METAPHORS OF ANGER, PRIDE, AND LOVE

Transcription

1 METAPHORS OF ANGER, PRIDE, AND LOVE

2 Pragmatics & Beyond An Interdisciplinary Series of Language Studies Editors: Herman Parret (Belgian National Science Foundation, Universities of Louvain and Antwerp) Jef Verschueren (Belgian National Science Foundation, University of Antwerp) Editorial Address: Department of Germanic Languages and Literatures University of Antwerp (UIA) Universiteitsplein 1 B-2610 Wilrijk Belgium Editorial Board: Norbert Dittmar (Free University of Berlin) David Holdcroft (University of Leeds) Jacob Mey (Odense University) Jerrold M. Sadock (University of Chicago) Emanuel A. Schegloff (University of California at Los Angeles) Daniel Vanderveken (University of Quebec at Trois-Rivières) Teun A. van Dij (University of Amsterdam) VII:8 Zoltán Kövecses Metaphors of Anger, Pride, and Love: A Lexical Approach to the Structure of Concepts

3 METAPHORS OF ANGER, PRIDE, AND LOVE A LEXICAL APPROACH TO THE STRUCTURE OF CONCEPTS Zoltán Kövecses L. Eötvös University JOHN BENJAMINS PUBLISHING COMPANY AMSTERDAM/PHILADELPHIA 1986

4 Library of Congress Cataloging in Publication Data Kövecses, Zoltán. Metaphors of anger, pride, and love. (Pragmatics & beyond, ISSN ; VII:8) Bibliography: p. I. Metaphor. 2. Semantics. 3. Concepts. 4. Anger. 5. Pride and vanity. 6. Love. I. Title. II. Series. P301.5.M48K Γ ISBN (European) / ISBN (US) (alk. paper) Copyright John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher.

5 CONTENTS Acknowledgements vii 1. GOALS AND METHODS 1 2. THE CONCEPT OF ANGER Some questions Metaphor and metonymy The other principal metaphors Some minor metaphors The prototype scenario Restatement of the prototypical scenario The non-prototypical cases Conclusions THE CONCEPT OF PRIDE Some additional issues Some metonymies for pride Some metaphors Causes of pride Scales, related concepts and the prototype Self-esteem Conceit Vanity Conclusion THE CONCEPT OF ROMANTIC LOVE Some further aspects of a concept The central metaphor The object of love Related concepts 74

6 VI CONTENTS 4.5. Intensity Passivity, lack of control, pleasantness The ideal model Towards the typical model The typical model IMPLICATIONS FOR THEORIES OF CONCEPTUAL STRUCTURE The structure of a concept Metaphorical aspects of concepts IMPLICATIONS FOR THEORIES OF LEXICAL STRUCTURE Polysemy Collocation Semantic fields 136 REFERENCES 145

7 ACKNOWLEDGEMENTS The work presented here was conceived during a stay at the Linguistics Department of the University of California, Berkeley, and was inspired in particular by interactions with Wallace Chafe, Charles Fillmore, Paul Kay, George Lakoff, Karl Zimmer, and others. I am greatly indebted to all of them. Thanks are also due to Ferenc Kiefer and Jef Verschueren for their critical comments on earlier versions of this work. Finally, I would like to express my deepest gratitude to the two persons who have helped me the most and in more than one way: George Lakoff and László T. András. Chapter 2, THE CONCEPT OF ANGER, was written jointly with George Lakoff and is a slightly revised version of the paper that appeared in the Berkeley Cognitive Science Report Series in May, 1983, under the title The cognitive model of anger inherent in American English. A revised version of this paper also appeared in D. Holland andn. Quinn (eds.), Cultural models in language and thought (published by Cambridge University Press) and in George Lakoff, Women, fire, and dangerous things: What categories reveal about the mind (The University of Chicago Press, 1987).

8 1. GOALS AND METHODS This study is an attempt to give an account of the structure of some abstract concepts as these structures are accessible through the study of the language we use to talk about the concepts. The concepts that I shall investigate are 'anger', 'pride' and 'love'. Since lexical meaning is inextricably bound up with concepts, the case studies to be presented (in chapters 2 through 4) will not only have implications for theories of conceptual structure (see chapter 5) but also for theories of lexical structure (see chapter 6). The force of my proposals is considerably weakened by the fact that the language materials on which they are based involve only three concepts. Much more research would be necessary to justify my claims. On the lexical semantic side, let it suffice for the time being to say that my investigations will not allow for a strict separation between sense and reference and, if the distinction can be handled at all, certainly not for a restriction of my approach to sense alone. Since it would be impossible to have deep interviews with every member of a speech community in the search for the structure of particular concepts associated with particular aspects of the world, we have to resort to a more practicable method. Although we cannot interview everyone, we can probe the language used by everyone. Since the goal is to get at our conceptual system, and the folk models within this system, through the lexicon of the English language, we can call such a method a 'lexical approach' (cf. Verschueren 1985). The various folk models corresponding to various areas of experience can in turn be called 'language-based' folk models. This lexical approach to our conceptual system can be regarded as a continuation of some respectable traditions in the study of cognitive systems. One tradition I have in mind is what is called 'linguistic analysis' as represented in the works of Wittgenstein (1963), Austin (1961), Ryle (1949) and others. These philosophers use ordinary language for discovering subtle conceptual distinctions among such philosophically interesting categories as 'action', 'meaning', 'mind', 'emotion', etc. The other tradition comes from ethnography and anthropology. Etnographers and anthropologists who are

