanings, and people may vary (more or less significantly) in their view of what individual words mean. For a nonessentialist, language is not fixed and meaning is not certain. The situation is just the opposite; language is open and meaning is largely various, flexible, and quite often fuzzy. More important, the opinions expressed by Shuy and Lakoff are true not only of many lay people, but also of some linguists.



A belief in the existence of ultimately correct definitions surfaces also in various academic discussions, and it materializes as a complaint. At a recent meeting I attended where PhD students presented their projects, many students, independent of how advanced they were in their projects, complained about the sort of definitional problems that they face with respect to some of the crucial concepts that they want to use in their studies. The complaints took what might be understood as a typical essentialist concern, for instance: Topic is the most frequently used, unexplained term in the analysis of discourse. Episode, which is related to if not identical to the notion of topic, is described as an imprecise term. Thus, there appear to be several terms in use that are closely related but not easy to define in any precise manner; Modality is, as opposed to mood, a semantic rather than a grammatical category. Consequently, it is not easy to find a precise definition of it. Many researchers tend to complain in this way. A general complaining atmosphere can be encountered in many branches of linguistics. Researchers often claim that they are not able to find precisely defined terms or that they themselves find it extremely difficult to define terms precisely. At the same time, however, these researchers give one the impression that only they (and not others as well) find themselves in this predicament. They also seem to indicate that it is the issue they are investigating (their research project) that makes the definitional task so difficult. They blame other researchers in the field for not having succeeded in defining precisely the terms in question. The researchers who complain in ways depicted earlier are right in the sense that it is indeed true that definitions are imprecise, but they are wrong in the sense that these definitions can ever be made ultimately precise. Consider the following definition of communication disruptions: “We define a communication disruption as occurring when mutual comprehension is impaired by one of the speakers misunderstanding the other or when the learner is manifestly in trouble in putting across what he/she wants to say” (Haastrup & Phillipson, 1983, p. 143). This definition of communication disruptions is obviously imprecise. It does not give us any precise information. It does not seem to make communication disruptions much easier to understand. In this sense, the researchers referring to or using Haastrup and Phillipson’s definition will be disappointed. The definition may have made things a bit clearer, but definitely not more precise. From the nonessentialist view, if these researchers complain, they seem to be wrong; or, we might say, their complaints are unjustified, because that predicament is not endemic to their own field. In any field that one considers, the situation is the same: There are imprecise terms everywhere. Mühlhäusler’s concern about pidgin and creole studies is a good example: “Pidgins studies have suffered for a long time from terminological and



definitional problems, as has been discussed often by Mühlhäusler (1974, pp. 11–25) and Samarin (1975)” (Mühlhäusler, 1986, p. 3). Having listed a number of definitions of the term pidgin, Mühlhäusler discussed in some detail a few of the problems that he saw with these definitions, rejected all the definitions discussed, and ultimately proposed his own. Because his definition is by no means precise,1 we may surmise that it has most likely been rejected by now by another scholar of pidgin languages or that it will be in the future. It is difficult to say what researchers see as the reason, or reasons, for these terms not having been successfully defined. Some may believe that the definitions available are not good for their own research purposes. Unfortunately, these researchers hardly ever say so explicitly (more important, if they did, their dissatisfaction with the existing definitions would be by all means justified, and they then should not be suspected of essentialist reasoning). Some may believe that after they have done more research, they or someone else will one day finally arrive at this one correct definition for which they have been striving. These researchers appear not to be discouraged by the fact that no successful definition of any term has been arrived at; some linguists continue to argue about and/or be dissatisfied with the terms that they use in their work. Surprisingly, these linguists do not get discouraged by the two millennia of continuous search (cf. the style and the man examples from before) for the right definition.

Essentialism in Linguistics (Covert) It is hoped that the cases quoted earlier show clearly how essentialists overtly ask what-is questions or variants thereof. As mentioned in chapter 1, some essentialism is covert and thus more difficult to detect. Consider the following two examples: Labov’s (1972a) definition of speech community: “The speech community is not defined by any marked agreement in the use of language elements so much as by participation in a set of shared norms . . .”(Wardhaugh, 1997, p. 118). Hymes (1974) disagrees with both Chomsky’s and Bloomfield’s definitions of a speech community. He claims that these simply reduce the notion of speech community to that of a language and, in effect, throw out speech community as a worthwhile concept. He points out that it is impossible to equate language and speech community when we lack a clear understanding of the nature of language (Wardhaugh, 1997, p. 120).



In both of these quotations, no straightforward what-is questions are asked (they would be, “What is a speech community?” and “What is language?”), but the authors of the reasoning presented in these passages by Wardhaugh are very likely essentialists anyway. Labov’s formulation indicates that he believes that one definition is ultimately correct. He says that defining the speech community by any marked agreement in the use of language elements is inadequate; the speech community should be defined by participation in a set of shared norms. In other words, only one definition is viewed as correct; others (including the one referred to) seem to be wrong. A belief in the correctness of only one definition of a term may be taken as an example of essentialism. Thus, for a nonessentialist, Labov’s formulation is not appropriate because it excludes the possibility of defining the speech community as “any marked agreement in the use of language elements.” According to the nonessentialist, this definition, as well as others, might in fact be useful to some researcher for some purposes. The nonessentialist will not argue against any definition, or for any definition, outside of a larger context (e.g., a problem to be solved), an analysis of which might require defining the concept of speech community. In the other example, the final “when we lack a clear understanding of the nature of language” should be of concern to us. The expression “the nature of language” may imply essentialism. “The nature of language” may easily be interpreted as “the essence of language,” and then the passage would clearly express essentialism. But, surely, giving the author the benefit of the doubt, that interpretation is not the only one possible. I may be wrong. If I insisted that “nature” must definitely be interpreted here as “essence,” I would be reasoning the essentialist way myself. In any event, the nonessentialist is alert to formulations including “the nature of. . . .” Especially when there is additional evidence that “the nature of . . . “ is to be interpreted as “the essence of . . . ,” the nonessentialist will reject it as expressing an unrewarding essentialist approach to the study of language. One other form of covert essentialism emerges when, for instance, the question is asked whether two varieties are considered to belong to one language or more. Before discussing this issue in some detail, let me stress that essentialists tend to argue about whether a particular contact situation involves separate dialects or separate languages. They tend to argue about what makes a language and what makes a dialect (see e.g., Zuengler, 1985). Such arguments do not seem to be very rewarding. This is because they involve the essentialist position that requires asking what-is-questions: “What is a dialect?” and “What is a language?” These questions are unanswerable in one way, and any serious attempt to define these terms precisely will lead us to an infinite regression of definitions.



One of the tokens of essentialism that surfaces in contexts where two or more languages or varieties are in contact is the tendency to broach the question of whether a particular variety belongs to language X or language Y. Jahr (1989) referred to the linguistic situation in Norway in the late 19th century when the struggle between New Norse and Dano-Norwegian began. The specific question was whether Dano-Norwegian was a variety of Danish or a variety of Norwegian. This issue became a political one; language planners considered Dano-Norwegian to be virtually a variety of Norwegian, whereas the political opposition considered it to be a special variety of Danish. Because there was certainly no one answer to this question (which, one might suspect, the competing parties would not accept), conflicts ensued. A similar dispute has recently been reported in the press. The Economist writes: Spain’s foreign minister, Miquel Ángel Moratinos, may have meant well when he asked the European Union to make four regional tongues—Catalan, Basque, Valencian and Galician—“official languages.” Yet Catalan nationalists went furious. They accused Mr Moratinos of making an “artificial” distinction between Catalan and Valencian, which they deem a mere dialect of Catalan. (The Economist, 2004, p. 32)

Interestingly, in response to the prior article, a reader wrote: Your article about the recognition of Catalan as an official language in Europe and the polemic about Valencian misses two essential factors—linguistics and history. . . . Experts unanimously recognize Valencian as a variety of Catalan (as many Valencians call their language). Traditionally, Castilian or Spanish nationalism has used the artificial division of Catalan and Valencian as a way to divide and rule people who until the 18th century had their own kingdom. . . . (The Economist, 2004, p. 16)

Applying the definition of essentialism used in this book, the reader’s response is a reflection of an essentialist frame of mind. The reader states categorically that Valencian and Catalan cannot be viewed as two different languages. Apparently the evidence for this has been provided by a unanimous verdict from the linguists. According to this reader, there is a natural relationship between Valencian and Catalan (he declares the former to be a variety of the latter), and no artificially introduced relationship between Valencian and Catalan (the former and the latter being different languages) should be considered. This kind of stance translates into the essentialist claim that the natural relationship is the real one; that is, according to this view, Valencian can be seen only as a variety of Catalan. By this essentialist view, calling Valencian a language, on a par with Catalan, is wrong in an absolute sense.



From the nonessentialist view, however, the reader’s reaction to the original article is incorrect. According to this view, there are no absolute criteria by which we can declare Valencian to be a variety/dialect of Catalan or a separate language. Both options can be argued, and some final solution should be arrived at as a result of a social agreement between parties participating in the discussion and not as a result of some authoritative statement by whomever might be willing to issue such statement, including linguists. Consider yet another example: “Whether Niederländish is a German dialect, or whether, in general, it is some part of German . . . or whether German is at all spoken in Elsass has not been clarified yet” (Löffler, 1985, p. 60; my translation). An observation that needs to be made here is that authors such as Löffler appear puzzled by the unexplained status of Niederländish. From a nonessentialist perspective, however, this status cannot be finally and unequivocally determined. The words not yet should be noted. They probably mean that, in principle, the problem of assigning a variety to a particular language is seen as soluble. The reasoning involved here seems to be that it is simply a difficult problem that has not yet been solved and that, given a sufficient amount of time, it is likely to be solved in the future. A nonessentialist will not accept such reasoning because it invokes conceptual essentialism; that is, in the three cases mentioned earlier, the belief is that concepts such as the Danish language, the Norwegian language, the Dutch language, the Catalan language, the Valencian language, and the German language are ultimately and unquestionably definable. The stance expressed in Löffler (1985), that is, the not yet stance, is to be found in the literature quite frequently. Mühlhäusler (1986) noted: “As regards the two principal questions addressed in this chapter, the definition of pidgins and creoles, and their identification as separate languages, no unequivocal answers can be given, as is only to be expected at this phase of inquiry” (p. 20). I believe that the final section of the quotation, “as is only to be expected at this phase of inquiry,” could be, in all fairness, taken to mean that unequivocal answers cannot be given as yet, but are, in principle, possible. In the nonessentialist view, unequivocal answers to definitional questions are not unavailable as yet or at this stage; they are always unavailable. In this view, such expressions as “as yet” or “at this stage” are misleading in the sense that they give the reader/researcher the expectation or illusion that some day a particular term of interest to him or her will be absolutely defined. However, no such thing is possible. Covert essentialism can be seen as the sort of essentialism that is harder to detect or, in other words, the sort of essentialism that is not expressed in a straightforward way in terms of what-is questions. Covert essentialism may be considered to be present in almost all those cases where an author seems to believe that an array of definitions will let the listener or the reader ulti-



mately understand what the author means by a given term. It needs to be reiterated that such an array of definitions usually leads to an infinite regression of definitions, but the author’s underlying belief is that at a certain point she or he will reach the stage at which the explanation/definition will be (or should be at least) utterly precise and thus completely clear to everyone; that is, at that stage, the essence of the concept will be grasped. Let us consider an example pertaining to the Speech Act analysis. Oleksy (1984) was concerned in his discussion with the question of Speech Act (SA) types. He tried to define Speech Act type: “The definition of the type of SA that the speaker has performed will be referred to as the assignment of the Speech Act Value (SAV) to the speaker’s utterance” (p. 352). The author feels immediately, however, that he needs to give more information so that “the assignment of Speech Act Value” could be understood. So, he continued: Suggested below are three possible steps leading to the assignment of SAV to S’s utterance: i. Illocutionary force (IF) of S’s utterance must be defined. ii. Felicity conditions (FC) for SA which S has performed must be specified. iii. Sociocultural context (SC) in which SA typically occurs must be specified. (p. 352)

At this stage, Oleksy (1984) felt an urge to explain the term illocutionary force (we are at Stage 3 in the array of definitions at this point). So, he continued, “IF expresses the communicative property of the utterance which was used by S in the given context. IF can be assigned to S’s utterance in the following circumstances: . . .” (p. 352). The author then produced three paragraphs that describe these circumstances. They include terms such as embedded performatives and formulations that might need further explanations. Further on, the author went back to Felicity Conditions and defined them: Felicity conditions (FC) are conditions which guarantee that an SA performed by S in communicative context (CC) is effective, that is, it expresses S’s intention to the addressee A who, in turn decodes S’s intention. Therefore, the essential constituents which are present in defining FC are the following: a. S’s assumptions about himself and about A b. S’s assumptions about the empirical setting in which the performance of SA takes place. (p. 353)

These definitions and explanations are very hard to follow. However, the author probably thought that the more definitions and explanations he



gave, the clearer things would be. The result may be just the opposite. The more definitions he provided, the more confusion there was. In fact he added still more definitions. For example: “FCs are what the speaker assumes about the addressee, about the content of what is being communicated, and about the empirical setting in which an SA occurs” (p. 353), and “Sociocultural context (SC) . . . (is) . . . those conditions which pertain to the felicitous use of speech act relative to social norms acceptable in the culture to which the participants speaking a language L belong” (p. 354). Many explanations (which may be considered as definitions) can be detected in the text of the pertinent article. For instance, the author assumed that the reader may not understand the definition of SC, so he added, “It is suggested that SC can be considered in terms of sociocultural restrictions imposed on the participants by social norms accepted in the given speech community and culture at the given time” (pp. 354–355), and “Closely connected with SC seems to be the role relationship that exists between the speaker and the addressee” (Oleksy, 1984, p. 355). The last two quotations may obviously be understood as further attempts on the part of the author to define the concept of sociocultural context even more precisely. There are many more simple and elaborate definitions in the article, and all of these definitions are meant to explain what Speech Acts really are. The author begins with the Speech Act type. In trying to define it, he gets deeper and deeper into an infinite regression of definitions. He first defined the Speech Act type in terms of the Speech Act value. He then defined the latter in terms of Illocutionary Force, Felicity Conditions, and Sociocultural Context. He subsequently gave detailed definitions of the last three and some further definitions of the terms occurring in the definitions of the three. It should be quite clear to the reader that the author believed that the more definitions he gave the more precise the whole discussion would be. The longer the list of definitions, the author appeared to believe, the closer he got to the essence of the Speech Act type. Nowhere in his discussion did the author say explicitly that his goal was to determine what Speech Act type is. In other words, he did not ask an explicit what-is question. He seemed to be exercising essentialism in a covert manner. It is not difficult to believe that the author would not mind providing still more definitions. From the point of view of the nonessentialists, the many definitions that Oleksy provided do not help us understand Speech Act type at all. On the contrary, the greater the number of definitions that are provided, the more complicated the concept in question appears to be. From the nonessentialist view, Oleksy should have stopped defining Speech Act type no later than the second, possibly third, stage in the attempt to provide a working definition, imprecise as this definition would actually be. More important, stopping at the second, or possibly third, stage in the definitional process



would have been much easier had Oleksy used less abstract words in his definition(s) than the ones he tried to define. Defining Speech Act type in terms of an “assignment of the Speech Act value,” as Oleksy did, does not help much. Likewise, Oleksy’s definition of Illocutionary Force in terms of “communicative property of the utterance” does not help either. As with Illocutionary Force and communicative property of the utterance, the concepts Speech Act type and Speech Act value are more or less at the same level of abstraction. Defining one in terms of the other does not seem to contribute to our understanding of what it is that we are actually talking about. To reiterate, a nonessentialist spends little time and effort on the defining process and, in the defining formula, tends to use terms that are significantly less abstract than the terms being defined. Only in this way can we significantly shorten the defining process, avoid getting into an infinite regression of definitions, and make our working definitions meaningful, tangible, and useful. To recapitulate the previous section, with regard to the nonessentialist view, it is unrewarding to give multiple layers of definitions of terms: definitions of definitions. It is very doubtful that we achieve anything significant by such efforts. In fact whomever these definitions are addressed to may get more confused than she or he was before they were offered. Very likely, in one of the definitions provided, we return to the term that we tried to define originally, and we thus have come full circle. Anyone who defines and explains terms as depicted by the prior Speech Act example is, by the definition of essentialism I have developed here, an essentialist. This form of essentialism may be called covert because the author may not realize at all or may not want to reveal that she or he, in fact, indirectly asks what-is questions. These questions assume a variety of different guises, but the fact that they get asked in whatever form quite clearly exhibits the author’s essentialist bias. Another form of covert essentialism is reflected in those studies that are primarily taxonomic. What is meant by primarily taxonomic is the kind of study that, either implicitly or explicitly, states that taxonomies should be the goal or an important goal of linguistic research (note that I have just given a working definition of primarily taxonomic and thus tried to make the notion a little clearer, but I have not managed and, obviously, I did not mean to make it absolutely precise). Consider an example from the field of language acquisition. In Tarone, Cohen, and Dumas (1983), the authors attempted to identify communication strategies in the process of second language acquisition. They arrive at the following list of communication strategies: (a) transfer from native language, (b) overgeneralization, (c) prefabricated pattern, (d) overelaboration, (e) epenthesis, and (f) avoidance. Interestingly, avoid-



ance is further subclassified into: (i) topic avoidance, (ii) semantic avoidance, (iii) appeal to authority, (iv) paraphrase, (v) message abandonment, and (vi) language switch. Subsequently, (i) and (iii) are further subcategorized. Topic avoidance has two subcategories: (a) change topic, and (b) no verbal response. Appeal to authority has three subcategories: (a) ask for form, (b) ask if correct, and (c) look it up. This is a proposal for an identification of six major categories of communication strategies and 11 subcategories at two different levels of classification. In the article, no significant discussion is offered as to what such a taxonomy may be proposed for. There is no research goal set outside the taxonomy. This is an extremely important point. As we might expect, the authors of the article express their concern over the fact that the categories they propose are not mutually exclusive. The tenor of the discussion is, however, that if more work is done, the taxonomy will be amended, and there will be no overlap and no fuzzy cases. The authors said: It is altogether likely that this framework is nowhere near all-inclusive of communication strategies. We welcome our readers to suggest further categories or modification of existing ones. As it is, we realize that the categories described in this paper are not always mutually exclusive one from the other. (Tarone et al., 1983, p. 12)

The prior passage can be read as an expression of the belief that when more work is done, we will be able to know exactly how many categories and subcategories of communication strategies there really are. It seems that the authors are looking for the essence of the many types of communication strategies; however, they never say that explicitly. Essentialists seem to believe that only one taxonomy/classification is the ultimately correct one, and it is usually that one taxonomy that they themselves have arrived at or that they hope to arrive at. Given that belief, one should not be surprised to observe that mere classifications are a frequent aim in various kinds of research. As Chilton and Schäffner (2002) stressed, “In text-linguistic research one aim has been the classification of texts” (p. 19; italics original). In my view, the belief that one taxonomy may be correct in an absolute sense, irrespective of some goal external to that taxonomy, can be perceived as a form of covert essentialism. This is because the claim that one such taxonomy is ultimately and objectively right implies the view that scientific observation or description may be theory-free—in a way pure and catching the essence of things. Only, or mainly, if theory-free observation and description are believed to be attainable does it appear gratifying to believe that some one taxonomy is ultimately correct, right, or true in the sense that it describes things (or classifies them) the way they really are.



If one believes, after Popper (1959) and others, that all observation and description are theory-impregnated or theory-laden, setting taxonomies as the final aim, or even as an important research aim, does not appear to be an especially profitable position (for a relevant general discussion of taxonomies/classifications supporting the reasoning presented earlier, see Kitcher, 2001). This reasoning should not be taken to mean that all taxonomies are useless. It should be kept in mind that there is a significant difference between, on one hand, proposing taxonomies so they can be used for something that stands beyond those taxonomies, and, on the other hand, proposing taxonomies as final classifications and as an aim in itself. Taxonomies that are meant to be used for some external purpose may be used, for example, for solving a problem—a practical or theoretical problem. Such taxonomies are clearly ancillary in the sense that they help us bring order to the phenomenon of interest to us. The final goal in such studies is placed, however, beyond the taxonomy. There is no argument here about whether the taxonomy is utterly correct in some absolute sense. Treating taxonomies as an ultimate research goal should be discouraged; covert essentialists who are involved in taxonomizing activities should be encouraged to locate a goal external to the taxonomy (ies) with which they are concerned. From the nonessentialist view, this way of approaching taxonomies will make their academic endeavor much more fruitful and rewarding. Let us consider the following example: If one wishes to distinguish between Londoners and non-Londoners (which should be taken in this example as two categories in our taxonomy), she or he will achieve little by trying to define each category and then trying to assign individual people to one of the two categories: Is a Londoner someone who was born in London? Where is the borderline of London? Should the address count? Is a Londoner someone who lives in London? What about people who were born in London but left the city shortly after birth; are they Londoners? Are people who were born outside of London but have lived in London for a considerable time Londoners? How long must these people have lived in London to be called Londoners? And so on and so forth. A nonessentialist would not be interested in distinguishing in some absolute sense between Londoners and non-Londoners (and semi-Londoners, perhaps, to give an example of a three-category taxonomy), believing that such a distinction based on some sort of objective criterion is not possible. A nonessentialist will need a reason (standing outside the taxonomy) to bother with the taxonomy in the first place. For example, if the city of London is planning a new transportation system (which should be treated as an external reason to bother with the taxonomy at all), defining the categories in the taxonomy (and thus distinguishing between the categories in the taxonomy) becomes quite easy. Given the goal in question, it would be sensi-



ble to define a Londoner as, for example, a person who lives in London now and who is planning to stay in the city for the foreseeable future. Accordingly, a non-Londoner would be defined as a person not meeting these two criteria. The taxonomy of Londoner/non-Londoner is then not final or absolute; it is simply useful for practical reasons that exist outside of this taxonomy. Related to the prior discussion of taxonomies is the question of how divisions into fields and subfields are to be treated. Let us return to linguistics. Some linguists are more strict about such divisions than others. We face no problems when we deal with clear cases. If a linguist studies French prepositions, he or she fairly obviously fits in what we now call a French department; if someone studies Old English, then his or her place is in the English department (although we can obviously think of other departments where these two scholars could be working; e.g., in the linguistics department). Problems arise, however, when we think of nontypical cases where people are interested in two, three, or more of what we normally think of as separate disciplines—that is, people whose interests and research go in different directions. Let us consider, for example, the biology of language. Does a researcher interested in the biology of language belong to one of the language departments or to one of the biology departments, or does that researcher perhaps belong to what should be a new department (a new category in our taxonomy of language fields)—the department of the biology of language? Or should we open interdisciplinary departments that would hire people with a variety of interests? Some people with an essentialist frame of mind are upset when they have to deal with researchers with borderline interests—that is, with interests that are not typical for a given discipline or subdiscipline without at the same time going clearly into some other discipline or subdiscipline. Difficulties arise when attempting to classify (place in the taxonomy) such researchers. To take another example, it may be difficult to classify, in terms of the existing departments, a researcher who is interested in the kinesic system of the German language/culture. One possibility is that a new kinesic section of a German department be created. It may not be rewarding, however, to open such a new kinesic section (as an administrative unit) in the German department; we might one day reach a stage at which multiplying such departments and subdepartments would be unmanageable. There could perhaps evolve as many departments as there are members of the staff. Even if someone agreed to open such a subdepartment for the German colleague, some other colleagues might immediately protest and claim that kinesics is not a part of linguistics. They might further claim that there is no place for our colleague with the kinesic interests in any language department. At this juncture, the reader should not be tempted to solve the problem by asking the question of whether kinesics is part of linguistics. If that ques-



tion were asked, it might be thought, on a positive answer, that the kinesics scholar would be included in the German department; receiving a negative answer, she or he would not. The way the question is asked (“Is kinesic behavior part of linguistic behavior?”), however, would indicate that one believes that one’s question can receive one ultimate definite answer. It would in fact be a what-is-question asked implicitly. Explicitly asked, there would have been two questions: What is kinesics? and What is linguistic behavior? In fact both these questions were asked implicitly. Namely, the question of whether kinesic behavior is a part of linguistic behavior can only be answered if we first know what linguistic behavior is. In other words, the original question forces you to ask the what-is-question: What is linguistic behavior? This, in turn, leads to one of the following options: Either linguistic behavior includes kinesic behavior or linguistic behavior excludes kinesic behavior. The question, “Is kinesic behavior a part of linguistic behavior?,” is answerable only if one believes that the question, “What is linguistic behavior?,” is answerable. The latter is clearly a what-is question, and, as argued earlier, it is not an ultimately answerable question. Let us paraphrase this reasoning. To be able to say whether kinesic behavior is a part of linguistic behavior, you would need to define linguistic behavior. You would need to state, as part of such a definition, whether kinesic behavior is included in language, in linguistic behavior, or neither. You can decide whether kinesics stands inside or outside language only through defining language, whichever way language is defined. So, for placing kinesics within or outside of linguistics, what matters is the definition of linguistics. Expressing an essentialist frame of mind with reference to field and subfield classification (taxonomy) may be seen as practically useful, but intellectually stifling and unrewarding. It may even be seen as harmful. Again, consider the university and the language departments. There are faculties, departments, or institutes at most universities. There are departments of French, English, German, linguistics, literature, comparative literature, phonetics, and tens of other departments, just to think of what traditionally have been called the philological departments. These departments employ researchers whose interests focus around, for example, German, literature, or phonetics. It is good to be able to tell the students what our interests are (more or less) and what our focal areas of research and teaching are. Problems arise, however, when these labels (i.e., names of departments) are taken very seriously. What I mean by this is that, for instance, in the case of the linguistics department, it is required that only linguistics be taught. The reader will certainly remember from our earlier discussion that words cannot be defined precisely, and that we can argue endlessly about the definitions of words. The word linguistics is no exception. There is no one, agreed-on definition of linguistics. On the contrary, there are many compet-



ing definitions in the literature, and linguists tend to disagree about what constitutes or does not constitute linguistics and related disciplines (see e.g., Levinson, 1983). Whether kinesics is part of linguistics could be taken as an example of such arguments. Some researchers would answer the question positively, others negatively, and still others would be in doubt. Not all linguists perceive disagreements about terms and their definitions as an important issue. It is primarily those linguists who exhibit an essentialist frame of mind who do so; only those linguists who believe that, in principle, for example, linguistics, pragmatics, or kinesics can be defined in one way tend to get involved in definitional disputes. Thus, an essentialist working in a linguistics department will tend to ask others, or demand (at worst), that his or her definition of linguistics be respected (it should be stressed that the essentialist believes that his or her definition of linguistics is correct), and that only research and teaching that adhere to or reflect the definition in question be allowed. In that way, we arrive at a narrow (if not narrow, then at least somewhat restrictive) definition of a discipline. Anyone not adhering to the stated definition will be considered to be wrong and will be stigmatized, perhaps, as not really belonging to the department in question. Other, more open, definitions of the field the essentialist will tend to reject. For such an essentialist, certain colleagues clearly do not belong to the group. Such a situation can be intellectually stifling and unrewarding. In such a case, we are faced with a situation in which a field is defined narrowly and where any activity outside the confines of that narrow definition tends to be treated as in fact belonging somewhere else. In this view, a department is treated as a discrete organizational unit to which an individual either belongs or does not belong. Departments of a university are thus like boxes. Each researcher is seen, in principle, as assignable to one of the boxes. From a nonessentialist view, the essentialist position supporting a rigid classification of fields and subfields (usually the one that is currently in place), whether it concerns linguistics or other departments, is stifling in the sense that it arrests free initiative, imaginative thinking, and academic freedom. Significant innovations in science (which may be assumed to successfully originate in free initiative, imaginative thinking, and academic freedom environments) have arisen, among other things, through bold leaps across disciplines. We have to keep in mind that departmental or institutional boundaries foster discipline boundaries. Feyerabend (1981a) praised the physicist and philosopher, Ernst Mach, for his neglect of discipline boundaries: A third interesting feature of Mach’s philosophy was its disregard for distinctions between areas of research. Any method, any type of knowledge could enter the discussion of a particular problem. In building up his new science,



Mach appealed to mythology, physiology, psychology, history of ideas, history of science as well as to the physical sciences. (p. 81)

Further: One outstanding feature of scientific research, especially of the kind envisaged by Mach, is its disregard for established boundaries. Galileo argued as if the distinction between astronomy and physics, which was a basic presupposition of the knowledge of this time, did not exist; Boltzman used considerations from mechanics, the phenomenological theory of heat and optics to determine the scope of the kinetic theory; Einstein combined specific approximations with a global and very “transcendental” of physical world views; Heisenberg got some of his basic ideas from the Timaeus and, later on, from Anaximander. (Feyeraband, 1981a, pp. 83–84)

Another related form of covert essentialism is to be found where questions concerning the boundary of disciplines are posed as serious academic issues. Aitchison (1994b) endorsed the view that such questions are indeed taken seriously: “Semantics and syntax therefore overlap, and linguists spend a good deal of time arguing where the boundary between them should be located” (p. 101). As Lundquist (1990) told us: “One of the prevailing, ever recurring, and—so it seems—everlasting metatheoretical questions in linguistics is where ‘to cut the cake of Meaning’ (Gazdar, 1979, p. 3). And more specifically, where to cut the semantic piece of the linguistic cake from the pragmatic” (p. 79). The distinction between semantics and pragmatics is one that no doubt many linguists are preoccupied with. Levinson (1983) also seemed to indicate that for him and others the question of discipline boundaries is not irrelevant: “In any case, it does not seem that the distinction between sentence-meaning and utterance-meaning can be relied upon to clarify the distinction between semantics and pragmatics” (pp. 20–21). The linguists’ concern with discipline boundaries is only one example of how academics in general tend to be concerned about where one discipline ends and another begins. The concern in question extends, as might be expected, to the question of how science differs from nonscience and how one type of science (e.g., social science) differs from another one (e.g., natural science). Committees are formed and reports are compiled to address such questions. As Fromkin (1993) told us: Those interested in the question of linguistics as a science may wish to look at the The Behavioral and Social Sciences report of the Committee on Basic Research in the Behavioral and Social Sciences of the National Research Council of the National Academy of Sciences, 1988, National Academy Press Wash-



ington, DC. This is a report on scientific frontiers in the behavioral and social sciences, including linguistics as well as psychology, anthropology, sociology, geography, political science, and economics. Interesting that the National Academy has no problem with including linguistics as a science nor does the AAAS. (p. 1)

Following Fromkin’s mail was a series of commentaries. One is this: Of course there are applied science and theoretical science. Is there a septum between the two? If not, then we must seek elsewhere for the elusive distinction between science and—shall we say—knowledge. Does the difference lie in the nature of the matter, in the uses to which it is put, in the manner in which it is acquired, in the abstractions ordering it, in the principles by which these abstractions are arranged and tested, or in some combination of these (or even of other characteristics as well)?

When is an auto mechanic—who has knowledge aplenty—a mechanic and when [is he] a scientist? When, say, is an automotive engineer a scientist? When is a linguist a practitioner of knowledge and when a scientist? When does a profession attain to the realms of science? Or are these ignes fatui of a meaningful discussion of the “science” of linguistics? (Purl, 1993, p. 1)

Haspelmath (1993) continued the discussion: I completely agree with Steven Schaufele that linguistics should be considered as science—but what follows from this? To a large extent, it may boil down to a question of prestige. The word “scientist” sounds much more prestigious than “scholar,” so it would be good for us in that respect if we could convince the world around us that we are scientists (though it would be of little help in many non-English speaking parts of the world, where the scientist/ scholar distinction is not made, e.g., Russian uchenyj, German Wissenschaftler). A much more interesting question, it seems to me, is what kind of science linguistics is. Is it more like biology, for instance, or more like physics? If it is like physics, then we should look for mathematical models (perhaps involving fairly simple mathematics). . . . On the other hand, if linguistics is like biology, then we should study language in connection with its environment and look at the linguistic strategies by which the organism attempts to cope with the imperatives of survival. (p. 1; see also chap. 7, this volume)

Williams (1993) was concerned with and about the following questions: “a. Is a study of puppet movements a ‘proper study’ of kinesics? b. Is a study of puppet movements a ‘proper study’ of linguistics? More specifically, would you accept this topic as a thesis proposal in linguistics?” (p. 1)



Mey (1993) was also concerned with discipline boundaries and definitions: “To define” means to impose a boundary (cf. the Latin word finis ‘end’; plural fines ‘frontier’). “Defining pragmatics” thus implies determining its frontiers with other adjoining fields of research within (and possibly also outside of) linguistics. Unfortunately, so far nobody has been able convincingly to postulate any such defining boundaries; nor have the definitions that have been offered given us any possibility of delimiting pragmatics clearly and neatly to everybody’s satisfaction. (pp. 42–43; italics added)

In this quotation, Mey (1993) clearly indicated that defining pragmatics and separating it from other disciplines has certainly been an issue. Although he said further, “A real ‘definition’ in this sense is thus just as impossible to provide as a ‘grammatical’ definition in the sense of the previous section” (p. 43), the reader is advised to note the words “so far” (italicized in the prior quotation), which may indicate that in principle distinguishing pragmatics from other disciplines is possible, although not achieved until the present moment for whatever reasons. As the prior quotations indicate, discussions about the nature of disciplines and discipline boundaries, such as those boundaries around linguistics, are a fact of life for some linguists. It is my contention that most, if not all, of these discussions, if taken as a serious attempt to define disciplines and their boundaries unambiguously, are an expression of essentialist reasoning. As follows from the nonessentialist agenda promoted in this book, disciplines and their boundaries, like all other concepts, will never be defined unambiguously, and ultimately no agreement may ever be reached about each individual study belonging to a certain discipline. For a nonessentialist, arguing about whether linguistics is a science or whether a particular study belongs to linguistics, sociolinguistics, psycholinguistics, pragmatics, and so on, is not necessary; it leads nowhere. For a nonessentialist, what is important is what you do (i.e., what kind of phenomenon you study, and not what you call the phenomenon or the study in question). Although labels such as sociolinguistics and psycholinguistics may sometimes be helpful in assigning a study to a tradition within linguistics or in assigning a researcher to an organizational unit such as a university’s department, nonessentialists do not ordinarily argue about discipline boundaries. These boundaries are usually not important to nonessentialists. Other examples of what might be called covert essentialism are the following: We should note that assimilation has many connotations, two of which are conformity to the dominant culture and complete homogenization. The first, in



the English-speaking world, is often termed anglo-conformity; here ethnic groups adapt themselves to the prevalent culture which, itself, remains stable. This is not a melting pot in the true sense, but it is what the melting pot has meant to many. (Edwards, 1985, p. 103; italics added)

Further, “Second, there is often confusion between belief and attitude; strictly speaking, attitude includes belief as one of its elements” (Edwards, 1985, p. 140; italics added) In the nonessentialist view, the formulations “in the true sense” and “strictly speaking” indicate that the author may believe that there exists one ultimately correct definition of the terms melting pot, attitude, and belief, which is a truly essentialist echo by my definition of essentialism. The former quotation especially testifies to this conclusion. On one hand, we have melting pot in the true sense, and, on the other, we have melting pot meaning different things to many people. The contrast between the apparently real meaning (the true sense) and the accidental meanings to many can hardly be denied. In the latter quotation, the formulation strictly speaking may be easily interpreted as meaning that once the notion of attitude is viewed in its true sense, that of belief needs to be included. However, when attitude is not viewed in its true sense, the inclusion of belief is not necessary. The presence of essentialism in linguistics should not be seen as entirely unexpected. Many linguists appear to exhibit an essentialist frame of mind in ways similar to those exhibited by very many lay people. Although we lack any reliable data as to how many people or what kind of people (e.g., representatives of certain professions) are exponents of essentialism, there is ample psychological evidence that people in general tend to believe that there is essence in things. Medin and Ortony (1989) contended that, although philosophical essentialism (roughly, the kind of essentialism criticized in this book) is untenable, psychological essentialism exists and is well attested. Referring to psychological essentialism, Medin and Ortony (1989) stated, “This would be not the view that things have essences, but rather the view that people’s representations of things might reflect such a belief (erroneous as it may be)” (p. 183). Further: We think there is evidence that ordinarily people do believe that things have essences. Many people behave as though they believed it, presumably because the assumption that things have essences is an effective way of viewing the world and making predictions about it. . . . The point about psychological essentialism is not that it postulates metaphysical essentialism but rather that it postulates that human cognition may be affected by the fact that people believe in it. In other words, we are claiming only that people find it natural to assume, or act as though, concepts have essences. (pp. 183–184)



In light of these claims about psychological essentialism, the many examples of essentialism quoted in this chapter certainly do not come as a surprise. Interestingly, when labeled essentialist, some essentialists do not seem to be able to defend their essentialist position cogently. This is probably because essentialism is in fact hard to defend; it is difficult to find convincing arguments for the essentialist position. Some essentialists behave like Harris’ segregationalists (who live by what Harris called the Western language myth; cf. chap. 3; this volume); that is, when challenged, “ . . . instead of trying to meet the challenge head-on and defend those assumptions, what they typically did was deny that they believed in the myth . . . and then carry on exactly as before” (Harris, 2002, p. 3). For lack of arguments perhaps, some people with an essentialist bias say explicitly that they simply believe that essentialism is a correct way of conceiving the universe, which is consonant with what Medin and Ortony (1989) claimed. Such belief needs to be respected partly because the nonessentialist’s own reasoning is also based on belief to some extent. Although the nonessentialist will not be convinced of the grounding of the essentialist’s belief, the former should have some tolerance for the latter’s position. If an essentialist tells a nonessentialist that the latter’s (relativist) view is unacceptable, and that she or he believes that, for example, the word good or evil can or, in principle, should be unambiguously defined, then the nonessentialist at least knows that different beliefs are at issue. Neither side accepts the other’s arguments. Although the two views differ significantly, the nonessentialist usually has tolerance for his or her essentialist interlocutor. It may be doubted, however, whether the essentialist interlocutor usually has tolerance for the nonessentialist in this respect. The essentialist position tends to be “a one-truth position.” That is why essentialists tend to think that they are definitely right; someone who has opposing views is usually seen as definitely wrong. This essentialist view is best reflected in the title of de Bono’s (1991) book, namely, I Am Right You Are Wrong. It should be noted that in the title “I am right and you are wrong,” and not the other way round. This is the way the essentialist tends to reason. “I am wrong you are right” is much less likely to be acceptable. A typical essentialist seems to give others the impression that she or he is always right. Obviously, nonessentialists also hope that they are right, in a way. More important, however, for a nonessentialist, the question of who is right or who is wrong is not very engaging. The nonessentialist’s mind, unlike the essentialist’s, is not set for thinking in terms of who is right or wrong, especially in terms of who is definitely right or definitely wrong, in some sort of absolute sense. The essentialist tends to think that the nonessentialist is definitely wrong, and the former tends to think that she or he can show to the



latter that this is the case. The nonessentialist, however, does not tend to think that she or he could ever conclusively show to the essentialist that the latter is definitely wrong. The nonessentialist is not really concerned with this question. While the essentialist tends to believe that she or he can prove she or he is right, the nonessentialist does not make any claims about proving anything. Most important, while the essentialist tends to think that she or he, being definitely right, has the right to impose his or her viewpoint on others, the nonessentialist does not at all feel that she or he has the right to impose his or her viewpoint on all those other people who think that his or her views are inappropriate. Essentialists tend to believe that they are infallible, whereas nonessentialists do not.

ENDNOTE 1. Pidgins are examples of partially targeted or nontargeted second-language learning, developing from simpler to more complex systems as communicative requirements become more demanding. Pidgin languages by definition have no native speakers, they are social rather than individual solutions, and hence are characterized by norms of acceptability (Mühlhäusler, 1986, p. 5).



Integrational Linguistics and Nonessentialism

INTRODUCTION The nonessentialist position argued for in chapters 1 and 2 is obviously not an entirely new approach in the field of linguistics. The nonessentialist position (often referred to also as the nonobjectivist, nonclassical, or indeterminist position, to mention only the three most frequently encountered labels) has been the choice of a number of linguists with a variety of interests within language and communication studies. Individual researchers have expressed more or less nonessentialist ideas in a variety of papers, books, and oral presentations. With the notable exception of the general semantics of the 1930s (see chap. 4, this volume), however, until about 25 years ago, nonessentialism had not been represented in linguistics by any single more or less coherent theory or any one individual or group of people having more or less similar views on the fundamentals of their discipline. A breakthrough occurred during the 1970s and 1980s: Eleonor Rosch’s two publications (Rosch, 1977, 1978) and Lakoff and Johnson’s seminal book on metaphor (G. Lakoff & Johnson, 1980) marked the beginning of cognitive linguistics as many people understand it today (see chap. 4, this volume). At more or less the same time, Harris’ (1981) publication of the Language Myth marked the beginning of what is now known as integrational linguistics. Of the two approaches (cognitive and integrational), the cognitive is certainly well known and appreciated by many linguists. This is one of the reasons that chapter 4 does not include much introductory material on cognitive linguistics. Rather it resorts to a selection of cognitive linguistics ideas 38



and uses them to develop an argument for applied cognitive sociolinguistics, which is my main objective in the book. It is my conviction that, as opposed to cognitive linguistics and linguists, integrational linguistics and the researchers who try to promote integrational linguistics are not widely known. For that reason, one of the goals of this chapter is to encourage linguists and other researchers to get acquainted with integrational linguistics and to consider making extensive use of at least some of its fundamental ideas. In what follows, I briefly introduce these ideas and indicate how they relate to the kind of nonessentialism that I promote in this volume. I also point to some aspects of integrational linguistics that the nonessentialist is likely to question; I state how my own views on these aspects differ from those of Roy Harris, who is commonly perceived to be the founder of integrational linguistics.

ROY HARRIS’ INTEGRATIONAL LINGUISTICS Harris (1981) launched an overall and overarching critique of most of contemporary linguistics, which he often referred to as “orthodox linguistics.” His initial assumption is that most of contemporary linguistics, and especially the transformational–generative framework, is based on a myth. Harris called this myth “the language myth” or “The Western language myth.” It is hoped that the passages quoted next and my own comments following them will allow the reader to connect Harris’ criticism of orthodox linguistics with my own criticism of essentialism. Although the two do not overlap entirely, their relationship should be evident. Simply stated, I wish to point out that Harris’ views may be seen as largely in agreement with my own reasoning advanced in this book. According to Harris, the Language Myth in its modern form is a product of post-Renaissance Europe. Harris made it explicit, however, that the myth in fact goes back to Aristotle, and that it has been passed on to us via Locke and de Saussure. In Harris’ opinion, the language myth springs from two related fallacies. These are the telementational and determinacy fallacies. The former concerns the function of language, and the latter concerns the mechanism of language (1981). These two fallacies, according to Harris, are significantly interrelated and associated with each other: According to the telementational fallacy, linguistic knowledge is essentially a matter of knowing which words stand for which ideas. For words, according to this view, are symbols devised by man for transferring thoughts from one mind to another. Speech is a form of telementation. (1981, p. 9)



The determinacy fallacy, or “fixed code” fallacy . . . provides the explanation of how the telementation process works, and indeed of how telementation is possible. . . . A language community is a group of individuals who have come to use the “same words” to express the “same ideas” supplied by Nature, and to combine those words in the same ways into sentences for purposes of connected discourse. (1981, p. 10)

According to the language myth, human communication works more or less in the following way: Suppose A has a thought that he wishes to communicate to B, for example, that gold is valuable. His task is to search among the sentences of a language known both to himself and to B, and select that sentence which has a meaning appropriate to the thought to be conveyed; for example, in English, the sentence Gold is valuable. He then encodes this sentence in its appropriate oral or written form, from which B is able to decode it, and in virtue of knowing what it means, grasp the thought which A intended to convey to him, namely that gold is valuable. (1981, p. 11; italics original)

Harris correctly associated this view of communication with that discussed by Reddy (1993) under the heading of the conduit metaphor; this metaphor (and expressions like None of these ideas came through to me, He didn’t manage to get his thoughts across) reflects the common belief that people insert thoughts and meanings into words and sentences, send them to their interlocutors, and the interlocutors take the thoughts and meanings out of these words and sentences. The conduit metaphor may well be taken to illustrate the Western language myth proliferated both by many professional linguists and lay people. In a related discussion of word meaning and definition, Harris expressed the view that it is a mistake to expect a single definition of a word and clear distinctions between related concepts (e.g., singular and plural, auxiliary and verb). At the same time, he argued for viewing linguistic concepts as continua. He discussed, for instance, grammatical indeterminacy and, invoking the concept of category squish, endorsed the treatment of syntactic categories “as ‘more nouny’ or ‘less nouny,’ ‘more verby’ or ‘less verby,’ and so on” (Harris, 1981, p. 83). Laying out his theory of language, Harris (1998) addressed the question of word meaning in a straightforward manner. Referring to classical truth conditional semantics (see Lyons, 1977), he maintained that it involves “the inevitable search . . . for some universal way of pinning down invariant, context-free meanings. These have to be meanings that are, as it were, permanently attached to vocal or written forms” (Harris, 1998, p. 68).



Criticizing this view, Harris rightly pointed out that “ ‘Meaning’ is the value which we seek to attribute to words” (Harris, 1998, p. 69; italics added). He stated further: Integrational semantics recognizes all this, while rejecting the assumption that words have meanings, at least in the sense in which that is maintained by segregational theorists. That is to say, for the integrationist meanings are not fixed semantic values which somehow attach to particular verbal (or other) forms irrespective of the communicational circumstances. (Harris, 1998, p. 85; italics original)

Referring to orthodox linguistics, in an important summarizing statement on meaning and understanding, Harris (1981) noted: The central indeterminacy of all communication is indeterminacy of what is meant. Orthodox linguistics is prepared to concede that on particular occasions of language use, it may be unclear what a speaker’s intentions were, or how a hearer understood what was said. But these uncertainties are treated as accidents or defects attendant upon the diverse circumstances in which words are used. What the words themselves mean, on the contrary, is held to be determinate for the linguistic community; as if the very multiplicity of possible individual doubts somehow cancelled one another out and produced collective certainty. (p. 167)

The earlier passages may certainly be taken to mean that meaning is not resident in words and that it is generated by people in real social situations. Discussing, as an example, the meaning of the word moshpit, Harris (1998) noted, “The great mistake the orthodox semanticist makes is . . . insisting that moshpit . . . must have a ‘real’ (= determinate) meaning known at least to those ideal speakers of whatever language (= fixed code) it belongs to” (p. 70). “In fixed code semantics, it is the code that determines meaning, not the speaker” (p. 71). In the integrationist view, there are no autonomous words in which meaning resides. Thus, the question of what the real meaning of moshpit is can never be resolved. Interestingly and more important, Harris’ reasoning applies not only to unusual and technical words such as moshpit; endorsing Hempel’s (1952) argument, Harris (1998) rightly noted that the same reasoning applies to commonly used words such as hat; it is not possible to define such words in any ultimate way. Harris endorsed the following statement by Hempel (1952):



The term “hat” is vague; i.e., various kinds of objects can be described or actually produced in regard to which one would be undecided whether to apply the term or not. In addition, the usage of the term exhibits certain inconsistencies both among different users and even for the same user of contemporary American English; i.e., instances can be described or actually produced of such a kind that different users, or even the same user at different times, will pass different judgements as to whether the term applies to those instances. (p. 10)

Harris’ (1981) view of communication, meaning, and understanding brought him to make a distinction between a democratic and an élitist view of meaning: It seems, then, that a “democratic” view of meaning will be challenged by an “élitist” view, and one of the attractions of the élitist view may be that it holds out the possibility that indeterminacy of meaning is only apparent. It is an uncertainty which arises from taking into account the conflicting usage of speakers who are not all equally expert. (p. 172)

Harris identified problems with both views. What is of special interest to the nonessentialist is the main problem of the élite view that Harris formulated in the following way: “If he [the theorist] is an élitist, he has to decide who the élite are: that is, which speakers in the community are those who do know what words mean and use them accordingly” (Harris, 1981, p. 173). For a nonessentialist, this formulation invokes the essentialist dilemma over the question of what the correct definition of “the élite” is. The nonessentialist, however, will certainly claim that such a definition is not available and that, by extension, the speakers who know the correct meanings of words can never be identified and defined in one indisputable way. Given the discussion in chapters 1 and 2, the similarity between Harris’ view of meaning and my own, as informed by Popper’s nonessentialism, should have become obvious to the reader by now. For further discussion of meaning and nonessentialism, the reader is referred to chapter 4. In his recent work, Harris (1997, 1998, 2002) referred to the main ideas of his earlier work, claims that they are still valid, and elaborated on some of them. Of this recent work, his Introduction to Integrational Linguistics (1998) may be taken to be his most salient expression of his views on language. Criticizing some of the fundamental assumptions of orthodox (segregationist) linguistics, Harris (1998) rightly contended that segregationists believe that the linguist has the privileged “. . . ability to determine in advance the boundaries of any given linguistic phenomenon and the appropriate methods to be employed in its investigation” (p. 15). According to Harris, part of the Western language myth is composed of the dialect myth and the standard language myth. The segregationist’s view



that languages are fixed codes includes the view that dialects (seen as units smaller than languages) are fixed codes, and the view that the standard languages/varieties are fixed codes as well. It is Harris’ opinion that the orthodox linguist posits that both dialects and standard varieties exist in the sense that they can be identified, circumscribed, defined, and, subsequently, described. According to Harris, most linguists still subscribe to this view and express it in their teaching: “. . . what most linguists still wanted to take as the basis of their teaching; namely, the objective existence of linguistic data, the objective existence of standard languages, and the objective existence of distinctions between correct and incorrect pronunciations, constructions, meanings, etc.” (Harris, 1997, pp. 240–241). To identify languages, dialects (including standard varieties), and objectively existing linguistic facts in general is, according to Harris, not plausible. A basic reason for this is that all the definitions of dialects and standard varieties, some of which Harris specified, appear to him debatable. For instance, to define the standard language in terms of prestige (e.g., “The standard variety is the most prestige variety of a language”) is debatable because the simple question that immediately emerges is, “To whom does a variety need to be prestigious to count as a standard variety?” Harris’ discussion of the points mentioned earlier can easily be interpreted as asserting that no one definition of dialect or standard variety can be provided. This conclusion parallels Harris’ view that neither dialects nor the standard variety actually exist. In connection with this, Harris (1998) drew our attention to the fact that some linguists make the mistake of treating languages as autonomous entities. Harris rightly noted that treating languages as autonomous entities is a token of treating languages as objectively existing objects that can be described in objectivist terms. The affinity between Harris’ reasoning and my own nonessentialist position should now be fairly clear; a nonessentialist will endorse Harris’ contention that languages and dialects (whether standard or not) cannot be viewed as autonomous entities primarily because they cannot be defined in one way. Any attempt at defining them in one way (i.e., trying to answer questions such as “What is German?” pushes the researcher into the essentialist orientation. Although many of Harris’ views and those of a nonessentialist (e.g., mine) are similar to a significant degree, it seems that there is also room for some disagreement between the two. Although I sincerely hope that I do not misrepresent Harris’ views, I would nevertheless like to dwell for a moment on a point or two of disagreement. One of Harris’ main claims throughout his work is that we should study communication rather than language—or, in slightly different terms, that we should study language as communication. It is to communication, then, that Harris addresses much of his attention. Harris used the term integra-



tional linguistics to refer to a different theoretical perspective from the mainstream (orthodox) one adopted by the majority of linguists, whom he called “segregationist” (Harris, 1998). The distinction between segregationism and integrationism is crucial. The former maintains that separating language from nonlanguage (and, we can add, by extension, linguistics from nonlinguistics) is possible and important; the latter maintains that such separation is neither possible nor important. A crucial part of Harris’ credo seems to be expressed in the following passage: The integrationist holds . . . that a theory of language(s) without a theory of communication is vacuous. . . . There is no autonomy for linguistics, because we cannot in practice segregate linguistic knowledge from extra-linguistic knowledge. The two domains are integrated, not segregated. . . . The study of that integration and its complexity is the proper study of language: there is no other. The integrationist therefore rejects the idea that verbal communication involves the kind of activity which allows the linguistic components to be distinguished from the non-linguistic and analysed systematically without reference to the latter. The integrationist claim, on the contrary, is that any such segregation is impossible. Not just difficult or dubious in certain cases, but impossible in principle. It is the recognition of that impossibility which is the cornerstone of integrational linguistics. (Harris, 1998, p. 10; italics original)

Further, There is consequently no way of divorcing linguistic analysis from questions of how we communicate with one another. In particular, there is no question of first describing a battery of linguistic instruments and only then inquiring how they work or what purpose they serve . . . It would be fallacious, in the integrationist’s view, to believe that the study of language can be divorced from the study of other forms of communication. (Harris, 1998, p. 54)

I welcome Harris’ plea for studying language as communication—that is, for studying language as a little-idealized phenomenon. The argument for applied cognitive sociolinguistics advanced in this book is no doubt sympathetic to treating language as a large communicational phenomenon rather than as a small, highly abstract and idealized one. The way Harris combines language studies and communication studies seems, however, to disagree with the non-essentialist perspective promoted in this book. Let us consider again Harris' claim that the integration of linguistic and extralinguistic knowledge should be (is!) the proper study of language. To my understanding, the prior claim (viewed in light of the two last quotations) stresses that there is only one correct way of doing linguistics— namely, by studying the linguistic and extralinguistic aspects of communication. Note the original italics of is. This passage implies that, in principle,



one correct definition of language study (in response to the what-is question, “What is the proper language study?”) can be expected. The definition is, in Harris’ view, a formulation like “The proper language study is a study of the complexity of the process of communication.” For a nonessentialist, no matter how valuable studying language as communication may be, it cannot be the only proper perspective. It may be very rewarding and it may be recommended (e.g., so that some language-related problems can be addressed), but it should not be viewed as the only perspective. Maintaining that there is only one way of viewing language study can be taken as a variant of the belief that there is only one correct definition of the words “language study,” which is of course unacceptable to a nonessentialist. It should be noted that Harris’ persistence that language be studied only as communication does not seem to be consistent with most of the rest of his views on how language works. Is he accidentally slipping into essentialism on the question of what language study should be? My suspicion that I may be unfair to Harris by suggesting that he might be slipping into essentialism may be determined in light of Harris’ other statements; his contention about what the proper study of language is does not seem to be reconcilable with his earlier statements, which, it should be stressed, a nonessentialist will fully endorse, for instance: The version of the language myth propounded by modern linguistics has it that there is only one descriptive standpoint which allows us to proceed to a systematic analysis of linguistic phenomena. Describing language must start by describing languages. And describing languages is not envisaged as describing anything that people do, but as describing what they are assumed to know. (Harris, 1981, p. 35; italics added)

In a related statement, Harris maintains, “But one of the basic messages of integrational linguistics is that the distinction between the linguistic and the non-linguistic can be drawn in a variety of different ways, from different perspectives, in different contexts and for different purposes” (Harris, 1997, p. 298). Could my suggestion that Harris may have slipped into essentialism be due to a misunderstanding? (See the discussion on misunderstanding in chapter 4 as well as Janicki, 1999.) As indicated before, for a nonessentialist, especially for a nonessentialist sociolinguist with interests in applied sociolinguistics, conceptualizing language as communication is obviously rewarding given the probable goal that an applied sociolinguist will have—solving practical problems. However, a nonessentialist will also accept other conceptualizations of language that may be rewarding for reaching other goals. The other conceptualizations (e.g., “Language as knowledge of grammatical structure”) may not be appropriate, interesting, or rewarding for a sociolinguist trying to under-



stand, for example, why people argue about whether Clinton had “a sexual relationship” with Monica Lewinsky (see chap. 5 for a detailed discussion of this topic), but they may be appropriate, interesting, and rewarding to other researchers trying to reach other goals. Although a nonessentialist will in general be very sympathetic to Harris’ integrating phenomena rather than segregating them, at times his views seem indefensible. His respect for studying language as communication and for studying it in context leads Harris to state the following: What is in contention as between the two approaches [segregationist and integrationist] is whether, or to what extent, linguistics is entitled to decontextualize human linguistic behaviour in order to isolate, describe and explain various aspects of it. For the integrationist, all decontextualization distorts, and therefore the resultant linguistic descriptions and explanations, to the extent that they rely on decontextualized “data,” are automatically suspect. (Harris, 1998, p. 13)

Further: The first thing an integrationist will insist on in linguistics is that any viable model of linguistic communication must treat all activities concerned—not just some of them—as integrated activities. This means that communication must be contextualized. Context is not an optional extra. It is not just a backdrop that can be changed at will, as in a theatrical performance. (Harris, 1998, p. 23; italics original)

While we should certainly accept the view that all decontextualization distorts, when read as part of the prior passage, this formulation seems to imply that it is possible to study communication without decontextualization. Neither this view nor the view that communication must treat all activities concerned seems to be compatible with the nonessentialist agenda endorsed in this book. This is primarily because both formulations seem to imply that what is communication and what is context can be ultimately defined; that is, that we can actually find out what communication and context really are. In other words, the implication of Harris’ views seems to be that communication can be distinguished from noncommunication and that context can be distinguished from noncontext on the basis of some principle and in a precise way. In my view, we could claim that we are able to study all the activities of contextualized communication only if such distinctions were possible. In the nonessentialist view, a clear distinction between communication and noncommunication is not possible, like a clear distinction between language and nonlanguage is not possible. Only if we could define communication precisely and infallibly could we state what activ-



ities come under its heading. Only then, to continue the nonessentialist reasoning, could we judge whether a linguist is considering all activities (aspects) of communication. Is Harris’ reasoning on this point another slippage into essentialism? Or is this another possible misunderstanding of Harris?

CONCLUSIONS It was not my objective in the last section to give an exhaustive presentation of Harris’ views of language. Given his thorough treatment of language and his long list of publications (including, e.g., the recent Harris, 2003), such a presentation would require a book. One goal was to draw readers’ attention to Roy Harris’ integrational linguistics (which, I daresay, is not widely known) and to show how it can be seen as largely compatible with the nonessentialist position that I develop in this volume. Given this goal, I presented only a selection of the main relevant points characterizing Harris’ view of language and linguistics, and I addressed them in light of the nonessentialist approach. Harris’ critique of orthodox segregationist linguistics is no doubt an allout attack on most of contemporary linguistics. Harris has been accused of trying to dissolve linguistics altogether, which he does not seem to have regarded as an insult at all. Having been accused of trying to annihilate linguistics and of making linguists unemployed, Harris responded to orthodox linguists’ criticism with flair: “I am not at all sure I am interested in the survival of the ‘linguistic industry’ as it operates at present: it generates too much intellectual pollution and too many cultural health hazards” (Harris, 2002, p. 303). Although Harris’ words may seem harsh and unsympathetic to many, I do not find them to be off the mark. I take my own presentation in chapter 2 to largely support Harris’ view that a lot of contemporary linguistics is unrewarding. Although I disagree with Harris on matters of detail, I believe that the significance of his views on language can hardly be overestimated. One can agree with Harris entirely; one can agree with him to a significant degree (as I do), or one can completely disagree with Harris. However, it is a mistake to disregard Harris. The relative silence around Harris’ work (which he himself mentioned), seems to indicate that many linguists, “indignant and hostile” (Harris, 1997, p. 231), have opted to do just that. Harris fully realizes that his agenda is a challenge that goes beyond linguistics. Referring to his 1981 book, The Language Myth, he stated the following:



From reactions to my book I also learned something that I had only dimly realized before. When you challenge the kind of assumptions that I had identified as the “language myth,” you are challenging much more than people’s beliefs about language. That is why they refuse to engage in debate, or reject the terms of debate out of hand, or say you are absurdly misrepresenting their position. . . . You are calling in question something about the way they conduct themselves not just in their professional capacity as linguists, but also as educated persons, law-abiding citizens, good neighbours, and in many other aspects of their social being. That is what offends them. You are challenging a whole cultural picture at a much more basic level than you might at first have supposed. (Harris, 2002, p. 6)

As the remaining part of my book shows, there is no reason whatsoever to disregard or denigrate at least some of the basic ideas about language promoted by Harris. These ideas, although expressed in different terms, can also be found in the core of cognitive linguistics, to which we now turn.



Variations on Meaning in Cognitive Linguistics

ON THE FUZZINESS OF CONCEPTS Like Harris’ integrational linguistics (see chap. 3, this volume), cognitive linguistics (CL), which emerged in the late 1970s and early 1980s (e.g., Johnson, 1987; G. Lakoff, 1982, 1987; G. Lakoff & Johnson, 1980; Rosch, 1977), may be seen as largely supportive of the nonessentialist, non-Aristotelian view of meaning, definition, and categorization. As stressed by Lakoff: The classical theory of concepts and categories has been accepted in the West for two thousand years. It has become so much a part of Western culture and education that it is hard for some of us to think in other terms. But the time for that has come . . . Changing our ideas about categories will require changing our ideas about rational thought, the nature of the mind and its relation to the body, and, in the process, changing our conception of man. . . . Academic disciplines, from my own field of Linguistics to Philosophy, Psychology, Anthropology and the other of Social Sciences, need to be rethought. (G. Lakoff, 1982, p. 99)

Cognitive linguistics is to a considerable extent nonessentialist. The qualification, “to a considerable extent,” indicates that cognitive linguistics does not consist of a group of people who all share the same beliefs and who all make exactly the same philosophical, possibly nonessentialist, assumptions. Many cognitive linguists, however, are nonessentialist in their attitude to meaning, definitions, and categorization (see e.g., Taylor, 2003). 49



One of the nonessentialist claims that most, if not all, cognitive linguists agree on is that meaning is not resident in words, but assigned to words by human beings (G. Lakoff & Turner, 1989). In the objectivist/essentialist view, words are conceived as containers in which meaning resides independently of the users of the words and in which it is transmitted from one language user to another. In the nonessentialist, cognitive linguistics view, the container metaphor expresses a misconception (i.e., in this view, meaning is created by the individual language user [in his or her head] basing on his or her experience, and attributed to words as they are produced in the spoken or written form). Lakoff was one of the first cognitive linguists to note that CL is a challenge to the 2,000-year-old tradition. We need to realize that CL is not a fad in linguistics or even a major research orientation, such as sociolinguistics. CL represents a major shift concerned with fundamental philosophical assumptions. This shift is away from essentialism and toward nonessentialism (see G. Lakoff & Johnson, 1999). As mentioned in chapter 3, cognitive and integrational linguists are not the first to have moved away from essentialism. Within what is normally thought of as linguistics, as early as in the 1930s, the general semanticists (e.g., Hayakawa, 1965; Korzybski, 1933) were also notably nonessentialist (see Janicki, 1990). In the 1930s and 1940s, when Popper presented his critical analysis of essentialism, quite independently, it seems, the general semanticists formulated their views. Unfortunately, Popper’s and the general semanticists’ work remains largely unacknowledged by cognitive linguists. This is true despite that what we are dealing with is a matter of a major shift, which concerns not only the practice of a single discipline such as linguistics or even the whole of science. What is at stake here is a view of the universe, with all kinds of crucial theoretical implications and practical social consequences that such a view may generate. Although Popper and the general semanticists may be seen as cognitive linguists’ philosophical predecessors with regard to some fundamental questions, the work of the former and the latter does not really overlap. Cognitive linguists aim their research in entirely different directions. For instance, they devote considerable time to such phenomena as metaphor, metonymy, and iconicity, which neither Popper nor the general semanticists were ever concerned with. The philosophical foundations on which cognitive linguists build are, however, similar in some respects to those on which Popper and the general semanticists built. This especially concerns the treatment of meaning and definitions. Of course one could claim that neither Popper nor the general semanticists were the first to have broached the nonessentialist, non-Aristotelian reasoning. The Sophists could be mentioned in this context (see Guthrie, 1971).



Cognitive linguistics is nonessentialist primarily in that it does not operate with clear-cut, discrete, discontinuous (often called classical, or objectivist) concepts. In the CL view, members of concept categories do not have to be either in or outside the category. They may be category members to varying degrees (see e.g., G. Lakoff, 1987; Taylor, Cuyckens, & Dirven, 2003). That is not possible in the classical Aristotelian view. The view of concepts and word meaning advocated in chapter 1 and chapter 2 connects significantly with the CL’s view of fuzziness. There is a simple and straightforward relationship between the two. Namely, fuzziness may be treated as an aspect of indiscreteness, or as the opposite of discreteness. It should be remembered that for people who believe in discreteness, something should be one thing or another; a chair or not a chair, a good boy or not a good boy, a stone or not a stone, and so on. These people are, obviously, in their daily lives, exposed to things, behaviors, or activities that are not easy to classify. They are then in a way forced to acknowledge the existence of fuzzy cases. However, they treat fuzziness as a deviation, as something abnormal, or something that we should avoid as much as possible. However, for the adherents of the fuzziness view, something may be a chair to some degree, one may be a good boy to some degree or in some sense, something may be a stone to some degree, or, shall we say—like a stone, and so on. In other words, the fuzziness view implies that concepts are fuzzy and that they may be so to varying degrees. Again, such a state of affairs is considered normal and very frequent. In connection with the idea of the degrees of fuzziness, let us consider the following three concepts: pity, bicycle, and meter. Let us consider pity first. A dictionary definition of pity says, for instance, “the feeling of sorrow and compassion caused by the sufferings and misfortunes of others” (The New Oxford Dictionary of English, 1998, p. 1414). This definition does not give an entirely clear picture of what pity means. One might certainly wonder about what feeling, sorrow, compassion, sufferings, and misfortunes really mean. The reader should remember from the earlier discussion of definitions (see chap. 1, this volume) that trying to further define these four terms would not help much. In the subsequent definitions, you would have further vague terms for which you would feel you need still more definitions. Let us pick one of the five. Compassion, for instance, is defined in The New Oxford Dictionary of English (1998) as “sympathetic pity and concern for the sufferings and misfortunes of others” (p. 374). This last definition has certainly not contributed much to our understanding of pity partly because the word pity reappears in it. Maybe a measure of clarity in some sense has been achieved, but at the same time this last definition has contributed to increased fuzziness. In the last definition, there are new terms, sympathetic and concern, which are quite abstract and, thus, highly ambiguous. The point here is that pity remains a highly fuzzy word (concept) no matter how



many levels one goes through in his or her definitional pursuit. It should be remembered that in this definitional pursuit, one quickly finds him or herself in an infinite regression of definitions. Let us now consider the second concept—namely, bicycle. A possible definition of it is the following: “a vehicle composed of two wheels held in a frame one behind the other propelled by pedals and steered with handlebars attached to the front wheel” (The New Oxford Dictionary of English, 1998, p. 170). The concept of bicycle is fuzzy because one may argue about what a wheel is, what pedals are, what a vehicle is, and so on. The concept is fuzzy also in the sense that one may think of (draw or make) a vehicle that will not clearly be the sort of bicycle that one normally thinks of. Consider a bicycle with three wheels, one with a round steering wheel, one with a top to cover the rider, or one with two or more of these characteristics. Such a bicyclelike kind of vehicle would not be clearly assignable to the bicycle category. When asked to classify such a vehicle, some people would classify it as a bicycle; others would not. Still others would hesitate for a while, and subsequently admit that they do not really know, and still others would feel pressed and ultimately, without being convinced, would classify the vehicle as a bicycle anyway. Despite all the doubt that one may have, the concept of bicycle is evidently less fuzzy than that of pity. One of the reasons for that is that a bicycle can be easily imagined. It is possible to have a mental image of a bicycle of some kind. It is not possible, or extremely difficult, to imagine abstract things such as pity. Even if one can envisage some situations, people, or events associated with the concept of pity, the images (mental pictures, mental descriptions) that one may have are significantly more blurred and nebulous. Many dictionaries include pictures of bicycles. There seem to be no dictionaries that include a picture of pity. Bicycle is then less fuzzy than pity in that the former is easy to imagine. Moreover, you can also touch a bicycle with your hand. Let us now consider the third concept, the meter, which should be taken to be the least fuzzy of the three. Webster’s dictionary defines the meter as follows: “the meter is the basic unit of length in the metric system, equal to 39.37 inches: originally meant to be one ten-millionth part of a meridian from the equator to the pole” (Webster’s New World Dictionary of the American Language, 1964, p. 926). As is well known, in Sevres, close to Paris, in the International Office of Measures and Weights, there is a platinum rod model of the meter. It is stored under specific physical conditions (0 degrees Celsius, 760 mm Hg). Thus, it might be supposed that it is known exactly what the meter is. It is defined in terms of physical analytical categories such as temperature and pressure, which are very detailed and highly discrete units of analysis. One might expect, given the model stored in Sevres, that any instance of a meter (a meter of fabric, a meter of thread, a meter of string) is



clearly an instance that is either like the model in Sevres or not. But, as we are told, in Sevres there are, in fact, three other models that were meant to be exact copies of the original. Periodical checks of the measures of the four models indicate, however, that the length of the four platinum rods changes probably due to structural changes in the material from which the rods are made. That is why in 1960 (at the XI General Conference of Measures) the meter was redefined in terms of wave length and vacuum radiation to increase the level of precision of the model measures. It should be noted that one speaks about an increase in precision, not about ultimate precision (see e.g., Russell, 1996). The discussion of the three concepts—pity, bicycle, and meter—was meant to show that concepts vary significantly as far as their level of precision (level of fuzziness) is concerned. This brings us to the question of whether all concepts are fuzzy. In other words, the question is whether there is no such a thing as a precise concept. Cognitive linguists vary on this extremely important question (see e.g., Wierzbicka, 1990). Some claim, out of modesty perhaps, that most concepts are fuzzy (Channell, 1994) and leave to others the discussion of whether all concepts may be considered as such. It is my contention that all concepts that we use to understand the universe are fuzzy. More important, all concepts are fuzzy in the following sense: A concept would be nonfuzzy (discrete) only if no example of that concept could be found that would be debatable in the sense that people would disagree about its membership in the concept category under discussion. In other words, a concept would be discrete (nonfuzzy) only if all examples of that concept would be judged as belonging, without any doubt or hesitation, either in or outside the concept in question. This view on the fuzziness of concepts can be seen as a testable and refutable theory. It is testable in a very simple way—namely, if you point to a concept and show that all examples of that concept clearly belong to it—that is, that there is no borderline, fuzzy example available—then it would have to be admitted that the concept in question is not fuzzy. This may be paraphrased in still another way: The reasoning advocated here is that no matter what concept you can mention (car, pity, mouse, giving, record, etc.), you will always be able to find examples that are fuzzy—fuzzy in the sense that somebody who speaks English will be in doubt as to whether the example that he or she is faced with is not a member/example of that concept. Of course, it must be remembered that by a different definition of fuzziness, the prior argument might not be coherent or credible. This means that if you understand fuzziness in a significantly different way, then the previous argument might not be correct or convincing. The reader should by now take it for granted that one should not accept the position that



there is something like a good, proper, ultimate definition of fuzziness, or of imprecision. As all definitions are understood in this book to be nonessentialist working definitions (and not essentialist definitions), defining fuzziness should be viewed in the same way. That is, a working definition of fuzziness is provided, and an argument about fuzziness is built on the basis of this definition. The working definition provided is intended to make the concept a little clearer, but obviously not entirely clear in the sense of “utterly precise.” The working definition of fuzziness is, in other words, intended to bring me, the author, and you, the readers of this book, closer to each other as far as the understanding of this notion is concerned. It is often thought that, unlike in the humanities (including linguistics), the concepts used in biology, physics, logic, and mathematics are clear, discrete, and unambiguous. Many social scientists believe that although in the social sciences, precisely defined concepts have always been an unreachable ideal, in biology or physics (and certainly in logic and mathematics), precise concepts are used. Many social scientists feel that although linguists, for example, mostly use fuzzy concepts, which occasionally makes linguists’ conversation difficult to follow, biologists and physicists use only, or mainly, discrete concepts so that when they talk about professional matters, they understand each other completely. That is not true. As Popper (1957) cogently showed, ultimate precision of concepts in physics is a sheer myth. Independent evidence to the same effect is provided by the physicist Heisenberg (1969), who rightly said that as long as words of natural languages are used, there is fuzziness everywhere, including in physics. Kitcher (2001), in turn, supported the view that biologists do not operate with entirely precise concepts. This includes the concept of the gene. “Matters are even worse when we move away from genetics and consider such important biological notions as cell, organism, Drosophila melanogaster, species, and predator” (Kitcher, 2001, p. 70). It is of utmost importance to remember that the conclusion (or conjecture) that all concepts are imprecise refers to a perception—to a human perception of real-world phenomena. Matters are different in logic and mathematics, but only to the extent that they are not about perception. In geometry, for example, concepts such as square and triangle are defined precisely, but only with respect to geometrical axioms, which are specific theoretical assumptions about the world. Neither the axioms nor the definitions of square and triangle are meant to be ultimate in any sense. Precision is understood differently here than it is understood in the main argument of this book, which, it must be repeated, concerns human perception. According to Kitcher (1984), even in mathematics, terms such as function, continuity, and series summation are not unambiguous. This demonstrates that even mathematics is no exception with regard to the imprecision of concepts.



Penrose (1997) noted: There are two other words I do not understand—awareness and intelligence. Well, why am I talking about things when I do not know what they really mean? It is probably because I am a mathematician and mathematicians do not mind so much about that sort of thing. They do not need precise definitions of the things they are talking about, provided they can say something about the connections between them. (p. 100; italics original)

Popper (1982) stated the following about mathematics: “Any attempt to make the meaning of the conceptual system «precise» by way of definitions must lead to an infinite regress, and to a merely apparent precision, which is the worst form of imprecision because it is the most deceptive form. (This holds even for pure mathematics.)” (p. 44; italics original). One interesting and significant difference between many social scientists, on one hand, and mathematicians, logicians, and physicists, on the other hand, is that although the former try to define the fuzzy concepts, which are in fact not possible to define ultimately, the latter are hardly concerned with fuzziness at all. They set research goals that do not involve having to define what appears to be the undefinable.

MORE ON THE IMPRECISION OF CONCEPTS The prior discussion may be taken by some readers to mean that all scientific disciplines make use of equally imprecise concepts. In fact, different disciplines make use of concepts that are imprecise/fuzzy to varying degrees. Compared to the social sciences, in the natural sciences, one can certainly talk about a lower degree of fuzziness. This point was illustrated earlier with the help of the concepts pity, bicycle, and meter, where the precision of the concept of meter corresponds to the natural rather than to the social sciences. To differentiate between equally imprecise and imprecise to varying degrees is fundamental. This is to mean that some disciplines use more precise concepts than others, but none uses ultimately precise concepts. Disciplines such as physics and chemistry use concepts that are much more precise than those used in the human sciences. This point can be very well illustrated with a metaphor. Let us consider what happens when a hard cake is cut with an extremely sharp knife. Not a crumb will fall off either of the pieces (the big cake and the small piece that has been cut out). When another piece is cut out with a very blunt knife, not only will the piece be a little crooked, but also tiny little pieces and crumbs are likely to fall off. People sometimes hesitate whether the little



pieces that have not been cut out properly belong to the big piece still on the plate or to that of the person currently helping him or herself. In the former case (i.e., when the cake is cut with a very sharp knife), the fact that one’s piece is not 100% straight is of no interest to anybody. For all practical purposes, a very nice looking and clearly straight piece has been cut out. In this case, the small crookedness that could certainly be detected by some sophisticated optic instrument is simply uninteresting to anybody present at the party. This crookedness is negligible. In the latter case (i.e., when the cake has been cut with a blunt knife), the crookedness is not negligible at all, especially when the cake is tasty. When it is exceptionally tasty, guests may be willing to “steal” a part of their neighbor’s piece, pretending that it is their’s. At least some concepts used in, for example, physics and chemistry, are comparable to the piece of cake cut out with a very sharp knife. The concepts are not utterly precise, but their imprecision is not a problem for the physicists or chemists. They do not have to be concerned about it. They can proceed with their work despite that the concepts they use are a little imprecise. Similar to the crookedness of the cake piece cut with a very sharp knife, the imprecision of the concepts is unimportant here. In the case of the human and social sciences, however, most concepts used are imprecise to a degree comparable to the piece of cake cut out with a very blunt knife. That is, most concepts in the human sciences (linguistics included) are very imprecise, and that fact makes theory formulating and testing, as well as all kinds of discussion, difficult. To return to the question of concepts and meaning, the reader should remember one of the main claims made in this book—namely, that all concepts are imprecise (fuzzy). What is equally important is that we should talk about all the consequences that follow from having made that claim. These consequences pertain to most, if not all, domains where human activity is involved. These consequences obviously apply to linguistics and other academic disciplines. As shown later, they apply to all kinds of conflicts, including political ones. If the claim is accepted that all concepts are fuzzy to some degree, it should be soon realized that many fundamental mistakes can be easily avoided—mistakes that are otherwise made in tackling the world in general or in academic disciplines such as linguistics. Considering the perpetual search for the right definition on the part of many academics, and considering the amount of time spent, effort invested, and possibly anguish experienced, abandoning such a search should be one clear and salient consequence of accepting the claim that all concepts are fuzzy. There are many ideas and areas of inquiry traceable to cognitive linguistics that appear attractive and yet, at the same time, are consistent with the nonessentialist reasoning promoted in this book. The most important one concerns concepts and the process of categorization (see e.g., Taylor,



2003). Categorization can be conceived of as cutting the universe into the little bits and pieces by which we investigate and understand it. We cut the universe into bits and pieces and we name them or give them linguistic labels (in English) such as: cars, lamps, mountains, valleys, computers, mice, silence, regret, killing, anger, shoes, patience, stones, learning, and thousands of others. As stressed earlier in various sections of this book, these concepts (or words labeling them) cannot be defined precisely so that we could know exactly what the words used as linguistic labels for them really mean. Most cognitive linguists do not view concepts/words as precisely definable. In the cognitive linguistics view, concepts are seen as prototypes (see e.g., Geeraerts, 1997; G. Lakoff, 1987; Rosch, 1977, 1978; Taylor, 2003). The prototype theory of the concept says, roughly, that concepts, these little bits and pieces in terms of which we understand the phenomena around us, are best understood when viewed as mental images or mental descriptions, which are nondiscrete. These mental images or descriptions are also typical images or typical descriptions, which the individual has learned in the ontogenetic process (i.e., during his or her life). In the prototype view of the concept, we organize our categorization of world phenomena around prototypes (i.e., around typical instances of concepts). However, we also have to tackle nontypical cases, peripheral, or borderline cases. When nontypical instances of concepts come to the fore, much freedom is left for the individual language user to assign the given instance to one concept or another. When we consider how individuals can couple that freedom with the significant fact that what is a typical image, or description (= prototype), for one person of tour, for example, is not necessarily a typical image, or description, of tour for another, we realize immediately that we are concerned here with a multitude of sets of concepts in terms of which individuals perceive the world. Consider an example to illustrate the reasoning presented earlier. Take the concept of a TV set, for example. When one says, “I bought a new TV set yesterday,” in the interlocutor, the concept of TV set is invoked, and only then does the interlocutor understand the words “TV set.” What comes to mind is a mental picture of what constitutes for the interlocutor (and maybe for other people as well) a typical image, a typical mental description of a TV set, with some kind of box (including shape, size, and color), some kind of screen (flat, or convex a little, rather than concave), some system of controls, cables, plugs, and so on. It needs to be stressed that the interlocutor does not really know what has been bought because she or he has not seen the object, but the mental picture that the mention of the words TV set invoked gave him or her a fairly good idea of what has been purchased. To continue the paraphrasing exercise, when the interlocutor visits in his or her house the person who has bought the TV, she or he will match



the typical image (his or her prototype) of the TV set with the example; that is, with the TV set that has actually been bought and that is now being displayed in the family living room. If a 28-inch Black Trinitron Sony™ set was bought, what was bought was what many would see as a typical TV set. In other words, the example will fit nicely the mental picture that the original interlocutor had. If what was bought was a 49-inch set of whatever make, the visitor (the original interlocutor) would be a bit surprised. The thing would still be a TV for him or her, but it would be a less typical example. Even less typical an example would be a 60-inch Bang and Olufsen™ set, which, as many may know, has a fairly unusual shape. It needs to be added that we can always think of examples of concepts that would be hard to classify as being in or outside of a particular concept. To return to the earlier example, if an 80-inch diameter flat box rectangular screen Bang and Olufsen (now often called plasma screen) were bought and if it were hung on a wall, the visitor would probably say that the host should have said earlier that she or he bought a small movie theater screen and not a TV set. Consider other examples: sofas, chairs, and tables. When you walk into a very modern furniture store and look around, you will find various types of furniture there. In addition to all the sofas, chairs, and tables that you will clearly and quickly identify as such, you will most likely come across pieces of furniture that will be hard to classify; in other words, they will not fit in your typical images of sofas, chairs, and tables. This may make you think whether, for example, a hanging bead curtain should be treated as a door. In turn, when we consider a three-wheeler, with a motorcyclelike steering equipment and a top roof, we might wonder whether it is a car. Also, we may ponder the question of whether toy guns are just toys, or both toys and guns. To take another example, is a gun in the form of a cigarette lighter (á la James Bond) a gun or a cigarette lighter? On the classical, essentialist/objectivist view of the concept, you would be expected to know immediately whether something is a chair or whether something is a sofa. You would simply need to state whether the object in question possesses the features defining, for instance, chair. On the prototype view of the concept, however, you match the given nontypical object with your prototype, and you then classify it as an instance of the concept. When you face a nontypical example, you obviously may be in doubt; you may not want to clearly state whether a chair is being considered, you may hesitate, or, finally, you may say that you do not know or wish to state whether a chair is being dealt with. If pressed, you may say that it is (or it is not) a chair. You can obviously think that you did not mean to make a commitment, and a minute later you can announce that you changed your mind. That kind of procedure is normal if one views concepts as prototypes; it is not possible in the objectivist classical view.



Concepts (prototypes) are learned in the ontogenetic process. However, they are not learned once and for all. Although most concepts probably involve fairly stable images that do not change much in the course of a lifetime, many are perpetually redefined in light of incoming new experience. This process of redefinition implies perpetual changes in concept boundaries, which, as should be remembered, remain indiscrete all the time. What is fundamental in the process of concept formation and concept redefinition is experience. Concepts are experientially grounded (G. Lakoff, 1987; Lakoff & Johnson, 1980), which means that the way one understands the concept of a TV set, for example, will depend on the experience that one has had in one’s life with instances of what people have called a TV set in one’s presence, of what one has read about TV sets, what one has seen (in shops) labeled as TV sets, and so on. In that sense, concepts are experientially grounded. When one’s experience changes, that is when one is in some way exposed to new instances of TV sets, and to new labels perhaps, a redefinition of a concept takes place. The concepts (and the words for them) that one has at one’s disposal at a given point in one’s life are always a function of one’s experience.

SLIPPING INTO ESSENTIALISM? Some researchers (including some cognitive linguists) seem to be inconsistent in how they handle fuzziness. They are probably not inconsistent with reference to their own definition of inconsistency or, in the context of their own views, assumptions and reasoning, which always remain to some extent tacit. They seem to be inconsistent, however, with respect to the claim made in this book that all concepts are fuzzy and, generally, in view of the nonessentialist argument promoted in this book. What is very likely is that some researchers, being in principle nonessentialists, occasionally slip into essentialism without realizing this. Let us briefly discuss a few such cases that can be found in the cognitive linguistics literature and that can be viewed, possibly, as slippage. Case 1 Langacker (1983) talked about the naturalness of linguistic analysis. He claimed that the term is elusive and largely unexplicated: . . . the term natural is elusive and largely unexplicated . . . I regard a linguistic description as natural to the extent that it deals with the data in its own terms, with full regard for the richness, subtlety, and complexity characteristic of all linguistic phenomena . . . Assessing the naturalness of a description presup-



poses some preliminary conception of the phenomenon described, and this conception is bound to vary from one linguist to another and be subject to considerable disagreement. I gladly leave to others the philosophical question of how or whether such disagreements can be resolved. (p. 8; italics original)

Langacker’s reference to the largely unexplicated concept of natural is of interest to us here. If unexplicated is synonymous to imprecisely defined, it is possible to argue that natural is an unexplicated concept because all concepts are unexplicated. If my reasoning is correct, Langacker might be seen as joining the group of researchers who tend to complain about having to tackle fuzzy concepts. This conclusion about Langacker holds unless by unexplicated concept he meant something entirely different from fuzzy, which is also a possibility. Most likely, however, by unexplicated Langacker did mean fuzzy, as later in the passage said that he gladly leaves to others the philosophical question of how or whether disagreements about different conceptions of the phenomenon described can be resolved. The kinds of disagreements mentioned by Langacker can be resolved by taking recourse to the nonessentialist view of definitions. Different conceptions (conceptualizations) of the same phenomenon are unavoidable, and the discrepancies that exist among scholars should be considered normal, perhaps even welcome. Arguing for one conception would only be tantamount to arguing for one definition of some phenomenon, which is clearly an essentialist stance on the part of the investigator. Different conceptions of the same phenomenon should be considered feasible and welcome with respect to different research goals (e.g., practical goals) outside of possible descriptions of the phenomenon. For instance, conceptualizing language as behavior (rather than or in addition to as knowledge) seems indispensable for the external practical goal of applying linguistics to solving legal questions. The view advocated in this book is that different conceptions of the same phenomenon may be argued for or against only in connection with different analytical goals that are external to that phenomenon. Arguing for one definite goal, one definite conception, one definite analytical method, and so on will always push us back into essentialism.

Case 2 Langacker (1983) made the following claim on idealization in research: For an idealization to be truly revelatory, of course, the factors eliminated must in fact be extraneous, and the remaining phenomenon must, in fact,



have some degree of autonomy. Not all idealizations are revelatory, unfortunately; nothing prevents the analyst from occasionally going astray and carrying out a simplification that is pernicious in the sense that it significantly distorts the nature of the phenomenon under investigation. (p. 27)

The claim that an idealization can distort the nature of the phenomenon seems to be a slip into essentialism. This claim seems to imply that there is something like the nature (the essence) of language that a good conceptualization (idealization) should grasp. In my view, however, the linguist’s idealizations may only be close, or not, to what various individuals’ prototypes of language are. Consider, for instance, kinesic behavior and the question of whether language studies could or should include or exclude this type of behavior. In other words, only some linguists’ idealizations of language correspond to the prototypes of language as possessed by many people (Chomsky’s idealization of language does not). A natural idealization, however, seems to be impossible. Case 3 Regarding the distinctions among the subdisciplines of linguistics (phonology, morphology, semantics, pragmatics, etc.), G. Lakoff (1990) stated the following: “Of course, no a priori commitment is made as to whether these are separate sub fields. It is an empirical matter, and empirical considerations suggest that they are not—that, for example, generalizations about syntax depend on semantic and pragmatic considerations” (p. 40). It seems that Lakoff’s position is mistaken. The question of the distinctions of subdisciplines (or disciplines) does not appear to be an empirical matter. If we follow the nonessentialist reasoning advanced in the present book, we need to consider data as being, for example, semantic or pragmatic only in view of the definitions of semantics and pragmatics adopted. Otherwise, we will endlessly argue over whether the data we are considering are semantic or pragmatic. As is well known, there is no one definition of semantics, nor is there one definition of pragmatics. On the contrary, the list of definitions of the two notions is very long (see e.g., Levinson, 1983). It needs to be kept in mind that no ultimate, unquestionable definitions of semantics and pragmatics are possible. Only theoretically can we distinguish between semantics and pragmatics and treat the incoming empirical data accordingly. In summary, Lakoff’s formulation may be, perhaps unjustly, understood as indicating that some sort of natural semantic or pragmatic data can be collected. Were this interpretation of Lakoff’s passage correct, it could be taken as a slip into essentialism.



Case 4 Let us consider the following passage from the first introductory textbook in cognitive linguistics: Even for prime cases of apparently discrete homogeneous categories, such as odd, number, square, and kinship categories (mother, uncle, etc.) a prototype structure cannot be completely excluded. As experiments have shown . . . informants do in fact distinguish between good and bad examples of odd numbers and squares, and similar reactions can be assumed for mother . . . and other kinship terms. However, this only applies to “everyday” categorization. In a mathematical or scientific context the logical or classical view comes into its own. In such a context, number, square, or kinship categories can be established as clear-cut and homogeneous categories by an act of definition. In other words, the classical paradigm of categorization has a wide field of application wherever there is a need for precise and rigid definitions, as in the domain of scientific categorization or in the legal field. And there is no reason why the discrete categories of science and the everyday prototype categories should not coexist in the mental lexicon and even influence each other. (Ungerer & Schmid, 1996, p. 40; italics original)

In this passage, the authors said that in technical mathematical and logical language, precision can be attained. If given the benefit of the doubt, the authors may be thought to be saying that in mathematics and logic things are different to the extent that perception of the “everyday categorization” type is not involved. When the reader reaches the part of the passage that talks about science, scientific context, and scientific categorization, she or he may begin to suspect that Ungerer and Schmid believe that the spectrum of precise concepts is quite large. After all, the domain of science is large, independent of how narrow or broad a definition of science you apply. We must, however, put in serious doubt the claim that scientific language is precise language. Furthermore, when the authors refer to the legal field as an example of where precise language can be achieved, our suspicion that they slip into essentialism becomes ever stronger. Legal language is not precise, as opposed to what Ungerer and Schmid (1996) claimed. For instance, Gibbons (2003), von Mettenheim (1999), Tiersma (1999), and Shuy (1993, 1998) showed that even carefully written and meticulously edited legal documents are open to various interpretations. Legal documents use natural language concepts, which are all fuzzy to some degree. I therefore remain in disagreement with Ungerer and Schmid (1996), who said that not only the language (concepts) of logic and mathematics is precise (see Kitcher, 1984; Penrose, 1997; Popper, 1957, 1982, for comments supporting my view), but also that of science and law. In other words, Ungerer and Schmid (1996) seemed to be saying that some



concepts are fuzzy and others are not (e.g., those in mathematics and law). I follow Popper, Kitcher, and others and remain convinced that Ungerer and Schmid err in their judgment and slip into essentialism. Case 5 In her discussion of understanding words and prototypes, Aitchison (1994a) stated the following: However, these differing rankings raise the following serious questions. On what factors is the selection of a prototype “normally based”?: And if it can be altered by people’s preferences, what exactly are “prototypes”?. Such questions raise serious difficulties for prototype theory, and the answers are still hotly debated. (p. 91; italics added)

Further: “All these difficulties show that prototype theory, while solving some problems, raises considerable difficulties of its own, in particular, what exactly prototypes represent” (p. 93, Aitchison, 1994a; italics added). The reader can easily guess that the reason I invoke the prior quotations is the straightforward questions, “What exactly are prototypes?” and “What exactly do prototypes represent?” I take the latter question to be a paraphrase of the former, which Aitchison described as raising serious difficulties. In my view, “What exactly are prototypes?” is not a difficult question at all primarily because it is the wrong question and because it should not be considered to be answerable in one ultimate way. Prototype is like any other concept; it can be defined in different ways; it can be defined with varying degrees of precision; it can be defined differently depending on the purpose of an individual study; and it can never be defined in such way that all researchers concerned with it will remain in full and ultimate agreement. Is Aitchison, whose work on words and the mind clearly fits in the cognitive linguistics agenda, slipping into essentialism here?