9 2 METAPHORS OF ANGER, PRIDE, AND LOVE interested in the cognitive systems of various peoples look at language as an important tool in the job of learning about these cognitive systems. Researchers in this area owe a great deal to the pioneering work of Sapir (1949) and Whorf (1956). One of the most influential recent advocates of this methodology, Charles Frake, sees the role of language in this connection in the following way: "The analysis of a culture's terminological systems will not, of course, exhaustively reveal the cognitive world of its members, but it will certainly tap a central portion of it. Culturally significant cognitive features must be communicable between persons in one of the standard symbolic systems of the culture. A major share of these features will undoubtedly be codable in a society's most flexible and productive communicative device, its language" (in Dil(ed.)1980:3). Similarly, my assumption in this study will be that it is possible to uncover a portion of our conceptual system by studying the way we talk about various aspects of the world. However, there are also some major differences between these traditions and the methodology that I am proposing. One point of divergence concerns the goals that the methods are aimed at. In the philosophical and anthropological approaches mentioned above the goal of researchers is to discover the most important conceptual distinctions that are coded into the language (especially into its vocabulary). That is, these researchers are searching for the main dimensions along which concepts can be shown to differ from each other. By contrast, my main emphasis will be on how a single concept, like 'anger", 'pride' or 'love', is structured internally. In short, while the main aim of the former approaches is to point out conceptual differences between items, my focus of interest is on the conceptual organization of a single item. This does not mean, however, that the lexical approach I am working with does not lead to some interesting 'inter-item' issues. It does, as will become obvious when such issues as 'related concepts', collocations and semantic fields are discussed. But the main emphasis is on 'intra-item' structure. It seems to me that it is this difference in emphasis that is mainly responsible for the, at least for me, impoverished view of meaning as predominantly sense. Concentrating on differences between items results in a skeleton of meaning for an item. On the other hand, I would like to believe that if we focus on the detailed internal structure of an item the conceptual features responsible for the differences will fall out as well and will do so in a natural way. Another difference between the approaches is closely related to the first.

10 GOALS AND METHODS 3 In accordance with their goals as given above the philosophers and anthropologists in question try to find those conceptual dimensions that provide systematic (preferably binary) contrasts that apply to a large portion of the lexicon. The lexical approach as conceived of in the present study does not seek such systematic contrasts. When it is proposed, for example, that the feature 'Self views himself or herself as forming a unity with the beloved', which characterizes love, is a part of the conceptual model of love, it is not expected that this feature has a (binary or any other kind of) counterpart and that it will show up in some other items in the lexicon. The idea of contrast is simply irrelevant here. Nevertheless, as we shall see, these non-systematic features play just as important a role as systematic sense components in the explanation of certain semantic phenomena. Finally, the philosophical and anthropological approaches are deficient in an important respect. This is the fact that they do not seem to recognize the special relevance of metomomy, metaphor and what will be termed 'related concepts' to the make-up of some concepts. I do not wish to claim that these are all relevant to each concept or that they are relevant to the same degree, but I certainly wish to claim that the analysis of a large number of concepts would be deficient without taking at least some of them into account. Since the notion of metaphor as employed here has special significance in this connection, I will say more about it in this chapter. One reflection of our conceptual system is language. Thus the examination of the linguistic expressions that have to do with anger, pride and love promises some success in the study of the conceptual model of anger, pride and love. Since I am primarily concerned with the everyday conception of anger, pride and love, as opposed to scientific and artistic conceptions, I wish to concentrate on what might be viewed as everyday linguistic expressions. The material on which this study is based is composed of those linguistic expressions that are commonly used by and are familiar to most, if not all, native speakers of English, that is, those expressions that do not belong to the sphere of either scientific or artistic discourse. As a result, the expressions we will be looking at will often be well-worn, clichéd or even hackneyed. However, this need not worry us in the least, since our goal is to 'dig up', or make explicit our most everyday conception of anger, pride and love. I will be referring to expressions of this kind as conventionalized (or "standardized", as Frake puts it (Dil, (ed.) 1980: 4)) linguistic expressions. Thus a twofold distinction is intended. First, I would like to draw a distinction between conventionalized language and conventional language. In a sense, most expressions

11 4 METAPHORS OF ANGER, PRIDE, AND LOVE of a language can be regarded as conventional (except, maybe, sound-imitating words). However, not all conventional expressions are also conventionalized, that is, worn-out, clichéd, etc. This distinction will become clearer as we go along. The second distinction is intended to capture the difference between conventionalized expressions on the one hand and creative, novel, unconventional or non-standardized expressions on the other. In the case of love, for example, such expressions could be used by a good poet when he or she writes about love. Let us continue with 'love' as an example. The number of conventionalized expressions about love in many, but perhaps not all languages, is considerable. In English it is roughly about 300. Here are some examples: He was given new strength by love. Love lasts forever. You are the most wonderful person in the world. We are one. He was burning with love. I am crazy about you. She gave him her heart. Don't ever let me go! The magic is gone. She pursued him relentlessly. My claim is that implicit in these and similar linguistic expressions is a conceptual model of love, which we can make explicit if we examine the way in which the expressions are related to each other and examine which aspects of love they capture. At this point at least two objections may be made. One is the doubt of the socialpsychologist who can argue that even if it is possible to discover such a language-based model of love, anger and pride nothing guarantees that people actually conceive of love, anger and pride in terms of such models. Admittedly, he is right. The methodology to be used in this study does not enable us to decide what is the psychological validity of the conceptual models that arise from conventionalized language use. Consequently, it cannot be my purpose to determine the social reality of these models. Secondly, the sociologist and the sociolinguist, who believe in the social stratification of society and language, might object that there are big differences in the social use of some of the above expressions. They could reasonably insist that expressions such as He was burning with love and / am crazy about you belong to very different social layers of language use, and to pretend that there are no

12 GOALS AND METHODS 5 differences beween them in this respect is to neglect important issues. One such issue could be that the conceptual models we bring to light are limited to particular social layers or social occasions. In response to this objection, I would like to say that, although I agree that the expressions do belong to different layers, it is most probable that, at least on the level of passive recognition, every native speaker of English knows these expressions. This would in turn mean that the conceptual models employed in understanding are at least somewhat similar among speakers. But even if this doubt is justified, it seems to me that it is more reasonable to consider our data homogenous at first, rather than worry about social complexity at this stage. Among the linguistic expressions that have to do with love, we can find several metaphors. Let us take some examples from the sentences we have used for illustration: He was burning with love She gave him her heart We are one As we shall see, metaphors play a very significant role in the conceptual model of love. Now I will briefly introduce those ideas and technical terms that we will rely on throughout this study. By metaphor I will mean 'conceptual metaphor'. The term conceptual metaphor comes from the work of Lakoff and Johnson (1980). Lakoff and Johnson make a distinction between a conceptual metaphor and the linguistic expressions that reflect, or make manifest, the conceptual metaphor. One of the expressions on our list of examples was He was given new strength by love. This expression makes us view love as some kind of nutrient, an idea which gains expression in other examples as well. Let us take some of these (in what follows, whenever necessary, concepts, either metaphorical or otherwise, will be written in CAPITAL letters): LOVE IS A NUTRIENT She's starved for affection. I need love. He's love-starved. I cant live without love. He thrives on love. She's sustained by love. He hungered for love. All you need is love.