ON FLEXIBILITY AND CREATIVITY The prototype view of the concept has significant implications for our understanding of how language functions. The prototype view allows us to understand immediately and treat as normal what the objectivists/essentialists avoid talking about or treat as a distortion that should be abandoned. Besides, and most important, the prototype theory of the concept allows us to understand flexibility and creativity in language use (see also Hudson, 1996). An important consequence of adopting the prototype theory of the concept is not only recognition, but also a potential exploita-



tion of the fact that specific instances of concepts may be partially members of one concept category and partially of others. For instance, when trying to categorize (assign to a concept category) some specific instance of cooking equipment, you are faced with a number of potential solutions. One may be that the “cooking thing” is classified as a pot. Another may be that it is classified as a nonpot (a saucepan, pan, skillet, etc.); still another is that you decide to attribute the object to at least two categories—for instance, a pot (because with respect to some features, it resembles the prototypical, for a given person, pot) and a pan (because with respect to other features, it resembles the prototypical, for a given person, again, pan). You may also hesitate—that is, at least one more possibility is that you hesitate as to how to categorize the cooking equipment in question (“What is it?,” “What shall I call it?”), and only later you decide to attribute it to one category/concept (e.g., a pot). There is also a third possibility, where there is room for flexibility. It is the possibility where partial attribution of an instance to two or more concept categories is being considered. Whenever the language user is faced with a nonprototypical instance of something (e.g., cooking equipment, writing equipment), it can be seen quite clearly that she or he is offered a significant amount of flexibility before she or he makes a decision as to the final classification. This kind of situation is not uncommon. The existence of this flexibility (whose degree may be perceived differently by different people) is one of the possible ways of accounting for the meaning discrepancies that often surface between language users. It should be kept in mind that these discrepancies occur not only across culturally remote communities, but also within small communities that share cultural patterns to a very high degree (see Tannen, 1986). Flexibility1 ties up closely with creativity. It is possible to claim that the former makes the latter possible. The two might also be viewed as two sides of the same coin. Still another possibility is that we refer to creative flexibility (Hudson, 1996). Under any circumstances, if you have choice of any sort, you can go one way or another. If the number of ways is not fixed, you can build your own. Facing the multitude of fuzzy cases that we are perpetually exposed to gives us the freedom to go different ways and to build our own. The prototype theory of the concept accounts for the fact that we are free to use language in a large number of conventional ways and in an infinite number of unconventionalized ways (i.e., “newly built” ways). We may want to visualize unconventionalized creativity as emerging with different intensity in different types of speech or writing. We may think of a continuum here. At one end there are all those cases of slightly “strange” uses; at the other end there are, for example, utterly incomprehensible pieces of modern poetry. To put it in a different way, we can say that flexibility and



creativity are made use of to varying degrees in the various types of spoken and written genres. It is not only genres, however, that lend themselves to varying treatments with respect to creativity. Obviously, people also differ tremendously as to whether they want to or are capable of making use of the freedom that the flexibility in question offers. Some people are keen to play with language with a fairly unpredictable outcome in sight. Others (mostly essentialists) treat language very seriously, and they meet any innovative unrestrained treatment of words and meanings with scorn or suspicion at best.

TREATING LANGUAGE “SERIOUSLY” Let us elaborate on the expressions “treating language seriously” as opposed to “treating it in an unrestrained manner.” Let us also keep in mind that essentialists tend to treat language seriously. If you believe that words have or should have one meaning, as essentialists tend to do, and if you believe that the meaning in question is, or should be, the original meaning, then you will try to use words as if they had only that very meaning. You will tend to be serious about words in the sense that you will try to use words as if they depicted that very single and original meaning that one believes is available. Essentialists, as a rule, are not terribly innovative linguistically. They tend to block new meanings. For essentialists, one of the few places where you can go to express your linguistic creativity is poetry. There you can make use of new unconventional metaphors and word meanings. In the essentialist view, metaphorical language is not considered valuable (or perhaps genuine) language. The real language is literal (precise) language, which essentialists seem to believe exists. Some essentialists tend to think that when we are outside of poetry or, generally, outside of art, we need to use words to mean what they really mean. Because most language use is beyond the domain of art, an essentialist is serious about language use most of the time. Essentialists tend to treat unrestrained language users (I take unrestrained to be the opposite of serious) as relatively unreliable people, probably also as comedians sometimes. For an essentialist, the nonessentialist’s relatively free, flexible, and creative use of language may be seen as irresponsible. This irresponsibility may cause some distress. When we read letters to newspapers, letters that complain about the misuse of language (see e.g., J. Milroy & L. Milroy, 1999), we can see quite clearly how essentialists are dissatisfied with being exposed to unrestrained language use, which, they claim, causes language decay.



ON UNCONVENTIONAL METAPHOR The prior discussion leads us to a consideration of metaphor, which constitutes a crucial aspect of cognitive linguistics inquiry. Cognitive linguistics has given metaphor a new treatment (note that “giving metaphor a new treatment” is an example of treating metaphors metaphorically). One of the main claims that the cognitive linguists make about metaphorical language is that it is present in most of the language that we use in our everyday interaction (see e.g., R.Gibbs, 1994; G. Lakoff & Johnson, 1980; G. Lakoff & Turner, 1989). This claim is coupled with the definition of metaphor that most cognitive linguists would accept—namely, that a metaphorical expression is an expression that talks about one thing (usually abstract, but not necessarily) in terms of another (usually concrete, physical). These two things come from what we normally conceive of as different domains. In the second to the last sentence of the previous paragraph, expressions are treated as people. The metaphor underlying “an expression that talks about one thing . . .” is Expressions are people. Many things that can be said about people can also be said about expressions. In such cases, cognitive linguists refer to mapping from one domain onto another (e.g., R. Gibbs, 1994; Lakoff & Johnson, 1980). One of these domains is called the source and the other the target domain. The source domain is the one that one usually knows well (usually the physical, concrete), the one that one maps, or imposes, on what is called the target domain, that is the domain that one knows less well (usually the abstract). In the example given before, people constitute the source domain and expressions the target domain. The topic of metaphor relates to our main argument about the essentialist misconception of language in a number of ways. Two questions, or aspects, of the cognitive linguistics treatment of metaphor are of special interest to us. One concerns the absence of literal meaning, and the other concerns open, unconventional metaphors. It is not necessary to dwell for long on the question of absence of literal meaning. Some linguists studying metaphor along the lines of cognitive linguistics (e.g., R. Gibbs, 1994) argued convincingly against the idea of literal meaning or literal language. These arguments translate easily into the claim advanced in the present book that there is no such a thing as precise (literal) meaning; there is no such a thing as essentialist, final, ultimately correct, indisputable meaning. As for unconventional metaphors, let us first invoke the distinction between conventional and unconventional (open) metaphors. Speakers of English constantly make use of what are called conventional metaphors. Consider time is money or ideas are people. The former refers to the idea that we normally think about time in terms of money or, to say it differently, as if time were money. We say, for instance, “I wasted a lot of time” ( just as “I wasted a lot of money”). We also say, “The short cut will save him two hours”



( just as “He managed to save twenty thousand dollars”); “I invested a lot of time in this book” ( just as “I invested a lot of money in my house”), and so on. Not only do we talk about time as if it were money, but we also explicitly express that underlying metaphor, and we often explicitly say in English, “Time is money.” It is often referred to as a saying. We do not normally say in English “Ideas are people” the way we do “Time is money.” We talk about ideas as if they were people—for instance, “His bright idea gave birth to a series of discussions on the topic,” or “Her idea is likely to live on for many years to come,” but the underlying metaphor—IDEAS ARE PEOPLE—does not get expressed explicitly in our daily use of English. In either case, that is, in the case of “Time is money” and “Ideas are people,” we deal with conventional mappings or, one might say, commonly rendered comparisons of what we know about the source domain (money, people) and about the target domain (time, ideas). The metaphors are conventional in the sense that we, as speakers of English, normally, regularly, and very frequently, put the two domains in question together. Nobody is surprised that we do it. This is because the mappings are conventional. Speakers of English have learned, as a fact of English, to put the two together. People have learned the convention that time is treated like money. Obviously, they have also learned the conventions that pertain to the typical, and possibly less typical, meanings of individual words and the connections among them. It needs to be remembered that when one produces a metaphor, one compares items that come from different semantic fields or, to put it differently, from different domains. When items are put together, the assumption is made that there is something that the two items share. It is usually a minor, although obvious characteristic. When one hears someone say, for example, “His new car is a bullet,” most probably the feature of speed will be immediately identified as that which motivates the comparison. Speed does not seem to be a critical feature of most modern cars. Other characteristics, such as the shape, size, color, fuel consumption, and comfort, seem to be much more important. Some cars are obviously faster than others, but given the speed limits in most countries and given that most cars run within a fairly narrow range of speed anyway, speed nowadays appears to be a minor characteristic. It is, nevertheless, an obvious characteristic of a car. In conventional metaphors such as “His new car is a bullet,” the comparison or connection invoked is easily understood, and so is the intended meaning of the whole metaphoric expression. In an unconventional metaphor, however, the comparison in question is novel, not usually made, possibly made for the first and only time. For example, in “He is like a frog,” “Her kiss was like cottage cheese,” and “His first idea was a cello concerto to me,” it is rather hard to make sense out of the examples. The connection that can be made between people and frogs



is very unclear. So is the comparison between a female kiss and cottage cheese. It is also difficult to find a feature (although minor) that puts ideas and cello concertos together. This task may be made even more difficult. Consider, “He looks like the front door of a town hall” or “The squirrel put me through to the wrong number” (about a telephone operator). There are several important facts about open metaphors that need to be stressed. First of all, such metaphors are actually used. The prior examples have been concocted, but, without doubt, such metaphors are really produced by some language users. Music magazines, such as Gramophone, provide numerous examples of unconventional metaphors, for instance, “Their sound could be described as bold, up-front, occasionally a little course, but above all, well, fruity . . .” (Kemp, 2000, p. 96), “the sugar turning into syrup with boneless wonders by Tavener, Messiaen, and Pärt” (Steane, 2000, p. 120), and “. . . including a full and bright ring to her tone” (Sadie, 2000 p. 130). Moreover, in my view, unconventional metaphors are usually produced by nonessentialists. This is because essentialists tend to believe that the connection between the source and the target domain is a natural one; that is, they seem to believe that the connection somehow resides in the words coming from the different domains. The nonessentialist, in contrast, believes that this connection is made by the human being, the speaker, and the hearer. The difference between these two positions is crucial, and it easily explains why essentialists generally dislike open metaphors and why nonessentialists like them. Namely, if one believes with the essentialist that the connection necessary for a metaphor to occur is resident in the words/items invoked, then she or he will also tend to believe that language users should rather not be flippant about what is outside of their realm of concern. However, one believes with the nonessentialist that the connection necessary for a metaphor to occur is made by the human being, then it will be very easy to see that the number and kinds of such connections is almost limitless. A relevant question emerging is whether new unconventional metaphors may be produced without any restrictions. On one hand, it might be said that language users are free to make any connection for any metaphorical expression, as they wish. On the other hand, there is the question of comprehensibility. If the connection is really far-fetched (as in the example, “The squirrel put me through to the wrong number”), one always risks not only being misunderstood, but also not being understood at all. If one is not concerned about not being understood, then she or he does not have to be concerned about using metaphors with far-fetched comparisons. Then, it must be realized, communication is likely to break down. Open metaphors may be treated as yet another reflection of the flexibility and creativity of language mentioned earlier. In the essentialist view, where the comparison (of items coming from two different domains) in-



volved in a metaphor is treated as natural, or fixed, the language user has no leeway, no latitude, and no linguistic freedom. Any distortion of, or deviation from, that natural comparison may be seen as unacceptable and may possibly be stigmatized. For a nonessentialist, however, using novel connections in unconventional metaphors is yet another way of expressing his or her general conviction that words do not have fixed meanings, that the original meaning of a word is by no means more important than its later semantic innovations, and that, finally, concepts represented in language by words have no essence. Unconventional metaphors are usually very hard to understand. The overwhelming majority of people want to be understood. The use of unconventional metaphors may thus be seen as irreconcilable with the fact that people normally want to be understood. However, even in cases of highly open metaphors, language users are most often able to come up with some connection between the items involved. This is exactly the outcome that needs to be counted on, on the part of one’s interlocutor. Obviously, whenever a person uses an open metaphor, she or he does make a connection (in his or her mind), and she or he obviously is unable to tell whether the same or a similar connection has been made by his or her interlocutor. That is part of the linguistic freedom that all language users have at their disposal. They may not want to make everything explicit. There is in fact a lot of evidence that, for various reasons, language users do not often want to make everything explicit and easily understood (see e.g., Bavelas, Black, Bryson, & Mullett, 1988; Bavelas, Black, Chovil, & Mullett, 1990; Brown & Levinson, 1987; Bull, 1998, 2000; Hamilton & Mineo, 1998; Janicki, 2002). It must be admitted that using open metaphors involves a measure of communicational danger. One danger is that one will not be understood at all. What may be even worse, however, is that the speaker will be grossly misunderstood. This danger always needs to be calculated when open metaphors are used. Misunderstanding is the price that may have to be paid for exercising the freedom of language use (for a more detailed discussion of misunderstanding, see the next section). When misunderstanding related to metaphor use occurs, the connection that one intends to evoke as the speaker will be different from that evoked in the hearer. That is similar to what happens when one, as the speaker, assigns to a word a meaning that appears to be different from that assigned by one’s interlocutor. If you say, for example, “Your cousin is a real turtle,” your interlocutor may think that the cousin is slow (the connection made concerns the speed with which the turtle moves), whereas you may have meant to say that the cousin is tough or difficult to break (the connection made concerns the hard shell of the turtle). In understanding open metaphors, the interactants’ background knowledge plays an important role. When you say “Your cousin is a real turtle,”



you assume that the interlocutor knows about turtles more or less what you know about them. In fact, it may turn out that one of the interactants knows much less, or much more, or that one of them knows different facts, compared with those that the other does. Given these possibilities, the connection that one of the interactants will make in the metaphor may be significantly different from that made by the other. People react to open metaphors in different ways. In Janicki (1997), reactions to open metaphors were studied. These reactions were collected in an interview. The list of metaphors studied included cases such as those in the following: 1. “ ‘Pulling out in a crisis and then wanting back in is the worst kind of signal to the Chinese.’ And when the crisis blows over, those 1.2 billion customers—and their heads, feet and gullets—will be wanting.” 2. “ ‘With guests so weird, so bizarre,’ says Young, ‘they can’t have a positive effect.’ She describes the American shows as a kind of multiple guest collision course between dented couples and their steamy stories piling up onstage.” 3. “There was a pause and another silence. Finally, the puzzled guest inquired, ‘The last of what exactly?’ Kohl sighed. ‘The last true European. The sense of weary struggle, of one very large man carrying an entire continent’s water, was everywhere in the room then, the visitor recalls.’ ” 4. “When Buchanan flashed his campaign’s 800 number last year on Crossfire, ‘there was no longer much difference between the talk-show world and the political battlefield,’ writes Howard Kurts . . . The media, once inmates in the political structure, had taken over the asylum.” 5. “The danger in this absorption of Hong Kong talents into the maws of Hollywood is homogenization. In the cuisine art of big studio production, will the wild flavors of Hong Kong survive?” (pp. 135–137). Interestingly, in many cases the respondents were unable to make any sense of the metaphor they were asked to interpret. Responses such as “I really don’t know what point they are trying to make here,” “I just don’t understand that at all,” “This is not clear to me,” and “I haven’t a clue” were frequent. On the whole, the respondents differed significantly in how they interpreted the metaphors. More important, although in some cases respondents gave similar general interpretations, when the interviewer discussed with them details of the metaphors, differences emerged; that is, there was more variation there in shades of meaning in each reading. Staying with the Janicki (1997) study, although the respondents’ primary task was to interpret the metaphors (and only secondarily to comment on



and evaluate them), most respondents preferred to evaluate the metaphors rather than interpret them. The respondents would often begin with an evaluation, and only after the interviewer’s persistent encouragement would they attempt to come up with an interpretation. Sometimes respondents provided an evaluation only. Some respondents really seized on the evaluative part of the exercise, despite that the interviewer reiterated the “meaning” and “unpacking” of the metaphor task before the interview began. The Janicki (1997) study also showed that there was something about open metaphors that irritated or put the informants slightly on the defensive. There were many comments about the manipulative nature of some of the metaphors and the editorializing they seemed to achieve. Most respondents thought this was inappropriate in a newsmagazine, from which the metaphors were taken. The irritation on the part of the informants may have been due to the fact that most informants were quite often in a predicament; that is, they could not interpret the metaphor in question. As is commonly known, most people are not willing to admit their fallibility and ignorance. This is perhaps why, to save face and not disclose what they may have thought to be their ignorance, respondents tried to say something. Despite having first admitted that they did not understand the metaphor, in some cases, respondents would proceed to offer some vague interpretation. In other cases, they went straight for evaluation, which may have been seen as a replacement of interpretation and, at the same time, a form of saying something. To paraphrase, the respondents may have felt that they were expected to know the one correct meaning of a metaphor. This brings us back to essentialism. It is assumed in Janicki (1997) that at least some of the respondents in the study were outright essentialists, or that they had a deeply rooted essentialist upbringing, of which they may have not been aware. It may be assumed that, as a byproduct of the essentialist frame of mind, at least some respondents felt they should be able to provide the one meaning of the metaphor apparently expected. Not being able to do so may have made them feel incompetent and ignorant. Therefore, they may have been dissatisfied and given the interviewer the impression of irritation. Given the prior discussion, the reader may assume that language users should avoid using open metaphors. It was not my intention to have the reader arrive at this conclusion. Although it must be admitted that open metaphors, with their high level of ambiguity, may be seen as a significant source of misunderstanding, although it must be admitted that we need to realize also that misunderstandings may lead to conflicts, and although it also must be admitted that caution must be taken and the risks considered whenever open metaphors are used, the positive aspects of the use of open metaphors should be emphasized. Open metaphors contribute to our cre-



ative use of language that, for many people, is a value in itself; open metaphors can be viewed as an expression of linguistic variation and freedom that, again, many people value highly; open metaphors are a spice of our linguistic behavior; open metaphors let us breathe linguistically with full strength. However, the potential danger of serious misunderstanding that open metaphors carry must be kept in mind.

ON MISUNDERSTANDING Misunderstanding, briefly referred to earlier in the context of open metaphors, is a phenomenon easy to understand once one resorts to the prototype theory of the concept. What follows from the view of concepts and categorization adopted in this book is the claim that significant misunderstanding between interlocutors is in fact a normal state of affairs. This is the opposite of what we can safely assume people usually believe—namely, that we normally understand (or should, in principle, understand) each other, and that we only occasionally do not understand each other. Let us elaborate on this question by turning things around. If language worked the way essentialists seem to think it does—that is, if concepts were precise and words had one correct meaning (one needs to keep in mind that essentialists tend to believe that meaning is in words)—understanding would be a very simple matter. In the essentialist view, one sends his or her meaning in words to his or her interlocutor and she or he picks up the words and takes the meaning out of the words (see Harris, 1981; Reddy, 1993). What is put into the word is the same as the interlocutor takes out of it. If language indeed worked that way, there would be no misunderstanding, provided, of course, that the words in question were used appropriately. It seems that essentialists tend to think that misunderstanding takes place when language users confuse things, when they have not learned the real meaning of the word, or when they are careless. However, when they have the proper language knowledge, and when they show language due respect, misunderstanding does not take place. In other words, for essentialists, misunderstanding is something extraordinary, and it may largely be avoided. This conclusion holds for all kinds of misunderstanding. That is, this conclusion pertains to the psycholinguistic situation in which the meaning intended by the speaker is different from that perceived by his or her interlocutor. Furthermore, this conclusion holds for the sort of misunderstanding that interlocutors are aware of, and possibly discuss openly as well as for the sort of misunderstanding that they may not realize at all, or that they may come to realize only in retrospect. Obviously, we most often do not know that misunderstanding has occurred; we have no evidence that it has occurred. However, the fact that we



are not aware of misunderstanding does not necessarily mean that misunderstanding does not create a problem. On the contrary, not being aware of misunderstanding may, in fact, make things worse for us because we may assume that we understand each other very well (or even fully), whereas, in fact, we do not. In other words, the problem is this: You assume that your partner in conversation understood you completely and you think and behave on the basis of this assumption. You do not realize that your assumption may have been wrong; that is, you do not realize that you may have been misunderstood to a greater or lesser degree. As a result of this lack of knowledge, in your daily life and activity subsequent to that conversation, your earlier conversational partner’s behavior may be difficult to interpret or seen as strange; you may be surprised that your partner says and does things for which you have no explanation. In this kind of scenario, you are never likely to find out that the strange behavior in question might, in fact, be traced back to the original conversation when you made the wrong assumption that you were understood the way you meant to be understood. You will then, very likely, think of all kinds of reasons for the person’s unexpected or strange behavior. You will speculate about whether she or he does not like you anymore, has changed, is ill, is unhappy, is envious of your success, has got some family problems, and so on. The simple reason for the unexpected behavior in question may be, however, the misunderstanding that occurred in the original conversation. But this fact will never be discovered and traced back to this original conversation. Unfortunately, there is nothing that can be done about the fact that we cannot know we have been misunderstood. We cannot really do anything about it. However, it is important for any language user to be aware of the fact that potentially we may be misunderstood all the time. This simple awareness may let us be more open-minded and help us understand other people’s behavior better. For some readers, the picture of misunderstanding presented earlier may appear very gloomy, and these readers might wonder how society can function at all. These readers are referred to later in text. There you can find a more optimistic version of the picture based on a discussion of shared experience and degrees of misunderstanding. To return to essentialism, my claim is that understanding the phenomenon of essentialism allows us to understand the phenomenon of misunderstanding. To consider things in a slightly different way, one might say that the nonessentialist view of language implies that misunderstanding occurs all the time regardless of whether interlocutors are aware of it. Within the broad rubric of nonessentialism, it is especially the prototype view of the concept and that of the experiential grounding of concepts that are found crucial for understanding how misunderstanding works. The reasoning is the following: If we make the assumption that concepts are experientially



grounded—that is, to put it in simpler terms, if we make the assumption that we learn our concepts (and the words that usually label them) in our daily experience (talking to people, playing, reading, working, eating, etc.)—then we have to admit that our concepts are all slightly or significantly different from those of other people. This is because there are no two people in this world who have had, throughout their lifetime, exactly the same experiences. Even twins who are brought up in the same family will always differ to some extent in what they do, where they go, how they are spoken to, and so on. In other words, the claim here is that even identical twins, who not only speak the same language but who also share an enormous amount of chemistry and experience (at least up to a certain point in their lives), misunderstand each other. If concepts are grounded in experience, if words reflect these concepts, and if people differ in experience (no matter how small the differences may be), then misunderstanding is a corollary. Lay people and scholars often talk about misunderstanding across cultures (see e.g., Bargiela-Chiappini & Harris, 1997; Yamada, 1997). Let us say, when a Chinese person talks to a French person, the likelihood is high that they will misunderstand each other even though they would be speaking their lingua franca (e.g., English) in a particular situation. This has been taken for granted for a long time. The claim I make here is, however, that even people who are close to each other culturally to the highest extent possible (e.g., twins) cannot avoid misunderstanding. This is what experiential grounding of concepts may be taken to mean with respect to misunderstanding. Obviously, the idea of prototype, in terms of which concepts seem to be best understood, contributes to the present reasoning about the ubiquity of misunderstanding. If one thinks of some of the issues discussed earlier in this book, such as fuzziness, flexibility, creativity, imprecision, and so on, then it will be still easier to accept the position that misunderstanding cannot be avoided. It needs to be stressed that the idea of experiential grounding of concepts is absolutely crucial for the understanding of misunderstanding. Here individual experience implies individual concepts, which in turn imply individual ways of comprehending words and the connections among them. This reasoning is fully consonant with the reasoning advanced by Tannen (1986, 1990, 1995). How can we function so well when there is so much misunderstanding? In other words, one major reservation might be brought up against the prior argument—namely, if the argument is correct, then, it might be supposed, there would, as a consequence, be total chaos as far as the use of language is concerned. We simply would not know what others are talking about. We could not get things done. We would not understand the newspapers, legal documents, our schoolteachers, employers, colleagues, and our fathers and mothers.



This is certainly not the case. Our linguistic interaction in fact usually appears to proceed rather smoothly. We do, indeed, understand most of the people around us with whom we have to deal in our daily affairs. There is no general chaos. Most of the time we do understand what our morning paper says, what our boss wants from us, how our children want to get money from us to buy toys, and so on. On the other hand, however, there are language problems often leading to social friction. Language, and especially the phenomenon of misunderstanding, is in practice more of a problem (worldwide) than most people are aware of or are willing to acknowledge. Admittedly, most people do, in fact, communicate quite efficiently, despite that misunderstanding is ubiquitous. To understand this point fully, we need to return to the question of experience. On one hand, we have to remember that each individual’s experience is, obviously, idiosyncratic. On the other hand, however, people’s experience is similar, sometimes, as in the case of brothers and sisters, significantly similar. Similarity of experience means that many language users have seen more or less the same chairs, cabinets, fields, cows, horses, cars, houses, tables, and so on. These same people have also had, one might surmise, more or less the same encounters with phenomena that people call pity, love, greed, insolence, sorrow, happiness, and so on. When these same people were children, their parents called more or less (the qualification “more or less” must be remembered all the time!) the same things toys, diapers, cots, cradles, teddy bears, rattles, and so on. It is this similarity of experience that accounts for the fact that people understand each other relatively well. It must be stressed, however, that people never understand each other completely, “down to the last drop,” as it were. After all, the toys the individual people played with as children and their attitudes toward them were, in fact, different, so each person may have a slightly different understanding of what is included in or excluded from the concept of a toy. The fact that we do not fully understand even the people closest to us culturally may not really be sociolinguistically important, but we should be aware of that fact (for a related analysis of cultural conceptualizations and understanding, see Sharifian, 2003). The reasoning presented herein generates the view that the degree of misunderstanding depends on the cultural proximity or distance between the interlocutors. The further away culturally one is from one’s interlocutor, the higher the degree of potential misunderstanding. More important, from the theoretical point of view, misunderstanding never disappears; it never reaches point zero. This means, among other things, that we cannot rule out that even in cases when we talk to people we know extremely well, misunderstanding (both overt and covert) may occur. From the practical point of view, however, cases that are close to the zero value may be totally negligible. The similarity of experience explains the fact that, on the whole, there is no linguistic chaos around us; that we usually feel we understand



our brothers and sisters, teachers, colleagues, and so on. Because our experiential differences, which underlie concepts, are often trivial for linguistic purposes, so may be the misunderstanding that occurs. Disregarding all those cases where misunderstanding is minute, insignificant, or trivial, misunderstanding may be a serious sociolinguistic problem. The nontrivial cases of misunderstanding are abundant, and that is exactly why it is important and useful to know why it occurs and how it functions. Misunderstanding may ultimately lead to conflict. It is my contention that by making people aware of why and how misunderstanding arises, we may be able to prevent at least some conflicts, which would be a very worthwhile achievement. It is quite obvious that misunderstanding is a very common phenomenon. Evidence for this claim can be collected in daily interaction without having to resort to any sophisticated methodological tools. We often hear people say, “Oh, I didn’t mean to say that,” “That’s not what I meant,” and so on. These are reactions to situations in which misunderstanding is overt. There are also frequent cases when people appear to semiunderstand what others around them say. Once one is aware of the ubiquity of misunderstanding, it is easy to spot and record all those everyday situations where one feels that conversation does not proceed smoothly, where there is doubt about what interlocutors mean. One will quickly feel or see quite clearly that problems are common indeed. Misunderstanding in the academic environment, where many people probably expect entirely clear messages, constitutes a good example of how rampant the phenomenon is. Karl Popper, who appears to many to be one of the easiest philosophers of science to read, paid a lot of attention to clarity in writing. He maintained that obscurity is intolerable (Popper, 1976, 1983). One might say that an easy writing style was one of his guiding principles. Despite all this, he claimed to have been misunderstood significantly. He asserted that most of the criticism raised against his philosophy is due to misunderstanding (Popper, 1983). Another prominent example is Ernest Mach, who is said to have been grossly misunderstood (Feyerabend, 1981a). One should not be surprised about this state of affairs. The same is true for the field of linguistics. Misunderstanding is rife, and linguists often complain about it. This becomes very clear when one goes to a linguistics conference and concentrates exclusively on the phenomena of misunderstanding and nonunderstanding. Conference participants often apologize for not being understood or for being misunderstood. During a discussion session, after a paper has been presented, speakers frequently do not understand some of the questions coming from the floor. The session participants who asked questions in the discussion session will often complain about understanding problems on the part of the session speaker and/or



the session participants. This is not surprising. If we try hard, if we care (as speakers), we may lower the level of misunderstanding to a significant degree, but we can never eliminate it completely. A significant role in misunderstanding is played by axiology—that is, our value judgments. As Krzeszowski (1991) rightly pointed out, the language user always assigns an axiological load—that is, some kind of value—to concepts. The fundamental axiological scale that we operate with daily is that of good–bad. Many concepts are axiologically marked in this respect either as positive (good) or negative (bad). Concepts such as killer, thief, illness, death, fail are, by most people, assigned a negative axiological load. As opposed to these, concepts such as victory, recovery, happiness, and help are by most people assigned a positive axiological load. The overwhelming majority of concepts remain axiologically neutral—that is, they are not thought of as either (sort of) bad or (sort of) good. Such concepts are, for instance, book, tape, paint, question, look, and so on. The formulation, “The language user assigns an axiological load to concepts,” was meant precisely that way. This formulation should be taken to be reminiscent of the earlier claim that meaning is assigned to words by people. When we are concerned with axiology, the situation is similar. The axiological load of a word is not in it; the load is assigned to words by language users. This is where the individual human factor comes in again. Here is where the similarity and difference of human experience enters the picture. Again here, to the extent that our experience is similar, we share the values that we normally associate with concepts. We also differ in experience, and the values that we assign to concepts may differ. An interesting and important fact is that language users normally do not know, or may make incorrect assumptions as to where the other party stands axiologically with respect to the concepts used in interaction. Of course it would not be feasible to check every time an axiologically marked concept is used whether one party’s assumptions (made mainly unconsciously) are similar to or different from those of the other party. This is how axiology could be another source of misunderstanding in linguistic interaction. We now need to elaborate the connection between the assignment of an axiological load to concepts, misunderstanding, and essentialism. Essentialists tend to believe that value is in words (concepts). This reasoning results in the view that the value a particular language user perceives as belonging to a particular concept is the value. It is thought that the value must be shared by everyone. So, if one is an essentialist, and if she or he thought that, for instance, cheating was bad (as many people do, of course) then she or he would tend to believe that cheating was inherently bad. In other words, she or he would believe that cheating is perceived as bad by everybody.



Cheating is indeed usually perceived as bad, but certainly not always. There are people who consider that there is nothing bad about cheating. Cheating may be their way of life. It might be axiologically neutral or, in fact, quite positive. The essentialist will reason that cheating is intrinsically bad, and that the people who think that cheating is neutral or good are simply wrong in some sort of absolute sense. The claim supported in this book is, however, that cheating is not inherently wrong. The value “bad” is not in the concept. It is assigned (by most people, it needs to be admitted) to the concept. This particular value is assigned by most speakers of English, but definitely not all. Let us now bring misunderstanding to the fore. Cheating is an example of a concept that is, by most speakers of English, assigned the value of bad. For all practical purposes, we may disregard the few people who assign the values “neutral” and “good” to it. In the case of “cheating,” we thus understand each other quite well. There is hardly ever any significant misunderstanding concerning this concept (its axiology, to be more exact). We deal with misunderstanding (nontrivial, overt, or covert) primarily in all those cases when we tackle concepts that are axiologically less salient than our example of cheating. When we consider concepts such as revolutionary, strike, prayer, preach, rain, entertainment, and decadent, the situation is much less clear. Here it is not so easy to say what is usual or typical with respect to axiology. We may safely assume that the axiological load in question may differ, not only from social group to social group and from person to person, but that it may also move back and forth along the bad–good continuum as the various examples of those concepts (being in, close to, or far away from the core) are tackled by the individual language user. In fact, people differ significantly with respect to the axiological load that they impute to concepts. We always make assumptions as to the axiological system that our interlocutor uses. Rarely, however, does the axiological load get explicitly expressed. Most of the time we make only quick guesses as the conversation proceeds. Whether our guesses are correct, or even, more or less correct, we never know. It follows that axiology can be treated as yet another source of misunderstanding. It should be stressed that when one adds the potential misunderstanding that springs from the imprecision/fuzziness of concepts to that springing from the freedom that language users have in assigning an axiological load to concepts, there arises much potential misunderstanding. It should have become clear to the reader by now that preventing misunderstanding altogether is not possible. It is possible, however, to lower the level of misunderstanding. A simple strategy may be used to achieve this goal. This strategy consists in a regular use of low abstraction level words. We need to climb down the ladder of abstraction as often as possible (Hayakawa, 1965). For example, when the word land is used in, for exam-



ple, “Robert owns a lot of land,” one is low on the ladder of abstraction because land refers to an easily imaginable part of the earth, and, for the interlocutor, for all practical purposes, there will be no doubt, or very little doubt, as to what is meant, and, thus, most probably, there will be no significant misunderstanding. When, in turn, the word property is used in, for example, “Robert has a lot of property,” we have climbed up one step higher up on the ladder of abstraction. Then one’s interlocutor does not know whether Robert owns land, houses, or lakes, perhaps. When, in turn, the word wealth is used in, for example, “Robert has a lot of wealth,” we have moved a further step up on the ladder of abstraction. One’s interlocutor does not know whether Robert has land, houses, lakes, cars, jewelry, or cash. The level of ambiguity has been raised. If we continue this exercise, we gradually move up the ladder of abstraction. Thus, one chain of possibilities is land, property, or wealth. More important, the higher one is on the ladder of abstraction, the broader the scope of reference of the words is used. The broader the scope of reference, the greater the likelihood of misunderstanding. When someone says “Robert is wealthy,” it can be guessed that Robert has money, houses, and jewelry perhaps. But it must be admitted that one can only guess because it might as well be cars, gold, castles, factories, shares, cows, and so on. So we must agree that the word wealthy may mean all kinds of things. When we say, however, “Robert has three houses and two cars,” the level of abstraction is lowered, and one knows better what is meant by wealthy. By lowering the level of abstraction (compared to “Robert is wealthy”), the degree of potential misunderstanding is also lowered. The previous examples show that when one says, “Robert has a lot of land,” the likelihood that she or he will be misunderstood is very low. When the word property is used, this likelihood increases. When wealth is used, one can see quite clearly that land may be only one of the things that are meant. Other types of property may of course be meant, in fact. At this level (wealth), what may be meant is movable assets such as sailboats, antique furniture, and possibly other things. In other words, the higher up one is on the ladder, the more things one may mean using a particular word. The lower down one climbs, the easier it will be to point to something fairly specific. People who talk all the time about substances, entities, realities, abstraction, rationalization, spirit, soul, and so on are very high up on the ladder of abstraction. This is what happens to many philosophers. They are very often misunderstood or not understood at all. Neither they nor their readers or listeners know whether the latter understand the former, or to what extent, or how the latter misunderstand the former. The writer or speaker operates high up on the ladder of abstraction, and only she or he knows which things she or he is referring to (even this optimistic view may be often questioned). The reader or listener may detect the things referred to by the



writer/speaker, perhaps, but she or he, in his or her own system of reference, might as well detect many other things instead. It is not only (or not mainly) philosophers who are often high up on the ladder of abstraction. Many linguists are as well. That is exactly why there is so much nonunderstanding of and misunderstanding about linguistics texts. The higher up the linguist is on the ladder of abstraction, the more difficult it is for others to know what she or he means. It is worthwhile to compare, for example, texts on transformational grammar with those on sociolinguistics. It seems generally easier to understand texts that treat the latter. This is primarily because sociolinguists often talk about people, countries, cities, sounds, and so on. People can be seen or easily imagined, cities can be visited, and sounds can be heard. Transformational grammarians, in contrast, talk about ideal native speakers, a language-acquisition device, feeding order, labeled bracketing, multiple-branching constructions, and so on. It is certainly difficult to imagine an ideal native speaker or a language acquisition device. The point about abstraction should not be exaggerated. Naturally, we must abstract, and we do abstract all the time. To put it in different terms, we always idealize or accept different levels of idealization. This process of idealization is especially evident in academic/scientific discourse. However, staying high up on the abstraction ladder all the time contributes significantly to misunderstanding and nonunderstanding. It is our duty in all kinds of discourse (including that on language) to walk down the ladder as often as possible. Otherwise, when we are high up on the ladder, we may start talking nonsense. The prior argument should by no means be taken to be an argument against abstract thinking and generalizations. This argument was not meant to raise negative feelings about high-level abstraction and generalizations at all. Constant activity on one of the low steps of the ladder of abstraction may be a problem as well. As we know, many people use mainly words at a very low level of abstraction. They can talk chiefly about babies, kitchens, food, houses, cars, computers, and so on. Such people find it hard to think abstractly. They feel safest when they can talk about the things they can see, touch, or taste. However, operating constantly at the highest step of the ladder of abstraction is not good either. One is then very likely to be frequently misunderstood. Some tasks require the use of highly abstract language and others do not. Scientific tasks obviously require the use of highly abstract language. It is crucial, however, that we be able and willing always to go down the ladder of abstraction (and provide tangible exemplification) whenever any trace of misunderstanding comes to the fore. The fact that we cannot rule out misunderstanding altogether does not at all mean that we should not try to reduce its level to the minimum. Coming down the ladder of abstraction is a good means to reduce this level.



There is an additional point to keep in mind: In a nonessentialist cognitive view, intelligibility and misunderstanding are all notions that should be associated with language users and not with words, sentences, linguistic expressions, varieties, or, generally, with language as abstracted from these users. Meaning is not in words, nor is understanding. Words or sentences are understood, misunderstood, or not understood by real people (see chap. 7, this volume, for an illustration of this point). As Hudson (1996) remarked, “Mutual intelligibility is not really a relation between varieties, but between people, since it is they, and not the varieties, that understand one another” (p. 35). Moreover, one should not forget that understanding should be viewed as a cline; that is, that understanding is not a question of all or nothing. It is also Hudson (1996) who rightly pointed out “Mutual intelligibility is a matter of degree, ranging from total intelligibility down to total unintelligibility” (p. 35). In summary, if one accepts (after the cognitive linguist) the view that concepts are fuzzy; if one accepts the view that meaning is assigned to words by language users; if one accepts the view that the meaning assigned by language users is a function of the language users’ personal (individual, idiosyncratic) and cultural (to varying degrees shared with others, but never shared completely) experience; if one accepts the view that concepts are best understood as prototypes; and if one accepts the view that an axiological load is assigned to concepts by language users, then misunderstanding should be viewed as unavoidable. By climbing down the ladder of abstraction all the way to the ground (i.e., by using low-abstraction words in examples), language users may successfully reduce misunderstanding, although it can never be eliminated altogether.

CONCLUSION The general conclusion that emerges from the prior discussion is that language is misconceived by many people, including some linguists. It is misconceived in the sense that many people think language works the essentialist way, whereas in fact it does not. Most specifically, many people think that concepts, which words, sentences, and utterances invoke, are definable in clear terms, whereas they are in fact fuzzy,2 although to varying degrees. Many people think that meaning resides or acts as if it resided in words, whereas in fact we, the language users, impute meaning to words. The consequences of misconceiving language the essentialist way may be far reaching. In linguistics as an academic field, essentialism may lead to confusion, frustration, and a lack of progress. In other domains (e.g., in politics and education), it may lead to conflict. It is to the discussion of the latter domains that we now turn.



ENDNOTES 1. For an excellent, inspiring, and convincing sociological discussion of flexibility as an alternative to extreme rigidity and extreme fuzziness, see Zerubavel (1993). Zerubavel’s notion of flexible mind ties up very nicely with the cognitive linguistics notion of experientialism. 2. For a very recent publication treating categorization, discreteness, and fuzziness, see Aarts, Denison, Keizer, and Popova (2004).




The argument presented in Part I of this book has a clear application to everyday life. Most important, understanding conceptual essentialism may help us understand and, possibly, alleviate conflict occurring in all walks of life. Essentialism and conflict are closely related. The question of the relationship between essentialism and conflict could be understood as a question of what kind of practical consequences can be brought about by the doctrine of essentialism. In what follows, one major assumption has been made—namely, that conflict may spring from misunderstanding and disagreements about meaning. This does not mean, of course, that all conflicts are languageridden conflicts. I claim, however, that language plays a significant role in conflicts or, to put it differently, that many conflicts are language-related. These conflicts are language-related in the sense that some people’s general beliefs about language, and some specific beliefs (e.g., beliefs concerning the ways in which meaning is generated) are conducive to conflict. More specifically, I claim in the present part of the book that essentialist beliefs concerning the ways in which meaning is generated are conducive to conflict. A related claim is, as expected, that nonessentialism is a more people-friendly doctrine, and that conflicts are less likely to arise among language users who have a nonessentialist frame of mind. It is crucial to note at this point that the reasoning about conflict given earlier cannot lead us to the essentialist groove, as some readers might think. These readers might get the impression that essentialists are meant to be presented as the bad people and the nonessentialists as the good ones. These readers might get the impression that what is being presented 83



here is a black-and-white picture. These readers might also get the impression that essentialists are meant to be presented as villains, and that essentialism leads to conflict and nonessentialism does not. It is not my intention at all to advance this somewhat simplistic picture of the state of affairs. Things are in fact more complicated. First, it should not be believed that we can make an ultimately clear distinction between an essentialist person and a nonessentialist one. If it were believed that we can, the position developed in this book would in fact be being undermined. It must be remembered then that essentialist is like any other concept; it is fuzzy. Similar to other concepts, we will be able to find, easily, typical cases (or prototypical cases) of an essentialist person as well as nontypical ones. The further we move toward the periphery of the concept boundaries, the more we may be in doubt as to whether we are dealing with an essentialist. The same reasoning obviously applies to a nonessentialist. It needs to be realized, then, that we encounter typical essentialists on one hand, and typical nonessentialists, on the other hand. In between, we experience many examples of nontypical cases of both the former and the latter. In other words, there exist not only essentialists and nonessentialists; there are typical essentialists and typical nonessentialists, and there are also nontypical essentialists (leaning on the nonessentialist side) and nontypical nonessentialists (leaning on the essentialist side). Second, I do not mean to claim here, in a deterministic fashion as it were, that essentialists (even if one thinks only of the typical cases) are all dangerous people; that is, I do not mean to claim here that if one is an essentialist then one is eo ipso a conflict-engendering person. All that is being claimed here is that the essentialist frame of mind is conducive to conflict. The main reason for this conclusion is that conceptual essentialism understood as the belief in the ultimate definability of concepts/words implies the belief that one’s own understanding of concepts is correct, as opposed to someone else’s understanding of the same concepts, which is viewed as incorrect or inappropriate. Within the essentialist frame of mind, it appears natural to believe one of two things: (a) that the real, good, ultimate, indisputable definition of a concept/word is somewhere to be found; this definition is in principle available; someone might know it, or someone is going to be able to find it out one day. This is the less dangerous position; or (b) that one’s own definition of the concept/word in question is the right one; other people’s definitions are wrong; these other people should learn the correct (i.e., one’s own) definition. This is the more dangerous position, as shown later in the book. It must be emphasized that a major claim in this book is that if one thinks one’s understanding of a concept is the right one, one will then tend to impose that way of understanding the concept on others. This is true in politics, for example. If a politician believes that she or he knows what justice, po-



litical dependence, cooperation, allegation, freedom, negotiation, and so on, really are, she or he is very likely to impose his or her way of thinking on others, with the effect that his or her opponents in political disputes (who will most likely have at least a slightly different understanding of the concepts in question) may be perceived as ignorant, strangely different, perhaps cunning, or largely irresponsible. In other words, our politician will tend to press the view, either overtly or covertly, that she or he is right and others are wrong. The thinking is in terms of right or wrong, plus or minus, black or white. This dichotomous way of looking at things certainly invokes the Aristotelian, two-valued orientation (see Hayakawa, 1965). We are thus brought to a further, related contention that essentialism may lead to a host of disagreeable phenomena, the most salient of which are dogmatism, unwarranted feelings of certainty, dislike of criticism, a sense of infallibility, conceitedness, intolerance, authoritarian argumentation, disregard of other people’s beliefs and opinions, and a sense of power. Most important, a combination of these traits is very likely to lead to conflict. The reasoning adumbrated earlier may thus lead us to the view that many conflicts can be thought of as language-related. This is because language permeates almost all human activities. One’s philosophical beliefs are expressed, to a significant degree, through language. It is difficult to get anything done without using language. Crucially then, what one believes about concepts, language, words, meaning, and so on is fundamentally important for how one deals with other people, how one approaches institutions, and how one perceives all kinds of activities. A question that the reader might want to pose at this point is how one can tell language-related conflicts from others. In other words, the question may be about how we know that the case of a conflict that one is describing is indeed a language-related conflict. Naturally, being faithful to the nonessential philosophy, we cannot give a fully satisfactory and final definition of a language-related conflict. What can be given is only a definition. Under any circumstances, it will only be a working definition (not some real, ultimate definition). A possible definition is that a language-related conflict is the kind of conflict whose roots go back to the language user’s essentialist beliefs. It is a philosophically induced conflict where beliefs about and the use of language play a very significant role. In other words, in a languagerelated conflict, people’s essentialist philosophical beliefs are directly reflected in their beliefs about and use of language. It should be noted that we can never say for sure that a conflict is, in fact, language-related. We can only believe/guess, suppose, or have strong evidence that it is languagerelated relative to the working definition that we use. We should use the label language-related conflict loosely. We will then know, roughly, what is meant by such a conflict. That is all we can have and all we need. Obviously, conflicts may also be traced to more than one source. In such a case,



essentialist beliefs about language may be thought of only as one of the many reasons for a conflict. We can view language-related conflicts in many different ways. One useful way is perhaps to think of these conflicts as placeable along the micro–macro scale—that is we may view some language-related conflicts as small scale or involving only small groups of individuals, whereas other conflicts can be viewed as involving larger groups of people and large-scale phenomena. The beginning of the scale can be exemplified by conflicts in the family (e.g., conflicts between married couples; see Tannen, 1986), where the conceptual differences between members of the community are the smallest. The remaining section of the scale may be exemplified by conflicts involving speakers of standard versus nonstandard dialects in all kinds of contact situations, such as the school (where standard dialect-speaking teachers interact with nonstandard dialect-speaking pupils; see Bex & Watts, 1999) and the ordinary workplace (see Tannen, 1995). This remaining section of the scale may also be exemplified by conflicts that involve two languages. In the last category, we may invoke the conflict pertaining to Kurdish and Turkish in Turkey; Kurdish is reported to have only recently been allowed for public use in Turkey. So we may view all language-related conflicts as conflicts of a different caliber, as it were, with significantly different social consequences for the parties involved. The conflicts mentioned previously do not only differ in terms of quantity or scale. They are also qualitatively different. They are not only microor macroconflicts (understood in terms of the number of people involved). They are, indeed, different as to their possible assignments to domains. For example, the Kurdish–Turkish conflict may be seen as primarily political, the school conflicts may be seen as primarily educational/ sociolinguistic, whereas the family examples may be seen as primarily private. All these conflicts involve language in some sense or other. Once we have agreed about that, we may say that these conflicts are of a different type, magnitude, or scale. In chapters 5 to 8, where we discuss conflicts in the political, legal, academic, and educational domains, it is contended that many participants in the activities pertaining to these domains exhibit essentialist behavior, which can most probably be traced to their deeply entrenched essentialist beliefs. Whether they indeed are conceptual essentialists by philosophical conviction or whether they only act as if they were conceptual essentialists, without being aware of it at all, is not possible for us to determine. Evidence for this contention is provided when several specific cases pertaining to the four domains are discussed in some detail. The conflicts described in chapters 5 to 8 do not come as a surprise, nor does our disclosure of many essentialists featuring in these cases. In addition to Medin and Ortony’s (1989) claim about the existence of psychologi-



cal essentialism, we find early independent statements that many people are essentialists; that is, they believe their understanding of concepts is the right one and that of others is wrong. For example, Hayakawa (1953) said: “We all tend to believe that the way we use words is the correct way, and that people who use the same words in other ways are either ignorant or dishonest” (p. 7). Also, much more recently, Tannen (1986), indirectly supported my view on the existence of essentialists. With respect to the concept of politeness, she said: “People instinctively feel that their ways of expressing things and of being polite or rude are natural and logical” (p. 188). Tannen’s work in general (see especially 1986, 1990, 1995, 1998) largely supports the argument advanced in the present book; Tannen’s work can be largely interpreted as indicating that disagreements and arguments between wives and husbands, parents and children, siblings (i.e., disagreements and arguments within a microcommunity), politicians, lawyers, and colleagues at work may be traced to people’s belief in having access to the only true interpretation of concepts and to the one true meaning of words rather than to overwork, lack of love, disrespect, sense of law, camaraderie, and so on. Handling silence is an example in point (see e.g., Tannen, 1990). In family conflicts, silence occurs frequently. Silence is not unambiguous, as many people think. One’s refraining from speech is for some people interpreted as a kind and conciliatory move; for others it is interpreted as unkind and belligerent behavior. Many of us tend to think, however, that the way we interpret the meaning of silence is correct. Recapitulating, it is reasonable to claim that many people feel that their own language, discourse strategies, customs, habits, ways of communicating, and, most important, ways of viewing concepts and meanings appear to be natural. Natural should translate here to be the one universally valid way. Evidence for this belief in the one natural way can be found around us in everyday life and in many professional settings. I trust I am allowed to make this claim on the basis of my own extensive observations, on the basis of what I hear people say, and on the basis of what I know they do and write. The additional evidence provided in chapters 5 to 8 should be significant and convincing to most readers and, hopefully, overwhelming to others.



Language in Politics

None of the conflicts discussed herein should be conceived of as strictly political. For instance, the Clinton–Lewinsky conflict may be seen as partly political and partly legal. The stem-cell research issue may, in turn, be seen as partly political and partly medical. It should have become clear to the reader by now that arguing for a conflict to be labeled as strictly political, medical, legal, and so on would in fact be getting oneself into the essentialist frame of mind. Political, legal, educational, and so on are only convenient labels that refer to cases that people normally group together. We should remember to treat these labels only as such.

THE CLINTON–LEWINSKY CASE As is well known to the general public, in 1998, there erupted a scandal featuring the president of the United States, Bill Clinton, and his intern, Monica Lewinsky. The president and the intern had been involved in an affair that was disclosed to the public. The president was accused of having had a sexual relationship with Lewinsky. He officially rejected the charge, stating explicitly, among other things, that he had never had a sexual relationship with the intern. “I did not have sexual relations with that woman” is a well-known and often quoted statement that Clinton made during his Jones deposition. Shortly after, it became very clear that Clinton had had oral sex with Lewinsky, which, according to Clinton, did not count as having had a sexual relationship, which, he still claimed, he had not had with Lewinsky. 89



The case became rather complicated as the evidence emerged. By insisting that he had not had a sexual relationship with Lewinsky, Clinton was accused of having lied. The argument against Clinton was that having had oral sex (which became evident at a certain point) equals having had a sexual relationship, a contention fully supported by the prosecuting side and vehemently rejected by the defending side. Ultimately, the question was raised whether the president should be impeached on the charge of perjury or, possibly, other crimes and misdemeanors. A conflict around the case erupted (for details, see also R. T. Lakoff, 2000; Morgan, 2001; Solan, 2002). There are two significant facts for the reader of this book to know about with respect to this case. One is that the legal argument focused around the definition of “having a sexual relationship,” which, crucially, may be seen as an argument about language. The other is that, to the best of my knowledge, no language expert was called on to express his or her opinion on the matter. The latter fact is significant, particularly as other kinds of experts (e.g., political scientists, lawyers) were present on numerous panels (e.g., CNN panels) organized at that time. As it became evident in course of the case, the prosecuting party wanted to come through with and impose their own definition of having sex, including oral sex. The defending party wanted to come through with and impose theirs—a definition of having sex excluding the oral touching of the sexual organs. So, there ensued a fight about definitions. As TIME magazine reported, At Clinton’s deposition, Jones’ legal team asked Judge Susan Webber Wright to approve a very precise, three-part definition of sexual relations. Clinton’s attorney, Robert Bennet, objected to the whole definition, but to the last two parts especially, as being too broad. Wright agreed to disallow parts 2 and 3, leaving only the first, narrowest definition of sex in place. (Lacayo, 1998, p. 16)

The three-part definition in question was the following: DEPOSITION OF WILLIAM JEFFERSON CLINTON Definition of Sexual Relations For the purposes of this deposition, a person engages in “sexual relations” when the person knowingly engages in or causes— 1. contact with the genitalia, anus, groin, breast, inner thigh, or buttocks of any person with an intent to arouse or gratify the sexual desire of any person; 2. contact between any part of the person’s body or an object and the genitals or anus of another person; or



3. contact between the genitals or anus of the person and any part of another person’s body. Contact means intentional touching, either directly or through clothing. (Lacayo, 1998, p. 16)

As TIME magazine further reported, With that, Clinton may have been given the room to offer a technically “true” denial to the question of whether he had sex with Lewinsky—even if she happened to perform fellatio on him. The truncated definition characterizes sex in terms of a checklist of body parts, including the genitals, breast, and thigh, Oral sex would not necessarily require the President to touch anything on Lewinsky that appears on that list. Strange as it may sound, under one reading of the definition, Lewinsky could have been having sex with him (because she was “touching the President’s genitals”) while at the same moment, he was not having sex with her. (At the deposition, Clinton wasn’t asked if she had sexual relations with him; just if he had them with her.) There are problems with the legalistic defense. For one thing, if Clinton and Lewinsky did have oral sex, is it really likely that he did not touch any body parts mentioned in the Jones definition? (Lewinsky testified that Clinton fondled her). And because that definition says that a person engages in sex if he or she “causes” contact with the genitals of “any person,” it could be argued that Clinton caused Lewinsky’s contact with his, even if he did not otherwise touch her. He could reply that she was the cause, or at least the active partner while he was merely the passive receiver, but that makes him seem like either an implausibly shrinking violet or a very cool customer. (Lacayo, 1998, p. 16)

Lacayo (1998) finally stated: “Hiding behind the ultimate tortuous legalism could help the President get through his testimony, but it won’t pass the laugh test with the American people—which is why Clinton won’t be parsing the meaning of ‘sexual relations’ in any public statements” (Lacayo, 1998, p. 16). Lacayo’s article was written in a derisive tone with the clear implication that, although everybody knows that oral sex equals having a sexual relationship, Clinton manipulated language to prove otherwise. In other words, Lacayo admitted that while Clinton may have scored a point on strictly definitional legal grounds, people (“we all”) know anyway what the definition of sex really is. Lacayo’s TIME magazine article largely reflected the view advanced by the prosecuting party. In the Kenneth Starr Report (Washington Post, 1998), we read the following: In the Jones deposition on January 17, 1998, the President denied having had “a sexual affair,” “sexual relations,” or “a sexual relationship” with Monica Lewinsky. . . . Testifying before the Grand Jury on August 17, 1998, The Presi-



dent acknowledged “inappropriate intimate contact” with Lewinsky but maintained that his January deposition testimony was accurate. . . . The President maintained that there can be no sexual relationship without sexual intercourse, regardless of what other sexual activities may transpire. He stated that “most ordinary Americans” would embrace this distinction. . . . The President also maintained that none of his sexual contacts with Lewinsky constitute “sexual relations” within a specific definition used in the Jones deposition. Under that definition: A person engages in “sexual relations” when the person knowingly engages in or causes—(1) contact with the genitalia, anus, groin, breast, inner thigh, or buttocks of any person with an intent to arouse or gratify the sexual desire of any person. . . . “Contact” means intentional touching, either directly or through clothing. According to what the President testified was his understanding, this definition “covers contact by the person being deposed with the enumerated areas, if the contact is done with an intent to arouse or gratify,” but it does not cover oral sex performed on the person being deposed. . . . In the President’s view, “any person, reasonable person” would recognize that oral sex performed on the deponent falls outside this definition. If Lewinsky performed oral sex on the President, then—under this interpretation—she engaged in sexual relations but he did not. (p. A29)

As can be easily seen, there was considerable awareness on the part of both the president and the prosecuting party that definitions matter. Interestingly, both parties explicitly or implicitly claimed that their definitions (which are obviously different, as mentioned earlier) were accepted by all or almost all Americans. In August 1998, a TIME magazine/CNN poll asked the question: “If Clinton and Lewinsky did the following, should it be considered sex?” With reference to oral sex, 87% of respondents stated they considered engaging in oral sex sexual relations. Seven percent of the respondents stated they did not consider such engaging sexual relations. The remaining 6% included those who were not sure or who refused to respond (N. Gibbs, 1998, p. 36). This shows that oral sex is not necessarily included in the definition of sexual relations. So, even if the statistics are taken very seriously (which they do not have to, be it only for the fact that the question asked was not a general question about the definition of sex, but a question clearly geared toward eliciting a response about the Clinton–Lewinsky affair), Clinton had the right to join the 7% of respondents stating that oral sex is not included in having a sexual relationship. Clinton was clearly wrong, however, in claiming that “ ‘any person, reasonable person’ would recognize that oral sex performed on the deponent falls outside this definition” (Washington Post, 1998, p. A29).



Interestingly, as Newsweek magazine reports, a Kingsley Institute study from 1991 defining sex showed that only 40% of the respondents defined oral–genital contact as sex (Cowley & Springen, 1999). Also interestingly, the respondents referred to themselves as “moderate to conservative.” Clinton may have then been part of a larger group of the same definition of sex than people commonly admitted. The question of whether Clinton was involved in a sexual relationship with Lewinsky is linked with that of perjury. If it could have been proved that Clinton used an illegitimate definition of sex, he could have been charged with perjury. Accepting the definition of sex that he used, he could not. As might be expected, the concept of perjury generates problems as well—that is, it is fuzzy. As mentioned in the Kenneth Starr Report, in Section VII A—The law of perjury: Perjury requires proof that a defendant, while under oath, knowingly made a false statement as to material facts. . . . The “knowingly” requirement is a high burden: the government must prove the defendant had a subjective awareness of the falsity of his statement at the time he provided it. It is beyond debate that false testimony provided as a result of confusion, mistake, faulty memory, carelessness, misunderstanding, mistaken conclusions, unjustified inferences testified to negligently, or even recklessness does not satisfy the “knowingly” element. (1998, p. A51)

In the Kenneth Starr Report, we read further: “The Supreme Court has made abundantly clear that it is not relevant for perjury purposes whether the witness intends his answer to mislead, or indeed intends a ‘pattern’ of answers to mislead, if the answers are truthful or literally truthful” (Washington Post, 1998, p. A51). Further: “First, answers to questions under oath that are literally true, but unresponsive to the questions asked, do not, as a matter of law, fall under the scope of the federal perjury statute. . . . The second clear rule is that answers to questions that are fundamentally ambiguous cannot, as a matter of law, be perjurious” (Washington Post, 1998, p. A51). Journalists and lawyers, being partly conditioned by their ideology, as most ordinary lay people, have their own preferences for one definition of sex or another. This is understandable. Both journalists and lawyers also seem to have a measure of awareness of the openness and ambiguity of meaning and the significance of definitions. As Newsweek (1998) reported: In any ordinary prosecutor’s office, and surely in the chambers of the House Judiciary Committee, the definition of such terms as “sexual affair,” “sexual relations,” and “sexual relationship” would be seen as vital to a determination whether some violation of law had occurred. The burden that must be met by the OIC (Office of Independent Council) extends beyond showing



that the President was wrong on the semantics, it must also show that, because perjury is a specific intent crime, he knew he was wrong and intended to lie—something that the OIC could not begin to demonstrate. In fact, all the OIC has is a witness who gave narrow answers to ambiguous questions. . . . One example will suffice to demonstrate the inherent weakness of the OIC’s claim. The OIC argues that oral sex falls within the definition of sexual relations and that the President therefore lied when he said he denied having sexual relations. It is, however, the President’s good faith and reasonable interpretation that oral sex was outside the special definition of sexual relations provided to him. The OIC simply asserts that it disagrees with the President’s “linguistic parsing” and that reasonable people would not have agreed with him. . . . (p. 46E)

The awareness of the openness of meaning extends to other concepts related to perjury and impeachment (e.g., high crimes and misdemeanors). Speculating about the grounds on which a presidential impeachment may take place, Cloud (1998) rightly stated that terms like treason, bribery, and maladministration are very ambiguous; “ ‘high crimes and misdemeanors’ sounds like it could mean anything, from murder to jaywalking. . . . The Constitution gives the House of Representatives sole authority to decide what constitutes grounds for impeachment” (p. 33). This discussion, including the numerous quotations, shows that both journalists and lawyers were indeed aware of the problem that language and definitions did, in fact, create a problem in handling the concepts crucial to bringing the scandal to a conclusion. These concepts include sex, a sexual relationship, perjury, knowingly, high crimes and misdemeanors, and many others. Most significant, however, what neither the journalists nor the legal professionals seemed to be able to do was offer a solution to the problem. They each argued for their definition of the crucial concept of a sexual relation, giving us the impression that the definition they endorsed was correct in some absolute sense. Whether they really believed that their definitions were correct in some absolute sense cannot be attested. On the surface, they acted as if they did. Recall both the OIC’s and Clinton’s claim that people who do not share their definitions of “a sexual relationship” (different as these definitions are) are unreasonable! When this discussion is now seen in the context of what was argued for in the first part of this book, the solution to the definitional problem in question is very simple. It may be formulated in the following way: There is no one absolute definition of sexual relationship, having sex, and the related concepts against which Clinton’s or anybody else’s statements could be evaluated. To consider the crucial concept of a sexual relationship of Clinton’s famous “I did not have sexual relations with that woman,” the fact is that neither Clinton nor the Kenneth Starr party is right in some absolute sense; nor is either of the parties wrong in some absolute sense. A sexual relation-



ship may be legitimately defined as either excluding or including oral sex. There is no arbiter that can objectively state that either of the two definitions, or any other, for that matter, is correct and needs to be adhered to. There is no such thing as an essentialist definition of a sexual relationship. There are only two other types of definitions of the concept. One is the real, or functioning, definition—that is, the definition or, rather, definitions with which language users operate. These definitions were quoted in TIME and Newsweek magazines. They are different and they change. Some include oral sex and others do not. The other is a working definition. Both the Clinton and Kenneth Starr parties worked, throughout the episode, with different working definitions (one including oral sex, and the other excluding it). Neither party had the right to claim (but they, in fact, did) that their definition was correct and used by reasonable people. When either definition is matched with some of the available functioning [real] definitions, it can be both accepted and rejected because ordinary people vary significantly in what they consider to be a sexual relationship. Even if one of the two definitions is the same or very close to the definition that most people operate with (e.g., the definition that includes oral sex), this fact does not constitute any proof that the majority definition is correct in any absolute sense. A majority definition is a majority definition; a minority definition remains a minority definition; and neither one is more correct in any absolute sense. As mentioned earlier, many concepts are best understood as prototypes, including typically central, peripheral, and borderline examples (see e.g., Geeraerts, 1997; G. Lakoff, 1987; Rosch, 1977, 1978). Central examples are those about which people normally agree. For instance, when the rain pours down, it is unlikely that anyone will claim that it is not raining. When it drizzles (a peripheral example of “rain”), a measure of disagreement about whether it is raining may be expected. When it drizzles very lightly, however, it is likely that some people will claim that it is raining, whereas others will claim that it is not raining. In other words, a very light drizzle may be taken to be a borderline case of rain. The concept of a sexual relationship is certainly eligible to join the group of concepts that can be easily handled in terms of prototypes. Like any other concept, a sexual relationship is a fuzzy concept; various activities have varying degrees of membership in the sexual relationship category. It seems reasonable to assume that many people view regular sexual intercourse as the most important and typical example of a sexual relationship. If this is so, then had Clinton and Lewinsky had sexual intercourse, there would have been agreement among lawyers and the lay public about their having had a sexual relationship. The case would have been typical, and the example central, and there would have been few if any arguments about whether we were dealing with a sexual relationship.



Oral sex, however, appears to be viewed as a significantly less indicative sexual relationship activity; it corresponds to the prior drizzle example. We may view it as corresponding either to an ordinary drizzle (a peripheral example of rain) or to a very light drizzle (a borderline example of rain). Oral sex seems to be taken by many as either a peripheral or borderline example of a sexual relationship. That is why people disagree about whether oral sex is to be judged as an example of a sexual relationship at all. The disagreement that often occurs around peripheral and borderline examples also occurred in the Clinton–Lewinsky case. The Clinton defense made use of this disagreement and argued for their view that oral sex was not an example of a sexual relationship. Recapitulating, a significant part of the Clinton–Lewinsky episode was a fight about definitions. Because neither party was able to prove the point that their definition was absolutely correct, as a nonessentialist would obviously expect, the openness of language and the principled availability of various definitions of one word or expression had to be acknowledged. What seems to be true in political cases is that in definitional disputes like the one under investigation here, the winning party is the one that manages to impose its definition on others by whatever means.

THE FLORIDA VOTE CONFLICT The Florida vote conflict of the 2000 U.S. presidential election provides us with another example of how definitions count in politics and how they can be decisive in a political struggle. As is commonly known, the Florida vote conflict revolved largely around the concept of the hole. The voting ballots were supposed to be punched. Holes that appeared on the ballot as a result of the punch each counted as a valid vote. The problem was, however, that very many votes were disputed because the cards had not been punched through properly and, as a result, the hole was not clear. The reasons for the punching problems (e.g., old and malfunctioning punching machines) are irrelevant to the present discussion. As The Economist (2000a) reported: Of the 462,000 votes cast, only 432,000 were found valid when the results were first counted by machine. Of the discarded 30,000 ballots (7% of the total), 19,000 were double-punched—that is, voters had picked two candidates. An additional 11,000 were not properly punched through. This introduced a new political vocabulary: if the punch leaves a flap of paper, it is called a “hanging chad”; if it does not break the skin, it produces a “pregnant chad” or “dimple.” (p. 66)

The question that this led to was what kind of punch constituted a valid vote or what was the definition of a valid vote.



Non-English-language newspapers have also noticed and reported on the definitional quagmire as regards the hole. For instance, on November 21, 2000, the Norwegian newspaper, Bergens Tidendei, carried a reprint of the Danish Jyllands-Posten article entitled “Hva er et hull?,” ‘What is a hole?’ (Ulveman, 2000, p. 10). The paper stated: “Definisjonen av hva et hull er kan tippe kampen om Florida til enten George W. Bush eller Al Gore. Florida høesterett har tatt fatt på saken” (Ulveman, 2000, p. 10, “The definition of what a hole is can decide about the fight in Florida either in favor of George W. Bush or Al Gore. The Florida Supreme Court sets to work on this issue”). Further complications emerged because stating whether the chad was pregnant or hanging was not always easy: “Republicans have mocked scenes of officials holding up punch cards to see how much light showed through and thus ascertain whether the chad is hanging or merely pregnant” (The Economist, 2000a, p. 66). When a chad is declared to be hanging, further doubts might be raised—namely, how firmly or “safely” does a chad need to be hanging to be declared hanging and not pregnant any longer. Decisions about this may be “taken,” although they may not be easy: Broward County will count dimpled chads and flaps of confetti attached by three corners. Palm Beach will do the same provided that there is some other indication of an intention to vote (such as evidence from other votes on the ballot paper). Indentation and intention are both in the eye of the beholder. The counties have settled on relatively “permissive” counting of dimples. (The Economist, 2000b, p. 72)

Finally, as the Norwegian newspaper rightly reported: “Spørsmalet har endt på høyesteretts bord. Det blir således de sju dommernes oppgave å definiere hva et hull er. Og hull-definisjonen kan tippe president-posten til enten Bush eller Gore” (Ulveman, 2000, p. 10, “The issue has ended up on the table of the Supreme Court. In this way, the task of the six judges is to define what a hole is. And the definition can decide about whether the presidential post goes either to Bush or to Gore”). Similar to the Clinton–Lewinsky case, in the Florida vote affair, there seems to have been a measure of awareness that “what makes a hole” is a crucial question, an answer to which is of utmost political significance. Again, similar to the Clinton–Lewinsky case, the pertinent parties argued about the definitions of valid ballots, and some ridiculed the procedure of detailed checking of the voting ballots for clear and unclear holes having been punched through. The various parties argued about dissimilar views of what makes a hole, and thus for what makes a valid vote. Again like in the Clinton–Lewinsky case, the definition (of a hole) was seen as a problem for which a solution was difficult to find.



Unlike in the Clinton–Lewinsky case, however, and not surprisingly, no recourse was taken to any common, nonspecialist views of the hole. In other words, the general public was not consulted on the meaning of the hole. Also unlike in the Clinton–Lewinsky case, there were no explicit arguments for any one working definition of the hole. The fact remained, no doubt, that the idea of what kind of punch does or does not make a hole is central. What is crucial for our present discussion is the fact that nobody (to the best of my knowledge) ever said explicitly how the definitional dilemma should be solved from the linguistic point of view. The solution was very simple—namely, one of the possible definitions (e.g., the one that required that no chad was hanging in any manner) should have been adopted as the working definition and adhered to in practice throughout the whole procedure of counting the votes. As long as several different definitions were used at the various voting stations, the results were unreliable and skewed in different directions. The confusion over the definition of the hole was exacerbated by the practical procedure of matching a voting card (punched through completely, punched through partly, the chad hanging firmly, the chad being pregnant, etc.) with whatever definition was valid at a particular voting station. This matching was done by human beings, and the differences in human perception certainly played a role. Although the matching problems in question could never be entirely eliminated, accepting one working definition of the hole would have made things significantly simpler and fairer. Instead the people concerned painfully argued about what really makes a hole, seeking what seemed to be an essentialist definition of the concept. Recognizing and acknowledging the fact that such a definition is unattainable as well as accepting one working definition for the whole counting procedure would have been a simple and rewarding solution to the problem.

ABORTION, EUTHANASIA, AND THE STEM-CELL RESEARCH CONFLICTS The three conflicts mentioned in the title of this section share at least one characteristic—namely, they all involve disagreements about the definition of life and a significant amount of essentialist reasoning. Let us first concentrate on the abortion conflict and illustrate it with data from the recent history of Poland. Between 1956 and early 1993, Polish law allowed abortion virtually on request. In 1989, when the communist system fell, there began a heated debate over the issue of legal abortion. This debate is, in fact, still going on in view of a highly restrictive law passed by the parliament late in 1993 and a new liberal law passed toward the end of 1996. As may be easily guessed, the



changes in the legal system concerning this issue reflect political vacillations, frequent parliamentary elections, and government changes. The abortion conflict involves die-hard opponents of legal abortion (primarily centered around the Polish Catholic Church); liberals, who are for legal abortion on request; and a relatively quiet group of semiliberals, who accept abortion with some restrictions. The first group is against abortion even if the pregnancy is a result of rape or when the fetus is genetically impaired. The third group accepts abortion, for instance, when the pregnancy results from rape. The way language comes into the conflict is primarily through the undeniable fact that most of the conflict is handled via language. The aspect of the conflict that is of interest to us here concerns the way arguments for any of the positions (especially the two extreme pro and anti positions) are formulated. What follows is a number of representative statements and my comments on them along the lines of the discussion of essentialism presented before. As might be expected, the numerous issues raised revolve around the fundamental questions of (a) What is life? and (b) What is a human being? Related questions such as these that are provided shortly can be answered only as a function of the answers to these two questions. Let us consider a few examples: “Is the doctor performing an abortion a murderer?” (Staniewicz, 1990, p. 4).1 In this question, the easily retrievable underlying what-is question is “What is a murderer?” Here the way the question is formulated indicates that only one of the two answers—namely, “A doctor performing an abortion is a murderer,” or, “A doctor performing an abortion is not a murderer”—is possible and thus true. No other answers are invited here. “What is of most fundamental concern for the human being?”2 Here the what-is question is formulated in a straightforward way. Only one answer is invited as well. “Whenever we pretend to be discussing the question of abortion, aren’t we, in fact, debating the question of the right to live?” (see Endnote 2). In this statement, essentialist reasoning is expressed much less explicitly than in the first two questions. The assumption is that there is one real problem under discussion (the problem of the right to live). The question of abortion is, at the same time, conceived as blurring the discussion of the real problem. The what-is question involved is, then, “What is the real problem under discussion?” The one essentialist answer implied in the question is: Abortion is not the real problem under discussion. The real problem under discussion is the right to live. It is especially interesting to study the answers to questions like the one before because, through such a study, one can quite easily identify several types of conflict participants. There are at least three types here: (a) people



who claim that they definitely know the answers via some illumination or teachings (e.g., life begins at conception); (b) those who claim that biological science has given the final binding definition of what life is; and (c) those who realize that biology has not yet given any final absolute definition of life. The last type of people assume, however, that biology will one day give the final definition, or that biologists are, in principle, capable of formulating such definitions, and that, simply, so far they have not been able to do so. Consider the following statement: “Science has not yet pronounced its last word as to when life begins” (Stawicka, 1990, p. 4; italics added). In this statement, we see a typical example of covert essentialism. The clue words here are “not yet.” The reasoning can be reconstructed as follows: “When life begins” is a paraphrase of the what-is question, “What is the beginning of life?” Science has not yet pronounced the final word on that question, but in principle it can do so. Thus, the reasoning runs further, the question of “What is the beginning of life?” is ultimately answerable. It is just that scientists need to work longer and harder to find out what the beginning of life really is. Another answer is the following: “Science has given the final answer as to what life is. Some people ask the question, however, of when the human being begins. This is as if I were asked if I get wet when I go out in the rain.”3 In this sentence, science is given credit for having provided one final, ultimate, indisputable answer to the question of what life is. One of the whatis questions involved here is obviously the question of “What is life?” The other question involved in this sentence is that of when the human being begins, which easily translates into the essentialist question of “What is the human being?” The jocular answer to the latter question may be taken as essentialist in that as, for the author, rain implies getting wet, life implies a human being. In other words, if we know what life is, we know what the human being is. This essentialist reasoning indicates that the one, true, ultimate (scientific) definition of life includes reference to the human being. Definitions that do not do so are wrong. Let us consider an example of what the prochoice proponents say: When I asked the question: “Do you think that abortion can be compared to a visit to a dentist?,” he answered in more or less the following way: “Abortion is no evil at all. The view, shared by many people, that abortion is evil, is a victory of the Church, and we have to fight that. Abortion is a medical operation like any other, and each woman, independent of the circumstances, can have it performed.” (Rogala, 1993, p. 1)

In this passage, the essentialist contends that abortion is not evil, which is seen in the response to the question, “What is evil?” Abortion is placed here, definitely outside evil. In other words, the correct ultimate definition



of evil may not or does not include abortion. If only one answer to the question of “What is evil?” is believed to be true (as is the case for the essentialist), the answers “Abortion is not evil” and “Abortion is evil” are equally essentialist. The following example further illustrates implicit covert essentialism relating to the discussion of abortion: . . . voting on legal destruction of human life not only violates human rights but is an attack on all human order. If the referendum is permitted—we read later—it may consequently lead to other vicious deeds, for example, euthanasia, i.e., killing sick people who are perceived as useless to the society. As we know, the most radical form of questioning the right to live, based on the above criteria, was concentration camps. (Rakowiecki, 1991, p. 3)

This is an opinion about a referendum on abortion, which was considered for a short period of time. Here the underlying essentialist reasoning is less obvious. To reconstruct it, one needs to go through the following chain of deductive steps: The embryo is definitely human life. Because voting on legal destruction of human life is assumed to violate human rights, the correct definition of human rights rules out reference to legal destruction of human life. Further, because legal destruction of human life is assumed to be an attack on the whole natural order, also the correct definition of natural order rules out reference to legal destruction of human life. Still further, euthanasia is killing (killing sick people who are perceived as useless in society), which is a clear answer to the essentialist question, “What is euthanasia?” Still further, the passage quoted earlier tells us that euthanasia is questioning the right to live, and concentration camps are questioning the right to live as well. So, euthanasia is like concentration camps. Eventually, the reasoning leads us to the conclusion that voting on legal abortion not only equals voting on legal destruction of human rights, but also equals (or is) voting on the legitimacy of Nazi-style concentration camps. What has been shown so far should be taken to indicate that parties in these conflicts are convinced that their definitions of what is good and what is evil are the only legitimate ones. These definitions are believed to be universally valid. It is important to observe that an essentialist politician will tend to conceive of his or her ideas (such as good and evil) as nonpolitical matters. Such ideas are believed not to be open to any questioning, referenda, political debates, or parliamentary voting procedures. Let us consider the following exchange: A: The majority of the public is against delegalizing abortion. You have accused the Mazowiecki government of disregarding the opinion of the majority.



B: I am not interested in what the opinion of the majority is. This is not about politics—as in the Mazowiecki government case—but about the fundamental distinction between good and evil. Abortion is evil, without any doubt. There is nothing to talk about.4

In this exchange, we see another prototypical case of an essentialist answer. The what-is question underlying B’s reaction to A’s comment is, again, “What is abortion?” The answer is a clear, “Abortion is evil” (“. . . without any doubt. There is nothing to talk about!”). Consider the following two examples: “In the Parliament, as a journalist from Gazeta Wyborcza reports, advocates of murdering innocent children fight with the advocates of murdering and mutilating women and with advocates of depriving women of civil rights” (Passent, 1991, p. 3). This passage paraphrases essentialist answers to the questions of “What is a prochoice person and what is a prolife person?” The answers are: (a) “A prochoice person is an advocate of murdering innocent children”; (b) “A prolife person is an advocate of murdering and mutilating women or an advocate of depriving women of human rights.” Finally, consider the following quotation, which is a question asked by a Member of Parliament (MP) of another MP. This question may be taken to be the epitome of the abortion conflict exchange of insult: You, sir, toward the end of your interview for òycie Warszawy, expressed your opinion of the people who are against anti-abortion law, in the following manner: if God wants to punish someone, he deprives him/her of reason. Sir! And how about yourself; what made God punish you in the same way?5

Here a prochoice person is referred to as a person who has been deprived by God of his or her reason. In this view, some people have reason (those that are against legal abortion), whereas others do not (those who are prochoice). Thus, the correct definition of a prochoice person excludes reference to reason; a prochoice person has no reason. Any definition of a prochoice person that will say that a prochoice person has reason will be treated as utterly and ultimately wrong. It is important to realize that the antiabortionist has by all means the right to call a prochoice person a murderer, just as a prochoice person has the right to call an antiabortionist unreasonable. Each individual has the right to live by his or her own axiological system, which allows any individual to use labels (words) she or he wishes to refer to any infractions of this axiological system. From the political point of view, it may not be a good idea to use these strong words in public, but in principle any label may be used. One obviously has the right to say that, for him or her, someone who is for abortion is a murderer. This would be a clearly nonessentialist expression. It should be noted, however, that the language used in the abortion



conflict is not such language. The earlier quotations show quite clearly that the harsh words used were not meant as individual, subjective labels. On the contrary, they were meant as universally valid and unquestionable designations. This is precisely what made the expressions essentialist. The abortion question is certainly valid outside of Poland. For instance, in the United States, it is referred to as “the war that never ends” (The Economist, 2003, p. 24) and perceived as shaping American politics to a considerable degree. It was one of the issues also addressed in the 2004 presidential election. The definitional problem discussed in the Polish context remains the same for the United States. Consider a case recently reported on in The Economist (2001a), pondering the question of whether a fetus is a person: Is a fetus a person? Regina Mcknight, a 24-year-old drug addict, has been sentenced to prison for 12 years in South Carolina. Miss McKnight smoked crack cocaine in 1999 while she was pregnant. Her unborn baby died in the eighth month. Miss McKnight, who has three other young children and is now pregnant again, has been convicted of homicide. . . . So far, nearly two dozen other states have rejected the criminal prosecution of pregnant women for behavior that harms their fetuses. This includes the Florida Supreme Court which ruled five years ago that a pregnant woman who had shot herself in the stomach could not be charged under a homicide statute. The thin-end-of-the-wedge argument against the South Carolina law is, however, less compelling than the unfair way it works in practice. It is mainly applied against drug addicts. Women who drink or smoke or drive recklessly while pregnant, and then deliver a stillborn child, are not being arrested; perhaps because many of them are well-to-do and White and not, like Miss McKnight, poor and Black. (pp. 54–55)

Let us now return to euthanasia. As the reader probably knows, in Holland, euthanasia is legal. Thus, euthanasia in that country is not treated as murder. People performing euthanasia do not go to prison. “Euthanasia is not murder” is the Dutch position. In contrast, in Poland, as well as in many other countries, euthanasia is illegal. In Poland, euthanasia is treated as murder. People performing euthanasia go to prison. “Euthanasia is murder” is the Polish position. The question that comes to the fore in the context of these facts is this: Although different, are the Dutch and the Polish positions equally essentialist? The answer to this question can be both positive and negative. The question can be answered convincingly only if one knows about the people who stand behind the formulations in question. The point is that, to the extent that either position (euthanasia is murder, and euthanasia is not murder) is believed to be absolutely (naturally, uncontroversially) true, it is essen-



tialist. However, if proponents of either position believe that it is the result of a social agreement, such a standpoint is clearly nonessentialist. In the latter case, members of a given community simply agree (without any necessary reference to some absolute authority or natural values) that euthanasia will or will not be treated as murder, and that people performing euthanasia will or will not be put to jail, respectively. What is being talked about, obviously, is the source of laws. Essentialists tend to believe that at least some laws (e.g., a ban on abortion) are natural—that is, they are prescribed by some force standing beyond the human being. According to essentialists, these kinds of laws are so obvious that they should not be discussed. Nonessentialists, in contrast, treat laws as nothing other than social agreements of clearly human provenance. For nonessentialists also, however, some behaviors should hardly be even talked about as legal problems (e.g., killing), not because they are prescribed or proscribed by some external power, but because the social agreement concerned is extremely wide, perhaps international. We need to say “wide, perhaps international” and not universal (which would equal in scope the essentialist’s “natural”) because it is difficult to find examples of laws with reference to which there is universal agreement. Even the earlier “killing law” example would not qualify here because, obviously, several groups of people allow killing widely (e.g., the Nazis, the Khmer Rouge), and others allow it under some circumstances. The latter, by the way, include orthodox antiabortion Catholics, who allow defensive killing in, for instance, wartime. Another issue related to those of abortion and euthanasia is the stem-cell research one. As The Economist (2001b) reported: George Bush has come face to face with the first policy dilemma of the new world of human genetic engineering: should the federal government finance research into stem cells derived from human embryos? Stem cells can transform themselves into the many different cell types that go to make up a body. They hold out the promise of new therapies for diseases like Parkinson’s or Alzheimer’s. But because such cells are derived from embryos, their extraction upsets many (though, crucially, not all) anti-abortion people. (p. 49)

Further: Critics of federal funding make three arguments. First, they say, stem-cell research is immoral. Embryos are human life. You cannot (yet) reliably extract stem cells from them without killing them. The fact that research into embryonic stem cells will initially be conducted on cells from embryos discarded in the course of in-vitro fertility treatments is irrelevant. These are still potentially human beings, and anyway they doubt whether scientists would content themselves with “discarded” embryos only. (The Economist, 2001b, p. 49)



Clearly, the issue at stake here is the definition of life and human being or person. The definitional quagmire involved in the stem-cell research question is to many people more of a conundrum than that involved in the abortion one. In The Economist (2001b), we read further: . . . the supporters differ on the moral issues. The stem-cell debate is much more complex than that hardy perennial of American ethical dispute, abortion. The abortion debate hinges solely on your attitude to the cells that make up an embryo or fetus. Supporters of abortion rights, who think embryos are not meaningfully persons, can carry the same logic forward to stem cells effortlessly. But for antiabortionists, the issue is harder, since it requires balancing two “pro-life” goods: the life of the embryo against the potential lives saved from research based on its stem cells. (p. 50)

A hot debate bordering on stem-cell research concerns cloning. Here again part of the issue pertains to the definition of the person: “. . . they [the procloning advocates] insist that an unimplanted ball of cells is not a person. Some argue it is even less of a potential person than a stem cell derived from in-vitro fertilisation, since cloning uses unfertilised eggs” (The Economist, 2002a, p. 46). Consider now all three conflicts referred to earlier. As far as the linguistic aspect of the conflicts is concerned, definitions of the relevant concepts (e.g., life, person, killing) become central. With reference to all three conflict cases, we may posit with a fair amount of certainty that the participants explicitly or implicitly maintain that their definitions are correct, whereas those of their opponents are wrong. Nevertheless, as some readers might claim, doubts may arise as to the actual beliefs of the people involved in the debates in question. For example, in one of the passages quoted before, prochoice people are referred to as murderers; couldn’t a label such as this be simply an individual’s label, which, as mentioned earlier, anyone has the right to use, although not necessarily in public? These harsh words could certainly be mere labels, not meant at all as an absolute designation. To refer the reader back to our discussion of the distinction between left-to-right and right-to-left definitions, it could be said at this point that if we give the author of a statement like those quoted earlier the benefit of the doubt, we can consider these definitions as right to left, which makes them immediately much milder. Of course there are no safeguards against misinterpreting a nonessentialist statement (i.e., a statement that looks stylistically like an essentialist one). Conceptual essentialism (not stylistic) should be seen as a belief to which access is certainly very limited for the linguist. As mentioned earlier, we should always give the speaker or the author the benefit of the doubt. However, when extensive extralinguistic



context provides overwhelming evidence that someone, indeed, reasons essentialistically, criticism is much easier to advance. By extensive extralinguistic context should be understood all the activities of the people concerned of which one might know. For instance, authors of or people referred to in the statements concerning the abortion conflict in Poland provide evidence for their essentialist philosophical commitment. They treat the adduced definitions as being true in some absolute sense independent of human judgment. Let us take, for example, “We must learn to distinguish hypocrisy and lies from truth.”6 This is one of the many statements of a similar kind formulated by representatives of the political right. Rightist MPs make perpetual claims in the mass media that only they are exponents of truth, justice, and morality. Consistently adhering to their commitments, they fiercely attack any form of moral relativism. Additionally, some of them give the impression of belonging to a small group of select few who have significantly more insight into moral questions than the others outside the Parliament. One of the MPs is reported to have said (during a debate on the possible referendum on abortion law) that an “accidental society” should not be allowed to decide about important matters. This person reasoned that she had access (by the very fact of having been elected an MP perhaps) to ultimate truth. Hence, the electorate (the accidental society) had nothing to say on the matter. Given such a state of affairs, there is very little doubt that “we must learn to distinguish hypocrisy and lies from truth” is an example of essentialist thinking. Consider another example—namely, “In the Parliament . . . advocates of murdering innocent children fight with the advocates of murdering and mutilating women and with advocates of depriving women of civil rights” (Passent, 1991, p. 3). This example reflects the thinking of two feuding sides that both seem to have fallen victim to essentialism. The antichoice proponents are well known to uncompromisingly oppose abortion and to call it murder. “Murdering innocent children” is a phrase frequently used in public statements, interviews, and newspaper articles. Antichoice advocates incessantly refer to both religious and scientific authorities that are reported to have infallibly defined or discovered the essence of life. “Life,” in this case, is identified with “the human being,” and thus abortion equals murder. Antichoice representatives are widely known to be strongly committed to their uncompromising position, which is certainly easy to understand given their religious orthodoxy. However, the prochoice advocates argue that a ban on abortion equals murdering or mutilating women or depriving them of human rights. Many members of the prochoice group are uncompromising in viewing the antiabortion law only as an act directed against women. Prochoice advocates claim that antiabortion law violates human rights; antiabortionists claim that prochoice law violates human rights. The way that the case is argued in daily



exchanges makes both sides equally essentialist. In other words, to claim that legal abortion is definitely a violation of human rights, or to claim that a ban on legal abortion is definitely a violation of human rights, is to make equally essentialist claims. Doubts may be raised of course as to whether the three conflicts discussed earlier should be seen as linguistic. There is, obviously, no way that anybody can prove that the conflicts are linguistic. The point being made is that if one insists that the conflicts are not language-related in any sense, then that is what one believes. If one believes that the conflicts are about exolinguistic, axiological, and moral systems, that is what the conflicts will be. It can be claimed, however, from a nonessentialist’s point of view, that viewing the conflicts as language-related is one legitimate perspective to take. So the conflicts can be taken as a moral issue, as an axiological issue, as a linguistic one, or possibly as other types of issues. Conceiving of the conflicts as necessarily of only one type would make you an essentialist. The cases that have been discussed can then be treated as any kind of conflict, not necessarily as language-related conflicts. This conclusion makes our argument very relativist, of course. This, in turn, may be seen as undermining its own position. Namely, if the cases discussed earlier can be treated as any kind of conflict, why do we argue in this book for treating them as language-related? This question will inevitably remind us of one of the major objections often raised against relativism in general: If any position is acceptable, then why should relativism, clearly being only one possible position, be the right one? My answer to this objection is this: Relativism is clearly only one possible position. Relativism is not meant to be the position that everybody is supposed to take. Relativism springs from a critical analysis of how the world works. To relativists, it seems the most rewarding position. Yet relativism allows for positions other than relativists’ to be conceived of as rewarding as well; rewarding to other people, not to relativists. Relativist claims do not have to entail any sort of absolutism. If a relativist claims to be absolutely right about his or her relativist claims, then, by the conception of relativism that I adopt in this book, she or he ceases to be a relativist.7 Returning to the question of whether the conflicts mentioned earlier should or could be treated as language-related conflicts, the following can be claimed: An analysis of the phenomena that have been discussed earlier suggests that it is very reasonable to claim that we, indeed, are tackling language-related conflicts. However, if anyone wants to investigate these phenomena in a different way and conceive of these conflicts as conflicts of a different kind, there is nothing that can be done about it. No one can ever prove that these conflicts are definitely and unquestionably languagerelated, and no one can ever prove that they are definitely and unquestionably not language-related.



To end this section, I would like to make three points clear. First, my prior reasoning is by no means to be taken as an argument for or against abortion, for or against euthanasia, for or against stem-cell research, or for or against cloning. These subjects are not at issue here. All I have been trying to indicate is that arguing for or against abortion, euthanasia, and so on, by resorting to allegedly clear and unambiguous definitions of concepts such as life, person, and human being is not a powerful way of arguing, and it is not convincing to a nonessentialist. Second, like essentialists and those who are rather uncommitted, nonessentialists may certainly have beliefs about whether abortion, euthanasia, stem-cell research, or cloning are morally right or wrong, and to what extent these are morally right or wrong. To be a nonessentialist does not equal having no beliefs, making no commitments, or having no opinion on important matters such as those discussed. On the contrary, to use the example of abortion, one can be a devout nonessentialist and a devout antiabortionist; that is, a nonessentialist can certainly voice the opinion that abortion is evil. If, however, an antiabortionist claims that she or he knows exactly what life is and when exactly the human being begins, she or he is an essentialist by the definition of essentialism accepted in the present book, and his or her position will not be seen by the nonessentialist as rewarding. Third, there is little doubt that one of the main reasons for the conflicts around abortion, euthanasia, stem-cell research, and cloning, as in the case of the Clinton–Lewinsky conflict, is that we are dealing here with borderline examples of the concepts crucial in the debates (see also the concept of science). It should be noted that there has never been any serious discussion or dispute about whether a healthy, 30year-old adult is a human being and about whether she or he could be killed, or about whether euthanasia could be performed on him or her. The case of a terminally sick, 30-year-old adult, which illustrates a borderline case between life and death, may, however, be discussed and viewed as a disputable case, and she or he may for some people be eligible for euthanasia. It seems that many conflicts arise especially where the peripheral or borderline examples of concepts are dealt with.

POLITICAL CORRECTNESS A phenomenon that clearly encompasses linguistic and political issues is the political correctness (PC) conflict. Although especially salient in the English-speaking world, the PC debate also reached some other, especially European, countries. When one reads the various publications on the topic (e.g., Cameron, 1995; Dunant, 1994; Suhr & Johnson, 2003), one clearly gets the impression that the PC debate is a conflict that goes significantly beyond language, but in which language plays an extremely important role.