13 6 METAPHORS OF ANGER, PRIDE, AND LOVE The idea LOVE IS A NUTRIENT, which involves two concepts, will be called a conceptual metaphor. The italicized linguistic expressions are the particular linguistic examples of the conceptual metaphor. To be more precise, we should say that we infer the existence of the LOVE IS A NUTRIENT conceptual metaphor because we know that these conventionalized expressions exist. Although the metaphorical expressions in the sentences I can't live without love and I need love do not directly reflect the NUTRIENT metaphor, I take them as examples of this metaphor on the basis of subcategorization. Subcategorization simply means that NUTRIENT is a concept which is subordinate to the concept of NEED. In such cases, following the practice of Lakoff and Johnson, I will use the more specific concept (which is NUT RIENT in the present case) to represent the appropriate conceptual metaphor. Lakoff and Johnson depart from the traditional view of metaphor according to which metaphor has merely an ornamental function. Instead, they claim that the primary function of metaphor is to understand difficult, complex, abstract, or less clearly delineated concepts. We do this by trying to understand such a concept (LOVE, for example) in terms of another, less complicated, physical, or more clearly delineated concept (for example, NUTRIENT). The concept we try to understand (LOVE) is called the 'target domain', and the concept which is used for this purpose (NUTRIENT) is called the 'source domain'. Thus in Lakoff and Johnson's view, metaphor has as its primary function the cognitive role of understanding one concept in terms of another. Representations such as LOVE IS A NUTRIENT are intended to capture this aspect of metaphor. The NUTRIENT metaphor focuses on only one aspect of love. It makes us see love as a need. Very different aspects of love are captured by the LOVE IS A JOURNEY metaphor. Here emphasis falls on the progress and the purpose of the love relationship, and the difficulties involved. That we indeed try to comprehend these aspects of love in terms of the concept of a JOURNEY is shown by the following linguistic examples taken from Lakoff and Johnson (1980): LOVE IS A JOURNEY Look how far we've come. We're at a crossroads. We'll just have to go our separate ways. We can't turn back now. I don't think this relationship is going anywhere.

14 GOALS AND METHODS 7 Where are we? We're stuck. It's been a long, bumpy road. This relationship is a dead-end street. We're just spinning our wheels. Our marriage is on the rocks. We've gotten off the track. This relationship is foundering. It may be objected that most of these conventionalized expressions are not limited to love relationships. For example, the expression go our separate ways is equally applicable to a friendship, and the expression if's been a long, bumpy road can just as well be used to describe a series of negotiations. Indeed, these expressions are not specific to romantic love. But this only shows that the concept of LOVE has certain aspects that are common to a great number of concepts, and that LOVE is built out of elements and in ways similar to other concepts. But we still have to answer the question of why it is that we try to understand the concept LOVE in terms of JOURNEY, rather than some other concept or concepts. According to Lakoff and Johnson, each conceptual metaphor has what they call an 'experiential basis'. This can be either physical or cultural. Furthermore, the experiential basis of a metaphor provides only some motivation for the metaphor but does not make it completely predictable: In other words, an experiential basis justifies the creation and existence of a conceptual metaphor, but it does not predict exactly what other concept (like JOURNEY) we will use in our effort to comprehend the concept in question (like LOVE). The experiential basis of the LOVE IS A JOURNEY metaphor seems to be that both love and the journey are situations or events that take place in this time. Obviously, it does not follow that the concept LOVE can only be conceptualized as a JOURNEY, since several other events could be used for this purpose. On the other hand, in the case of concepts that have a time dimension the chances are that certain aspects of these concepts will be conceptualized in terms of the JOURNEY metaphor. Such concepts, in addition to LOVE, are LIFE and FRIENDSHIP. It has been mentioned that the JOURNEY metaphor addresses certain aspects of love: Namely, the progress of the relationship, the difficulties involved, and the goal of the relationship. The examples in the JOURNEY metaphor that indicate the aspect of progress are we've come a long way, go our separate ways, etc. The difficulties of love life are expressed by such

15 8 METAPHORS OF ANGER, PRIDE, AND LOVE examples as we're stuck, bumpy road, dead-end street. Before pointing out the expressions that exemplify the aspect of goal in love however, I have to make a short digression. We can make a distinction between 'ontological' and 'epistemic correspondences' that characterize a number of relationships between a source domain and a target domain. Let us first take ontological correspondences. These are correspondences that obtain between entities in the source domain and corresponding entities in the target domain. We find the following ontological correspondences between JOURNEY and LOVE: Source domain: JOURNEY Target domain: LOVE - the journey is love - the travellers are the lovers - the road covered is the progress of the love relationship - the difficulties along the way are the difficulties of the love relationship These are the correspondences that constitute the frame, the basis of the LOVE IS A JOURNEY metaphor. This is why we call them ontological correspondences. Epistemic correspondences are those that obtain between our knowledge of the source domain and our knowledge of the target domain. They arise as a result of a process whereby we carry over some of our knowledge of the source domain onto the target domain. For example, we know that journeys typically have a destination, a goal. This concept of GOAL.appears in the JOURNEY metaphor when we say things like This relationship isn't going anywhere. In other words, the concept of GOAL as it relates to journeys, and which is implicit in this example, creates a GOAL concept in LOVE as well. It is debatable whether love has an inherent goal which is independent of the metaphor. We have seen that the NUTRIENT and JOURNEY metaphors capture various aspects of the concept LOVE. The main issue that the NUTRIENT metaphor addresses is that love is a need, while the JOURNEY metaphor elaborates the aspects of progress, difficulty and goal. That is, different metaphors emphasize different aspects of a concept and at the same time they hide or downplay certain other aspects of the concept. However, as we shall see, overlaps between the aspects highlighted by different metaphors are also possible. At this point it could be asked why this particular theory of metaphor has been singled out for the purposes of this study. Since the vast existing literature on metaphor contains a large number of rival theories, it requires some justification if one gives preference to any one of these. It cannot be my pur-