Many will associate the PC movement with attempts to promote tolerance and sensitivity. It is, however, often “characterized as an excessive attention to the sensibilities of those who are seen as different from the norm (women, gays and lesbians, Black people, the disabled)” (Mills, 2003, p. 89). In fact, the PC label debate is interpreted very widely these days. Sometimes it may seem that the PC debate is simply about correct behavior, no matter who or which political group behaves or who or which political group evaluates behavior (linguistic and nonlinguistic). Some evidence of this extremely wide interpretation of PC comes, for instance, from the political fights and arguments in Poland. In a telling article, a rightist author in a right-wing newspaper scolded the political left for promoting abortion. He derisively referred to promoting abortion as an example of politically correct behavior, which he subsequently called politically correct murder (Sakiewicz, 1995). However, in a left-wing magazine, a letter to the editor was published in which the author referred to Catholic political correctness. The term is used here to ridicule those statements and teachings of the Polish Catholic Church that the author disapproves of (e.g., a statement made by one of the bishops, who said, “. . . for me it would be better if contraceptives were not available” [Król, 1995, p. 11]). Later in a list of further examples, the author continued illustrating what he considers to be Catholic political correctness. His list includes: anti-Semitism, disdain for people who are different, punishing the poor, rewarding the rich, materialism, and so on. It seems that the PC debate may be about almost anything. Our discussion of the PC debate is restricted in that the discussion is only to show that the philosophical question of essentialism is highly relevant to the PC debate. To put it differently, the view is promoted that our awareness of the essentialist/nonessentialist distinction and our knowledge of the basic tenets of cognitive linguistics may help us understand some of what the PC conflict is about, and, more important, this understanding, in turn, may help us resolve the conflict. We need to remember that the political right supports the view that words (language) must be taken seriously. The political right gives us a lot of evidence that they adhere to the essentialist view of the world in general and of concepts in particular. Staying with Poland, that position is clearly evidenced by data available from its recent history; the political right has been arguing for one indisputable definition of terms, especially those that pertain to the question of morality. The belief in the immutability of word meanings and the necessity to protect the threatened status of a language goes, in the case of the conservatives, far beyond morality or religionrelated terms. Conservatives are against change as such partly because they think that change is, in principle, against some natural state of affairs. With respect to language, that natural state of affairs translates into the belief that, among other things, a word has in principle one immutable meaning that can be clearly defined.



If one considers the PC proposals from this conservative point of view, then when the PC advocate proposes, for example, the term disabled to replace handicapped or African American to replace Black American, the conservative essentialist of the political right will tend to disapprove of the proposal because she or he believes that the relationship between the challenged word and the reality to which it refers is at least to some extent natural, and thus the change or innovation proposed is seen as an act of intervention into what should be left untouched. The essentialist tends to believe that language change occurs only, in a way, on its own, and that human intervention of the PC kind is an unnecessary and unwelcome act; it is an incursion into the natural state of linguistic affairs. Most probably, not all anti-PC conservatives reason like this. If one did a detailed study of the enemies of PC on the political right, one would probably find a variety of positions and views on the issues concerned. It may be safe to posit, however, that it is the essentialist objectivist view of concepts and meaning that significantly unites most, if not all, of those people who are anti-PC. The other side (not to speak of the group in between)—that is, the pro-PC group—seems to be much less homogeneous. The PC side includes at least two groups of people. One group consists of the PC advocates, who, like the PC enemies, reason in the essentialist way. This reasoning is that the terms/words challenged (i.e., the words to be replaced) are simply wrong, inadequate, and intrinsically denigrating, as it were. Such terms, this PC advocate would argue, must or should be replaced by other terms/words, which are the correct, nondenigrating words. For instance, hair disadvantaged would be seen as having to replace bald because hair disadvantaged is simply a nondenigrating and thus a better term. The term is seen, from the axiological point of view, as neutral in some absolute sense, as if the axiological load of a word were in the word, again, like meaning, forever, and as if it were the same for all users of English. This type of PC reasoning is evidently essentialist, like that of the enemy of PC. One orthodoxy is replaced with another. One word or expression, which is believed by one group of people to be the right word or expression to refer to some nonlinguistic reality, is replaced by another word or expression that is believed by another group of people to be the right word or expression to refer to the same nonlinguistic reality. As a result of this kind of operation, a conflict evolves in which one person (or group of people, rather) says, “Our word is correct and the only good one,” and the other person (or group of people) says, “No, your word is wrong, mine (or ours) is the right one, and the only right one.” To the extent that the two competing sides are essentialist with respect to meaning, they both appear to be wrong, doing roughly the same thing. Not all PC movement seems, however, to be essentialist. Not all PC advocates believe that one word must be replaced by



another. There is obviously a way of subscribing to some PC assumptions without having to be essentialist. One question may be posed in passing—namely, whether PC involves conflict-ridden language, language-ridden conflict, or both. An answer to this question gradually emerges and becomes clear as we proceed with our discussion. To suggest an interim answer, we may say that the view that PC involves conflict-ridden language is hardly debatable. In other words, the conflict seems to have come first. As is well known, PC arose as a sociopolitical movement that was intended to combat social injustice, in general, and discrimination against minorities, in particular. It is quite obvious, in other words, that we are dealing with a primarily nonlinguistic, sociopolitical conflict. However, if one accepts the views on essentialism promoted in this book, then it goes without saying that people’s beliefs about language and meaning play a crucial role in this conflict. These beliefs may exacerbate, intensify, or alleviate the conflict. It needs to be repeated that essentialism as a philosophical stance, so obviously characterizing the political right, should not be combated with like essentialism, such as frequently characterizes the political left. What we get as a result is two totalitarian linguistic states, as it were, which are, in many respects, not unlike two totalitarian sociopolitical states or two dictatorships. It is just that the banners and the colors under which the two totalitarian states are ruled are different. In other words, political correctness can be looked at both as a conflictridden language and language-ridden conflict. More important, the PC conflict will remain a language-ridden conflict even if the PC camp ceases to be essentialist. This is because we would then face a situation in which we have an essentialist on the one side of the barricade (an anti-PC person) and a nonessentialist on the other (an enlightened PC advocate). This means that as long as one of the parties is essentialist, the conflict can be viewed as language ridden. The anti-PC group member will continue to believe that she or he is definitely right, that she or he knows what words should be used. In other words, the essentialist’s view of language will perpetuate the conflict irrespective of the position that the PC advocate takes. We might be wondering about the PC advocate’s position—whether it is stronger or weaker—when she or he ceases to be an essentialist. It may seem that she or he would lessen the chances of winning the argument with an essentialist. As a matter of fact, this position is probably both stronger and weaker. It is stronger intellectually (from the point of view of the philosophy advocated in this book at least), but at the same time much weaker, unfortunately, for all practical purposes. The typical nonessentialist will have many answers to one question (whereas the typical essentialist will have one); the typical nonessentialist will often be in doubt about many



things (and the typical essentialist will not). Incidentally, this conclusion holds for all kinds of activities, and not only those having to do with political correctness. The important concern is whether the intellectual gain is enough. The intellectual gain has to be enough for the time being. There seems to be no other solution. Our nonessentialist intellectual gain is not unlike other intellectual gains. It can only be hoped that spreading the message on the nonessentialist thought will bring about large-scale practical gain as well. Let me now illustrate how, from the nonessentialist point of view, the PC advocate should proceed. Let me do that by addressing one of the most crucial concepts in the PC struggle—namely, the concept of insult. By scrutinizing closely this central PC concept, a solution to the PC dilemma can be offered. It is hard to say whether the PC movement would then be more successful than it has been until now. What can be said, however, is that the view promoted next gives at least a clear intellectual, philosophical, and linguistic alternative to the PC opponent. First of all, it needs to be noted that some of the core questions that stand out in the PC debate are “What language is insulting language?” and “What language/word use is insulting to the hearer?” Insult, then, may be taken to be the central concept in the debate. The first question that can be raised about these two questions is whether these questions are answerable at all. The reader will certainly recall that they are not, if meant to search for one final and correct answer. That is, these questions, being what-is questions, are not answerable in any one way. Any serious search for one unambiguous answer to the question of what insult is will certainly press us into an infinite regress of definitions. For example, in The New Oxford Dictionary of English (1998), insult is defined in the following way: “disrespectful or scornfully abusive remark or action” (p. 948). One can see clearly that insult is defined in terms of disrespectful and scornful behavior. Disrespect, in turn is defined as “lack of respect or courtesy” (p. 534). We subsequently ask the question of, for instance, what courtesy means. What we read is this: “the showing of politeness in one’s attitude and behavior towards others” (p. 422). Further, what does respect mean? What we read about it is this: “a feeling of deep admiration for someone or something elicited by their abilities, qualities, or achievements” (p. 1580). We certainly do not know what admiration, deep, abilities, and so on are exactly. So we proceed. What does admiration mean? What does deep mean? The point being made here is that insult is just like any other concept/ word. There is no essence in the concept and, as a result, in the word behind which the concept stands. If we continue asking what-is questions about further definitions, we will most likely come full circle. To put it more explicitly, the word “insult” is highly ambiguous. What may be an insult to one person or a group of people may or may not be an insult to another



person or another group of people. If we adhere to the view that concepts are experientially grounded, we need to accept the view that the meaning of insult is individual to the extent that one’s experience in learning and using that concept is individual. Let us view things in a slightly different way: As our experience is, in fact, individual and only group-based or grouprelated, there will in principle be no ultimate or large-scale agreement as to what kind of behavior (verbal or nonverbal) will be insulting or not insulting, or as to what extent such behavior will be insulting. That is how the nonessentialist would reason. The essentialist, in contrast, being sure that his or her definition of insulting behavior is correct, will tend to think that we all use or should use the word “insult” in the same or at least in a very similar way. If faced with a deviation from this view, the essentialist will try to impose his or her definition on others. As opposed to the essentialist, the nonessentialist is aware of the fact that insult is no more than a word—a label that may be used and, in fact, is used to refer to significantly varied behavior. Moreover, the nonessentialist, following the argument of the cognitive linguist, will be fully aware of the fact that insult is a graded category. That is, the nonessentialist is fully aware of the fact that some instances of behavior may be more insulting than others. Consider the often quoted words nigger, Negro, Black, for example. These words are, for most people, words of varying degrees of insult. Also, as individual users of English, we vary with respect to how each one of us will estimate the degree of insult attachable to each of these three words. Put differently, although for some of us nigger and Negro are very insulting, Black is less insulting; for others, only nigger is very insulting, whereas the two remaining words are fully acceptable; and still for others, nigger is very insulting, Negro is not insulting at all, and Black is less insulting than nigger. While talking about the three concepts, one should not forget about fuzziness. We may say, then, that although nigger is clearly an insulting word for the present author and Black a clearly noninsulting one, Negro is a fuzzy case; it is kind of insulting or it is hard to say whether it is actually insulting. One may hesitate about whether Negro is to be viewed one way or the other or one may simply remain undecided and ambivalent about the term. To recapitulate, the nonessentialist reasons that any of the four words, nigger, Negro, Black, and African American (the inclusion of African American should be noted), may be insulting. Interestingly, some African Americans find the term African American insulting. The essentialist selects one or another of these terms and declares it insulting or noninsulting. With reference to our example, the PC side (the essentialist faction) declares that nigger is definitely insulting, whereas African American is not. Likewise, the essentialist on the other side of the barricade declares Black as clearly nonoffensive (noninsulting). In terms of the axiological load of the concepts, the essentialist believes that Black is inher-



ently bad (insulting), whereas the nonessentialist believes that Black may be bad to some people for some reasons and not bad to other people for other reasons. The conclusion about the central PC notion of insult is that any behavior may be insulting, any behavior may be non-PC behavior, and any behavior may be interpreted as discriminating against some minority. Any expression, word, joke, or comment may be taken as an insult to somebody. The designation, to somebody, should be noted because, as will be remembered, the view promoted in this book is that meaning and pragmatic function (insult, in our case) are imputed to words and not resident in them. It is crucial to keep in mind that the nonessentialist’s position is not totally relativist. Although the nonessentialists will acknowledge the speaker’s and the hearer’s right to his or her ultimate verdict as to what counts as insulting or noninsulting behavior (in general, always, or on a particular occasion), we are restricted in our verdicts (as to what counts as an insult) by the experiences that we share with others and by the definitions of concepts that these shared experiences lead to. The relativist aspect of the matter is the fact that no one can be denied the right to label a particular behavior as insulting. The nonrelativist (experientialist in cognitive linguistics terms) aspect of the matter is that although labeling some behavior as insulting is seen by others as proper labeling as such (e.g., calling a university colleague “a filthy pig”), labeling other kinds of behavior as insulting would be seen as utterly odd (e.g., calling a university colleague “my dear neighbor”). Consider Johnson’s (1993) opinion about the issue, with respect to the notion of morality: There are universal human experiences of pain, pleasure, suffering, joy and fulfillment. There are universal human needs, such as shelter, food, love, and protection from harm. There are widely, if not universally, shared moral prototypes (e.g. the bully, fair distribution, undeserved kindness, etc. (p. 259)

Building on this opinion, one may easily surmise that we share with others our view of what constitutes insulting behavior. We share, to some extent at least, the prototype of insult. That is, within the rough limits of what constitutes our culture, there will be a significant amount of agreement (but never total, 100% agreement) as to what is an insult or what constitutes insulting linguistic and nonlinguistic behavior. In this sense, nonessentialism does not equal total relativism (although it is fairly robust relativism, one has to admit). It needs to be remembered that this prototype view always leaves the final decision as to what is insulting or not to the speaker/ hearer. There is no such a thing as inherently insulting language or insulting words. Whether something is insulting is decided on by the living human being. That human being is in his or her decision conditioned (but



never deterministically determined) by culture. This prototype view of insult obviously also leaves room for certain kinds of behavior to be labeled partly insulting, and for others never to be labeled in any way at all. Another controversial issue within the PC debate is that of the canon ( Jardine, 1994). The analysis of insult presented earlier can be easily expanded to the canon question. As is known, the crux of this part of the PC debate is which pieces of literature should be included in the canon recommended to the students. Interestingly, the canon question could, in turn, be easily expanded to cover the debate about what courses to teach. For a nonessentialist, the situation is very simple: All the arguments about what to teach and what items to include on reading lists reduce to simple what-is questions: “What is valuable literature?” and “What is a valuable course?” If one applies the relativist and experientialist reasoning that we applied for insult before, then one is likely to arrive at the following conclusions concerning courses and canons: Arguments for any one course or canon being definitely better than others, in some absolute terms, clearly reflect essentialist thinking. There is no such thing as a good (valuable) course or a good (valuable) canon in its own right. Any argument for or against either of the two is relative to our understanding of the concepts included in the questions. A valuable course is a course valuable to you, as individual readers of this book or to me, the author of this book, whatever the reasons are. We all obviously tend to agree, to some extent, with others as to what a valuable course is because we all share with others, to some extent, the experiences that underlie the concepts concerned. Ultimately, however, the basis for deciding whether a course is valuable or what to include in a canon is an individual’s own opinion, just as only individuals decide what counts as insult. The typical essentialist thinks, however, that certain novels, plays, or poems are inherently valuable and that that is why they should be known to everybody. Trying to better understand the essentialist’s line of reasoning, let us concentrate briefly on a related example—namely, that of music. Essentialist classical music lovers often feel that classical music is inherently valuable, whereas nonclassical music is not (there of course arises the question of telling the difference between the two, which the essentialist finds very difficult to define). They express their respect for people’s interest and expertise in classical music. When they are told that all kinds of music other than classical can be equally valuable, inspiring, educating, entertaining, relaxing, and so on, they tend to disagree. They claim that one should not treat heavy metal music as in any way equal or as equally valuable as classical music, for example. For an essentialist classical music lover, the argument that for some people heavy metal music may be valuable inspiring, educating, and relaxing is strange and unconvincing. For an essentialist, heavy metal music cannot be valuable regardless of whether



others listen to it and like it. In other words, the essentialist classical music lover will insist that his or her interests are more precious than those of a heavy metal fan. The essentialist deeply believes that classical music has inherent positive value that other types of music do not. This reasoning about music corresponds to that about the canon and curricula. Our previous discussion should certainly be seen to some extent as integrating judgments about language and art. Given the assumption that the cognitive linguistics version of nonessentialism is a rewarding view of language, two questions may be posed at this juncture. One is whether it is a legitimate procedure to try to introduce new words the way the PC advocates do. The other question is whether these new words can, if introduced, indeed change the state of affairs—that is, whether they can change the inequality that they are intended to change. A straightforward answer to both of these two questions is positive. To elaborate on this answer, we need to return for a moment to our earlier discussion of the axiological load of concepts. What may be assumed many PC advocates try to do is to replace some of the words that are axiologically marked (by groups of people) as bad, or close to bad, with words that are neutral or positive (marked good). If we assume, for instance, that fat is marked negatively on the axiological scale of good–bad, we can assume that differently sized is neutral on this scale. It needs to be remembered, however, that the word fat and the concept that stands behind it may by all means be marked positively by some people—that is, it may carry the axiological load, good. As was mentioned earlier, we may share a prototype with other people, but we may not share the axiological load of that prototype. In simpler terms, an individual may share with others his or her idea (mental description, mental image, prototype, meaning) of what a fat person looks like, how she or he behaves, how she or he dresses, and so on, but this individual may differ from those other people as to whether the label fat is positive, negative, or neutral. It must be remembered, both positive and negative are scales. Keeping all these facts in mind, the reader will hopefully agree that culturally (i.e., with respect to a group, with respect to the group constituted by the majority of the speakers of English) garbage collector is more negative than refuse collector. If that is indeed the case, the proposal to replace garbage collector with refuse collector is sensible because the culturally more positive axiological load of refuse collector may lead to, turn into, or generate a more positive attitude on the part of the language user toward the people who are labeled refuse collectors. Obviously, there are no safeguards against refuse collector taking over the negative axiological load from its negatively charged predecessor. This is, in fact, what does happen with some new proposals. This is what first happened to Negro (which took over the negative load from nigger) and subsequently to Black. We may wonder whether African American is next in line.



The basic question being tackled here is that of whether language can change social reality. It is clearly one of the crucial questions within the PC debate. It may seem that language certainly does not change social reality if each newly introduced word gradually assumes the negative load of its predecessor. It can be argued, however, that at least to some extent language can change (influence) social reality. For example, consider the change in social reality that switching to first-name terms in some languages (e.g., German, Polish) involves. In Polish, switching to first-name terms generates a number of social rights and obligations (e.g., the right to making a late-night phone call) significantly different from those characteristic of the period prior to the switching. As for possible social changes instigated by PC proposals, new (PC) words should be seen only as proposals. Indeed, a social change may be brought about by promoting (and never imposing) a word that is axiologically more positive than other words having the same reference. The matter may in fact be more complicated. The word cripple (as opposed to handicapped and disabled) is a telling example in this respect. Namely, one can claim, that, on one hand, the term disabled has been used until recently as a relatively neutral word. On the other hand, one can claim that there has also been a change in attitudes at the same time. One can claim that the relatively neutral axiological load of disabled is due to the attitudes that changed things rather than the use of new words. We may certainly be tempted to ask the question of which came first. The significance of attitude change should certainly be recognized here. Which comes first may remain a question never to be resolved. Not all PC proposals fit in the pattern outlined earlier, and not all such proposals make sense. If one accepts the reasoning presented before, the proposal to call, for instance, a girl a prewoman can only be taken as a joke. In fact it is treated as a joke by the PC language advocates (see Beard & Cerf, 1992). Very likely, this example and other similar examples were coined by the enemies of PC language to make the whole idea of PC language look ridiculous. Let us briefly pause at the question of why prewoman should be treated as a joke. In other words, let us ponder the question of how it differs from other words proposed as PC substitutes. Girl and prewoman differ from pairs such as Black and African American in that, in the former, girl is not negatively charged, whereas in the latter Black often is. So, with respect to girl, the substitute, prewoman, is clearly redundant. In case of the latter pair, there is an axiological proposal for a change to be made. This kind of proposal cannot be made in the case of the former pair. This difference may indeed be crucial, and it may constitute an important criterion for many people to treat PC proposals as jokes or serious endeavors. Consider other examples, such as extra large versus generously cut, dead versus terminally



inconvenienced, pet versus animal companion (Beard & Cerf, 1992). There are obviously other types of proposals that are equally seen as jokes for different reasons. Examples are stupid versus cerebrally challenged and lie versus terminological inexactitude, where the humorous effect is perhaps brought about by the technical overtone of the proposed PC expression. Another question is whether PC words/expressions differ from the nonPC ones in ways other than with respect to the axiological load. The answer seems to be positive. Let us consider categorization in general. All categorization involves highlighting an aspect of the object that we are referring to (Aitchison, 1994b). No word can describe or refer to all the aspects of a given object or phenomenon. For example, if we say, “How much did you pay for this jalopy?,” we highlight the fact (aspect of the object) that the vehicle is old and looks bad. At the same time, we disregard the fact (aspect) that the car may be very comfortable, or other facts/aspects, such as that it is fast or cheap to run. When we ask the question, “Where did you get this guzzler?,” we highlight the fact (aspect) that the car uses a lot of gasoline, and we also disregard other facts, aspects, or characteristics of the car. It may be argued then that new word proposals are primarily new proposals for highlighting aspects of a person or thing that were not highlighted with the old term to be replaced. This is obviously the case with Black (color of skin highlighted) as opposed to African American (place of origin highlighted). The significance of the shift of the highlighted aspect should be appreciated irrespective of the fact that, in cognitive linguistics terms, the newly proposed word (and the mental image behind it) will be grappled with partly in terms of or relative to the mental image invoked by the old word. The main claim with respect to the question of highlighting is that use can be made here of the highlighting options that the language offers to use. In that way, words may be thought of as attempting to rechannel or redirect views, opinions, and values. That rechanneling may contribute to social equality and tolerance. Yet it must be remembered that no intention at all is being expressed in this book to encourage any imposition on the language user (in any form of prescription or proscription) of any one of the options available. Such proposed word replacements, which intend to have the speaker focus on one highlighted aspect of a person or an object rather than on another, are in line with the nonessentialist view of language. Manipulating both the culturally shared axiological load of a concept and the aspects highlighted by a particular word are consonant with nonessentialism. These two types of language engineering may obviously be combined. Whether such proposals will, indeed, work in a particular social situation is another question, but it can be trusted that the chances that they will work are quite good provided, it must be repeated, that no imposition of any sort is involved.



The prior discussion of and proposal for the solution of the PC conflict indicates that changing names for people and things may be a rewarding task. It may work, and it does work, at least sometimes. The earlier proposal is not only philosophically sound, but also practically feasible. We have, in fact, quite a lot of evidence from areas other than the PC debate that changing the name of a language, for example, may bring about a significant change in the social sphere. Although names are only names, they may be important. The Albanian Greeks speak what is now known as Arvanitika (clearly a new name for Albanian). Nowadays, speakers of Arvanitika are no longer considered to be Albanian, although they used to be thought of as Albanian about 150 years ago. The name change and active work toward changing the identity did work (Trudgill, 2002). Let us consider another example—Afrikaans. About 70 years ago, Dutch was spoken in South Africa. Now they speak Afrikaans there. In fact, the major difference between Dutch and Afrikaans is the name! The name of the language appeared to be quite important, however. Afrikaans has managed to gain autonomy (Trudgill, 1996, 2003). Consider also the following passage: What’s in a name? A great deal, apparently, if the subject in question is experimenting on people. When America’s Advisory Committee on Human Radiation Experiments (ACHRE) asked 1800 patients recently what they thought of human medical experiments, the answer was clear: they were against them. Such experiments were thought of as highly risky and potentially harmful. If the term clinical trials was used instead of experiments, however, the same group voted in favor of them. (The Economist, 1997, p. 94)

It should be remembered that a name change (i.e., proposing a new word for something) and the fact that this linguistic maneuver may bring about a social change does not at all alter the fact that the name is merely a name or a label and nothing else. In summary, although the PC conflict does not seem to be primarily a language conflict, language plays a significant part in it. Furthermore, essentialism, exercised by both sides of the conflict, may only make things worse. Still further, the conflict may be solved by promoting a PC nonessentialist position, which suggests changes in the domain of word use that in no way involve any imposition. Besides, we need to always keep in mind that the use of words prescribed by fiat obviously has nothing to do with nonessentialist philosophy. Even those new words that the nonessentialist reasoning fully allows (see the prior discussion of axiology and the highlighting options) should by no means be officially prescribed, while the old ones proscribed. The distance from such prescription and proscription to totalitarian state censorship and other calamities is short. It should be stressed that one of the crucial implications of nonessentialism is that rules, laws, and regulations, of whatever kind, should be kept to a minimum, and



that they should never be introduced in any authoritarian fashion. This last statement also applies to linguistic matters.

MISCELLANEOUS If one keeps track of political events throughout the world, it becomes clear that some of the events have a linguistic aspect to them, a fact that is rarely acknowledged. Even when such a fact is acknowledged, no recourse to any linguistic expertise is taken. What follows in this chapter is a listing, with brief comments, of three recent political events with a linguistic aspect— events that could be well understood and addressed were they only approached in a nonessentialist fashion. The penultimate section takes up the question of how essentialism breeds a feeling of superiority felt to naturally belong to one language, or dialect, rather than to another, which may have severe political consequences. After the dissolution of the Yugoslav state in the early 1990s, Macedonia, which was part of the former Yugoslavia, became independent and wished to retain the name of Macedonia. As is known, the northern part of Greece is also called Macedonia. As Newsweek reported about the Greek Macedonia: The 16-pointed star of Macedonia has sprouted everywhere in Greece. Businessmen display it on suit lapels. Jewelers fashion it on rings, brooches, and watches. Bakers hawk Macedonia crackers. Billboards display the star with the slogan, MACEDONIA—GREEK FOR 3000 YEARS. All this because the former Yugoslav republic of Macedonia has declared itself an independent state with the same name. The Greeks are outraged: Macedonia to them is the ancient empire that was built by Phillip II and his conquering son Alexander the Great—not the land now peopled mostly by ethnic Slavs and Bulgars. The Greek government is fighting to head off international recognition of the new nation until it finds some name other than Macedonia. . . . Greeks oppose any use of the word “Macedonia” by Skopie, as in such suggested names as “Slavic Macedonia” or “North Macedonia.” (Stanger, 1992, p. 20)

This conflict appears to have an essentialist slant/flavor on the part of the Greeks. The reasoning behind the Greeks’ objections is that only one part of the land formerly, in the two countries, called Macedonia (theirs) has the right to be called Macedonia. In this reasoning, the word Macedonia is understood to be appropriately used for one plot of land and not for another. The essentialism underlying this reasoning reduces to the belief that there is a natural connection (and not arbitrary and conventional) between a given plot of land (the Greek one) and the name, Macedonia. The Greek plot of land must be called Macedonia, and no other plot of land can be called so; this is the Greek essentialist position. The historical argument on



the part of the Greeks certainly invokes an etymological fallacy; that is, the view that the original meaning of a word is the only correct one (Lyons, 1981). The nonessentialist solution to this conflict seems simple, at least in theoretical terms—namely, no country or political group should be seen as having an exclusive right to use any one name (as much as it might be historically justified). It must be remembered that words (Macedonia should be seen as a word) are only labels to refer to other things, and these labels are used largely arbitrarily and conventionally. Therefore, there is no reason that one should insist that the word Macedonia cannot be used to refer to the Yugoslav part of the ancient Macedonia. For practical reasons, however, to avoid having two countries with the same name, the parties concerned could have certainly negotiated various alternative solutions instead of the Greeks insisting that only their Macedonia can be called so. After some of the al-Qaeda terrorists were transported to Guantanamo Bay where the Americans kept them prisoners, a legal question arose as to what these people should be called in legal terms and, accordingly, how they should be treated. This question arose partly due to some objections raised against the way these prisoners were treated by the American authorities. As The Economist (2002b) reported: The Pentagon says they are “unlawful combatants,” but there is no such category in international law. Detainees are either ordinary criminals and should be charged as such, or prisoners of war, protected by the Geneva conventions. There is no third way. With the al-Qaeda prisoners, the Pentagon has a case, though it is not making it particularly well. The attacks on New York and Washington were acts of war by any normal definition. So was the conflict in Afghanistan (the congressional resolution that preceded it was, in effect, a declaration of war). The al-Qaeda detainees are not ordinary criminals. But neither are they ordinary prisoners of war. International law gives POWS special protection on the grounds that the responsibility for soldiers’ actions rests with the state they are fighting for. Al-Qaeda is not a state. Many in Guantanamo Bay are not Afghans, so if they are prisoners of war, which state was responsible for their actions: Pakistan? Britain? Of course not. (p. 46)

For a nonessentialist, the situation seems relatively clear, and the problem arising from it is easy to solve. The Guantanamo prisoners do not fit in the categories criminal and prisoner of war. Neither of these two linguistic labels grasps the nonlinguistic reality (the people and their terrorist deeds) in question. The nonlinguistic reality is new (the type of attack on the World Trade Center on September 11, 2001), and there is no legal name for the perpetrators of the act involved. This legal and linguistic situation may indeed be initially difficult. However, although the essentialist might be concerned about this linguistic predicament, or, although she or he might be



pressing the new nonlinguistic reality into one of the two existing legal categories, for the nonessentialist, the solution appears to be less of a problem: A new legal category of international criminal must be devised and a new name given—that is, a new linguistic label must be devised. Another possibility is that one of the two originally existing labels (criminal and prisoner of war) gets redefined in legal terms, which would then suffice to include the New York City type of terrorists. The preparations for the second war with Iraq and President Bush’s disputed right to strike Iraq involved an interesting linguistic aspect. Namely, Under the constitution, the president is commander-in-chief and can direct American forces on his own, so long as war is not declared. The trouble is that the constitution gives the president power to “repel sudden attacks,” but not commence an offensive war. Thus the crucial question is whether a “preemptive war” is offensive or defensive. Mr. Cheney seemed to assert the latter: “We realize that wars are never won on the defensive. We must take the battle to the enemy.” (The Economist, 2002c, p. 34)

In other words, the question focused around the definition of preemptive war. Some people claimed that it is defensive and others that it is offensive. For the former, an American strike on Iraq would be justified; for the latter, it would not. As the reader will know by now, there is no one good and absolutely correct definition of preemptive war, as the essentialist might think. The nonessentialist, however, will be aware that different definitions may be promoted, and, like in the Clinton–Lewinsky case, the different parties involved will argue for and press their own definition to reach their political goals. The winning party will have imposed their definition on others by whatever means. One of the points made earlier was that essentialism generates a feeling of power and superiority. Power is clearly reflected in language contact and political-conflict situations. When language and politics are discussed, one of the fundamental facts that needs to be remembered is, “Military conquest appears to be one of the most important factors in accounting for the spread of language” (Wardhaugh, 1987, p. 7). Military action, the feeling of one’s power, of one’s infallibility, superiority, righteousness, the stance that my views and ideals are right, definitely right—all of these come directly from and do not fortuitously co-occur with conceptual essentialism. Strong evidence for this view comes from the history of the spread of French: “To the French an important part of their imperial mission was the mission civilisatrice (civilizing mission), the desire to create yellow, brown, and black French people having the same ideals and views as those of metropolitan France” (Wardhaugh, 1987, p. 7; italics original). The attempted imposition of the one, good and natural way of the speakers of one language on the speakers of another language has been a



familiar trend in the history of language contact. In 1784, the French writer, Antoine Rivarol, won a prize offered by what was then called L’Académie de Berlin for a text in which he glorified French and its virtues. Part of the text runs as follows: There has never been a language in which you could write more purely and more precisely than in ours, which is more resistant to equivocations and every kind of obscurity, more serious and more gentle at the same time, more suitable for all kinds of styles, purer in its phrases, more judicious in its expressions, which has a greater liking for elegance and ornament, but which is more fearful of affectation. It knows how to moderate its strengths with the modesty and restraint it must have in order to avoid the monstrous expressions in which our neighbors today are fixed. There is none among them which is more attentive to number and rhythm in its declarations, which is the true mark of perfection in languages. (cited in Wardhaugh, 1987, p. 101)

Such opinions about one’s language are by all means possible today; the previous quote, which is totally unacceptable to the overwhelming majority of modern linguists, has more than historical value. The probability is high that very many lay people would now subscribe to Rivarol’s views, except that they would substitute the name of their own native language for French. Wardhaugh also thinks that the French have not learned much since Rivarol’s time, and they still view their language as inherently superior to others. He ironically summed up his view about the French in the following way: The particular attitude that the French have toward their language also encourages them to want others to learn it. A language praised for its clarité, beauté, perfection, stabilité, utilité, and simplicité, just to name a few of the «virtues» which have been ascribed to it, obviously deserves widespread dissemination. (Wardhaugh, 1987, p. 118; italics added)

In light of what has been written about French, it is not surprising to see that some people express quite explicitly their inner conviction about the superiority of English. For instance, Newmark (2002) said: “I think that the English language is superior to all others in its range and resourcefulness in rendering thoughts, emotions and observations . . .” (p. 64). Phillipson (2003) believed that many English-speaking people have such a conviction about English. Political conflict situations that involve the belief that one language or dialect is viewed as intrinsically better than others are not infrequent. Examples of such situations may easily be provided. One comes from the contact situation involving Catalan and Spanish. Catalan, a language much denigrated and suppressed until Franco’s death (Franco is reported to have



called it “the language of dogs”; Wardhaugh, 1987), has only recently been recognized as a language of culture with long traditions. Observations and conflicts about the use of pidgins and creoles also give us relevant insights. As is known, pidgins and creoles usually meet with contempt motivated by the conviction that they are inherently (essentialistically) inferior to West European languages. Interestingly, this contempt sometimes goes hand in hand with a feeling of pride that the denigrated pidgins and creoles derive from the writer’s or speaker’s own superior native tongue. Thus, while complaining that the “natives” have rendered the European language a “debased mongrel jargon,” “a crude macaronic lingo,” “a perversion,” or one of the many similar expressions found in the relevant literature, even in its debased form the European language remains superior to what the “primitive natives” had before. (Mühlhäusler, 1986, p. 27)

This kind of superiority complex is unfortunately frequent. In the political context, views that stress the superiority of one language over another are not rare. Obviously, some people may voice such opinions strictly for political reasons without really believing them. However, with the evidence that we have, we may safely assume that many a speaker of whatever language will, indeed, believe that his or her own language is inherently good, better, or superior. This position is significantly essentialist because believing that one language is superior to another implies, among other things, beliefs that one knows the right way, that his or her language is proper, and that other people’s languages are not, and that his or her language expresses things in an adequate way, whereas other people’s do not. To conclude the present section, let us briefly dwell on how the question of the superiority of one language/dialect over others relates to the essentialist way of viewing symbols and the reality to which these symbols refer. Most important, the superior position that one gives to one’s own language or dialect stems from the essentialist belief that if symbols are reference systems, there is only one correct way of referring. This is obviously closely related to the misconceived idea that the map is the territory (Hayakawa, 1965). More important, however, the essentialist belief that there is only one way of referring is different from the belief that the symbol used to refer is actually the thing to which it refers. The two beliefs should not be conflated and confused, but they are related. Namely, it seems that people who believe in only one correct way of referring may ultimately be led to believe that the symbol that they use to refer is, in fact, the thing itself. It is some form of animism—a belief in the power of words, in word magic (see Popper, 1972). What is relevant to this part of our discussion is the belief in the one symbol—one thing relationship rather than the belief in the symbol is the thing



view (the map is the territory). So we need to keep in mind that a symbol, in this essentialist view, is taken to be the symbol. Other symbols (another word, e.g., used by different people) are understood to be entirely wrong, partly wrong, less adequate, strange, unfortunate, imprecise, and so on. Such other symbols are taken to be not the symbol that the speaker believes to be the appropriate one for the thing to be referred to.

CONCLUSIONS: FROM ESSENTIALISM TO CONFLICT AND WAR First of all, we should not be tempted to believe that language alone (or rather specific, essentialist, beliefs about it) directly causes political conflicts, including wars in which people get tortured, humiliated, and killed. Wars are waged because people can be greedy, selfish, and self-righteous. Wars are waged because people can want more territory. Wars are waged for economic reasons. Wars are waged because people can want to control other people. Wars are waged for religious reasons. Of course, all of this is true. One could go on listing other related reasons for which wars are waged and millions of utterly innocent people suffer and die. Given all of the causes of wars, what is claimed in this book is that people’s essentialist beliefs about how language works may contribute to conflicts, including political conflicts, and to the warlike frame of mind. Essentialism enhances the likelihood of any conflict, including real wars. It is not possible to establish a clear connection between languagerelated conceptual essentialism, on one hand, and dogmatism, on the other hand. Dogmatism is mentioned at this juncture because dogmatism is an important reason for conflicts. Dogmatism is defined as a positiveness in assertion of opinion, especially when unwarranted or arrogant. It would, however, be implausible to try to establish the relationships among conceptual essentialism, dogmatism, intolerance, self-righteousness, and a set of other related concepts. One needs to remember that they are all only highly abstract words, and only words, that may mean many different things to different people. If we attempted to establish the relationships in question, we would get into a verbal argument very quickly. We would easily climb high up the “ladder of abstraction” without referring to any tangible reality. In other words, it is impossible to establish an absolute difference between conceptual essentialism and dogmatism. One might suggest that dogmatism be viewed as a label more general than conceptual essentialism. But obviously, the relationship could be seen differently. For instance, it could be claimed that there is no or hardly any difference between the two. Recapitulating, the contention advanced here is that establishing a specific number and kind of conflict sources is a futile task. In this book, we are con-



cerned with conceptual essentialism as a source of conflict, and not with other sources. To repeat, the main claim of this part of the book is that the essentialist beliefs, especially those that pertain to language, significantly increase the likelihood of conflicts of various kinds, including political conflicts leading to wars. It needs to be added that people are usually unaware of the fact that their beliefs about how language works are so important, maybe even crucial at times. Philosophical assumptions concern most or all of human activities. Language should be seen only as one of these activities. However, as language permeates most other activities, beliefs about it bring about critical consequences. The fact that political conflicts and wars are usually waged by politicians might lead us to the claim that this is because politicians are often essentialists. Indeed it seems to be the case that in the essentialist group, politicians have a very strong representation. If this is indeed so, the close relationship between essentialism and war becomes evident. That is, politicians, like any other essentialists, will tend to think that their understanding of words is the correct one. They will tend to think that their understanding of freedom, democracy, allegation, cooperation, negotiation, independence, allegiance, loyalty, and many other political concepts is the correct one. This does not mean, of course, that all wars are sparked by conceptual essentialism. Obviously, as mentioned before, wars are often waged for blatantly economic and for other reasons. Some may be sparked by the insanity of leaders. It is nevertheless claimed here that essentialism might be seen as one of the causes of war or warlike behavior. If we put together two politicians, two groups of politicians, or any group of negotiators, then very likely we will be dealing with more or less significantly different understandings of all the relevant concepts. Given the inevitable differences, if any one of the parties believes that his or her understanding of the concepts is the right one, political conflicts are likely to ensue. They may first be mild and insignificant, but may then gradually, step by step, assume more acute forms, including major military confrontations. With reference to power, which was mentioned earlier, it needs to be realized that if one is convinced that she or he is right, definitely right, and that someone else is definitely wrong, in the political domain, she or he may decide to fight—to attempt to seize power and to rule over others. As far as nonessentialist politicians are concerned, one may seriously wonder whether there are many; most politicians appear to be essentialists expressing dogmatic positions. There seems to be little room for nonessentialists in politics. However, politicians do not need to be essentialists (i.e., they do not need to be philosophically committed essentialists) for some form of essentialism to play an important role in politics. Intelligent politicians may be aware of the fallacy of essentialism, but at the same time



they may take advantage of the fact that many of those that they represent are essentialists. Such politicians may easily manipulate the meaning of words, and the uninformed and/or confused electorate may arrive at unrewarding beliefs, not necessarily short of the Orwellian “Peace is war.” This manipulation and its consequences might be possible because of the widespread belief that definitions may give us crucial information about the essence of things. Asking for the definition of peace, “What is peace?” may bring us to “Peace is war”! Obviously, politicians would not be elected if they did not have the answers. But the answers that they have should be understood and promoted as the best ones under the circumstances; the best answers arrived at in a social process and agreed on by a group of people. This is very different from having the answers understood as absolute and universally valid. A political stance that fits in the picture of essentialist politics depicted earlier is ethnocentricism or chauvinism. Both generally spring from the conviction that one group of people is definitely better than another group. If one claims that she or he knows what “better” is, she or he also needs to be able to claim what “good” is, and here is precisely where chauvinism closely relates to essentialism. If, one day, someone were successful in persuading the chauvinists that good, better, and worse may take hundreds of different meanings, they would most likely cease to be chauvinists. Needless to say, in connection with chauvinism, one of its consequences is conflict leading, possibly, to war. It seems that essentialist politicians win over nonessentialists not only in totalitarian systems, but also in democratic states. This may be a bit surprising. It is probably so because the fundamental distinction between totalitarian and democratic systems concerns the way one governs and comes to govern and not the type of people that govern. In other words, the distinction between totalitarian regimes and democratic systems is one not so much about people as about formal ways of coming to govern and of governing a country. Also here, it must be admitted, we may acknowledge the existence of degrees of essentialist thinking. Some leaders in a democracy have more of an essentialist frame of mind than others. The same seems to be true for totalitarian countries. The changes in Eastern Europe toward the end of the 1980s come to mind in this context, and the recent history of Poland may again serve as a good example of how switching from communist totalitarianism to Western-style democracy does not at all entail moving from essentialism to nonessentialism. Poland was first strongly, and later moderately, totalitarian between 1945 and 1989. Since 1989, it has been a democratic country, where both the parliament and the president are chosen in free elections. Interestingly, in both the former and latter periods, there have been many politicians who expressed firm opinions about what “good” was, what “the truth” was, what “equality” was, what “justice” is, and



so on. Likewise, in countries that are reputed to have been fully democratic for ages (e.g., Great Britain), essentialist politicians claiming that they are definitely right hold important posts. It seems that no matter how politicians gain power (e.g., through democratic elections or through a coup d’état), once they have obtained it, they will tend to reveal and make use of essentialist reasoning. A glimpse at the world’s political systems will bring one to the conclusion that, in all systems, those in power tend to claim to know what is true, what is moral, what is good, what is evil, and so on. We can see quite clearly that almost every group that has political influence claims to know for sure what the crucial concepts really are and what the words that refer to them really mean. Let us reiterate: The belief in knowing something definitely (e.g., knowing what is good, what is bad, etc.) promotes the belief in one’s significance, in one’s greatness, which in turn may lead to one trying to seize and/or overuse power. It may lead to fights, to persecution, to war. One might wonder again about what proportion of the conflicts that we are discussing here are language related. As mentioned earlier, it is impossible to completely separate language from other aspects of conflict. Still it is very hard to disregard language as a crucial factor; the role that it seems to play in some types of conflicts is crucial. Although we need to remember that language cannot be clearly and entirely separated from other aspects of the conflicts, it often may be seen as a vital and perhaps critical factor. In connection with this, let us briefly refer again to the sociopolitical situation in Poland. Countries such as Poland are very interesting places for a linguist to study because of the recent sociopolitical turmoil. Until 1989, Poland was a communist country where bipolar distinctions of the blackand-white kind were the order of the day. As the reader might guess, in the official propaganda, the communists were good, and the anticommunists (mostly perceived as Catholics) were bad. That was obviously true for the communists. For the anticommunists, however, as might be expected, the communists were bad and the noncommunists were good. Either way the situation was perceived, the distinction was usually bipolar. The people who were neither of the two, neither communists nor Catholics, were (and many still are) perceived as strange and noncommitted. There is very little room, if any, for a gray character in a black-and-white world. In a black-andwhite world, there are “us” and “them.” Anyone who does not fit one of the two categories may be stigmatized and viewed as suspicious. This way of viewing people led, as might be expected, to conflicts. It may then be reiterated that conceptual essentialism, which promotes bipolar categorization, may be one of the reasons for conflicts in general and for political language-related conflicts, in particular. To dispense with such conflicts, one has to dispense with the essentialist view of the world. However, that is much easier said than done. Because essentialist thinking



goes back to Plato and Aristotle or earlier, and because it is deeply ingrained in us, changing the thinking in question is a very hard task. We cannot be too optimistic about that change in the near future. Most presentday institutions (e.g., schools, ministries, cultural institutions) are built on the basis of essentialist philosophy. Most everyday functioning all over the world is based on essentialism and two-valued orientation. Language-related conflicts are, at least partly, a function of that orientation. Doing away with language-ridden conflicts would thus involve changing people’s minds, which, as is well known, is very difficult. Thus, we should not be too optimistic about the feasibility of a possible change of mind. However, this book is a modest attempt at changing the present state of affairs. To what extent this attempt will be successful remains to be seen. Unfortunately, the reasons for the restrained optimism, or rather pessimism, mentioned earlier are clear. They can be traced to the following reasoning: When two essentialists disagree on the definition of concepts/words, they usually get involved in a linguistic battle. A physical battle to follow cannot be ruled out. In such a situation, one of the essentialist sides is forced to give in or is eliminated. In a real war, which can be taken to be an extreme example of a conflict, one of the two essentialist sides dies or surrenders. If, in turn, an essentialist fights a nonessentialist, it is usually the nonessentialist who is in a more difficult situation. The essentialist is usually certain, determined, and convinced of his or her righteousness. The nonessentialist is in doubt. She or he allows alternative solutions, and she or he is willing to compromise. The nonessentialist is not allowed to say that she or he is definitely right, which, for all practical purposes, weakens his or her position. As is well attested, the essentialist tends to claim that she or he is definitely right. The nonessentialist, as opposed to the essentialist, tends to be much more openminded and tolerant. In an intellectual battle with an essentialist, the nonessentialist cannot even attempt to impose his or her viewpoint on the adversary. That would be a betrayal of the nonessentialist philosophy. Having reviewed this material, it would be logical to ask whether a nonessentialist can be a successful politician, director, or manager. All these positions seem to require the exercising of at least a certain amount of power or authority. If one appoints a nonessentialist politician at the top of a government, for instance, in the position of the prime minister, we may suspect that the person will be in doubt all the time, that she or he will hesitate all the time, that she or he will be claiming all the time that things can be done in one way, in another way, or still in another way. In other words, we may suspect that the person will not be perceived as a person who knows what to do. She or he may then be perceived as incompetent. In the meantime, a determined essentialist politician, who will not hesitate in his or her decisions, could appear and replace our dithering nonessentialist politician.



This scenario repeats a point made earlier—namely, that if a nonessentialist fights an essentialist, it is the nonessentialist who is likely to lose; to die, perhaps. In a somewhat ideal world, where we would all, or the majority of us at least, be nonessentialists, things would be easy to handle, and no danger would exist. That is, “power jobs” not only may, but in fact should be taken by nonessentialists. Then, of course, they would cease to be power jobs. The nonessentialist would direct, govern, or rule by a moral or formal code that will be understood by him or her as one of the many codes available. One particular code will be arrived at by social consensus, for instance, by majority rule. Although our nonessentialist politician would believe that the code should be followed for the benefit of others under given circumstances, she or he could not believe and try to persuade others in the society that this code is the only good one, the correct one in some absolute sense. As opposed to the essentialist politician, manager, or director, our nonessentialist will claim that his or her political proposals or suggestions for practical implementation are formed on the basis of his or her best knowledge and analysis of the domain in question, and that they should ultimately be traced back to a social judgment. The critical part of this stance is that if the proposals are rejected by others, the nonessentialist will resign from the post rather than shoot the opponents!

ENDNOTES 1. All the English-language quotations in this part of the book are my own translations from the original Polish. Note that some of the quotations do not reflect the opinions of the authors; in such cases, journalists are simply reporting on what others have said. Unfair as it may seem to present an opinion that the journalist does not share along with his or her name, I cannot think of a better way to provide the necessary bibliographical information. 2. A question raised during the parliamentary debate on abortion held on May 16, 1991. 3. A statement made during the parliamentary debate on abortion held on May 16, 1991. 4. Polish Radio Broadcast, January 25, 1991. 5. A statement made during the parliamentary debate on abortion held in January 1993. 6. A statement made during the parliamentary debate on abortion held on May 16, 1991. 7. For a recent discussion of moral relativism, see Levy (2002). For a discussion of morality significantly informed by cognitive linguistics, see Johnson (1993).



Language in Law

Essentialist reasoning often surfaces in the everyday practice of the legal profession. This reasoning is reflected in many different ways. One is a set of misconceptions about defendants, and another is a set of misconceptions about language. Still another materializes as a set of instructions available in the handbooks on interrogation commonly available in legal institutions; and still another manifests itself through the ideological affiliation of the lawyer. The following sections discuss each of these ways in which essentialist reasoning in law is realized. Most of the material illustrating the main claims comes from Roger Shuy’s (1993, 1998) work. MISCONCEPTIONS ABOUT DEFENDANTS As Shuy (1993) rightly pointed out, a person preparing to listen to another person on a tape approaches the task with a number of preconceptions about the event recorded and about the people whose deeds are to be analyzed. The three misconceptions about the defendants are the following: 1. If they are on tape at all, they must be guilty of something; otherwise the police would not have been after them. 2. If they are guilty of one of the charges, they are probably guilty of the other charges as well. 3. The defendants hear, understand, and remember everything said by the agent or other persons in the taped conversation. (Shuy, 1993, p. 3) 131



Although these three misconceptions may be viewed as springing from essentialist reasoning, it is Misconceptions 2 and 3 that clearly stand out in this respect. Misconception 2 in particular illustrates very well the binary opposition that essentialism promotes: Someone is either guilty or not guilty; someone is either guilty of no charges or of all charges; all or nothing; black or white. This reasoning is no doubt a token of the essentialist clear-cut view of concepts. This view leaves no space for people being guilty on some charges and not on others; this view leaves no space for fuzziness and guilt to a degree, or on some charges only. The “all-or-none” essentialist reasoning may be, as expected, harmful to the defendant. As Shuy (1993) reported, defendants are often judged innocent on all charges when the evidence clearly indicates that they are guilty on one or more charges. More disturbingly, perhaps, defendants are frequently found guilty on all charges when the evidence clearly shows that they are guilty only on one. Misconception 3 may be analyzed in a similar manner: Once the defendants hear something at all; once they understand something at all; once they remember something at all—and surely when exposed to a tape they do hear, understand, and remember something (under all kinds of ordinary circumstances)—they hear, understand, and remember everything. In other words, the reasoning is that the defendant either hears, understands, and remembers nothing or everything. As any ordinary person will always hear, understand, and remember something, the essentialist-grove thinking lets one conclude that she or he hears, understands, and remembers everything. As Shuy (1993) noted, the facts in legal practice do not at all lend themselves to the reasoning depicted earlier. “The fact that an agent says something that the microphone can record is clearly no guarantee that the target [the suspect], or for that matter anyone else, could hear” (p. 7). A lot of what is being said and recorded under most circumstances, including the legal context, is at least partially missed by at least some of the participants involved. Hearing and understanding are most often partial and selective and remembering time-constrained and incomplete.

MISCONCEPTIONS ABOUT LANGUAGE IN THE LEGAL CONTEXT The most salient misconceptions about language to be encountered in the legal context are, according to Shuy (1993), the following: 1. Meaning is found primarily in individual words. 2. Listening to a tape once will be enough to determine its content.



3. Reading a transcript of a tape is as good as hearing the tape itself. Transcripts are accurate and they convey everything that is on the tape. 4. All people in a conversation understand the same things by their words. 5. People say what they mean and intend. (p. 8) All these misconceptions are salient manifestations of essentialist reasoning. This clearly indicates that essentialist thinking in the legal environment may be rampant. Misconception 1 takes us back to one of the most fundamental essentialist beliefs about meaning. Meaning is understood to be resident in words, in some sort of abstract sense, and not to be imputed to them by people. When meaning is understood to be found in words, clear unambiguous meaning is expected to be produced and perceived. The consequences of this position in legal practice are significant, as we see in the next section. Misconception 2 may be understood as relating to essentialism in the following way: When we listen to a tape recording, we certainly understand something (rather than nothing); if we understand something, then we most likely understand everything; because words can be understood basically in one way (because meaning is understood to be in them), understanding anything is understanding everything. Hence, there is no need to listen to the tape more than once. Misconception 3 may be taken to mean that the lawyers, and especially the jury, usually believe that the written transcript (to which the legal profession assigns most significant value) is an accurate replica of what was said in the original text having been transcribed (Shuy, 1993). This leads to a related “all-or-none” type of conclusion—namely, that whatever is not part of the transcript did not take place. This essentialist reasoning does not, of course, reflect reality. On the contrary, transcripts are often selective and inaccurate, not reflecting and often dramatically distorting the events that actually took place (Shuy, 1993). Transcripts can be understood to be accurate only on essentialist reasoning—namely, only when it is believed that there is in principle only one (and correct) way of interpreting events and utterances. This essentialist way of thinking clearly ignores that any perception is a function of one’s preconceptions; that is, in Popperian terms, that any observation is theory laden (Popper, 1959, 1972). This latter Popperian view will be no doubt supported by the nonessentialist. The essentialist thinking underlying Misconception 4 is the same as that underlying Misconception 1—namely, that meaning is resident in words. Meaning is understood to be clear-cut, stable, and fixed. It is also understood not to be dependent on the individual speaker. For these reasons, people in conversation must understand (or should understand) the same



things by the words that they use. The logical connection in question is very clear. In Misconception 5, we can trace essentialism to the link between words and meaning again. The reasoning expressed in Misconception 5 implies that a particular meaning accompanies a word or, to put it differently, that it cannot be separated from the word. A meaning and a word are perceived on this view as co-existing and co-occurring in the speaker’s or the hearer’s mind. Because meaning cannot be separated from the word with which it co-exists and co-occurs, it cannot be changed. It follows that anytime a word is uttered, the meaning attached to it comes with it. As a consequence, one says what one means. Interestingly, as Shuy (1993) maintained, “It has been claimed by prosecutors in tape cases that ‘the tapes speak for themselves’ and that ‘all the jury has to do is listen to the tapes and they can easily determine what they say’ ” (pp. 3–4). We can posit very little doubt that what Shuy observed results from the pervading essentialist thinking in the legal environment. The five misconceptions listed earlier lead to serious problems. One clear and most tangible realization of these misconceptions is the form taken by instructions on interrogation. They are discussed in the next section.

INSTRUCTIONS ON INTERROGATION MacDonald and Michaud (1992) listed the following clues to deception: Brief answers Excessively delayed answers Repeating the question Rephrasing the question Hesitation in answering Memory problems Qualified answers References to honesty References to religion Softening terms of violence and theft Speaking in the third person Overpoliteness or irritability Short-lived anger. (pp. 36–38)



As Shuy (1998) rightly pointed out, these clues may be accurate in helping us disclose deception. However, as he also rightly stressed, in many cases when suspects exhibit any one or some of the behaviors listed earlier, they do not deceive at all. For instance, . . . the suspect who gives only brief answers is not almost certainly lying through concealment of information. The interrogator needs to know considerably more than this clue might provide. For example, one needs a baseline of the suspect’s normal answering style, as well as information about his or her background that would tell whether or not he or she becomes brief while under pressure, anxiety, or fear. (pp. 46–47)

If understood to provide unerring clues on deception, any one of the items on the prior list could be questioned along similar lines. What should be understood as an “excessively delayed answer”? In other words, how much delayed is excessively delayed? Delayed (by any definition) answers may be due to deception, but they may be equally well due to thoughtfulness and an attempt at answering the question accurately. Why should repeating the question be necessarily a clue to deception? Repeating the question may be an attempt at making sure, on the suspect’s part, that the question was heard and understood properly. Hesitation in answering may spring from the suspect’s cultural background and may have nothing to do with deception whatsoever. If qualified answers should be seen as deceptive behavior, most academics, who typically use qualified language, as Shuy (1998) noted, should be considered liars. Still other clues, such as references to honesty and religion, are even more difficult to accept as clues to deception. Some people feel it is appropriate to resort to the notion of honesty and others do not. Some use expressions such as “to tell you the truth” (not necessarily imputing much meaning to these words), and others do not. Some people feel that religion is a private matter not to be mentioned in any sort of public circumstances. Others allow for religion to be discussed in public. Still others flaunt it. It becomes obvious that lists of instructions like the one discussed before are highly questionable. More important, what underlies them is, in my view, essentialist thinking. It may have indeed been the case that the authors of lists of instructions such as the one earlier collected evidence for the view that their lists are accurate. That is, these authors may have compiled their list on the basis of cases where indeed the clues mentioned were clues of deception. The mistake that they are making seems to be, however, that they promulgate their list as valid for all other cases. In other words, these authors seem to believe that the way they interpret certain kinds of behavior (e.g., “reference to honesty is deception”) is correct in some absolute sense. In their reasoning, there seems to be no room for variation and



alternative understandings, perceptions, and interpretations. This reasoning may be seen as a typical token of essentialism. Consider another list of instructions offered to law enforcement agencies to help their employees find out whether their suspect is being truthful or deceitful. Providing overly detailed statements Repeating oneself spontaneously Complicating the story unexpectedly Providing marginally relevant details Giving related external associations Displaying subjectivity Correcting spontaneously Admitting memory loss Hedging Self-referencing excessively Manifesting verbosity Pausing excessively Using unnecessary connectors Using pronoun deviations such as you for I Producing disproportionate amounts of language in the prologue, central action, or epilogue portions of the narrative Producing low lexical diversity by means of type–token ratio. (Shuy, 1998, pp. 76–77) As in the previous list, any item on the current list may be subject to doubt. Although any one of them could indeed be a clue for deception, this is by no means necessarily so. Consider the fact that hedging, for example, very frequently occurs in language (see e.g., Channell, 1994). Why should it be a clue for deception? To take another example—pausing excessively— some people obviously produce and tolerate longer pauses than others. These differences are both intra- and intercultural. What is an excessive pause, then? Clearly there is no one answer to this question. Finally, when we invoke yet another example from the prior list—displaying subjectivity—the essentialist thinking underlying the list becomes obvious. The essentialism (and thus objectivism) is expressed in an entirely explicit way. The reasoning here is that the suspect is supposed to be objective. When she or he displays subjectivity, she or he deceives. In the nonessentialist view of language, this argument will certainly be seen as spurious. We should not be surprised to learn then that Porter and Yuille (1996) found, in a study



that they performed on university students, that only three of the features listed earlier show any significance. Similar to the previous list, the likelihood is extremely high that the authors of the items on this list believe that the clues are reliable universally. In my view, any such belief is a token of essentialism. The evidence that we get from a multitude of studies on variable behavior testifies to the fact that it is indefensible to treat this list as universally or even intraculturally valid. A well-known technique applied in interviewing and interrogation is the Reid technique, which is primarily a set of instructions including 15 numbered questions the interrogator should ask (Shuy, 1998). Amazingly, this technique does not consider cultural variation with respect to asking and answering questions as well as with respect to nonverbal behavior. For instance, as Shuy reported, the Reid technique disregards that most Asians and Africans avoid eye contact and that Indians avoid body contact publicly. Such avoidance is then interpreted as deceptive behavior, which of course may be a wrong interpretation. “When Indians are accused of something, they evidence extreme nervousness, body twitching, and downward glances. They are ashamed of being so ill-thought of and become quiet, passive, and helplessly humble” (Shuy, 1998, p. 148). Again, such behavior is most likely to be wrongly interpreted as deceptive. As Shuy (1998) further explained, the Reid technique also disregards intracultural social variation. For instance, encouraging the investigators to elicit brief answers, and imputing positive value to such answers, it overlooks that some professionals, such as academics, tend to elaborate their answers, especially in cases when they are familiar with the topics raised. According to the Reid technique, such elaborate answers are to be viewed also as deceptive behavior. Similarly, lengthy pauses, which happen to be normal for the Indians, are to be viewed as deceptive behavior. In addition, the Reid technique requires from the investigator that she or he clearly state on the spot whether the suspect’s behavior is deceptive. The investigator is not allowed to be in doubt. The investigator is thus presented with a set of ethnocentric instructions that are to clearly determine which behaviors are deceptive. The Reid technique leaves no space for assessing variables such as culture, profession, or personal character. Needless to say, this technique may not only be useless, but also harmful because it encourages the evaluation of culturally diverse behaviors in terms of only one set of norms (that of the interrogator’s). Behaviors assessed as deceptive, when they are not meant as such, may easily turn the suspect into a person declared guilty. The philosophy reflected in the Reid technique as well as in the other sets of instructions discussed earlier is markedly essentialist. It is essentialist in the sense that all these sets of instructions say either explicitly or implicitly that certain types of behaviors (such as lengthy answers, lengthy pauses,



lack of eye contact) are definitely deceptive, whereas others (such as short answers to yes or no questions) are not.

JUDGES’ IDEOLOGIES AND LANGUAGE It should follow from my prior discussion that essentialism is well rooted in the legal profession. Another example of how essentialism surfaces in law is provided next. The evidence for essentialism in law given so far should not give the reader the impression, however, that all law is essentialist and that all representatives of the legal profession are essentialists. On the contrary, like in any other walks of life, the legal domain seems to include typical essentialists, typical nonessentialists, and borderline or mixed cases of the two. Nevertheless, again similar to other domains, typical representatives of the essentialist and the nonessentialist categories can be easily identified. An informative study that reveals such typical representatives of essentialism and nonessentialism in the legal profession is Philips’ (1998) investigation of diversity in judge behavior. This investigation led her to identify two principal groups of judges differing in both their behavior and reasoning. One of these groups elaborates the judicial procedure and involves the defendant more. The other shortens the procedure and involves the defendant very little. These groups are referred to by Philips as procedure-oriented and record-oriented, respectively. In my own theoretical framework, the former corresponds, roughly, to nonessentialism and the latter to essentialism. Philips’ data allow her to claim that record-oriented judges are most preoccupied with making a good, precise record. “Their individual routines show little variation from instance to instance of handling their procedure, and it is their intention to handle their procedure exactly the same way each time unless special circumstances call for special measures” (Philips, 1998, p. 77). The record-oriented judges openly acknowledge the positive value of variation and flexibility, but in practice their flexibility in pursuing the judicial procedure is exceptional. In fact, they want to rely on a fixed script. “Their main reason for not varying the procedure is to make sure they get everything said that must be said . . .” (Philips, 1998, p. 77). They give relatively little information to the defendant and tend to limit the number of the questions addressed to him or her. The record-oriented judges are not interested in giving the defendant an opportunity to provide a wider context that may shed light on the case. They often engage in asking yes or no questions, which are expected to provide clear answers. The record-oriented judges do not appear to be concerned about the defendant’s possible comprehension difficulties. Hence, no comprehension checks are



offered by these judges. More important, they also tend to make the defendant agree with their own framing of the conversation—that is, to make the defendant see things the way they want these things to be seen and talked about. All in all, record-oriented judges do not leave much space for variation and imprecision, and they do their best to reduce the two to the minimum. In their belief and insistence on precision (e.g., through asking frequent yes or no questions or insisting on using a fixed script), the record-oriented judges exhibit typical essentialist characteristics. As opposed to the record-oriented judges, the procedure-oriented judges leave much space for variation, openness, and flexibility. The routines these judges are involved in are significantly variable. They adapt the legal procedure to the needs of the individual defendant and encourage the defendant to provide a wider social context of the case. Interestingly, by asking numerous wh-questions, they elicit from the defendant significantly more information. This is, of course, risky; by being allowed to speak more, the defendant may divert the procedure off track. This may result in more openness of interpretation. However, the procedure-oriented judges involve more comprehension checks, which are likely to reduce this openness. All in all, compared with record-oriented judges, procedure-oriented judges are much more flexible and tolerant of, if not encouraging, variation. By asking numerous wh-, rather than yes or no, questions, they encourage multidimensional views. They avoid yes or no questions because they probably know that these questions often simplify reality, pressing complex issues into binary oppositions. In the procedure-oriented judge, the reader should easily recognize a judge sympathetic to nonessentialist philosophy. As Philips rightly pointed out, the two types of judges exhibit contrasting political ideologies. More important, “Judges are not supposed to allow their political ideologies to affect their behavior, and these judges say they adhere to that standard. They see themselves as trying and largely succeeding in keeping their political ideologies from affecting their judicial behavior” (Philips, 1998, p. 79). As might be expected, judges do not really succeed in keeping their political ideologies away from their profession. While the record-oriented view clearly reflects a politically conservative agenda, the procedure-oriented view clearly reflects a politically liberal one. This is not surprising. The connection between homogeneity, strictness, yes or no questions, clear-cut distinctions, belief in clear meaning, and nonflexibility, on one hand, and the conservative agenda, on the other hand, surfaces in many walks of life and most saliently in politics (see e.g., G. Lakoff, 1996; R. T. Lakoff, 2000; Miller, 2001; Silberstein, 2002). The liberal agenda, in contrast, encourages attendance to the openness of meaning, open-ended questions, heterogeneity, flexibility, and compromise.



CONCLUSIONS There is ample evidence to believe that essentialism permeates the legal profession. The views and doings of many representatives of the law indicate that one correct interpretation of other people’s behavior and viewing things in terms of black and white are very common. These essentialist views and doings are significantly reinforced by the various types of manuals and sets of instructions that the lawyers and the law enforcement officers receive as part of their professional training. In addition, as argued by Tannen (1998) about the American legal system, courtroom proceedings are generally governed, at least in America, by the litigation is war metaphor. Handling legal proceedings in terms of the litigation is war metaphor reinforces the view that there must always be two sides to each case and that one side must be the definite winner while the other is the definite loser. There is no space for partial guilt, partial innocence, partial winning, and partial losing. Given facts such as those mentioned earlier, it becomes obvious that legal professional training should open up for nonessentialist thinking. Such thinking would promote open-mindedness, attention to linguistic and cultural variation, and general flexibility in carrying out judicial procedures and in making judicial decisions. Both the representatives of the legal profession and all those people who happen to be involved in legal matters would profit immensely. At least some of the guilty who happen to be free might justly go to jail, and at least some of the innocent who happen to be in jail might justly be set free.



Language in Academia

INTRODUCTION Some academics express great dissatisfaction with other academics’ work. Some even seem to take great pleasure in denigrating it. Negative and/or challenging opinions about what kinds of research others are pursuing or their methodology are occasionally made public in professional publications (e.g., Harris, 1998) and formal lectures. More often, however, such negative opinions are voiced in private conversations. They also sporadically appear in the press. A frequent basis for the negative opinions seems to be academics’ different views about what science should be concerned with and the value of particular research. Nevertheless, we may safely assume that for most people, science is a positive concept and people concerned or associated with it enjoy considerable prestige. Within science, however, a measure of tension related to how the various sciences are perceived by insiders and outsiders to a discipline can be readily observed. For instance, some natural scientists look down on the social sciences and humanities. The reason for this is that the latter are viewed as imprecise and, sometimes, as hocus-pocus endeavors. The human scientists, in contrast, occasionally look down on the natural scientists, accusing them of cold, inhuman, and heartless research often detrimental to the human existence. We also witness situations in which the human scientists look up to the natural scientists and try to appropriate the latter’s methods (see e.g., Sokal & Bricmont, 1997, for a detailed discussion of this point). In general, there is much intolerance of the kinds of research other academics (within or outside of a given discipline) are involved in 141



and of the methods they use. Although this intolerance is seldom officially expressed, it would be hard for an academic to deny that it exists. Such intolerance is not accidental. One of the reasons for it, in my view, is the rampant essentialism reigning in many academic communities. In both direct and indirect ways, many academics express the opinion that only their field is worth studying and that only their methods are correct. This stance reflects the essentialist belief that only their definitions of science and of scientific methods are appropriate.

IS JULIA KRISTEVA A SCIENTIST? Many linguists view linguistics as a science (e.g., Crystal, 1997; Laver, 2003; Lyons, 1981). Julia Kristeva, the well-known Bulgarian-French intellectual, is sometimes referred to as a linguist (in addition to semiotician, literary critic, psychoanalyst, political philosopher); she is perceived at least by some people to be a scientist. Can Kristeva’s work rightly be called “scientific”? Is Kristeva really a scientist? Is Kristeva really a linguist? What is linguistics? What is science? As the reader certainly knows by now, these can be and often are understood as essentialist questions. As might be expected, although there is much disagreement about how these questions should be answered, some make the essentialist argument that only one answer to each of these questions is correct. On December 3, 2004, Kristeva received the first Holberg International Memorial Prize for outstanding scholarly work in the arts and humanities, social sciences, law, or theology. The day before the prize was awarded, she delivered a public lecture that was received with a long, standing ovation. After the award had been handed over to the laureate at a festive celebration, a series of articles appeared in Bergens Tidende, the main newspaper of the city of Bergen, Norway, where the prize is awarded. In the first article in the series, Flåm (2004a) attacked the laureate directly! He claimed, among other things, that Kristeva’s work does not fall under the category of science. In subsequent articles, a number of authors severely criticized Flåm for his views. Flåm’s critics claimed, among other things, that Flåm’s view of science was flawed and that Kristeva’s work can certainly be labeled scientific. The tenor of the polemic was rather aggressive and hostile; the opponents used irony and denigrating language. It became obvious toward the end of the first stage of the debate (after about a week) that both sides were inflexibly convinced that they were right. In the present chapter, I focus on this debate and analyze aspects of it from a nonessentialist point of view. It is my claim that the central bone of contention, the question of what is and what is not science, should have been approached in a different manner. Instead of exchanging arrogant in-



sults, the linguistic and philosophical aspects of the conflict should have been discussed. To the best of my knowledge, these were not discussed. It is my intention to analyze the debate in some detail and offer a solution to the question of whether Kristeva’s work should be treated as scientific. The Concept of Science The history of scientific methodology has been one long debate, with multiple points of view, about the question of what does or should count as science. Notable figures in this debate have been Kuhn (1962), Popper (1945, 1957, 1959, 1972, 1982, 1983), Lakatos (1974, 1978a, 1978b), and Feyerabend (1975, 1981a, 1981b, 1987, 1991). It is not my intention to give a survey of views on the questions of scientific method and the demarcation of science and nonscience. The literature on this topic is enormous (see e.g., also Nagel, 1961, on the distinction between science and common sense). Let me just mention, by way of a brief illustration, the outstanding and often-debated positions expressed by Popper, Kuhn, and Feyerabend. Popper is widely known to have proposed falsifiability as the criterion of demarcation between scientific and nonscientific statements. He proposed that a system be called “empirical or scientific only if it is capable of being tested by experience” and that “the falsifiability of a system is to be taken as a criterion of demarcation” (Popper, 1959, p. 40). Kuhn’s idea of science is different. According to him, a field is a science if it can endorse and conform to a normal science tradition: “. . . it is hard to find another criterion that so clearly proclaims a field a science” (Kuhn, 1962, p. 22). For Feyerabend (1975): . . . science is much closer to myth than a scientific philosophy is prepared to admit. . . . The idea that science can, and should, be run according to fixed and universal rules, is both unrealistic and pernicious. It is unrealistic, for it takes too simple a view of the talents of man and of the circumstances which encourage, or cause, their development. And it is pernicious, for the attempt to enforce the rules is bound to increase our professional qualifications at the expense of our humanity. . . . It makes our science less adaptable and more dogmatic; every methodological rule is associated with cosmological assumptions, so that using the rule we take it for granted that the assumptions are correct. (p. 295; italics original)

The following may be taken to be Feyerabend’s (1975) answer to the question of what is science or the scientific method: “Case studies such as those reported in the preceding chapters show that such tests occur all the time, and that they speak against the universal validity of any rule. All methodologies have their limitations and the only ‘rule’ that survives is ‘anything goes’ ” (p. 296; italics original).



The three approaches to science mentioned earlier illustrate the wide range of existing positions on the question of what kind of inquiry can or should be considered scientific. Most of these positions can be grouped under two rubrics: rationalism and relativism. According to Chalmers (1996), The extreme rationalist asserts that there is a single, timeless, universal criterion with reference to which the relative merits of rival theories are to be assessed. Whatever the details of a rationalist’s formulation of the criterion, an important feature of it is its universality and ahistorical character. (p. 101)

The rationalist believes that a universal criterion for distinguishing science from nonscience exists and can be identified. However, it may be difficult to find this criterion. The typical rationalist also believes: . . . that theories that meet the demands of the universal criterion are true or approximately true or probably true. The distinction between science and non-science is straightforward for the rationalist. Only those theories that are such that they can be clearly assessed in terms of the universal criterion and which survive the test are scientific. (Chalmers, 1996, p. 102)

Relativism, in contrast, promotes an entirely different approach to the concept of science. It “. . . denies that there is a universal, ahistorical standard of rationality with respect to which one theory can be judged better than another” (Chalmers, 1996, p. 102). According to Kuhn, for example, it is the community of the researchers that decides what is to be counted as science: “There is no standard higher than the assent of the relevant community” (Kuhn, 1962, p. 94). Chalmers (1996) noted further that, according to the relativist point of view: . . . decisions and choices made by scientists or groups of scientists will be governed by what is valued by those individuals and groups. Since for the relativist, the criteria for judging the merits of theories will depend on the values or interests of the individual or community entertaining them, the distinction between science and non-science will vary accordingly. For the extreme relativist, the distinction between science and non-science becomes much more arbitrary and less important than it is for the rationalist. (p. 103)

The Debate About Kristeva’s Work In the first of a series of articles on Kristeva’s work, Flåm (2004a) objected to a number of Kristeva’s ideas. His criticism is two-pronged: He criticized Kristeva for her use of incomprehensible language, and he claimed that her



work does not deserve to be called science. As for the former, he referred to Sokal and Bricmont’s (1997) devastating criticism of postmodernist writings in general and of Kristeva’s in particular, and he quoted examples of what for him is incomprehensible language. Let us first consider the question of whether Kristeva’s work can be labeled scientific. Flåm noted: The Holberg prize concerns science (Norwegian: vitenskap), and not literature or expository writing. Certainly, the notion of science has gradually been strongly diluted and is often used as a label for various writing activities, but certain demands must ultimately be made. How does she score on method, empirical research, experiment or formal analysis? I can hardly find her scientific. (Flåm, 2004a, p. 37; italics added)1

In a further section of the same article, Flåm complained: “Young people are tricked into avoiding crafts, natural science and practice. Instead, in front of a computer in an institute, they seek a purely verbal existence, say, in media science, the science of science, or text and documentation science” (Flåm, 2004a, p. 37). In Flåm’s view, what Kristeva does is not science. That the debate contention begun in Bergens Tidende is about what constitutes science is confirmed by Røyrane (2004): “Awarding the international Holberg Memorial Prize to Julia Kristeva illustrates the ongoing academic debate about what constitutes science. Those who think that we should be able to count and measure scientific results are critical of the decision” (p. 8). Referring to the committee chairman’s statement, Røyrane (2004) stressed, “What is good science can always be discussed.” She further emphasized: “I think Flåm reveals an unbelievable myopia as regards research methods that cannot be documented or supported empirically. The Holberg Prize is clearly founded on a view of science different from Flåm’s” (p. 8). Further: Kristeva’s cultural research cannot be documented through laboratory tests and empirical investigations, which Flåm seems to require. The members of the academic community that are behind the award are of the opinion that it is also necessary to use methods other than the classical methods of the natural sciences if we want to understand human life and culture. Some social scientists are of the opinion that science should deal only with things that can be documented empirically. They do not recognize argumentative research which is concerned with analyzing human understanding and communication, nor the assumptions this research is based on. The sciences of culture, for instance, law and literary studies, would not be complete if we were to concentrate only on the empirical. (Røyrane, 2004, p. 8)



Bråten (2004) stated that the discussion focuses on the question of what is science: “But it is important to remember that the exchange [of opinions and insults] reflects two central underlying questions: first, what is ‘science’ . . .” (p. 1). Also Steffens (2004) gave us additional evidence that the discussion concerns the notion of science: “Let me immediately make it clear that several of [Kristeva’s] views on politics and society seem okay and sensible to me, but I don’t think that Kristeva’s ideas are based on the kind of precise, analytical intelligence that would reasonably justify calling her work scientific” (p. 37). Tunstad (2004) left no doubt whatsoever that the central point of the discussion concerns the notion of science: “What is science? And this is precisely where the rub is: What is science? And is Kristeva concerned with this—i.e., science?” (p. 1). Most of the remaining participants in the discussion refer to notions such as science, empirical research, experiments, argumentation, and scientific qualifications (e.g., Gripsrud, 2004; Sandanger, 2004), which are usually central notions in discussions about what constitutes science, and they argue for or against the view that Kristeva’s work is scientific. In the final response to his critics, Flåm (2004b), admitting variation in views about scientific endeavor, reiterated his initial standpoint and paraphrased his own definition of science: On science—including the human sciences and law—there are many opinions, appalling confusion, and innumerable approaches . . . but . . . science is an arena for well-formed presentation, clear analysis, clear focus, clear argumentation, unbiased knowledge, and bold criticism. To put it bluntly, the field is demanding, and the chances of making mistakes are high. So, when Kristeva, conceitedly—and celebratedly—fails on all these points, what am I to say? (p. 35)

I hope it is apparent that, in the discussion of the value of Kristeva’s work, the definition of science featured as a central point. Some authors argued for the view that Kristeva’s work was scientific, whereas others supported the position that it was not. We now view the Kristeva debate in the context of the philosophy of science and evaluate it in light of the nonessentialist framework adopted in this book. There is little doubt that the two parties involved in the Kristeva debate represent, roughly, the rationalists on one side and the relativists on the other. Flåm seems to belong to the former, Røyrane to the latter. The rationalists (some of whom I would call essentialists) argue for one definition of science (including experiments, empirical research, and formal analysis); the relativists (some of whom I would call nonessentialists), arguing for a much more open definition of science, were of the opinion that experiments, empirical research, and formal analysis are not defining characteris-



tics of scientific activities and almost any intellectual activity, including purely verbal analyses of the Kristevan type, can certainly be called science. The former think the latter position leads to charlatanry, and the latter think the former position leads to narrow-mindedness and one-sidedness in academic endeavor. Who is right? The answer to this question depends on one’s philosophical assumptions. Both sides obviously believe they are right. The discussion in the philosophy of science referred to earlier was reflected in the discussion over the Kristeva case. How can a nonessentialist contribute to this discussion? In my view, the theoretical position adopted by many general semanticists, integrationist linguists, and cognitive linguists is consistent with the relativist view expressed by Kuhn (1962) and Popper’s (1945) treatment of what-is questions. Having accepted this fact, the nonessentialist approach that I promote may be understood to support those participants in the debate who defend the award to Kristeva, argue against Flåm, and argue for a wide-open definition of science. To be more specific, my nonessentialist reasoning is, as the reader will expect, that there is no essence in the concept of science. The question of “What is science?” cannot be answered in any absolute way. When we start defining science—that is, when we start answering the question of what science is—we quickly end up having to define the concepts/words in the defining formula (which concepts/words are not unambiguous), and in this way we quickly get involved in an indefinite regression of definitions. In a nonessentialist view of meaning, it is sufficient to claim that science can never be defined in any absolute terms. As opposed to the nonessentialist, the essentialist believes that science can in principle be defined unambiguously and we do not necessarily need to get into an infinite regression of definitions. The essentialist tends to believe that, at one point, the defining will reach a satisfactory stage, at which we will agree on what science is. In a nonessentialist view, however, such a stage can never be reached. In the nonessentialist view, asking the question, “What is science?” is not rewarding; it always leads to an infinite regression of definitions. In the nonessentialist view, science is best viewed as a label—something, in this case, we call a type of research activity. As we know from analyses such as Kuhn’s, Feyerabend’s, and Popper’s, these activities have been heterogeneous, and the research methods involved in them have been numerous. It may easily be argued that virtually no intellectual activities can in principle be barred from being referred to by the word science. All in all, the position on meaning generated by the theories endorsed in this book allows Kristeva and her supporters to use the word science to refer to the activities in which she has been involved. In this sense, I contend that those who criticized Kristeva and insulted the scientific character of her work were unfair. There remain two points to be made. One concerns the prototypical



structure of the concept of science; the other has to do with the social acceptance of the type of science in which Kristeva is involved. As mentioned earlier, many concepts are best understood as prototypes, including typically central, peripheral, and borderline examples. As indicated earlier, central examples are those about which people normally agree. The concept of science is certainly eligible to join the group of concepts that can be easily handled in terms of prototypes. Science is a very fuzzy concept. To put it in a slightly different way, we may posit that various academic (scientific) activities have varying degrees of membership in the category of science. To the best of my knowledge, no analyses of the meaning of the word science along cognitive linguistic lines (as in the case of the word lie; see Coleman & Kay, 1981) have been carried out; we may thus only speculate as to what science and scientific (and their rough equivalents such as German Wissenschaft and Norwegian vitenskap) mean to academics and to ordinary users of English, German, or Norwegian. I think, however, that it is reasonable to assume that large groups of people (academics and lay persons alike) view the natural sciences (defined along the lines of Flåm) as the most representative (scientific) sciences. Social sciences, on the whole, appear to be viewed as less typical sciences. The human sciences, in turn, are viewed by many as peripheral; one piece of evidence to substantiate this view is that the word scholarship rather than science is often used to refer to these sciences. Finally, there are borderline cases. In my view, for many people, the work of the French postmodern intellectuals in general and that of Kristeva in particular qualifies as a borderline case. The location of Kristeva’s work at the border of the concept of science may then be one of the main reasons that the debate in question took place, with some participants arguing for and others arguing against Kristeva, and why it was at times so fierce. Interestingly, this kind of debate would have been unlikely if a typical case were to be taken up—for instance, work in nuclear physics or molecular biology; it is unlikely that anyone would have questioned the legitimacy of this work as scientific. However, equally unthinkable is viewing floor sweeping as scientific work. In other words, while nuclear physics for most speakers of English is definitely within the category of science, floor sweeping is definitely outside of this category. It is the borderline cases that are most often widely discussed and disputed. To conclude this section, the discussion and fierce disagreements about whether Kristeva’s work is scientific seem to originate from the fact that Kristeva’s work can be located in the borderline area (or at the periphery at best) of the concept of science as understood by many academics and lay persons. If we agree with the relativists that there is no one criterion for distinguishing science from nonscience, if we agree with Kuhn that it is the academic community that decides what is to be defined as scientific, and if we agree with Popper and some cognitive linguists that what-is questions can-



not be answered in any absolute way, Kristeva’s work may certainly be referred to as scientific. Kristeva, as well as such postmodern colleagues as Lacan, Deleuze, Guattari, and Irigaray, has been severely criticized for her use of incomprehensible, convoluted language and for her pointless importation of natural science terminology (e.g., Flåm, 2004a; Sokal & Bricmont, 1997). Sokal and Bricmont (1997) did not hesitate to refer to many statements by Kristeva, Lacan, and others as sheer nonsense and truisms. Sokal and Bricmont (1997) claimed that intellectuals such as the ones mentioned earlier merely want to impress the reader; that they indulge in fancy scientific jargon; that the scientific and pseudoscientific terminology they use is devoid of meaning; that they often “give an appearance of profundity to trite observations about sociology and history”; and that some of their texts are perfect examples of “diarrhea of the pen,” the text meaning precisely nothing. As part of their devastating criticism of postmodernist work, Sokal and Bricmont (1997) stated: “Many people are simply irritated by the arrogance and empty verbiage of postmodernist discourse and by the spectacle of an intellectual community where everyone repeats sentences that no one understands” (p. 192). Criticizing the postmodernists’ use of incomprehensible language, which, in their view, the postmodernists do not understand, Sokal and Bricmont (1997) claimed, irritatedly, it seems, “It’s a good idea to know what one is talking about” (p. 176). The prior opinions expressed by Sokal and Bricmont (1997), and at least partly endorsed by Flåm (2004a, 2004b), do not at all undermine our earlier conclusion that Kristeva’s work may be referred to as scientific. That many people find her writing totally incomprehensible, stilted, or nonsensical is another matter. Admirers of Kristeva’s work claim, often in personal communication, that they do understand her and that her work is indeed profound and extremely valuable. How are we to judge whether Kristeva and her colleagues are in fact comprehensible? As mentioned, in a nonessentialist view, a text, whether written or spoken, may be comprehensible only to people; a text is never in itself comprehensible or not. If some colleagues claim that they understand and profit immensely from Kristeva’s writings, we have no tools to show them that they do not. No matter, then, how many people find postmodernist work to be incomprehensible and worthless; as long as there are people who find it comprehensible and valuable scientific research, we should not denigrate or reject the opinion of the latter. Admittedly, many members of the academic community seem to have derived utmost pleasure and intellectual satisfaction from Kristeva’s written work, from her lecture in Bergen on December 2, 2004, and from the award ceremony the following day. The standing ovation that Kristeva received after her Bergen lecture (although, in my view, forced on some at-



tendees by the standing-ovation dynamics) may be taken to be a token of this pleasure and satisfaction. Kristeva’s work appears very meaningful and useful to some academics regardless of whether they are called scientists. Whether her work is of any use to people outside of this group of admirers is an entirely different matter. For whom else should work such as Kristeva’s be useful, or to how many people (in terms of real numbers) this work should be meaningful and useful might ultimately be for local communities, academic communities, or for taxpayers to decide (French, Norwegian, or others).

CONCLUSIONS Academia is a good institution for nonessentialists to look for essentialist examples. The essentialist view of concepts and word meaning is reflected not only in scholarly and scientific work, as shown in chapter 2 for linguistics, but also through academics’ frequent arrogance, pomposity, and pontificating about their work, the way they understand science, and the way they carry out their research. Essentialism breeds intolerance and hostility in academia. We have witnessed much discussion in the past few years (e.g., in Germany, Switzerland, Poland, and Norway) about the autonomy of the university as an academic institution. In an essentialist view, one of the reasons (perhaps the main reason) that the university should be autonomous is that only the academics know with what the university should be concerned. This, in my view, is a reflection of an essentialist frame of mind. First of all, this position entails the view that academics and nonacademics can be easily defined and distinguished from one another, which of course is not possible. More important, however, arguing for the autonomy of the university also implies that the academics know best what science is, what good science is, what correct scientific methods are, and, most important, what the good things are (the good and valuable knowledge) that the academics can provide for the society at large. In general, essentialists support the autonomy of the university. Many academics (members of the university community) view themselves as having access to exquisite (essential) knowledge that nonacademics (nonmembers of the university community) are understood to be barred from by definition. This position leads to the arrogance and pomposity mentioned earlier. In a nonessentialist view, the autonomy of the university is a highly debatable idea, to say the least. One of the main reasons for this is that it is not true that academics have access to some special knowledge that nonacademics have no access to. As the Kristeva example shows, academics argue about the basic concepts (including the concept of science) and basic val-



ues; there is no overall consensus among academics whatsoever about what kind of research is good, useful, necessary, and so on. This lack of consensus alone might be a good reason for treating the autonomy of the university with utmost suspicion. We academics may have read many books on a particular topic, but our knowledge will remain uncertain and we will always remain fallible.

ENDNOTE 1. All the English-language quotations in this part of the book are my own translations from the original Norwegian.