16 GOALS AND METHODS 9 pose here to survey all of these rival views (for an overview, see Fónagy 1982). It suffices to say that, broadly speaking, two main trends can be distinguished. One trend views metaphor as a linguistic means of describing certain preexisting similarities between two things in the world. This can be called the 'traditional or classical view' of metaphor. The other major trend is the 'interactional view' (Black 1962), which emphasizes that in many cases metaphors are conceptual devices used for understanding or creating reality, rather than merely describing it. The three case studies will provide overwhelming evidence in favor of this latter view. Already in this chapter we have seen some convincing examples. It would be difficult to find any preexisting similarity between LOVE and NUTRIENTS or JOURNEYS. Rather, these latter two concepts are best seen as being used for understanding or creating some aspects of the concept of LOVE. Lakoff and Johnson (1980) work within the interactionalist tradition. So the question now becomes: why have I chosen their theory from among the other interactionalist theories as given, for example, in Sacks (1978) and Ortony (ed.) (1979)? I have done so because it is their theory which, at least to my mind, presents us with the clearest, the most elaborate and comprehensive version in this tradition. They provide a large amount of linguistic and non-linguistic evidence, they set up a detailed and consistent terminological system (a part of which has just been introduced), and link up their work in an explicit way with a large number of fundamental issues in linguistics and philosophy. But one could perhaps simply say that the reason was that it was this approach that proved most useful as a technique in the study of the three concepts under investigation. So far we have surveyed the most important concepts that I will use in the three case studies. These include conceptual models, folk theories, conceptual metaphors, epistemic and ontological correspondences, source and target domains, etc. Mention has also been made of such conceptual tools as metonymies and 'related concepts', which will be discussed in more detail in the appropriate places in the case studies. However, there is a further notion on which the analyses to be presented will draw a great deal. This is the notion of prototype as it is used in the work of Berlin and Kay (1969), Coleman and Kay (1981), Rosch (1973, 1975, 1977), Fillmore (1975) and Lakoff (1973, 1987). I will not try to defend the prototype view of categorization here because I think this has been done successfully by these authors. The usefulness of the prototype approach will be assumed. Nevertheless, I hope that the case studies will also demonstrate that the notion of prototype, augmented by the other conceptual tools, can take us beyond traditional componential

17 10 METAPHORS OF ANGER, PRIDE, AND LOVE analysis in the study of lexical meaning. An account of reference depends on how we can give an account of the knowledge (concept) that we need in order to pick out certain aspects of the world. One of the major reasons for the neglect of the study of reference (as opposed to that of sense) has been that we have had no reliable method with which to investigate the structure of this knowledge (concept). Lyons writes in this connection: "Until we have a satisfactory theory of culture, in the construction of which not only sociology, but also both cognitive and social psychology, have played their part, it is idle to speculate further about the possibility of constructing anything more than a rather ad hoc practical account of the denotation of lexemes". (Lyons 1977: 210) The case studies that follow are intended to be the beginning steps in the construction of just such a theory of culture. Lyons takes the notion of prototype as relevant to this job, but, as we have seen in this section, in addition to this notion several other conceptual tools are needed to accomplish the task.

18 2. THE CONCEPT OF ANGER 2.1. Some questions - Are emotions just amorphous 'feelings' or do they have a cognitive content? - If they have a cognitive content, how can we find out what it is? - When people speak about anger, are they invoking a coherent folk theory? That is, are the conventionalized ways of talking about anger actually based on some cognitive model of what anger is? - Could a mere analysis of the language used to talk about anger actually uncover something real about the way we understand anger? At first glance, the conventional expressions used to talk about anger seem so diverse that finding any coherent system would seem impossible. For example, if we look up anger in, say, Roget's University Thesaurus, we find about three hundred entries, most of which have something or other to do with anger, but the thesaurus doesn't tell us exactly what. Many of these are idioms, and they too seem too diverse to reflect any coherent cognitive model. Here are some example sentences using such idioms: He was foaming at the mouth. You're beginning to get to me. You make my blood boil. He's wrestling with his anger. Watch out! He's on a short fuse. He's just letting off steam. Don't get a hernia! Try to keep a grip on yourself. Don't fly off the handle. When I told him, he blew up. He channeled his anger into something constructive. He was red with anger. He was blue in the face. He appeased his anger.

19 12 METAPHORS OF ANGER, PRIDE, AND LOVE He was doing a slow burn. He suppressed his anger. She kept bugging me. When I told him, he had a cow. What do these expressions have to do with anger, and what do they have to do with each other? One thing is clear: they are not random. There seems to be a systematic relationship among these constructions, but it is not immediately obvious what it is. How do we know, for example, that someone who is foaming at the mouth has lost his cool? How do you know that someone who is looking daggers at you is likely to be doing a slow burn or be on a short fuse? How do we know that someone whose blood is boiling has not appeased his anger? How do we know that someone who has channelled his anger into something constructive has not had a cow? What we will try to show is that there is a coherent conceptual organization underlying all these expressions, and that much of it is metaphorical and metonymical in nature Metaphor and metonymy Let us begin with the folk theory of the physiological effects of anger: THE PHYSIOLOGICAL EFFECTS OF ANGER ARE INCREASED BODY HEAT, INCREASED INTERNAL PRESSURE (BLOOD PRESSURE, MUSCULAR PRESSURE), AGITATION, AND INTERFERENCE WITH ACCURATE PERCEPTION. AS ANGER INCREASES, ITS PHYSIOLOGICAL INCREASE. EFFECTS THERE IS A LIMIT BEYOND WHICH THE PHYSIOLOGICAL EFFECTS OF ANGER IMPAIR NORMAL FUNCTIONING. Given the general metonymic principle, THE PHYSIOLOGICAL EFFECTS OF AN EMOTION STAND FOR THE EMOTION the folk theory given above yields a system of metonymies for anger: BODY HEAT: Don't get hot under the collar. Billy's a hothead. They were having a heated argument.