Language in Education

The present chapter illustrates essentialism in the domain of language education and offers solutions to some of the problems that essentialism engenders. Language education is to be understood here as a general rubric under which several topics are discussed. Specifically, essentialism is disclosed and criticized with reference to discussions about standardization, stigmatized language, and language purism; sections of these discussions link education and politics. Finally, suggestions are made as to how the language user can possibly be educated toward an approach to the use of words engendering fruitful innovation. THE STANDARD VARIETY One of the most salient issues in language education is that of speaking, learning, and teaching the standard variety of a language. In most countries, the standard variety of the language (or languages) of that country is focused on and assigned much social prestige. A typical language with a focused and prestigious standard variety is Polish. Norwegian, in contrast, seems to be one of the few languages (if not the only one) where the standard has a uniquely weak position. Although Norwegian has two written standards, a spoken standard understood as a variety adhering to a set of prescriptive rules does not exist (see Jahr & Janicki, 1995). While arguments for and attitudes about standard Polish exhibit typical essentialist thinking, those to Norwegian’s lack of a spoken standard exhibit typical nonessentialist thinking. In what follows, we begin with quoting arguments 152



for standard English, which are roughly like those for standard Polish and like those for many other standard languages (e.g., French, German, Spanish, Hungarian). Having identified essentialist thinking in them, we then briefly argue for the Norwegian solution, which clearly illustrates a nonessentialist stance. A most often expressed stance toward the standard variety of a language is that the standard is both clearly definable and inherently better than the nonstandard varieties of the language in question, which is a clear misconception. This stance is expressed by both the lay public and many writers on language, some of whom are professional linguists. We find much evidence for the prior statement, for instance, in Lippi-Green (1997), J. Milroy and L. Milroy (1999), and Bex and Watts (1999). The stance extolling the standard variety is of course not new. On the contrary, this stance has been with us for some time. For example, Wyld (1934) treated standard English with special reverence and esteem. This special esteem led some other authors to say that standard English does or should have special authority. As put by Honey (1997): “What the English language needs is a form of authority that can easily be appealed to for guidance as to the uses which are acceptable compared with those which are not—an authority based not on an individual’s irrational likes and dislikes but on the genuine consensus of educated opinion” (p. 163). Authors like Honey believe that there is only one right (correct) way of saying things, and it is standard English that is the exclusively correct way. What makes standard English utterly correct, it is believed, is a set of objective criteria that identify the educated speakers of this variety. As Crowley (1998) rightly pointed out, what these criteria are remains, however, unknown. Importantly and not unexpectedly, the questions of what is exactly standard English, and of what an educated speaker of English is remain unanswered (see e.g., Lippi-Green, 1997, on this point). The essentialist belief in the feasibility of these questions, however, as might be expected, remains strong. “Several years ago there was an entire conference devoted to one subject: a precise definition of Standard American English. This meeting did not succeed in satisfying everyone as to what SAE should be” (Fromkin & Rodman, 1993, p. 284). To put things differently, many people (including some professional linguists) believe that the linguistic world is bipolar in the sense that some linguistic forms are definitely correct (the standard forms) and others definitely incorrect. The correct standard language forms are believed to be correct in some absolute sense. The fact that these standard forms are simply linguistically different, as many linguists continuously stress, remains ignored. This attitude to the standard language is clearly an expression of essentialism; one language form is considered definitely right—another, or others, definitely wrong. More important, also in accordance



with essentialist reasoning, objective criteria (unspecified as they are) are invoked. In contrast to most other languages, the sociolinguistics of Norwegian exemplifies a significantly nonessentialist stance toward the standard and toward variation in general. The two Norwegian written standards aside, the language does not have a spoken standard in the form of a set of prescriptive rules indicating which grammatical or phonological renditions are correct and which are not. Local vernacular varieties are fully acceptable in schools; a school pupil hardly ever gets corrected as long as his or her Norwegian falls within the confines of what will generally be conceived of as Norwegian. The sporadic teacher corrections might be an expression of the teacher’s personal preference or prejudice. This correction is not, however, a correction toward some one standard form, which does not exist. The sociopolitics of spoken language standardization may be seen as a token of essentialism, leading to conflict, oppression, and injustice in the domain of language education. On this aspect of the essentialist reasoning, people who speak the standard are generally considered inherently better and superior; they are the ones who should be given privileges, for instance, in the form of better paid jobs. Those who do not speak the standard are denigrated and considered inherently inferior. The sociopolitics of no spoken language standardization, however, may be seen as a token of nonessentialism, where everybody is given an equal chance in life independent of the variety that she or he speaks. If this philosophy is not fully implemented in reality, as some observers of the Norwegian sociolinguistic situation claim, it is at least a clearly formulated goal, and it is a goal nearly reached. A sociolinguistic situation in which there is a focused, spoken, standard variety is conducive to conflict. When speakers of a standard variety of a language come into contact with those of a nonstandard variety of the same language, the former and also sometimes the latter claim that the standard variety is inherently (i.e., in some natural way) better than nonstandard varieties. Outright denigration of nonstandard dialects is a frequent practice, especially among lay people. It is perhaps useful to remember that for many lay people, the term dialect invokes only nonstandard varieties and is often conceived as something incorrect, bad, wrong, and something to be eradicated and avoided (at least by educated people). It is unfortunately true in many educational settings that standard-variety-speaking teachers view their students’ nonstandard dialects as inherently bad. Although there have been significant attempts to change this state of affairs, the degree of success has not been spectacular. Even in Norway (which is a special case, as mentioned earlier), where sociopolitical egalitarianism is presumably much higher than in many other countries, some speakers still think that the variety they speak makes them better people. As



Jahr and Trudgill (1993) reported, before World War II, Nynorsk (which may in sociolinguistic terms be best characterized as a range of local dialects) users were often viewed by themselves as better Norwegians. The link that I have been trying to make between essentialism, on one hand, and the distinction between standard and nonstandard varieties, on the other hand, is that many speakers of standard varieties feel they know for certain what concepts such as proper language, good language, appropriate expression, wrong dialect, and so on are. In other words, these people seem to feel that they know what these words really mean. Obviously, the words really mean what they mean to them. This brings us to the question of language purism (see Thomas, 1991), which often emerges in the educational and political settings. It is easy to see that the conflicts that relate to standard and nonstandard dialects reach a high point for those involved in language purism. Although we should not be critical of all prescriptive linguistics, language purism can be viewed as a salient reflection of essentialist thinking, and thus as an activity that has a potential for generating significant social conflict. This is because language purism reflects categorical judgments. For language purists, as opposed to descriptive linguists, the difference between standard and nonstandard varieties is or should be clear. A linguistic form is either wrong or right. It is either in or out of the category of proper language. Language purists claim to know what proper language is. This kind of claim is, in fact, a good example of two-valued orientation. Nonstandard dialects are perceived by language purists as definitely wrong, whereas standard dialects are perceived as definitely right. In such reasoning, there is no room for a third possibility; there is no room for fuzzy cases. This is certainly a dangerous position to take. It divides people into two groups: insiders and outsiders; those who speak the (allegedly clearly defined) standard and those who do not; friends and foes. As a result, the hostility level rises, and open conflict may result. Evidence to support this analysis may be easily collected. Let us analyze in some detail, for example, the sociolinguistic congress for Polish held in Szczecin, Poland, in 1984.1 As the reader may remember, at that time, Poland was still a nondemocratic country, where a congress on the sociolinguistic aspects of the national language, as expected, had political significance. Many contributions made during the congress clearly reflected the bipolar, black-and-white essentialist mode of thinking, which was just one reflection of the general bipolar attitude exercised by the communist states. The communist rhetoric was full of clear-cut distinctions. As the congress showed, the rhetoric pertaining to language was no exception. The congress showed quite vividly how language purism may be harmful. Consider four quotes from the proceedings. These examples concern linguistic form (Quote 1), linguistic form and content (Quotes 2 and 3),



and content (Quote 4). (1) “. . . linguistic skill and correctness, without which it is difficult to think logically and creatively, and impossible to convey thoughts to others” (Urba½czyk, 1987, p. 110).2 This quote may safely be taken to mean that speakers of non-standard Polish are mentally deficient. (2) “. . . hatred makes language impure. If we look into those formal and informal, legal and illegal attacks on our statehood, (and) the conditioning of this statehood, then we would see how horrifying the language with which these attacks are formulated is” (Urba½czyk, 1987, p. 259). This contribution was made by Por“ba, who was a film director known in communist Poland not only for making films, but also for his outspoken support for the communist regime. The “formal and informal, legal and illegal attacks on our statehood” should be understood as a reference to the then-banned political opposition, whose activities, Por“ba claimed, were motivated by hatred. The passage may be understood as meaning that the political opposition in Poland corrupted language. To put it in slightly different terms, this demonstrates the following attitude: “One political side’s language is good, and the other side’s (the political enemy’s) is bad, or even ‘horrifying.’ ” Finally, (3) “It is extremely important how one speaks. But incomparably more important is what one says” (Urba½czyk, 1987, p. 25; italics original). This view leads the speaker to another conclusion: (4) “Our responsibility is enormous—not only for the beauty and the purity of Polish, but also for the content that it expresses” (Urba½czyk, 1987, p. 6). One needs to stress that Quotes 2 and 3 come from Mieczys»aw Rakowski, who represented the government and the communist party at the congress. “Our” in the second quote refers to “them”—that is, the communist party. The political consequences of these purist arguments were very obvious. One of them was censorship imposed on the spoken and written language, which was not seen to be in line with the communist party and official government position. More important, the essentialism-laden language purist movement is also harmful because it encourages making straightforward connections between linguistic phenomena and morality. As Shapiro pointed out: Every society is involved to some degree with identity politics, with separating people into groups with identities which form a hierarchy of worthiness, and one’s language group membership is an important part of many of these identity politics processes. Clearly, then, attempts to “purify” a language implicitly promotes those who can most closely identify themselves as belonging to the language base toward which the change is aimed to a position of moral superiority. And because purification implies getting rid of stain and thus evil, purification movements imply at some level that the impure language elements belong to impure persons. This impurity ascription makes it then possi-



ble to put people who cannot claim affiliation with the privileged language in a lesser moral space. (Shapiro, 1989, pp. 22–23)

Language purism, relating significantly to standardization, is a serious sociopolitical and educational matter. EBONICS The Ebonics issue, discussed at length by Rickford (1999), Baugh (2000), R. T. Lakoff (2000), and Romaine (2000) and largely reported on at Web sites such as and topics/ebonics/, constitutes another excellent example of how people in the educational setting misconceive language, how they fruitlessly argue about definitions, and how easily these activities lead to unnecessary conflict. The reader may recall that in December 1996, the Oakland School District Board announced in its resolution that it would treat Ebonics (otherwise referred to as Black English or African-American Vernacular English) as the primary language of its students, and that it would perceive Ebonics as the language that the students should resort to to better master standard English. In other words, the idea was, roughly, that Ebonics and standard English should be contrasted, and that the latter should be learned as a useful means of communication to, for instance, increase job opportunities, whereas the former should be resorted to as a genuine linguistic system, rather than a conglomeration of errors and deviations from standard English. As might have been expected, the Oakland District School Board’s announcement was largely received, by the general public and some educators, as a provocation or fundamental educational mistake at best. The Californian newspapers were flooded with critical articles and protests (cf. Fillmore, 1996) from people who view Ebonics as a degenerated variety of English (including masses of mistakes), whose speakers deserve neither respect nor attention. What they deserve, it was argued, is a general blame for not having been able to master what they thought was the inherently correct standard English. The attacks on Ebonics (whether understood by the attackers as a nonstandard dialect or anything else) are only an example of the common attacks on nonstandard varieties witnessed in many countries and in many contexts. In this sense, these attacks were not a new phenomenon and might not be worth our special attention. What was new, however, and no doubt very salient in the midst of the Ebonics conflict, and what is clearly relevant to the main point of our discussion of this book, was the fierce arguments and disagreements about whether Ebonics was a language or a dialect. As might be expected, some people argued for the former and



others for the latter. Each side argued, again, as might have been expected, that their definition (of Ebonics) was correct. Interestingly, informed opinions of nonessentialist linguists (e.g., Fillmore, 1996) were not taken into account. Fillmore correctly explained that deciding about whether Ebonics was a dialect or a language should be seen as deciding about whether Greenland is a small continent or a big island (Fillmore, 1996). Most of the essentialist public insisted, however, that Ebonics was a dialect of English. Interestingly and importantly, declaring Ebonics a dialect of English or declaring it a separate language had financial significance. Speakers of Ebonics understood as a dialect of English do not qualify for funds usually allocated within bilingual education programs. Speakers of Ebonics understood as a separate language do qualify for such funds. Thus, it can easily be seen that the educational problem in question is also a political problem, partly in that whether Ebonics is viewed as a dialect or a language is a sociopolitical decision and partly in that political fights about allocating financial resources are at issue. The Ebonics conflict is a good illustration of essentialist reasoning in the educational setting because it is largely about definitions and what is seen to be correct (standard English) versus incorrect (e.g., Ebonics) English. There is, however, an additional essentialist twist to this conflict. Namely, as R. T. Lakoff (2000) reported, as the conflict developed in 1997, the topic was taken up, as expected, by the American TV shows, where the general view seemed to be, “Ebonics . . . is a low and dirty talk by people who don’t know any better” (R. T. Lakoff, 2000, p. 242). What was very characteristic of these shows was that usually only two types of views were accepted: either for the Oakland Board or against it. Views against the Board were very welcome. Views for the Board were immediately argued against by the opposing side, backed by the TV moderator. Most significantly, moderate views, or attempts at explaining the complexities of the issue, were immediately hushed and ignored. The simple, nonessentialist solution to the problem— that is, that Ebonics is neither definitely a language nor definitely a dialect and that it can be defined as either—was never seriously considered. ON NEW WORDS—EDUCATION TOWARD NONESSENTIALISM Having read the preceding pages of this book, the reader may have painted for him or herself a rather gloomy picture of the world around us; a world full of essentialists who know best, who know the truth, who are selfrighteous, and who are authoritarian, imposing, and so on. It is in fact difficult to disagree with this description of the state of affairs. Many people around us do, indeed, seem to be essentialists, at least to some degree. Do we need to be worried? Is there any way out? Is there anything we can do to



persuade people to abandon essentialism? How can we educate people against essentialism? The task is a Herculean one, and there are important reasons for this. One is that the two-millennia-old Aristotelian tradition has imprinted on our civilization a marked feeling of the truth of essentialism. Many students whom I have educated against essentialism report that, although they accept the nonessentialist argument intellectually—that is, although they find no faults with it—emotionally they feel that something is wrong with it. It is terrible to realize, they say, that there is no final supreme moral code to live by; no one truth; no one idea of good that we should attempt to achieve, no one correct definition of concepts/words, and so on. They feel, and rightly so, that there is nothing firm or stable to attach to as far as language is concerned. This disturbing dissonance, on the part of many people, between deeply rooted, emotionally experienced essentialism and intellectually accepted nonessentialism is, with little doubt, an Aristotelian heritage. Essentialism is so deeply rooted in us that a cogent philosophical argument, which may first appeal to us, will be hard to accept after a while. The feeling described earlier is not unlike that which we have about the customs and habits of another cultural group. We accept them intellectually as culturally different, while emotionally these customs and habits still remain strange or totally unacceptable to us. The tension between intellectual nonessentialism, on one hand, and emotional essentialism, on the other hand, may be quite a problem for many people. That essentialism is very deeply rooted in us and is one reason that education against essentialism is a very hard task. Another reason that the nonessentialist’s task is a Herculean one is the power of dogmas. The influence of the various sets of dogmas throughout the world certainly varies. With markedly dogmatic people, however, wherever they live, telling good from evil, moral from immoral, right from wrong, and so on is an extremely easy task. The truth is equally easy to locate. Of course not all dogmatic people are dangerous and imposing. Nevertheless, given the impact of dogmas all over the globe, the strength of essentialism is no surprise. Dogmatic essentialists may be religious, but it has to be remembered that many atheists are also essentialists—for instance, Marxists (Popper, 1945, 1957). What we see, then, is that essentialism is ideologically realized in ways that are seemingly incompatible. Marxists and Catholics often fight each other, as they did, for example, in pre-1989 Poland. In fact, however, both Marxists and Catholics argued for one essentialist truth and one essentialist definition of words; it is only that the definitions promoted were different. So, no matter where the essentialism of an individual person is philosophically anchored, the result is likely to be comparable, with a host of disagreeable phenomena such as intolerance and prejudice likely to be coming to the fore.



It is difficult to be optimistic on the question of nonessentialism gradually gaining ground. Considering politics, for example, the present sociopolitical situation in most countries of the world does not seem to inspire nonessentialists with much optimism for the future. This is despite the recent collapse of Marxist ideology in Eastern Europe. For whatever reasons (and those reasons differ depending on the country), nonessentialists seem to be in the minority. When one’s ends are at odds with those of most people around, the task of educating the overwhelming majority is an extremely hard one. The academic community does not supply us with much optimism either (cf. chap. 7, this volume, for a recent example). Like any social group, academics have been under strong Aristotelian influence. This can be documented especially in the social sciences. Significant antiessentialist writings have not really had much impact on the workers in most social scientific disciplines. For example, Popper’s detailed criticism of essentialism, which was first openly voiced in the mid-1940s, does not really seem to have had much effect. Neither does the general semanticists’ antiessentialist work, which goes back to an earlier time (1930s); neither do Feyerabend’s cogent and admirable writings in favor of relativism, which go back to the 1970s and 1980s. Unfortunately, all these writings have had little influence at best. The general semanticists’ work is hardly known among and/or referred to by linguists. Popper’s and Feyerabend’s work is surely known among philosophers, but hardly among linguists. Despite the growing popularity of cognitive linguistics among linguists, which, the reader will remember, may be seen as a nonessentialist linguistics, education against essentialism remains an enormous task. One of the most important things to do to change this state of affairs, however, is to make the general public realize how important our beliefs about language are for all walks of life. That is the real task. In other words, we have to do our best to make lay people realize that linguistics is not only for linguists. The general awareness of the significance of language for all kinds of human activities seems very low. We may assume that most lay people feel the same way as do some arrogant and ignorant doctors and lawyers, who claim that they know their native language and therefore do not need any help from superfluous linguists. In other words, linguistics or applied linguistics should play an important role in trying to have people switch from essentialism to nonessentialism. Potentially, the role is enormous. We have to keep writing and talking to people about the importance of beliefs about language for whatever one does. One might wonder whether there is anything specific and concrete that can be done so that essentialism may become weaker, whether there is perhaps something that can be done about language so that it could better



serve our nonessentialist objectives. One possible answer to this is that new words should be coined and their use encouraged. In a brief article, Fairclough (1994) addressed the question of the war in Bosnia. He rightly claimed that the way the British media report on the war “harmonises with the perverse logic of barbarism despite their antipathy towards it” (p. 431). One of the main points that Fairclough made is that the media resort to terminology that emphasizes difference (Catholic vs. Protestant, Blacks vs. Whites, etc.) and to labels such as Croat, Muslim, Serbian, which do not reflect the complexity of the sociopolitical reality involved. This complexity, the hybridity and the de facto mixing of races, religions, and cultures, should be reflected in the language used to refer to the phenomena in which the various races, religions, and cultures are involved. In Fairclough’s (1994) view, “The media have a particular responsibility for developing identificational practices which avoid reductive identification and misleading or irrelevant difference” (p. 432). Fairclough’s discussion translates into two pleas. One is for the avoidance of binary oppositions (black–white, good guys–bad guys types of distinctions). The other is for an attempt to linguistically reflect the complexity of the nonlinguistic reality to which language refers. The plea against binary oppositions translates into a plea against essentialism. We might be tempted to ask the question of who or what is causing the situation about which Fairclough is talking. In other words, the question is that of why it is that opposites such as progovernment–antigovernment are so prevalent and why it is that labels such as Muslim, Croat, Catholic, and so on so inadequately reflect the complexity of the underlying reality. Still in other terms, the question is whether it is people who mishandle language, and thus cause the problem, or whether it is perhaps language that does not allow them to express the complexity of reality. Is it perhaps a combination of both? Once we have identified the reason(s) for all this, a related question will be that of how we can possibly change the existing state of affairs. The common practice of resorting to binary terms such as progovernment–antigovernment can obviously be traced to essentialism. The binary oppositions in journalistic reports are very common indeed. To a large extent, they reflect the thinking and cater to the needs of the general reader. We have to keep in mind that it is the tabloids, depicting the world in blackand-white terms, that sell best around the globe. The binary opposition language used to talk about Bosnia and Northern Ireland is in no way anything special or exceptional. This language is ubiquitous as is the conceptual essentialism that underlies it (see also Tannen, 1998, for a discussion of binary opposition in American media). Descriptions of the sociopolitical turmoil in Poland at the turn of the 1980s and 1990s provide another excellent example of how gross linguistic



labels get perpetually contrasted, how they are tossed around, and how they at the same time inadequately depict the complex reality in the country. Terms such as Catholic, communist, postcommunist, Solidarity person (i.e., someone associated with Solidarity), post-Solidarity person, progovernment person, antigovernment person, prochurch, antichurch, Marxist, opposition member, and so on were in daily use. These terms were usually conceived as belonging to binary oppositions—for example, Catholic versus communist,3 progovernment versus antigovernment, Solidarity person versus anti-Solidarity person, opposition member versus opportunist, and so on. Thus, a person who was (before 1989) not a Catholic was perceived by many people as a communist; a person who was not a Solidarity person was perceived as an antiSolidarity person;4 and a person who was not an opposition member (before 1989) was perceived as an opportunist. The reality to which all these labels refer was and is still significantly more complex than those labels seem to indicate. Let us consider in some detail the terms Catholic and communist. These two labels, which still occasionally appear as binary opposites, were until recently two crucial terms in Polish politics. Calling someone a Catholic (or a religious person) may have invoked at least one, several, or all of the following characteristics, and possibly others: born into a Catholic family, baptized in the Catholic church, took first communion, regularly attends mass on Sundays, regularly attends mass also on days other than Sundays, goes to church only occasionally, does not go to church at all, goes to confession regularly, rarely goes to confession, does not go to confession at all, takes communion regularly, takes communion irregularly, does not take communion at all, uses contraceptives, uses some contraceptives but not others, does not use contraceptives at all, speaks for prochoice, is against abortion, is for restricted abortion, says daily prayers, says prayers occasionally, or does not say prayers at all. Calling someone a communist in Poland, in turn, may have invoked an almost equally long list of characteristics and conditions. Only one of them or several of them may have been present for the label to be used. The list in question may include a member of the communist party before it was dissolved in 1989, formal nonmember but advocate of the communist ideology, formal member but advocate of anticommunist ideology, communist party activist, communist party apparatchik, pre–World War II communist party member, someone informally associated with the communist party, member of the pre-1989 government, and so on. With respect to the labels Catholic–communist, things have recently been much more complex compared to what we may safely assume was the case before 1989, and much earlier when the words were first used in the language. Especially in the period of the last 25 years (1980–2005) we have been witnessing important meaning shifts and an increase of ambiguity. It has obviously been the changes in the sociopolitical domain (e.g., those



caused by the political events in 1980 and 1981—i.e., the time of the outburst of Solidarity activities and the declaration of marshal law in 1981) that have influenced the scope of phenomena to which Catholic and communist pertain. These changes have given rise to meaning shifts and extensions. More important, we are dealing here with radially structured concepts that include typical examples (e.g., a prototypical Catholic who has been baptized, regularly attends mass, regularly goes to confession, regularly takes communion, does not use contraceptives, is an antiabortionist) and nonprototypical examples—that is, those who radiate away from the core of the prototype (e.g., a Catholic who has been baptized in the Catholic church, but who does not go to mass at all, does not go to confession, uses contraceptives, does not respect the Pope, and is a prochoice person). As to the axiology of the two concepts in question, we may safely assume that the axiological load currently imputed by speakers of Polish to the two concepts varies a lot. All in all, a single instance of use of the words Catholic or communist adequately reflects neither the intricacies of the underlying reality nor the axiological variance that the reality involves. As a result of all this, we get two simplifying opposites that invoke a significantly high level of idealization. This idealization is imposed on the language user by the labels (words) currently in circulation in Polish. We are no doubt faced here with words that are highly ambiguous. We may pose two interesting questions: “Why is it that language users tolerate words exhibiting a very high level of ambiguity?,” and “Do language users have to tolerate such words?” To answer the first question, at least one of the reasons for tolerating highly ambiguous words is the underlying essentialist belief that words have only one correct meaning and that the “deviant” uses (departures from this one correct meaning) should or will eventually disappear. More important, all these highly ambiguous words are not so highly ambiguous to the essentialist. They are ambiguous to the nonessentialist. Our prototypical essentialist is not aware that many words are highly ambiguous, so she or he can hardly be said to be tolerant of such words. However, it needs to be admitted that essentialists are aware that many words have different meanings; it is just that these meanings (except for one) are not perceived as legitimate. As mentioned earlier, essentialist language users are usually skeptical about change and innovation. People are fearful of, or skeptical about, new words and new meanings imputed to old words. If words are believed to have one correct meaning, that meaning cannot legitimately vary or change, it is believed. Innovation may thus be conceived of as a threat to the essentialist way of reasoning. The prior argument could also be taken to mean that language users talk about reality as if it were much simpler than it actually is. Many words frequently conceived of as binary oppositions (such as Catholic vs. commu-



nist) force categorizations on speakers and hearers that they might not want to make at all. For lack of better means, however, those speakers and hearers resort to the opposites that their language expresses. This is exactly the fact in the case of Catholic and communist. These gross categories force high-level idealizations on the speaker of Polish that she or he might not be willing to make. But again, for lack of better means, the language user resorts to words involving such high-level idealization. It might be contended that, unfortunately, these words tend to freeze one’s perception, although clearly they do not do so entirely, and once and for all. The idea of freezing one’s perception is not unrelated to the Sapir–Whorf hypothesis (see Carroll, 1956). We need to remember, however, that such freezing is not total, as some claim the Sapir–Whorf hypothesis says. We can always liberate ourselves from the words that we use. We can do that, for example, by learning another language (Popper, 1983). The fact that all words involve a certain degree of idealization is not tantamount to saying that we as language users have to accept every level of idealization that words impose on us. If one believes that some words freeze perception to a dangerous degree (as is the case with the Polish Catholic and communist; Polish: katolik and komunista), we may ask ourselves the question of whether there may, perhaps, be something that we can do about it. This brings us to the other question asked earlier—namely, whether language users have to tolerate words that promote binary opposition contrast and those that significantly blur the complexity of reality. The answer to that question is clearly negative. Language users do not at all need to, or, to put things differently, they are not at all forced to use words the way these words have developed to cut the universe into manageable chunks. Language use is not predetermined. Language users can act on language and try to change it so that it could better serve their needs and aspirations. This can be done by coining new words (see also de Bono, 1991). We need new words that can reflect the complexity of the reality in ways that the presently existing words do not. These new words should reflect to a higher extent the gradience of concepts. As for the binary oppositions, we should try to work on language to turn these oppositions into continua. Obviously, languages do try to reflect gradience and thus fight oppositions. In English, for instance, as in Polish, we may speak of a good Catholic, an orthodox Catholic, a lapsed Catholic, a real Catholic, and a pseudo-Catholic alongside with a real communist, an orthodox communist, a former communist, a postcommunist, and so on. However, these labels still hardly depict the intricacies and complexity of the reality underlying them. The discriminations that these labels make are still too gross. In other words, I claim that language users have at their disposal a considerable number of concepts for the expression of which they do not have lexical means. They obviously



have other linguistic means; they can talk about and eventually describe in much detail the complex reality in question. But these language users do not have lexical means—that is, single words; the concepts in question do not get lexicalized. Having general linguistic means is very different from having specific lexical means. One might wonder why lexicalizations are so important. Lexicalizations are extremely useful because they allow linguistic shortcuts. This appears to be true in terms of sheer time. We can hardly afford the “speaking time” each time we use the term Catholic or communist (not to mention the many other terms that we constantly use) to describe in detail the kind of Catholic or communist that we have in mind. Such descriptions would not be psycholinguistically feasible—that is, the descriptions would be lengthy, repetitive, and difficult for the hearer to process. What we do is simply cut a long story short and say, for example, Catholic. The advantage of using a single word rather than a set of words (such as a very orthodox Catholic) does not reduce only to the question of time. On the whole, the use of single words (rather than long descriptive chains) seems to make people more satisfied. Language users seem to generally feel that a short label (mostly one word) is somehow better than a long one—that is, a number of words. In this context, let us invoke translation or interpreting: When a translator or interpreter is able to provide word-for-word translation, she or he is usually perceived as a better translator or interpreter compared with when she or he provides multiword paraphrases. Also, to mention another example, proposals for politically correct language that are the most funny are multiword expressions (e.g., “temporarily metabolically abled” for “alive,” “temporarily inconvenienced” for “dead,” “spontaneous display of community dissatisfaction with prevailing socioeconomic conditions” for “riot,” etc.). Single-word proposals are viewed either as significantly less funny or as serious, for instance, “nonsighted” for “blind,” “text” for “book,” “houseless” for “homeless,” “diseasism” for “discrimination against the ill by the temporarily healthy” (Beard & Cerf, 1992). For some reason then, single words are seen as better, more serious, and more appropriate than chains of words. Additionally, it seems that it is especially nouns that carry some special value. If one has a noun at his or her disposal to refer to something, then she or he is likely to be perceived as having the right tool at his or her disposal; with it one can get the point expressed easier. When one uses many words, especially words other than nouns, she or he will more likely be perceived as equivocating. For some reason, the noun appears to be more valuable and taken more seriously. The value assigned to single words, and especially nouns, may have something to do with our essentialist heritage—that is, the belief in only one single word for one concept relationship. We might, in fact, try to exploit this essentialist belief for our own nonessentialist purposes, and we



might encourage people to use new single words instead of multiword descriptions. This is exactly what happens in science and in the technical domain. In the technical domain, we hear and begin to use new words all the time. It seems to be true that new words are easily coined in the technical domain, whereas the process seems to lag behind quite significantly in the social domain. With reference to the computer, for example, new terms that reflect new phenomena are coined continuously. We are flooded with gateways, esps, dsls, information superhighways, e-mails, cyberspaces, Web sites, bit bangs, and so on. Some of these terms are debated, and opinions are expressed as to whether they do or do not appropriately reflect the nonlinguistic reality they are intended to reflect. For instance, discussions are held about whether e-mail should be used as a verb. We obviously need new words to refer to new computer-related phenomena, which are fairly complex. We get them. They get coined. However, we also need new words to refer to new social phenomena, which seem to be even more complex than those that pertain to the technological domain. Although it seems fairly easy to coin words that pertain to the technological domain, we do not seem to be able to coin new words with comparable ease with respect to the social domain, where the need is so acute. We may speculate about the reasons for that. One possible explanation is that language users are reluctant to coin new words relating to social phenomena because social phenomena, as opposed to technical phenomena, reflect an abstraction rather than a physical invention. Discriminations in the abstract world seem to be more difficult to make than those in the physical world. Social life is more elusive, variable, and changeable. It may be that it is harder in the social domain, again, compared with the technical domain, to grasp and hold the new constellations of relationships. In other words, it may be hard for a language user to conceptually retain and give a name to a Catholic who has been baptized, has made his or her first communion, does not take communion regularly afterward, goes to church only occasionally, loves the Pope, is prochoice, and does not use contraceptives. We would then need another name (word) to refer to a Catholic who may be characterized as exhibiting all the characteristics just mentioned except for the last one, does use contraceptives. Of course, we would still need another word to refer to a Catholic for whom the second to the last characteristic on the list would be antichoice. The reader should be cautioned that the argument presented earlier is not an argument for promoting Bushisms—that is, for new coinages of the President Bush type (see Cienki, 2003). In fact, our proposal meets Bushisms in a head-on collision. Bush’s new coinages such as “misunderestimated,” “resignate,” “embetterment,” and “securitize” may be a result of creativity and unorthodox thinking (involving blends such as “misunderstood” and



“underestimated,” yielding “misunderestimated”) or a confused mind. This book certainly encourages creativity and unorthodox thinking. Bushisms, however, go beyond the types of words encouraged in the prior argument. This is because the types of word innovations that I encourage in this book are supposed to improve understanding and communication. Bushisms do not contribute to these. On the contrary, understanding Bushisms often requires access to Bush’s mind; that is, only he can understand what he says. The reason for this is that the cognitive connections that understanding his language requires are not available to most speakers of English. Understanding new words may require a measure of cognitive effort, but understanding should be possible in principle. New words need to build on cognitive bridges, leading the interlocutor to concepts that she or he is familiar with, which makes understanding possible. Bushisms hardly build on such bridges. A measure of uncertainty in interpreting new words may be unavoidable, but it should not lead to pure guesswork, as is often the case when Bushisms are produced. More important, as Cienki (2003) indicated, the use of Bushisms may indicate a lack of empathetic thinking about the audience (because the audience does not understand what is being said). On the contrary, my own argument should lead to treating the audience with much empathy and respect. It should definitely lead to making communication easier and more efficient, not more difficult and less efficient. To recapitulate, to promote nonessentialism, we should teach, first of all, that words are not sacred things to be treated with awe. We should teach that definitions of words are not indisputable. We should teach that words do not have one correct meaning, and deviations from what might be the typical meaning of a word should not be seen as deplorable departures from one central meaning. We should teach that words are not the things to which they refer. We should teach that words are best understood as convenient tools for handling reality. If the tools appear to be bad, outdated, or inconvenient, we should try to invent new ones. We should think of modern cars as an analogy. Modern cars cannot be repaired with an ordinary screwdriver, a simple wrench, or a hammer, as was the case years ago. New cars require new tools to maintain and fix them. Old linguistic tools (words) may become outdated in a similar fashion. In addition, we should teach that the ambiguity of words is normal, and that although reducing ambiguity of words may be a desirable goal, doing away with ambiguity altogether is not an attainable goal. Finally, we should teach widely that new words can and should be invented if the need arises. It is important to realize that the prior proposal should be seen as encouraging the invention of new words only when such invention is motivated. In other words, the proposal should not be seen as encouraging the invention of new words just for the sake of inventing new words. The encouragement to introduce new words serves a



clear purpose: New words should be introduced to decrease ambiguity and transform conceptual and linguistic binary oppositions into continua. Coining new words and proposing them for use needs to be done with caution. One thing is certain: We should not propose words that are unlikely to be accepted by the language users. Of course we will never know for sure that a new coinage will be accepted or rejected, but some proposals will obviously be more realistic than others. In connection with this, we need to remember that some attempts at language change are vehemently rejected by language users (e.g., the 18th-century spelling reform proposals in American English). Other proposals (such as those relating to computer language) are accepted without much resistance. Some resistance must always be expected. Even computer language innovations are not entirely smooth. Some people reject e-mail as a verb on the grounds that we already have verbs such as type and write, which can be combined with the noun e-mail, and that e-mailing someone involves typing and writing. There is no procedure or strategy that will make our new coinages proposals accepted. There does not seem to be, in fact, any one way or any one set of procedures that would guarantee success for the introduction of new words. Any way is worth trying. First of all, whenever we try to introduce a new word, we should be fully cognizant that it may be by all means advantageous to do so. One important guideline then is not to feel constrained, restrained, scared, or intimidated. We should use new words whenever we feel there is a need to do so, and we should not feel at all that we are doing anything wrong. In other words, the primary thing in our using new words is overcoming a psychological barrier. This psychological barrier was erected by essentialism a long time ago, and it is high time that the barrier be overcome. Once the average language user feels that she or he is entirely free to be creative linguistically, new words may start entering the lexicon in different ways and directions. We do not need to propose ways in which new words should be introduced. Suffice it to say that language users should feel free to introduce new words whenever they sense the need to do so. It is extremely important to realize that the previous argument for new words has nothing to do with “prescriptivism.” The argument is not meant to advocate an introduction of new words that language users would be expected to consult and respect as a standard. The reader should not assume that some kind of language police is being proposed here (the way that some PC language advocates want things to be controlled), which would stigmatize or attempt to block the publication of adjectival or nominal opposites, as well as terms that would appear to some as too all-encompassing. The thinking proposed here is supposed to lead to coinages that language users would sincerely feel to be useful for their handling of everyday matters. Moreover, and



most important, language users are encouraged to coin new words especially when they feel that the new words and expressions launched in the “language space” can possibly contribute to alleviating conflicts. This brings us back to one of the major concerns of this book—namely, conflict. We should and, in fact, can do a lot to avoid or at least alleviate them. There is every reason to believe that binary opposition categorization and the use of highly abstract words (underpinned by essentialist thinking) contribute to confrontational and belligerent human behavior (see also Tannen, 1998, on this point). Should we succeed in our endeavor to change our language along the lines proposed earlier, conflicts of various kinds, including large-scale international ones, can be lessened. To recapitulate, essentialism leads to an extensive use of binary oppositions, an extensive use of highly abstract words that pertain to social domains, resistance to new words, and resistance to meaning innovation. The final recapitulating question in this chapter is why the somewhat restricting word use discussed before can be seen as contributing to conflicts. The answer seems to be that this is because a language user who, to refer to a phenomenon, uses few words at a high level of abstraction tends to perceive reality in a significantly oversimplified form. Language users are often unaware of the complexity of the phenomena around them. It is that lack of awareness and the perceptual oversimplifications that derive from it that encourage simple answers to difficult questions. Simple answers to questions are promoted when mental categorizations are multifaceted and the corresponding lexical reflections of those categorizations simple. We refer here primarily to cases such as the Polish Catholic–communist dichotomy, where words offer merely binary opposition categorization and where they do not allow low-level abstraction. To reiterate, especially highly abstract binary oppositions encourage confrontational behavior. Due to binary opposition words, the world tends to be perceived in terms of us and them, black and white, friends and enemies, good guys and bad guys. There is no clearly indicated space here for other options. CONCLUSIONS As in the other sections of this book, a crucial goal of the present education section has been to raise the reader’s level of consciousness. This part of the book is also meant to encourage the reader to talk about essentialism to others, to show its fallacy, and to show what kind of pejorative consequences it may lead to. It is hoped that education against essentialism may, in the long run, clear the way for education for permanent peace, not to speak of more immediate, equally desired, and related goals such as freedom and tolerance.



The social groups that should be approached first and most seriously about the problem of essentialism reflected in language are teachers and politicians. These two groups have enormous influence on our lives. With respect to these two social groups and with respect to their potential function in promulgating nonessentialism, some optimism may be expressed. In fact, optimism is likely to intermingle constantly with pessimism. On one hand, we might agree with Popper, who once expressed the view that Aristotle’s ghost will haunt us still for many years to come.5 We should obviously wish Popper were wrong in his prediction. Some of the recent observations in the field of linguistics, for example, may make us a bit optimistic. Education against essentialism may take place in various institutions and places. It should be most salient, and it is probably the easiest to implement, however, in the educational institutions. This is because education is realized mainly through language and because much of it is about language. These are two reasons that the teacher’s position is crucial in promoting ideas about language. Of the many linguistic issues prominent in the educational setting, two that relate closely to the core topic of this book—essentialism—and that clearly stand out are standardization and the treatment of words. With reference to the former, an essentialist view translates into a strong argument for viewing standard varieties as inherently valuable and worth learning and teaching at any cost. Nonessentialism, on the other hand, although possibly acknowledging certain advantages of standardization (mainly those of possessing a written standard), by all means rejects that standardized varieties are better in any absolute sense. Nonessentialists will point to the fact that the essentialist view of standard varieties is detrimental to the speakers of nonstandard varieties and that this view promotes intolerance, inequality, and basic injustice. With reference to the latter—that is, the treatment of words—essentialism translates into viewing words, and especially word meaning, inflexibly. Nonessentialism, in contrast, treats word meaning with much flexibility and creativeness. This leads to a positive stance toward new coinages. These coinages are encouraged because many of the existing words, especially in the social sphere, are seen as drastically simplifying reality. In the nonessentialist view, the new words are encouraged so that perception in terms of continua rather than bipolar extremes should be promoted.

ENDNOTES 1. A detailed report on the congress can be found in Janicki and Jaworski (1993). 2. All translations from the Polish in Urba½czyk (1987) by Janicki and Jaworski.



3. Opposites such as Catholic–communist are obviously culture/country-specific. It seems that in most countries other than Poland, Catholic was not at all seen as the opposite to communist. 4. That particular binary opposition pertains to the pre-1989 Poland when Solidarity was a huge anticommunist organization. 5. This view was expressed in Popper’s private correspondence to me.




In this final part of the book, in chapter 9, I summarize the differences between the essentialist and nonessentialist views and state the reasons for embracing the nonessentialist position for handling language, in general, and for doing applied cognitive sociolinguistics, in particular. I reiterate the main points made in Parts I and II, which explain how language can be viewed and studied in a nonessentialist way.




Viewing and Studying Language a Nonessentialist Way

Throughout the preceding eight chapters, I have argued for a nonessentialist view of language, in general, and for doing what might be called applied cognitive (nonessentialist) sociolinguistics, in particular. Applying some of the central ideas of cognitive linguistics for analyses of sociolinguistic and sociopolitical phenomena is not entirely new. G. Lakoff’s (1996, 2004) work on metaphors in political discourse and his popular most recent book are excellent examples. I put some nonessentialist cognitive linguistics ideas together with politics in the beginning of the 1990s ( Janicki, 1991). Applying cognitive nonessentialist reasoning to describe and analyze sociolinguistic phenomena is not, however, a widely exercised activity. It is hoped that the present book can be taken as a further encouragement to this kind of activity. I would like to summarize the nonessentialist view of language advocated in this book in the form of a set of instructions or pieces of advice. This advice is intended to benefit a student of language committed to adopting a nonessentialist approach to language study. Although the pieces of advice listed next have been given with a student of language in mind, they can in fact be found useful by anyone involved in academic research, especially within the social sciences and the humanities. The list warns against the most important essentialist pitfalls and guides the nonessentialist into smooth and rewarding research. 1. Accept the view that making a philosophical commitment is important in doing academic research. In other words, decide for yourself what 175



you believe exists, what you believe language should be studied as, whether you want to study it as knowledge or as behavior or as a combination of the two, and so on. When you have done so, be consistent. For example, do not let yourself be caught studying something that you earlier posited did not exist. Your philosophical commitment may (but does not need to) be written about. However, it needs to be spelled out intellectually. That is to say, as a researcher, you need to be fully aware of what your philosophical assumptions are on fundamental ontological and epistemological issues. I hope you will make a nonessentialist philosophical commitment. Then consistently follow the advice listed in Points 2 through 21. 2. Do not assume, like the essentialists, that meaning can be found in a word; words are not containers into which meaning is put at one end of a communicative chain and taken out at the other. Words do not have fixed essential meaning that is transmitted with them when they travel from a language speaker to a language hearer. It is much more realistic and rewarding to posit, as the nonessentialist does, that meaning is assigned to words by individual language users. Although a word’s meaning is often similar or very similar for many speakers of a particular language, always be prepared to accept that language users may differ in the meaning they assign even to such very commonly used words as car, father, or lamp. 3. Avoid what-is questions in an essentialist fashion. That is, do not ask what-is questions with the view in mind that they can in principle be answered in one absolute correct way. For a nonessentialist, the only way to approach questions such as “What is freedom?” and “What is sociolinguistics?,” which, as I have shown earlier, are often asked by essentialists or semiessentialists, is to consider the question a request for information as to what kind of social situation, or what kind of phenomenon, people call, or may want to call, freedom and what kind of activities, or study, people call, or may want to call, sociolinguistics, respectively. For a nonessentialist, a what-is question requiring one ultimate answer, the way the essentialist would have it, is simply the wrong kind of question; it is not answerable. What-is questions may also be very confusing. You are well advised not to pose them at all! 4. Do not fall victim to etymological fallacy; that is, do not believe, as some essentialists do, that the original meaning of a word or expression is the correct meaning. In the nonessentialist view, the original meaning is the original meaning; it is not a privileged or ultimately correct meaning. Any meaning deviating from the original meaning is equally legitimate no matter how different from the original meaning it may be. 5. Accept (after Popper) the view that all concepts are imprecise and that examples of the fuzzy meanings of all words in natural languages may always be provided. Accept (after cognitive linguists) the prototype theory



of the concept, rather than the classical checklist theory. Accept the view that concepts are best viewed as typical mental images/descriptions. Accept further the view that meanings of words may be illustrated by central, less central, peripheral, and borderline examples. Take this to be a normal state of affairs in natural language, and do not believe that words used in such fields as science and law, for instance, are exceptions. 6. Do not complain, as the essentialists often do, that the concepts/ terms you are dealing with are especially difficult to define, which allegedly puts you in a predicament; do not say that terms such as preposition, sentence, and communicative competence (with which you as a linguist might currently be working) are difficult to define. If you say so, you give others the impression that the term(s) you are trying to define are exceptional in that they are difficult to define, whereas other terms (with which you are not currently working) are not difficult to define. The terms that you are working with are not exceptional; like all other terms, they are not ultimately definable. For instance, the fact that you cannot ultimately answer the question, “What is communicative competence?,” is not surprising or special in any sense at all. If you want to approach language in a nonessentialist way, what you need to do is simply state what your working (operational) definition for a term is; that is, you need to say how you will be using the term(s) in your project. Having made a decision as to how you will be using the term(s), be consistent throughout your project. Do not depart from the working definition that you have adopted. Disregard other definitions until your project is completed. 7. Your working definition should not be selected inadvertently. The definition that you will choose as your working definition must be useful for your project; it must be useful for tackling a question/problem standing outside your definition. In a nonessentialist’s view, definitions are never a or the goal of a study. Definitions should be provided only when they may help tackle some task (e.g., a problem; see Point 21) external to the definition at hand. For example, if you are involved in a sociolinguistic project dealing with medical language, you are likely to find useful and will most probably adopt the following definition of register: “. . . is used to describe a language variety that is associated with a particular topic, subject or activity” (Trudgill, 2003, p. 110). You are less likely to find useful and thus less likely to adopt the following definition: “used to refer to variation according to the context in which language is used” (Swann, Deumert, Lillis, & Mesthrie, 2004, p. 261). Even less likely, or rather unlikely, are you to find useful and very unlikely to adopt the following definition of register: “an official list or record, for example, of births, marriages, and deaths” (The New Oxford Dictionary of English, 1998, p. 1562). In all probability, you would find completely useless for your project the last definition as well as many others.



8. Be prepared to accept working definitions that seem counterintuitive and that are likely to seem so also to many of your colleagues. Gibbons (2003) reported on two, what might appear intuitively to be outrageous definitions, which have, however, been accepted for legal purposes: (a) “A person . . . an individual, partnership, corporation, limited liability company, association, or other group, however organized” (Consumer Legal Remedies Act, California; quoted in Gibbons, 2003, p. 48); and (b) “. . . day-old poultry is deemed to be: any poultry of an age of 72 hours or less” (New Zealand Poultry Act, 1968; quoted in Gibbons, 2003, p. 48). These definitions may be unusual in their deviation from the ordinary understandings of the words person and day-old poultry. They are nevertheless fully legitimate working definitions. 9. Your working definition should include terms that are less abstract— preferably significantly less abstract—than the term defined. Your working definition is to help understand the term defined, to make it clearer (but never ultimately precise). That is why the terms in the definitions should be less abstract than the term being defined. For instance, defining cline as a “continuum with an infinite number of gradations” will not be very helpful. Cline and continuum seem to be at the same (high) level of abstraction. However, to define gesture as “a movement of a bodily part for the purpose of communication” may be taken as a good example of how a definition offers a significantly lower level of abstraction compared with the term defined. Your working definition should not only be useful; it should also be easy to understand. For instance, if you chose to adopt the following definition of a gypsy—“anybody who calls himself/herself a gypsy”—you will be easily understood; this definition can hardly be improved as far as comprehensibility is concerned. 10. Consider the function of working definitions mainly to be an attempt to make things clearer (see Point 9), as an attempt to help the reader/hearer follow the argument of a paper, book, project, and so on. Defining terms may not always be necessary. Do not take “define your terms” as a prerequisite for any kind of study. However, appreciate that providing working definitions of terms is often very useful and that it may be important in a particular project. For a nonessentialist, working definitions are by all means welcome as long as they are viewed as such and not as attempts to grasp the essence of things, as the essentialist would have it. 11. Do not use expressions such as “not yet,” “strictly speaking,” “in the true sense,” or “exactly” in sensitive sections of your discussion of working definitions. The following formulations may be taken to imply that ultimate essentialist meanings of the terms are being sought: “This term hasn’t been defined as yet,” “Strictly speaking, this is not a sentence,” “This is not liberty in the true sense of the word,” “This is not really optimism.” Such formulations should be avoided at all costs.



12. Making distinctions between or among disciplines or subdisciplines should not be an important goal of your study or of some considerations. For example, do not seriously engage yourself in discussions about whether a study falls into pragmatics or sociolinguistics. In a nonessentialist’s view, a distinction, for instance, between pragmatics and sociolinguistics can never be stated once and for all because the meanings of the words pragmatics and sociolinguistics cannot be stated once and for all; this is just the opposite of what an essentialist tends to think. In light of some definitions of the two terms, pragmatics is a superordinate term to sociolinguistics; in light of others, pragmatics is more or less the same as sociolinguistics. In light of still others, pragmatics and sociolinguistics refer to significantly different types of activities. For a nonessentialist, it is impossible to give one clear answer to the question of how pragmatics relates to sociolinguistics. The relationship between the two terms/disciplines depends on how they are defined. There are two things that a nonessentialist can do when faced with making a distinction between the two notions/terms/disciplines: One is to point to what people calling themselves pragmaticians or sociolinguists actually do; that is, what kind of research they actually carry out. The other is to give working definitions of the two terms, if this is advisable for some reason (see also the discussion in Point 7). 13. Do not argue about discipline boundaries, and feel free to cross them in research activities. Discipline boundaries are not natural and are not there to stay once and for all. They are imposed by human beings, by human categorization. The essentialist tends to respect discipline boundaries and avoid crossing them. The nonessentialist, in contrast, acknowledges the existence of structure and boundaries (fuzzy as they are), but is encouraged to cross them. Discipline boundaries can be moved, crossed, or discarded; new boundaries may be created. The nonessentialist encourages flexibility with respect to treating boundaries. The nonessentialist will endorse Zerubavel’s (1993) statement, “Flexible people notice structures yet feel comfortable destroying them from time to time” (p. 120). 14. Attend to discipline verges and exploit them in a creative manner. “Creativity usually flourishes on verges. . . . Creative scholars . . . defy the parochialism of insular academic disciplines” (Zerubavel, 1993, p. 117). An essentialist stays away from verges. A nonessentialist resorts to them and avails him or herself of them. A nonessentialist does not need to formally move, redefine, discard, or cross boundaries. A nonessentialist is comfortable positioning him or herself on the verge of disciplines and doing research whatever the disciplines around him or her are called and wherever other researchers may want to draw the lines for the disciplines in question. In principle, for a nonessentialist, discipline boundaries are important only as conveniences, and verges of disciplines are enticing.



15. In general, do not argue about the use of words. Essentialists often say that this or that word is used inappropriately, meaning that the word is used to refer to what it should not be used to refer to. As a nonessentialist, you should not subscribe to this kind of reasoning. For example, if your interlocutor uses the term discourse analysis where you would rather use the term conversational analysis, do not make an issue of this. One should remember that the nonessentialist treats the conventionality of language seriously. That is why she or he can treat the use of words lightly. It is not important what term is used to refer to some object or state of affairs (a referent). What is important, especially in academic discussion, is recognizing the object or state of affairs (the referent) in question. Reaching agreement with an interlocutor with respect to the former may be helpful, but is not really crucial or even important. Reaching agreement with respect to the latter is crucial because only after agreement has been reached with respect to the latter can the interlocutors know, at least more or less, what it is that they are talking about. 16. Do not make taxonomies (classifications) an important goal of your study. Taxonomies are never final; they can always be questioned and challenged. According to the nonessentialist, the main reason for this is that no category of a taxonomy can ever be finally and unambiguously defined; categories are labeled by words, which, as is now obvious to the reader, in the nonessentialist view cannot be precisely defined; nor can the words included in the examples, of course. If that is the case, assigning examples to categories of a taxonomy may always be questioned either because the various categories of a taxonomy may be interpreted in different ways or because examples (i.e., words in the examples) may be interpreted in different ways. For a nonessentialist, no taxonomy is correct in itself. It may only be useful for a goal standing outside the taxonomy (see Point 17). Consider the following example. Discussing metaphors, Kövesces (2002) stated: “Conceptual metaphors can also be classified according to the cognitive functions that they perform. On this basis, three general kinds of conceptual metaphor have been distinguished: structural, ontological, and orientational” (p. 33). Subsequently, he defined the three categories of his taxonomy: in structural metaphors, “the source domain provides a relatively rich knowledge structure for the target concept. In other words, the cognitive function of these metaphors is to enable speakers to understand target A by means of the structure of source B ” (p. 33; italics added). Ontological metaphors are these that “provide much less cognitive structuring for target concepts than structural ones do. Their cognitive job seems to be to ‘merely’ give an ontological status to general categories of abstract target concepts” (p. 34). Finally, Kövesces (2002) defined the category, orientational metaphor, of his taxonomy in the following way: “Orientational metaphors provide even less conceptual structure for target concepts than onto-



logical ones. Their cognitive job, instead, is to make a set of target concepts coherent in our conceptual system” (p. 35). A quick and superficial look at these three definitions is sufficient to make us realize immediately that many of the central terms in the definitions provided are open to many interpretations. For example, the terms knowledge structure, cognitive function, general categories, and abstract target concepts are, obviously, not precise. We can understand them in many different ways. Because the terms defining the categories of the taxonomy are never precisely worded, we will always be able to find examples whose assignment to more than one category of the taxonomy is possible and whose assignment to any of these categories may be doubtful. This is one important reason that taxonomies can always be challenged and, thus, why setting them as an important goal of a study does not appear to be particularly fruitful. 17. Do not discard taxonomies altogether. Realizing that taxonomies can never include final and watertight categories and realizing that they can always be challenged should not, however, lead us to the conclusion that they are entirely useless. On the contrary, taxonomies may also be useful for a nonessentialist. They introduce a measure of structure to our data; they let us group examples that are similar to one another with respect to some criteria. We can assume Kövesces was motivated to list the three categories of conceptual metaphor (see Point 16) to provide some structure and to group examples. This may be a good reason to set up taxonomies as long as we always remember that no ultimately correct taxonomy can be produced. For a nonessentialist, however, setting up a taxonomy is best suited for the kind of studies where a goal external to the taxonomy is clearly identified. For instance, bothering about types of foreign language learners is seen especially useful in the context of trying to solve some foreign language-learning problems. For a nonessentialist, identifying types of foreign language learner (providing a taxonomy of the foreign language learner) without some context is not an interesting and rewarding research activity. 18. What is important is what you do (what your study is about) and not what you call what you do or study. A nonessentialist will not argue about whether a study is to be labeled, for instance, sociolinguistic or pragmatic. A nonessentialist will tend to show what it is that she or he studies, rather than argue about what the study is to be called or insist that the study be labeled or called sociolinguistic, pragmatic, linguistic, or anything else. The main reason for this nonessentialist position is that these terms, like all others, cannot be defined precisely. A nonessentialist will expect that a study may be labeled as a typical sociolinguistic study or a typical psycholinguistic study, but will not engage in any serious discussion of whether some particular study is definitely a study in sociolinguistics, pragmatics, linguistics, psycholinguistics, and so on. Even disputes over whether a study is, for instance, a typical



sociolinguistic study, which may occur, will not be of particular interest to a nonessentialist. If labeling what a nonessentialist would call a typical sociolinguistic study is challenged, and if another label (e.g., a psycholinguistic study) is offered or argued for by an opponent, the nonessentialist will easily give in; the question of labeling or naming activities is not important to him or her. 19. Accept the view that misunderstanding in linguistic communication is not a marginal phenomenon. On the contrary, it is ubiquitous, and it should be seen as a corollary of how the nonessentialist assumes language works. Misunderstanding may be seen as a function of the imprecision of concepts and the ambiguity of words. Misunderstanding, like concept imprecision, can be of varying degrees. More important, because ultimate concept precision and word meaning can never be reached, misunderstanding can never be ruled out. 20. Make sure to treat essentialism–nonessentialism as a continuum. It is easy to find typical examples of essentialism (e.g., what-is questions asking for some ultimately correct answer) as well as typical examples of nonessentialism (e.g., formulations clearly defying what-is questions as asking for some one final answer). However, it is also easy to encounter formulations about which people may disagree as to whether they are examples of essentialism or nonessentialism. For instance, the formulation, “It remains debatable whether the question ‘what is life’ is answerable,” does not allow us to assign it, without much doubt, either to essentialism or to nonessentialism; on the basis of this formulation, it is difficult to state whether its author believes that the question, “What is life?,” is ultimately answerable. We are dealing here with a fuzzy, borderline case. Likewise, make sure to treat overt–covert essentialism as a cline. An example of overt essentialism can be a straightforward what-is question expressed in formulations such as, “What is life exactly remains to be a very important question.” An example of what I call covert essentialism can be a formulation such as, “We do not know yet the nature of life”; in this case, no straightforward what-is question is asked. The latter example can be treated as a borderline case because it may be argued that “the nature of life” easily translates into “the essence of life,” in which case the essentialism expressed is quite apparent. Overt and covert essentialism may sometimes be useful labels, but a nonessentialist will never vehemently argue about whether some example is definitely an example of overt essentialism or covert essentialism. 21. Try to locate and address a problem that will constitute the starting point for your research. This recommendation does not spring directly from nonessentialist philosophy, but may be understood to be fully compatible with it. Although the nonessentialist appreciates descriptive and ex-



planatory endeavors and although she or he actually gets involved in them, she or he values addressing problems, and especially practical problems, which may need attention. We may want to ponder the question of what is to be taken to be “a problem”—that is, we may want to discuss possible working definitions of a problem. The notion of a problem may, obviously, be defined in different ways— that is, language users differ in what they conceive of as a problem. One possible way to use the word problem, which I believe is shared by many language users, is to have it refer to a situation in which something goes wrong. For instance, a house that collapses or a car that breaks down will be conceived of by many people as a problem. In other words, people normally do not face a problem when their house stands and their car runs; they do when their house collapses or when their car breaks down. I suggest that one try to identify problems in the sociolinguistic domain comparable to those mentioned in the earlier examples. The nonessentialist should treat addressing problems (and subsequently trying to solve them) as a high-level, methodological guiding principle. There are at least two reasons for this: One is that merely descriptive and especially taxonomic studies are easy to criticize. This is true because, in the nonessentialist view, no description or taxonomy may be ultimately correct. The nonessentialist assumes that no description can grasp the essence of things and that all description, being theoretically impregnated (Popper, 1972), expresses a perspective. The other reason is that the nonessentialist sociolinguistics developed in this volume is meant to be socially sensitive and socially significant sociolinguistics. As indicated in the subtitle of this volume, what I have been promoting in this book is an applied cognitive sociolinguistics. I take such an enterprise to be geared toward identifying and offering solutions to practical human problems of a largely linguistic (and at least partly philosophical) provenance (see our earlier discussion of all the topics related to politics, law, academia, and education). If one takes the label applied seriously, locating and addressing practical problems where the use of language plays a role may be a very rewarding overarching anchor. Needless to say, what will be judged to be a sociolinguistic problem worth addressing is ultimately up to the individual researcher or to a group of researchers to decide (see e.g., Maxwell, 1984; Kitcher, 2001, on this point). The last piece of advice may be addressed to all researchers no matter what their philosophical commitment: 22. Try to find out what your interlocutor’s (opponent’s, critic’s) philosophical position is. If you are successful, compare it with yours and try to discover whether any disagreement between you and your opponent is perhaps due to some high-level, fundamental philosophical differences, rather than to any lower level differences surfacing in viewing a linguistic phenom-



enon. If you discover a fundamental philosophical difference between you and your opponent, state the difference. Disclosing and stating such a difference may be seen as a very civil way of ending a discussion; if people differ as to their fundamental philosophical assumptions, any further discussion may be unproductive and disadvantageous. For instance, if you and another researcher discuss and find it hard to come to an agreement as to whether American English should be viewed as a language or a national dialect, immediately check whether your opponent believes that the terms language, dialect, variety, lect, and so on can be unambiguously defined. You, as a nonessentialist, will not believe that, but your opponent may if she or he is an essentialist. If that is indeed the case, your opponent’s view and yours are not reconcilable philosophically, and any further discussion will most probably be pointless or utterly frustrating!


The main objective of this book has been to show that some professional linguists and many lay people exhibit significant misconceptions about language. The misconceptions discussed in this book can all be subsumed under the heading of essentialism. The professional and social consequences of these misconceptions may be grave. To the extent that we succeed in fighting essentialism, we can alleviate the social consequences it generates. The nonessentialists have a Herculean job to do. Because the essentialists generally have the upper hand, it can only be hoped that the job is not a Sisyphean one.



Aarts, B., Denison, D., Keizer, E., & Popova, G. (Eds.). (2004). Fuzzy grammar: A reader. Oxford: Oxford University Press. Aitchison, J. (1994a). Understanding words. In G. Brown, K. Malmkjær, A. Pollott, & J. Williams (Eds.), Language and understanding (pp. 81–96). Oxford: Oxford University Press. Aitchison, J. (1994b). Words in the mind (2nd ed.). Oxford, England: Blackwell. Aristotle. (1983). Metafizyka [Metaphysics] (K. LeÑniak, Trans.). Warsaw: Polish Scientific Publishers. Bargiela-Chiappini, F., & Harris, S. (Eds.). (1997). The languages of business. Edinburgh: Edinburgh University Press. Baugh, J. (2000). Beyond Ebonics: Linguistic pride and racial prejudice. Oxford: Oxford University Press. Bavelas, J. B., Black, A., Bryson, L., & Mullett, J. (1988). Political equivocation: A situational explanation. Journal of Language and Social Psychology, 7, 137–145. Bavelas, J. B., Black, A., Chovil, N., & Mullett, J. (1990). Equivocal communication. Newbury Park, CA: Sage. Bealer, G. (1987). The philosophical limits of scientific essentialism. Philosophical Perspectives, 1, 289–365. Beard, H., & Cerf, C. (1992). The official politically correct dictionary and handbook. New York: Villard Books. Bex, T., & Watts, R. J. (Eds.). (1999). Standard English: The widening debate. London: Routledge. Brown, P., & Levinson, S. (1987). Politeness: Some universals in language usage. Cambridge, England: Cambridge University Press. Bråten, E. (2004). Kristeva, Flåm og universitetet [Kristeva, Flåm, and the university]. På Høyden. Nettavis for UiB, 18, 12.04, 1. Bull, P. (1998). Equivocation theory and news interviews. Journal of Language and Social Psychology, 17, 36–51. Bull, P. (2000). Equivocation and the rhetoric of modernization. An analysis of televised interviews with Tony Blair in the 1997 British General Election. Journal of Language and Social Psychology, 19, 222–247. Cameron, D. (1995). Verbal hygiene. London and New York: Routledge.




Carroll, J. B. (Ed.). (1956). Language, thought and reality: Selected writings of Benjamin Lee Whorf. Cambridge, MA: MIT Press. Chalmers, A. F. (1996). What is this thing called science? Buckingham, England: Open University Press. Channell, J. (1994). Vague language. Oxford: Oxford University Press. Chilton, P., & Schäffner, C. (2002). Introduction. Themes and principles in the analysis of political discourse. In P. Chilton & C. Schäffner (Eds.), Politics as text and talk: Analytic approaches to political discourse (pp. 1–41). Amsterdam: John Benjamins. Cienki, A. (2003, May 2). Will new word order embetter us? Times Higher Educational Supplement, Internet version. Cloud, J. (1998, September 21). What exactly are “high crimes and misdemeanors”? TIME Magazine, p. 33. Coleman, L., & Kay, P. (1981). Prototype semantics. Language, 57, 26–44. Cowley, G., & Springen, K. (1999, January 25). A second opinion on sex. Newsweek, p. 16. Crowley, T. (1998). Curiouser and curiouser: Falling standards in the standard English debate. In T. Bex & R. J. Watts (Eds.), Standard English: The widening debate (pp. 271–282). London: Routledge. Crystal, D. (1997). The Cambridge Encyclopedia of Language (2nd ed.). Cambridge: Cambridge University Press. de Bono, E. (1991). I am right you are wrong. Harmondsworth: Penguin. Downes, W. (1998). Language and society (2nd ed.). Cambridge: Cambridge University Press. Dunant, S. (Ed.). (1994). The war of the words: The political correctness debate. London: Virago Press. The Economist. (1996, September 14). What is man, p. 13. The Economist. (1997, February 1). Knowledge or certainty?, p. 94. The Economist. (2000a, November 18). Unleashing the dogs of law, pp. 65–67. The Economist. (2000b, November 25). Whatever will they think of next?, pp. 71–77. The Economist. (2001a, May 26). Is a fetus a person?, pp. 54–55. The Economist. (2001b, July 7). The cutting blob of ethical politics, pp. 49–50. The Economist. (2002a, May 11). Biology and politics. The great cloning debate, pp. 45–46. The Economist. (2002b, January 26). The prisoners dilemma, p. 46. The Economist. (2002c, August 31). Putting his cards on the table, pp. 33–34. The Economist. (2003, January 18). The war that never ends, pp. 24–26. The Economist. (2004a, October 9). Regional languages. Spanish practice, p. 32. The Economist. (2004b, October 16). Language barrier, p. 16. Edwards, J. (1985). Language, society and identity. Oxford: Basil Blackwell. Escobar, A. (1999). After nature: Steps to an antiessentialist political ecology. Current Anthropology, 40(1), 1–30. Fairclough, N. (1994). “Mainly Muslim”: Discourse and barbarism in Bosnia. Discourse and Society, 5(3), 431–432. Fasold, R. (1984). The sociolinguistics of society. Oxford: Basil Blackwell. Ferguson, C. (1959). Diglossia. In A. Dil (Ed.), Language structure and language use. Essays by Charles A. Ferguson (pp. 1–26). Stanford, CA.: Stanford University Press. Feyerabend, P. (1975). Against method. London: Verso. Feyerabend, P. (1981a). Philosophical papers: Problems of empiricism. Cambridge, England: Cambridge University Press. Feyerabend, P. (1981b). Philosophical papers: Realism, rationalism and scientific method. Cambridge: Cambridge University Press. Feyerabend, P. (1987). Farewell to reason. London: Verso. Feyerabend, P. (1991). Three dialogues on knowledge. Oxford: Basil Blackwell. Fillmore, C. (1996). A linguist looks at the Ebonics debate. Available at ebonics/ebfillmo.html



Fine, K. (1994). Essence and modality: The second philosophical perspective lecture. Philosophical Perspectives, 8, 1–16. Fishman, J. (1970). Sociolinguistics: A brief introduction. Rowley, MA: Newbury House. Flåm, S. D. (2004a, December 4). Høgt prisa skodde [Highly prized fog]. Bergens Tidende, p. 37. Flåm, S. D. (2004b, December 11). Julia de France, Philosophus udi egen Indbildning [ Julia of France, philosopher in her own imagination]. Bergens Tidende, p. 35. Fromkin, V. (1993, October 12). Linguistic history, science. Linguist List, p. 1. Fromkin, V., & Rodman, R. (1993). An introduction to language (5th ed.). New York: Holt, Rinehart & Winston. Geeraerts, D. (1997). Diachronic prototype semantics: A contribution to historical lexicology. Oxford: Clarendon. Gibbons, J. (2003). Forensic linguistics: An introduction to language in the justice system. Oxford: Blackwell. Gibbs, N. (1998, August 10). Tick, tock, tick . . . talk. TIME Magazine, pp. 31–38. Gibbs, R. (1994). The poetics of mind. Cambridge: Cambridge University Press. Gray, B. (1969). Style: The problem and its solution. The Hague, Netherlands: Mouton. Greimas, A. (1987). On meaning. London: Frances Pinter. Gripsrud, J. (2004, December 16). Flammende egghoder [Blazing eggheads]. Bergens Tidende, p. 37. Guthrie, W. K. C. (1971). The Sophists. Cambridge, England: Cambridge University Press. Haastrup, K., & Phillipson, R. (1983). Achievement strategies in learner/native speaker interaction. In C. Færch & G. Kasper (Eds.), Strategies in interlanguage communication (pp. 140–158). London: Longman. Hallett, G. (1991). Essentialism: A Wittgensteinian critique. Albany: State University of New York Press. Hamilton, M., & Mineo, P. (1998). A framework for understanding equivocation. Journal of Language and Social Psychology, 17, 3–35. Harris, R. (1981). The language myth. London: Duckworth. Harris, R. (1997). From an integrational point of view. In G. Wolf & N. Love (Eds.), Linguistics inside out (pp. 229–310). Amsterdam: John Benjamins. Harris, R. (1998). Introduction to integrational linguistics. Oxford: Elsevier Science Ltd. Harris, R. (2002). The role of the language myth in the Western cultural tradition. In R. Harris (Ed.), The language myth in Western culture (pp. 1–24). Surrey, England: Curzon Press. Harris, R. (2003). Saussure and his interpreters. Edinburgh: Edinburgh University Press. Haspelmath, M. (1993, October 14). What kind of science? Linguist List, p. 1. Hayakawa, S. I. (1953). Symbol, status and personality. New York: Harcourt, Brace & World. Hayakawa, S. I. (1965). Language in thought and action (2nd ed.). London: George Allen & Unwin Ltd. Heisenberg, W. (1969). Der Teil und das Ganze: Gespräche im Umkreis der Atomphisik [Part and whole: Conversations about atomic physics]. Munich: R. Piper & Co. Hempel, C. G. (1952). Fundamentals of concept formation in empirical science. Chicago: University of Chicago Press. Honey, J. (1997). Language is power: The story of Standard English and its enemies. London: Faber. Hudson, R. A. (1996). Sociolinguistics (2nd ed.). Cambridge, England: Cambridge University Press. Jahr, E. H. (1989). Limits of language planning? Norwegian language planning revisited. International Journal of the Sociology of Language, 80, 33–39. Jahr, E. H., & Janicki, K. (1995). The function of the standard variety: A contrastive study of Norwegian and Polish. International Journal of the Sociology of Language, 115, 25–45. Jahr, E. H., & Trudgill, P. (1993). Parallels and differences in the linguistic development of modern Greece and modern Norway. In E. H. Jahr (Ed.), Language conflict and language planning (pp. 83–98). Berlin: Mouton de Gruyter.



Janicki, K. (1990). Toward nonessentialist sociolinguistics. Berlin: Mouton. Janicki, K. (1991). Applying linguistics for peace education. International Journal of Applied Linguistics, 2, 164–173. Janicki, K. (1997). On unconventional metaphors as a source of misunderstanding. In H. Ramisch & K. Wynne (Eds.), Language in time and space (pp. 127–138). Stuttgart, Germany: Franz Steiner Verlag. Janicki, K. (1999). Against essentialism: Toward language awareness. Munich: Lincom Europa. Janicki, K. (2002). A hindrance to communication: The use of difficult and incomprehensible language. International Journal of Applied Linguistics, 12(2), 194–217. Janicki, K., & Jaworski, A. (1993). Language purism and propaganda: The first congress for Polish. In J. Fishman (Ed.), The earliest stage of language planning: The “first congress” phenomenon (pp. 219–231). Berlin: Mouton de Gruyter. Jardine, L. (1994). Canon to left of them, canon to right of them. In S. Dunant (Ed.), The war of the words: The political correctness debate (pp. 97–115). London: Virago. Johnson, M. (1987). The body in the mind. Chicago: The University of Chicago Press. Johnson, M. (1993). Moral imagination: Implications of cognitive science for ethics. Chicago: The University of Chicago Press. Johnstone, B. (2000). Qualitative methods in sociolinguistics. Oxford: Oxford University Press. Kemp, L. (2000, December). Bach. Gramophone, p. 96. Kitcher, P. (1984). The nature of mathematical knowledge. Oxford: Oxford University Press. Kitcher, P. (2001). Science, truth and democracy. Oxford: Oxford University Press. Korzybski, A. (1933). Science and sanity: An introduction to non-Aristotelian systems and general semantics. Garden City, NY: Country Life Press Corporation. Kövesces, Z. (2002). Metaphor: A practical introduction. Oxford: Oxford University Press. Król, J. (1995). Katolicka poprawnoу [Catholic correctness]. Wprost, 47, 11. Krzeszowski, T. (1991). The axiological parameter in preconceptual image schemata. In R. Geiger & B. Rudzka-Osty½ (Eds.), Conceptualization and mental procession in language (pp. 315–338). Berlin: Mouton de Gruyter. Kuhn, T. (1962). The structure of scientific revolutions. Chicago: The University of Chicago Press. Labov, W. (1975). What is a linguistic fact? Lisse, Belgium: The Peter de Ridder Press. Lacayo, R. (1998, August 24). When is sex not “sexual relations”? TIME Magazine, p. 16. Lakatos, I. (1974). Popper on demarcation and induction. In P. A. Schlipss (Ed.), The philosophy of Karl Popper (pp. 241–273). La Salle, IL: Open Court Publishing. Lakatos, I. (1978a). Mathematics, science and epistemology. Philosophical papers. Vol. 2 ( J. Worrall & G. Currie, Eds.). Cambridge, England: Cambridge University Press. Lakatos, I. (1978b). The methodology of scientific research programmes. Philosophical papers. Vol. 1 ( J. Worrall & G. Currie, Eds.). Cambridge, England: Cambridge University Press. Lakoff, G. (1982). Categories and cognitive models. LAUT (Series A, No. 96). Lakoff, G. (1987). Women, fire, and dangerous things. Chicago: The University of Chicago Press. Lakoff, G. (1990). The invariance hypothesis: Is abstract reason based in image-schemas? Cognitive Linguistics, 1(1), 39–74. Lakoff, G. (1996). Moral politics. Chicago: Chicago University Press. Lakoff, G. (2004). Don’t think of an elephant. Know your values and frame the debate. White River Junction, VT: Chelsea Green Publishing. Lakoff, G., & Johnson, M. (1980). Metaphors we live by. Chicago: The University of Chicago Press. Lakoff, G., & Johnson, M. (1999). Philosophy in the flesh: The embodied mind and its challenge to Western thought. New York: Basic Books. Lakoff, G., & Turner, M. (1989). More than cool reason. Chicago: The University of Chicago Press. Lakoff, R. T. (2000). The language war. Berkeley: University of California Press. Langacker, R. (1983). Foundations of cognitive grammar. LAUT (Series A, No. 99).



La Porte, J. (1997). Essential membership. Philosophy of Science, 64(1), 96–112. Laver, J. (2003). Linguistic phonetics. In M. Aronoff & J. Rees-Miller (Eds.), The handbook of linguistics (pp. 150–179). Oxford: Blackwell. Levinson, S. (1983). Pragmatics. Oxford: Oxford University Press. Levy, N. (2002). Moral relativism: A short introduction. Oxford: Oneworld. Lippi-Green, R. (1997). English with an accent: Language, ideology, and discrimination in the United States. London: Routledge. Löffler, H. (1985). Germanistische Soziolinguistik [German language sociolinguistics]. Berlin: Erich Schmidt Verlag. Lundquist, L. (1990). Argumentation—semantics or pragmatics? In L. Lundquist & L. S. Rasmussen (Eds.), Pragmatics and its manifestations in language (pp. 79–102). Copenhagen: Handelshøjskolens Forlag. Lyons, J. (1977). Semantics (Vols. 1 and 2). Cambridge: Cambridge University Press. Lyons, J. (1981). Language and linguistics. Cambridge: Cambridge University Press. MacDonald, J. M., & Michaud, D. L. (1992). The confession: Interrogation and criminal profiles for police officers. Denver, CO: Apache. Maxwell, N. (1984). From knowledge to wisdom: A revolution in the aims and methods of science. Oxford: Basil Blackwell. Medin, D., & Ortony, A. (1989). Psychological essentialism. In S. Vosniadou & A. Ortony (Eds.), Similarity and analogical reasoning (pp. 179–195). Cambridge, England: Cambridge University Press. Mey, J. (1993). Pragmatics: An introduction. Oxford: Blackwell. Miller, M. C. (2001). The Bush dyslexicon. London: Bantam Books. Mills, S. (2003). Caught between sexism, anti-sexism and “political correctness”: Feminist women’s negotiations with naming practices. Discourse and Society, 14(1), 87–110. Milroy, J., & Milroy, L. (1999). Authority in language: Investigating standard English. London: Routledge. Morgan, P. (2001). The semantics of impeachment: Meanings and models in a political conflict. In R. Dirven, R. Frank, & C. Ilie (Eds.), Language and ideology: Vol. II. Descriptive cognitive approaches (pp. 77–105). Amsterdam: John Benjamins. Mühlhäusler, P. (1986). Pidgin and creole linguistics. Oxford: Blackwell. Nagel, E. (1961). The structure of science. Problems in the logic of scientific explanation. London: Routledge & Kegan Paul. Newmark, P. (2002). April–May (41.2). Translation now -17. The Linguist, pp. 62–64. Newsweek. (1998, September 21). The White House rebuttal, pp. 46–47. The New Oxford Dictionary of English. (1998). Oxford: Oxford University Press. Oleksy, W. (1984). Towards pragmatic contrastive analysis. In J. Fisiak (Ed.), Contrastive linguistics: Prospects and problems (pp. 349–364). Berlin: Mouton. Passent, D. (1991). Partia nienawiÑci [A party of hatred]. Polityka, 18, 3. Penrose, R. (with A. Shimony, N. Cartwright, & S. Hawking). (1997). The large, the small, and the human mind. Cambridge, England: Cambridge University Press. Philips, S. (1998). Ideology in the language of judges: How judges practice law, politics, and courtroom control. Oxford: Oxford University Press. Phillipson, R. (2003). Linguistic imperialism. Oxford: Oxford University Press. Popper, K. (1945). The open society and its enemies: Vol. 2. London: Routledge & Kegan Paul. Popper, K. (1957). The poverty of historicism. London: Routledge & Kegan Paul. Popper, K. (1959). The logic of scientific discovery. London: Hutchinson. Popper, K. (1972). Objective knowledge: An evolutionary approach. Oxford: Clarendon. Popper, K. (1976). Unended quest. Glasgow: Fontana. Popper, K. (1982). Quantum theory and the schism in physics. London: Unwin Hyman. Popper, K. (1983). Realism and the aim of science. London: Hutchinson.



Porter, S., & Yuille, J. (1996). The language of deceit: An investigation of the verbal clues to deception in the interrogation context. Law and Human Behavior, 20(4), 443–457. Purl, D. (1993, October 13). Linguistic science. Linguist List, p. 1. Rakowiecki, J. (1991). Biskupi przeciwko referendum [The bishops against the referendum]. Gazeta Wyborcza, 103, 3. Reddy, J. M. (1993). The conduit metaphor: A case of frame conflict in our language about language. In A. Ortony (Ed.), Metaphor and thought (2nd ed., pp. 164–201). Cambridge, England: Cambridge University Press. Rickford, J. R. (1999). African American vernacular English. Oxford: Blackwell. Rogala, K. (1993). Spór o brzuch [Argument over the belly]. Wprost, 4, 1. Romaine, S. (2000). Language in society: An introduction to sociolinguistics (2nd ed.). Oxford: Oxford University Press. Rosch, E. (1977). Classification of real-world objects: Origins and representations in cognition. In P. Johnson-Laird & P. Wason (Eds.), Thinking: Readings in cognitive science (pp. 212–222). Cambridge, England: Cambridge University Press. Rosch, E. (1978). Principles of categorization. In E. Rosch & B. Lloyd (Eds.), Cognition and categorization (pp. 28–46). Hillsdale, NJ: Lawrence Erlbaum Associates. Røyrane, E. (2004, December 5). Synes Flåm, er utrolig nærsynt [I think Flåm is unbelievably myopic]. Bergens Tidende, p. 8. Russell, B. (1996). Vagueness. In R. Keefe & P. Smith (Eds.), Vagueness (pp. 61–68). Cambridge, MA: MIT Press. Sadie, S. (2000, December). Handel. Gramophone, p. 130. Sakiewicz, T. (1995). Wo»anie o obron“ óycia. Najwaóniejsza encyklika Jana Paw»a II [A cry in defense of life: John Paul II’s most important encyclical letter]. Gazeta Polska, 16(92), 1–6 [electronic version]. Sandanger, J. (2004, December 9). Avslørt og avkledd [Disclosed and naked]. Bergens Tidende, p. 37. Shapiro, M. J. (1989). A political approach to language purism. In B. H. Jernudd & M. J. Shapiro (Eds.), The politics of language purism (pp. 21–29). Berlin: Mouton. Sharifian, F. (2003). On cultural conceptualizations. Journal of Cognition and Culture, 3(3), 187–207. Shuy, R. (1993). Language crimes: The use and abuse of language evidence in the courtroom. Oxford: Blackwell. Shuy, R. (1998). The language of confession, interrogation and deception. London: Sage. Silberstein, S. (2002). War of words: Language, politics and 9/11. London: Routledge. Sokal, A., & Bricmont, J. (1997). Intellectual impostures. London: Profile Books. Solan, L. (2002). The Clinton scandal: Some legal lessons from linguistics. In J. Cotterill (Ed.), Language in the legal process (pp. 180–195). Basingstoke, England: Palgrave Macmillan. Stanger, T. (1992, October 19). Crusade for Macedonia. Newsweek, pp. 19, 20. Staniewicz, E. (1990, July 4). Dziecko w drodze [A child on the way]. Gl-os Wielkopolski, pp. 7–8. Stawicka, J. (1990, July 4). Bez wyboru [With no choice]. Gl-os Wielkopolski, pp. 7–8. Steane, J. (2000, December). Bluebird. Gramphone, p. 120 Steffens, K. E. (2004, December 9). Julia Kristeva—ein verdig prisvinnar? [ Julia Kristeva—a worthy prize winner?]. Bergens Tidende, p. 37. Suhr, S., & Johnson, S. (2003). Re-visiting “PC”: Introduction to special issue on “political correctness.” Discourse and Society, 14(1), 5–16. Swann, J., Deumert, A., Lillis, T., & Mesthrie, R. (2004). A dictionary of sociolinguistics. Edinburgh: Edinburgh University Press. Tannen, D. (1986). That’s not what I meant: How conversational style makes or breaks relationships. New York: Ballantine. Tannen, D. (1990). You just don’t understand. New York: William Morrow & Co. Tannen, D. (1995). Talking from 9 to 5. Reading, PA: Virago.



Tannen, D. (1998). The argument culture. New York: Random House. Tarone, E., Cohen, A. D., & Dumas, G. (1983). A closer look at some interlanguage terminology: A framework for communication strategies. In C. Faerch & G. Kasper (Eds.), Strategies in interlanguage communication (pp. 4–13). London: Longman. Taylor, J. (2003). Linguistic categorization (3rd ed.). Oxford: Oxford University Press. Taylor, J., Cuyckens, H., & Dirven, R. (2003). Introduction: New directions in cognitive lexical semantic research. In H. Cuyckens, R. Dirven, & J. R. Taylor (Eds.), Cognitive approaches to lexical semantics (pp. 1–27). Berlin: Mouton de Gruyter. Teller, P. (1975). Essential properties: Some problems and conjectures. The Journal of Philosophy, 72(9), 233–248. Thomas, G. (1991). Linguistic purism. London: Longman. Tiersma, P. M. (1999). Legal language. Chicago: The University of Chicago Press. Trudgill, P. (1996, November). Language and identity. A guest lecture held at the University of Bergen, Norway. Trudgill, P. (2002). Sociolinguistic variation and change. Edinburgh: Edinburgh University Press. Trudgill, P. (2003). A glossary of sociolinguistics. Edinburgh: Edinburgh University Press. Tunstad, E. (2004, December). Kristeva krangel [Kristeva row]. Available at http:// Ulveman, M. (2000, November 21). Hva er et hull? [What is a hole?]. Bergens Tidende, p. 10. Ungerer, F., & Schmid, H. J. (1996). An introduction to cognitive linguistics. London: Longman. Urba½czyk, S. (Ed.). (1987). Sl-owo pie3kne i prawdziwe: Material-y kongresowe (Kongres Kultury Je3zyka Polskiego 7 i 8 grudnia (1984) roku w Zamku Ksiaza 3 & 3 Pomorskich w Szczecinie) [Beautiful and true word. Congress materials (Congress of Polish language culture, December 7–8, 1984, Ksiót Pomorskich castle in Szczecin)]. Warsaw: Pa½stwowe Wydawnictwo Naukowe. van Dijk, T. A. (1997). What is political discourse analysis? In J. Blommaert & C. Bulcaen (Eds.), Political linguistics (pp. 1–52). Amsterdam: John Benjamins. von Mettenheim, C. (1999). The problem of objectivity in law and ethics. In I. Jarvie & S. Pralong (Eds.), Popper’s open society after 50 years: The continuing relevance of Karl Popper (pp. 111–127). London: Routledge. Wardhaugh, R. (1987). Languages in competition. Oxford: Blackwell. Wardhaugh, R. (1997). Introduction to sociolinguistics (3rd ed.). Oxford: Blackwell. The Washington Post. (1998, September 12). Special report. Referral to the United States House of Representatives Pursuant to Title 28, United States Code, 595 (c), pp. A27– A52. Webster’s New World Dictionary of the American Language. (1964). New York: The World Publishing Company. Wierzbicka, A. (1990). “Prototypes save”: On the uses and abuses of the notion of “prototype” in linguistics and related fields. In S. L. Tsohatzidis (Ed.), Meanings and prototypes: Studies in linguistic categorization (pp. 347–367). London: Routledge. Williams, G. (1993, May 14). Shadow plays: Summary of responses. Linguist List, p. 1. Winford, D. (1985). The concept of “diglossia” in Carribbean creole situations. Language in Society, 14, 345–356. Wodak, R. (2001). What CDA is about—a summary of its history, important concepts and its developments. In R. Wodak & M. Meyer (Eds.), Methods of critical discourse analysis (pp. 1–13). London: Sage. Wyld, H. C. (1934). The best English: A claim for the superiority of Received Standard English. Proceedings of the Society of Pure English, 4(Tract 39), 603–621. Yamada, H. (1997). Different games, different rules. Oxford: Oxford University Press. Zerubavel, E. (1993). The fine line. Making distinctions in everyday life. Chicago: The University of Chicago Press. Zuengler, J. (1985). English, Swahili, or other languages? The relationship of educational development goals to language of instruction in Kenya and Tanzania. In N. Wolfson & J. Manes (Eds.), Language of inequality (pp. 241–254). Berlin: Mouton de Gruyter.

Author Index

A Aarts, B., 82 Aitchison, J., 32, 63, 118 Aristotle, 4

B Bargiela-Chiappini, F., 74 Baugh, J., 157 Bavelas, J. B., 69 Bealer, G., 4 Beard, H., 117–118, 165 Bex, T., 153 Black, A., 69 Bråten, E., 146 Bricmont, J., 141, 145, 149 Brown, P., 69 Bryson, L., 69 Bull, P., 69

Channell, J., 53, 136 Chilton, P., 27 Chovil, N., 69 Cienki, A., 166–167 Cloud, J., 94 Cohen, A. D., 26–27 Coleman, L., 148 Cowley, G., 93 Crowley, T., 153 Crystal, D., 142 Cuyckens, H., 51

D de Bono, E., 36 Denison, D., 82 Deumert, A., 177 Dirven, R., 51 Downes, W., 17 Dumas, G., 26–27 Dunant, S., 108

C Cameron, D., 108 Carroll, J. B., 164 Cerf, C., 117–118, 165 Chalmers, A. F., 144

E Edwards, J., 35 Escobar, A., 4




F Fairclough, N., 161 Fasold, R., 15 Ferguson, C., 15 Feyerabend, P., 31–32, 76, 143 Fillmore, C., 157–158 Fine, K., 4 Fishman, J., 14 Flåm, S. D., 142, 144–146, 149 Fromkin, V., 32, 153

G Geeraerts, D., 57, 95 Gibbons, J., 62, 178 Gibbs, N., 92 Gibbs, R., 66 Gray, B., 17 Greimas, A., 15 Gripsrud, J., 146 Guthrie, W. K. C., 50

H Haastrup, K., 19 Hallet, G., 4, 6–7 Hamilton, M., 69 Harris, R., 36, 38–48, 72, 141 Harris, S., 74 Haspelmath, M., 33 Hayakawa, S. I., 50, 78, 85, 87, 124 Heisenberg, W., 54 Hempel, C. G., 41 Honey, J., 153 Hudson, R. A., 18, 63–64, 81

J Jahr, E. H., 22, 152, 155 Janicki, K., 45, 50, 69, 70–71, 152, 170, 175 Jardine, L., 115 Jaworski, A., 170 Johnson, M., 38, 49–50, 59, 66, 108, 114, 130 Johnson, S., 108 Johnstone, B., xi

K Kay, P., 148 Keizer, E., 82 Kemp, L., 68 Kitcher, P., 28, 54, 62, 183 Korzybski, A., 50 Kövesces, Z., 180 Król, J., 109 Krzeszowski, T., 77 Kuhn, T., 143–144, 147

L Labov, W., 17 Lacayo, R., 91 Lakatos, I., 143 Lakoff, G., 38, 49–50, 51, 57, 59, 61, 66, 95, 139, 175 Lakoff, R. T., 18, 90, 139, 157–158 Langacker, R., 59–60 La Porte, J., 4 Laver, J., 142 Levinson, S., 31–32, 69 Levy, N., 130 Lillis, T., 177 Lippi-Green, R., 153 Löffler, H., 16, 23 Lundquist, L., 32 Lyons, J., 40, 121, 142

M MacDonald, J. M., 134 Maxwell, N., 183 Medin, D., 35, 86 Mesthrie, R., 177 Mey, J., 34 Michaud, D. L., 134 Miller, M. C., 139 Mills, S., 109 Milroy, J., 65, 153 Milroy, L., 65, 153 Mineo, P., 69 Morgan, P., 90 Mühlhäusler, P., 20, 23, 37, 124 Mullett, J., 69



N Nagel, E., 143 Newmark, P., 123

O Oleksy, W., 24–25 Ortony, A., 35, 86

P Passent, D., 102, 106 Penrose, R., 55, 62 Philips, S., 138, 139 Phillipson, R., 19, 123 Popova, G., 82 Popper, K., 4–5, 7–8, 14, 28, 54–55, 62, 76, 124, 133, 143, 147, 159, 164, 183 Porter, S., 136 Purl, D., 33

R Rakowiecki, J., 101 Reddy, J. M., 40, 72 Rickford, J. R., 157 Rodman, R., 153 Rogala, K., 100 Romaine, S., 157 Rosch, E., 38, 49, 57, 95 Røyrane, E., 145 Russell, B., 53

S Sadie, S., 68 Sakiewicz, T., 109 Sandanger, J., 146 Schäffner, C., 27 Schmid, H. J., 62 Shapiro, M. J., 157 Sharifian, F., 75 Shuy, R., 18, 62, 131–137 Silberstein, S., 139 Sokal, A., 141, 145, 149 Solan, L., 90 Springen, K., 93 Stanger, T., 120

Staniewicz, E., 99 Stawicka, J., 100 Steane, J., 68 Steffens, K. E., 146 Suhr, S., 108 Swann, J., 177

T Tannen, D., 64, 74, 86–87, 161, 169 Tarone, E., 26–27 Taylor, J., 49, 51, 56–57 Teller, P., 4 Thomas, G., 155 Tiersma, P. M., 62 Trudgill, P., 119, 155, 177 Tunstad, E., 146 Turner, M., 50, 66

U Ulveman, M., 97 Ungerer, F., 62 Urba½czyk, S., 156, 170

V van Dijk, T. A., 16 von Mettenheim, C., 62

W Wardhaugh, R., 14, 20, 122–124 Watts, R. J., 153 Wierzbicka, A., 53 Williams, G., 33 Winford, D., 15 Wodak, R., 17 Wyld, H. C., 153

Y Yamada, H., 74 Yuille, J., 136

Z Zerubavel, E., 82, 179 Zuengler, J., 16, 21

Subject Index

A Abortion, 98–103 Abstraction, see ladder of abstraction Animism, 124 Applied cognitive sociolinguistics, 44, 183

B Binary oppositions, 139, 162, 168–169 Bushisms, 166–167

C Canon, 115 Categorization, 56–58, 82, 118, 128 bipolar, 128 highlighting in, 118 Classification of fields, 29–31 see also discipline boundaries, 31–34, 179 Clinton–Lewinsky case, 89–96 Cloning, 105 Cognitive linguistics, 49–51 Complaints about definitions, see definitions and complaints Concept, 56, 58–59

classical view of, 51, 58 experientially grounded, 59 imprecise (fuzzy), 56 objectivist view of, 51, 58 and ontogenetic process, 59 prototype theory of, 57, 72, 176–177 of science, 143–144 Conduit metaphor, 40 Conflict-ridden language, 111 see also language-related conflict, 85–86 Creativity, 63–65 see also flexibility, 63–65, 82, 139

D Definitions, 3, 7–9, 14, 19, 21, 24, 35, 50, 178 and complaints, 19–20, 177 exactly in, 16–17, 178 in general semantics, 50 in the true sense in, 35, 178 infinite regress of, 7–9, 21 left to right, 8–9 misconceptions about, 3 not yet in, 23, 178 operational, 9 right to left, 8–9, 14 in sciences, 14




Definitions (cont.) strictly speaking in, 35, 178 ultimately correct, 19 utterly precise, 24 working, 9–10, 177–178 Discipline boundaries, 31–34, 179 see also classification of fields, 29–31 Discreteness, 51, 82 see also fuzziness, 51, 53–54, 59, 82

E Ebonics, 157–158 Essentialism, 4–7, 10, 13–14, 20–23, 26–27, 34–36, 77, 125–127, 129, 138, 150, 154–155, 170, 182 and autonomy of university, 150 and chauvinism, 127 conceptual, 23, 84, 105 and conflict, 125–130 as continuum, 182 covert, 6–7, 21, 23, 26–27, 34, 182 criticism of, 7–10 and dogmatism, 125 and ethnocentricism, 127 explicit/overt, 5–6, 182 and judges’ ideologies, 138–139 and language purism, 155 in linguistics, 14–37 covert, 20–37 overt, 14–20 and misunderstanding, 77–78 nontypical, 84 philosophical, 35 and politicians, 170 psychological, 35–36 and standardization, 154 and teachers, 170 and two-valued orientation, 129, 155 typical, 84 and war, 125–129 Essentialist frame of mind, 84–85 Euthanasia, 103–104 Experiential grounding of concepts and misunderstanding, 73–75 Extensive extralinguistic context, 105–106

F Fallacy determinacy, 39–40

etymological, 121, 176 telementational, 39–40 Falsifiability, 143 Fixed code semantics, 41 Flexibility, 63–65, 82, 139 see also creativity, 63–65 Florida vote conflict, 96–98 Fuzziness, 51, 53–54, 59, 82 in biology/mathematics/physics, 54–55 of concepts, 51, 53–59 working definition of, 53–54

G General semantics, 50

I, J Idealization, 80 Imprecision, see fuzziness Imprecision of concepts, see fuzziness of concepts Indiscreteness, see fuzziness Instructions on interrogation, 134–138 Insult, 112–115 prototype of, 114 Integrational linguistics, 38–48 and nonessentialism, 43 Integrationism, 44 Intelligibility, 81 Judges procedure-oriented, 138–139 record-oriented, 138–139

L Ladder of abstraction, 78–80 and linguists, 80 Language and judges’ ideologies, 138–139 Language and social reality, 117 Language as communication, 43–45 Language myth, see Western Language Myth Language-related conflict, 85–86 working definition of, 85



M Meaning, 3, 41, 50, 69, 176 assigned to words, 50, 176 and general semantics, 50 misconceived, 3–4 in words, 41, 50, 176 Metaphor, 66 conventional, 66–67 reaction to, 70–72 source domain in, 66 target domain in, 66 unconventional/open, 66–69 Misconceptions and defendants, 131–132 and definitions, 3 and language in the legal context, 132–134 about meaning, 3–4 Misunderstanding, 72–81, 182 and axiology, 77 and essentialism, 73 and unconventional metaphor, 69–72

N New words, 161–169 and binary oppositions, 168 and continua, 168 and lexicalizations, 165 Nonessentialism as continuum, 182 and judges’ ideologies, 138–139 and new words, 170 nontypical, 84 typical, 84 and word meaning, 170

Philosophy for linguists, 12 Political correctness, 108–119 and jokes, 117–118 Postmodernism and incomprehensible language, 149 Precision, 53–55 of concepts, 54–55 increase of, 53 ultimate, 53 see also fuzziness, 51, 53–54, 59, 82 Problem, 182–183 Prototypes, 57, 63, 72–73, 95, 148

R Rationalism, 144 Relativism, 107, 144

S Segregationism, 44 Segregationist linguistics, 42 Slipping into essentialism, 45, 59–63 Sophists, 50 Standard, 43, 152–157 definition of, 43 in Norway, 154–155 Stem-cell research, 104–105

T, U Taxonomies, 26–29, 180–181 and external research goal, 28, 180–181 Theory-impregnated/laden observation, 28 Two-valued orientation, 85 Understanding as cline, 81

O Open metaphor, see metaphor, unconventional Orthodox linguistics, 42

P Philosophical commitments, 13

W Western Language Myth, 38–40, 42–43, 48 What-is questions, 5–6, 16, 176 philosophically essentialist, 6, 16 stylistic, 6, 16 Word meaning, 11–37 Working definitions and comprehensibility, 178