20 THE CONCEPT OF ANGER 13 When the cop gave her a ticket, she got all hot and bothered and started cursing. INTERNAL PRESSURE: Don't get a hernial When I found out, I almost burst a blood vessel. He almost had a hemorrhage. Increased body heat and/or blood pressure is assumed to cause redness in the face and neck area, and such redness can also metonymically indicate anger. REDNESS IN FACE AND NECK AREA: She was scarlet with rage. He got red with anger. He was flushed with anger. AGITATION: She was shaking with anger. I was hopping mad. He was quivering with rage. He's all worked up. There's no need to get so excited about it! She's all wrought up. You look upset. INTERFERENCE WITH ACCURATE PERCEPTION: She was blind with rage. I was beginning to see red. I was so mad I couldn't see straight. Each of these expressions indicate the presence of anger via its supposed physiological effects. The folk theory of physiological effects, especially the part that emphasizes HEAT, forms the basis of the most general metaphor for anger: ANGER IS HEAT. There are two versions of this metaphor, one where the heat is applied to fluids, the other where it is applied to solids. When it is applied to fluids, we get: ANGER IS THE HEAT OF A FLUID IN A CON TAINER. The specific motivation for this consists of the HEAT, INTER NAL PRESSURE, and AGITATION parts of the folk theory. When ANGER IS HEAT is applied to solids, we get the version ANGER IS FIRE, which is motivated by the HEAT and REDNESS aspects of the folk theory

21 14 METAPHORS OF ANGER, PRIDE, AND LOVE of physiological effects. As we will see shortly, the fluid version is much more highly elaborated. The reason for this, we surmise, is that in our overall conceptual system we have the general metaphor: THE BODY IS A CONTAINER FOR THE EMOTIONS He was filled with anger. She couldn't contain her joy. She was brimming with rage. Try to get your anger out of your system. The ANGER IS HEAT metaphor, when applied to fluids, combines with the metaphor THE BODY IS A CONTAINER FOR THE EMOTIONS to yield the central metaphor of the system: ANGER IS THE HEAT OF A FLUID IN A CONTAINER You make my blood boil. Simmer down! I had reached the boiling point. Let him stew. A historically derived instance of this metaphor is: She was seething with rage. Although most speakers do not now use seethe to indicate physical boiling, the boiling image is still there when seethe is used to indicate anger. Similarly, pissed off is used only to refer to anger, not to the hot liquid under pressure in the bladder. Still, the effectiveness of the expression seems to depend on such an image. When there is no heat the liquid is cool and calm. In the central metaphor, cool and calmness corresponds to lack of anger. Keep cool. Stay calm. As we will see shortly, the central metaphor is an extremely productive one. There are two ways in which a conceptual metaphor can be productive. The first is lexical. The words and fixed expressions of a language can code, that is, be used to express aspects of, a given conceptual metaphor to a greater or lesser extent. The number of conventionalized linguistic expressions that code a given conceptual metaphor is one measure of the productivity of the metaphor. In addition, the words and fixed expressions of a language can elaborate the conceptual metaphor. For example, a stew is a special case in

22 THE CONCEPT OF ANGER 15 which there is a hot fluid in a container. It is something that continues at a given level of heat for a long time. This special case can be used to elaborate the central metaphor. 'Stewing' indicates the continuance of anger over a long period. Another special case is 'simmer', which indicates a low boil. This can be used to indicate a lowering of the intensity of anger. Although both of these are cooking terms, cooking plays no metaphorical role in these cases. It just happens to be a case where there is a hot fluid in a container. This is typical of lexical elaborations. Let us refer to the HEAT OF FLUID IN A CONTAINER as the source domain of the central metaphor, and to ANGER as the target domain. We usually have extensive knowledge about source domains. A second way in which a conceptual metaphor can be productive is that it can carry over details of that knowledge from the source domain to the target domain. We will refer to such carryovers as metaphorical entailments. Such entailments are part of our conceptual system. They constitute elaborations of conceptual metaphors. The central metaphor has a rich system of metaphorical entailments. For example, one thing we know about hot fluids is that, when they start to boil, the fluid goes upward. This gives rise to the entailment: WHEN THE INTENSITY OF ANGER INCREASES, THE FLUID RISES His pent-up anger welled up inside him. She could feel her gorge rising. We got a rise out of him. My anger kept building up inside me. Pretty soon I was in a towering rage. We also know that intense heat produces steam and creates pressure on the container. This yields the metaphorical entailments: INTENSE ANGER PRODUCES STEAM She got all steamed up. Billy's just blowing off steam. I was fuming. INTENSE ANGER PRODUCES PRESSURE ON THE CON TAINER He was bursting with anger. I could barely contain my rage. I could barely keep it in anymore.

23 16 METAPHORS OF ANGER, PRIDE, AND LOVE A variant of this involves keeping the pressure back: I suppressed my anger. He turned his anger inward. He managed to keep his anger bottled up inside him. He was blue in the face. When the pressure on the container becomes too high, the container explodes. This yields the entailment: WHEN ANGER BECOMES TOO INTENSE, THE PERSON EXPLODES When I told him, he just exploded. She blew up at me. We won't tolerate any more of your outbursts. In an explosion, parts of the container go up in the air: WHEN A PERSON EXPLODES, PARTS OF HIM GO UP IN THE AIR I blew my stack. I blew my top. She flipped her lid. I went through the roof. When something explodes, what was inside it comes out: WHEN A PERSON EXPLODES, WHAT WAS INSIDE HIM COMES OUT His anger finally came out. Smoke was pouring out of his ears. This can be elaborated in terms of animals giving birth, where something that was inside causing pressure bursts out: She was having kittens. My mother will have a cow when I tell her. Let us now turn to the questions of what issues the central metaphor addresses and what kind of ontology of anger it reveals. The central metaphor focuses on the fact that anger can be intense, that it can lead to a loss of control, and that a loss of control can be dangerous. Let us begin with intensity. Anger is conceptualized as a mass, and takes the grammar of mass nouns, as opposed to count nouns. Thus you can say: How much anger has he got in him?

24 THE CONCEPT OF ANGER 17 but not: *How many angers does he have in him? Anger thus has the ontology of a mass entity, that is, it has a scale indicating its amount, it exists when the amount is greater than zero and goes out of existence when the amount falls to zero. In the central metaphor, the scale indicating the amount of anger is the heat scale. But, as the central metaphor indicates, the anger scale is not open-ended; it has a limit. Just as a hot fluid in a closed container can only take so much heat before it explodes, so we conceptualize the anger scale as having a limit point. We can only bear so much anger before we explode, that is, lose control. This has its correlates in our folk theory of physiological effects. As anger gets more intense the physiological effects increase and those increases interfere with our normal functioning. Body heat, blood pressure, agitation and interference with perception cannot increase without limit before our ability to function normally becomes seriously impaired, and we lose control over our functioning. In the folk model of anger, loss of control is dangerous, both to the angry person and to those around him. In the central metaphor, the danger of loss of control is understood as the danger of explosion. The structural aspect of a conceptual metaphor consists of a set of correspondences between a source domain and a target domain. These correspondences can be factored into two types: ontological and epistemic. Ontological correspondences are correspondences between the entities in the source domain and the corresponding entities in he target domain. For example, the container in the source domain corresponds to the body in the target domain. Epistemic correspondences are correspondences between knowledge about the source domain and corresponding knowledge about the target domain. We can schematize these correspondences between the FLUID domain and the ANGER domain as follows: Source: HEAT OF FLUID IN CONTAINER Target: ANGER Ontological Correspondences: - The container is the body. - The heat of fluid is the anger. - The heat scale is the anger scale, with end points zero and limit. - Container heat is body heat. - Pressure in container is internal pressure in the body. - Agitation of fluid and container is physical agitation. - The limit of the container's capacity to withstand pressure caused by heat is

25 18 METAPHORS OF ANGER, PRIDE, AND LOVE the limit on the anger scale. - Explosion is loss of control. - Danger of explosion is danger of loss of control. - Coolness in the fluid is lack of anger. - Calmness of the fluid is lack of agitation. Epistemic correspondences: Source: The effect of intense fluid heat is container heat, internal pressure, and agitation. Target: The effect of intense anger is body heat, internal pressure, and agitation. Source: When the fluid is heated past a certain limit, pressure increases to the point at which the container explodes. Target: When anger increases past a certain limit, pressure increases to the point at which the person loses control. Source: An explosion is damaging to the container and dangerous to bystanders. Target: A loss of control is damaging to an angry person and dangerous to other people. Source: An explosion may be prevented by the application of sufficient force and energy to keep the fluid in. Target: A loss of control may be prevented by the application of sufficient force and energy to keep the anger in. Source: It is sometimes possible to control the release of heated fluid for either destructive or constructive purposes; this has the effect of lowering the level of heat and pressure. Target: It is sometimes possible to control the release of anger for either destructive or constructive purposes; this has the effect of lowering the level of anger and internal pressure. The latter case defines an elaboration of the entailment WHEN A PERSON EXPLODES, WHAT WAS INSIDE HIM COMES OUT: ANGER CAN BE LET OUT UNDER CONTROL He let out his anger. I gave vent to my anger. Channel your anger into something constructive. He took out his anger on me.

26 THE CONCEPT OF ANGER 19 So far, we have seen that the folk theory of physiological reactions provides the basis for the central metaphor, and that the central metaphor characterizes detailed correspondences between the source domain and the target domain correspondences concerning both ontology and knowledge. At this point, our analysis enables us to see why various relationships among idioms hold. We can see why someone who is in a towering rage has not kept cool, why someone who is stewing may have contained his anger but has not got it out of his system, why someone who has suppressed his anger has not yet erupted, and why someone who has channeled his anger into something constructive has not had a cow. Let us now turn to the case where the general ANGER IS HEAT metaphor is applied to solids: ANGER IS FIRE Those are inflammatory remarks. She was doing a slow burn. What you said inflamed him. He was breathing fire. Your insincere apology just added fuel to the fire. After the argument, Dave was smoldering for days. That kindled my ire. Boy, am I burned up! He was consumed by his anger. This metaphor highlights the cause of anger (kindle, inflame), the intensity and duration {smoldering, slow burn, burned up), the danger to others {breathing fire), and the damage to the angry person (consumed). The correspondences in ontology are as follows: Source: FIRE Target: ANGER - The fire is anger. - The thing burning is the angry person. - The cause of the fire is the cause of the anger. - The intensity of the fire is the intensity of the anger. - The physical damage to the thing burning is mental damage to the angr * person. - The capacity of the thing burning to serve its normal function is the capacity of the angry person to function normally. - An object at the point of being consumed by fire corresponds to a person whose anger is at the limit.

27 20 METAPHORS OF ANGER, PRIDE, AND LOVE - The danger of the fire to things nearby is danger of the anger to other people. The corrrespondences in knowledge are: Source: Things can burn at low intensity for a long time and then burst into flame. Target: People can be angry at a low intensity for a long time and then suddenly become extremely angry. Source: Fires are dangerous to things nearby. Target: Angry people are dangerous to other people. Source: Things consumed by fire cannot serve their normal function. Target: At the limit of the anger scale, people cannot function normally. Putting together what we've done so fat, we can see why someone who is doing a slow burn hasn't hit the ceiling yet, why someone whose anger is bottled up is not breathing fire, why someone who is consumed by anger probably can't see straight, and why adding fuel to the fire might just cause the person you're talking to to have kittens The other principal metaphors As we have seen, the ANGER IS HEAT metaphor is based on the folk theory of the physiological effects of anger, according to which increased body heat is a major effect of anger. That folk theory also maintains that agitation is an important effect. Agitation is also an important part of our folk model of insanity. According to this view, people who are insane are unduly agitated they go wild, start raving, flail their arms, foam at the mouth, etc. Correspondingly, these physiological effects can stand, metonymically, for insanity. One can indicate that someone is insane by describing him as foaming at the mouth, raving, going wild, etc. The overlap between the folk theories of the effects of anger and the effects of insanity provides a basis for the metaphor: ANGER IS INSANITY I just touched him, and he went crazy. You're driving me nuts! When the umpire called him out on strikes, he went bananas. One more complaint and I'll go berserk. He got so angry, he went out of his mind. When he gets angry, he goes bonkers.

28 THE CONCEPT OF ANGER 21 She went into an insane rage. If anything else goes wrong, I'll get hysterical. Perhaps the most common conventional expression for anger came into English historically as a result of this metaphor: I'm mad\ Because of this metaphorical link between insanity and anger, expressions that indicate insane behavior can also indicate angry behavior. Given the metonymy INSANE BEHAVIOR STANDS FOR INSANITY and the metaphor ANGER IS INSANITY, we get the metaphorical metonymy: INSANE BEHAVIOR STANDS FOR ANGER When my mother finds out, she'll have a fit. When the ump threw him out of the game, Billy started foaming at the mouth. He's fit to be tied. He's about to throw a tantrum. Violent behavior indicative of frustration is viewed as a form of insane behavior. According to our folk model of anger, people who can neither control nor relieve the pressure of anger engage in violent frustrated behavior. This folk model is the basis for the metonymy: VIOLENT FRUSTRATED BEHAVIOR STANDS FOR ANGER He's tearing his hair out\ If one more thing goes wrong, I'll start banging my head against the wall. The loud music next door has got him climbing the walls! She's been slamming doors all morning. The ANGER IS INSANITY metaphor has the following correspondences: Source:INSANITY Target: ANGER - The cause of insanity is the cause of anger. - Becoming insane is passing the limit point on the anger scale. - Insane behavior is angry behavior. Source: An insane person cannot function normally. Target: A person who is angry beyond the limit point cannot function normally. Source: An insane person is dangerous to others. Target: A person who is angry beyond the limit point is dangerous to others.

29 22 METAPHORS OF ANGER, PRIDE, AND LOVE At this point, we can see a generalization. Emotional effects are understood as physical effects. Anger is understood as a form of energy. According to our folk understanding of physics, when enough input energy is applied to a body, the body begins to produce output energy. Thus, the cause of anger is viewed as input energy that produces internal heat (output energy). Moreover, the internal heat can function as input energy, producing various forms of output energy: steam, pressure, externally radiating heat, and agitation. Such output energy (the angry behavior) is viewed as dangerous to others. In the insanity metaphor, insanity is understood as a highly energized state, with insane behavior as a form of energy output. All in all, anger is understood in our folk model as a negative emotion. It produces undesirable physiological reactions, leads to an inability to function normally, and is dangerous to others. The angry person, recognizing this danger, views his anger as an opponent. ANGER IS AN OPPONENT (IN A STRUGGLE) I'm struggling with my anger. He was battling his anger. She fought back her anger. You need to subdue your anger. I've been wrestling with my anger all day. I was seized by anger. I'm finally coming to grips with my anger. He lost control over his anger. Anger took control of him. He surrendered to his anger. He yielded to his anger. I was overcome by anger. Her anger has been appeased. The ANGER IS AN OPPONENT metaphor is constituted by the following correspondences : Source : STRUGGLE Target : ANGER - The opponent is anger. - Winning is controlling anger. - Losing is having anger control you. - Surrender is allowing anger to take control of you. - The pool of resources needed for winning is the energy needed to control anger.

30 THE CONCEPT OF ANGER 23 One thing that is left out of this account so far is what constitutes 'appeasement'. To appease an opponent is to give in to his demands. This suggests that anger has demands. We will address the question of what these demands are below. The OPPONENT metaphor focuses on the issue of control and the danger of loss of control to the angry person himself. There is another metaphor that focuses on the issue of control, but whose main focus is the danger to others. It is a very widespread metaphor in Western culture, namely, PASSIONS ARE BEASTS INSIDE A PERSON. According to this metaphor, there is a part of each person that is a wild animal. Civilized people are supposed to keep that part of them private, that is, they are supposed to keep the animal inside them. In the metaphor, loss of control is equivalent to the animal getting loose. And the behavior of a person who has lost control is the behavior of a wild animal. There are versions of this metaphor for the various passions desire, anger, etc. In the case of anger, the beast presents a danger to other people. (For some more discussion of the issue, see section 8 in chapter 4.) ANGER IS A DANGEROUS ANIMAL He has a ferocious temper. He has a fierce temper. It's dangerous to arouse his anger. That awakened my ire. His anger grew. He unleashed his anger. Don't let your anger get out of hand. He lost his grip on his anger. His anger is insatiable. An example that draws on both the FIRE and DANGEROUS ANIMAL metaphors is: He was breathing fire. The image here is of a dragon, a dangerous animal that can devour you with fire. The DANGEROUS ANIMAL metaphor portrays anger as a sleeping animal that it is dangerous to awaken; as something that can grow and thereby become dangerous; as something that has to be held back; and as something with a dangerous appetite. Here are the correspondences that con-

31 24 METAPHORS OF ANGER, PRIDE, AND LOVE stitute the metaphor. Source: DANGEROUS ANIMAL Target: ANGER - The dangerous animal is the anger. - The animal's getting loose is loss of control of anger. - The owner of the dangerous animal is the angry person. - Sleeping for the animal is anger near the zero level. - Being awake for the animal is near the limit. Source: It is dangerous for a dangerous animal to be loose. Target: It is dangerous for a person's anger to be out of control. Source: A dangerous animal is safe when it is sleeping and dangerous when it is awake. Target: Anger is safe near the zero level and dangerous near the limit. Source: A dangerous animal is safe when it is very small and dangerous when it is grown. Target: Anger is safe near the zero level and dangerous near the limit. Source: It is the responsibility of a dangerous animal's owner to keep it under control. Target: It is the responsibility of an angry person to keep his anger under control. Source: It requires a lot of energy to control a dangerous animal. Target: It requires a lot of energy to control one's anger. There is another class of expressions that, as far as we can tell, are instances of the same metaphor. These are cases in which angry behavior is described in terms of aggressive animal behavior. ANGRY BEHAVIOR IS AGGRESSIVE ANIMAL BEHAVIOR He was bristling with anger. That got my hackles up. He began to bare his teeth. That ruffled her feathers. She was bridling with anger. Don't snap at me! I was growling with rage. He started snarling. Don't bite my head off! Why'd you jump down my throat?

32 THE CONCEPT OF ANGER 25 Perhaps the best way to account for these cases would be to extend the ontological correspondences of the ANGER IS A DANGEROUS ANI MAL metaphor to include: - The aggressive behavior of the dangerous animal is angry behavior. If we do this, we can account naturally for the fact that these expressions indicate anger. They would do so via a combination of metaphor and metonymy, in which the aggressive behavior metaphorically corresponds to angry behavior, which in turn metonymically stands for anger. For example, the snarling of the animal corresponds to the angry verbal behavior of the person, which in turn indicates the presence of anger. Aggressive verbal behavior is a common form of angry behavior, as snap, growl, snarl, etc. indicate. We can see this in a number of cases outside of the animal domain: AGGRESSIVE VERBAL BEHAVIOR STANDS FOR ANGER She gave him a tongue-lashing. I really chewed him out good! Other forms of aggressive behavior can also stand metonymically for anger, especially aggressive visual behavior: AGGRESSIVE VISUAL BEHAVIOR STANDS FOR ANGER She was looking daggers at me. He gave me a dirty look. If looks could kill,... He was glowering at me. All these metonymic expressions can be used to indicate anger. As in the case of the OPPONENT metaphor, our analysis of the DANGEROUS ANIMAL metaphor leaves an expression unaccounted for 'insatiable'. This expression indicates that the animal has an appetite. This 'appetite' seems to correspond to the 'demands' in the OPPONENT metaphor, as can be seen from the fact that the following sentences entail each other: Harry's anger is insatiable. Harry's anger cannot be appeased. To see what it is that anger demands and has an appetite for, let us turn to expressions that indicate causes of anger. Perhaps the most common group of expressions that indicate causes of anger consists of conventionalized forms of

33 26 METAPHORS OF ANGER, PRIDE, AND LOVE annoyance: insects, minor pains, burdens placed on domestic animals, etc. Thus we have the metaphor: THE CAUSE OF ANGER IS A PHYSICAL ANNOYANCE Stop bugging me! Don't be a pain in the ass. Get off my back! You don't have to ride me so hard. You're getting under my skin. He's a pain in the neck. Don't be a pest! These forms of annoyance involve an offender and a victim. The offender is at fault. The victim, who is innocent, is the one who gets angry. There is another set of conventionalized expressions used to speak of, or to, people who are in the process of making someone angry. These are expressions of territoriality, in which the cause of anger is viewed as a trespasser. CAUSING ANGER IS TRESPASSING You're beginning to get to me. Get out of here! Get out of my sight! Leave me alone! This is where I draw the line! Don't step on my toes! Again, there is an offender (the cause of anger) and a victim (the person who is getting angry). In general, the cause of anger seems to be an offense, in which there is an offender who is at fault and an innocent victim, who is the person who gets angry. The offense seems to constitute some sort of injustice. This is reflected in the conventional wisdom: Don't get mad, get even! In order for this saying to make sense, there has to be some connection between anger and retribution. Getting even is a form of balancing the scales of justice. The saying assumes a model in which injustice leads to anger and retribution can alleviate or prevent anger. In short, what anger 'demands' and has an 'appetite' for is revenge. This is why warnings and threats can count as angry behavior:

34 THE CONCEPT OF ANGER 27 If I get mad, watch out! Don't get me angry, or you'll be sorry. The angry behavior is, in itself, viewed as a form of retribution. We are now in a position to make sense of another metaphor for anger: ANGER IS A BURDEN Unburdening himself of his anger gave him a sense of relief. After I let out my anger, I felt a sense of release. After I lost my temper, I felt lighter. He carries his anger around with him. He has a chip on his shoulder. You'll feel better if you get it off your chest. In English, it is common for responsibilities to be metaphorized as burdens. There are two kinds of responsibilities involved in the folk model of anger that has emerged so far. The first is a responsibility to control one's anger. In cases of extreme anger, this may place a considerable burden on one's 'inner resources'. The second comes from the model of retributive justice that is built into our concept of anger; it is the responsibility to seek vengeance. What is particularly interesting is that these two responsibilities are in conflict in the case of angry retribution: If you take out your anger on someone, you are not meeting your responsibility to control your anger, and if you don't take out your anger on someone, you are not meeting your responsibility to provide retribution. The slogan 'Don't get mad, get even!' offers one way out: retribution without anger. The human potential movement provides another way out by suggesting that letting your anger out is okay. But the fact is that neither of these solutions is the cultural norm. It should also be mentioned in passing that the human potential movement's way of dealing with anger by sanctioning its release is not all that revolutionary. It assumes almost all of our standard folk model and metaphorical understanding, and makes one change: sanctioning the 'release' Some minor metaphors There are a few very general metaphors that apply to anger as well as to many other things, and are commonly used in comprehending and speaking about anger. The first we will discuss has to do with existence. Existence is commonly understood in terms of physical presence. You are typically aware of something's presence if it is nearby and you can see it. This is the basis for the metaphor:

35 28 METAPHORS OF ANGER, PRIDE, AND LOVE EXISTENCE IS PRESENCE His anger went away. His anger eventually came back. My anger lingered on for days. She couldn't get rid of her anger. After a while, her anger just vanished. My anger slowly began to dissipate. When he saw her smile, his anger disappeared. In the case of emotions, existence is often conceived of as location in a bounded space. Here the emotion is the bounded space and it exists when the person is in that space: EMOTIONS ARE BOUNDED SPACES She flew into a rage. She was in an angry mood. He was in a state of anger. I am not easily roused to anger. These cases are relatively independent of the rest of the anger system, and are included here more for completeness than for profundity The prototype scenario The metaphors and metonymies that we have investigated so far converge on a certain prototypical cognitive model of anger. It is not the only model of anger we have; in fact, there are quite a few. But as we shall see, all of the others can be characterized as minimal variants of the model that the metaphors converge on. The model has a temporal dimension, and can be conceived of as a scenario with a number of stages. We will call this the 'prototype scenario'; it is similar to what De Sousa (1980) calls the 'paradigm scenario'. We will be referring to the person who gets angry as S, short for the Self. Stage 1: Offending Event There is an offending event that displeases S. There is a wrongdoer who intentionally does something directly to S. The wrongdoer is at fault and S is innocent. The offending event constitutes an injustice and produces anger in S. The scales of justice can only be balanced by some act of retribution. That is, the intensity of retribution must be roughly equal to the intensity of offense. S has the responsibility to perform such an act of retribution.

36 THE CONCEPT OF ANGER 29 Stage 2: Anger Associated with the entity anger is a scale that measures its intensity. As the intensity of anger increases, S experiences physiological effects: increase in body heat, internal pressure, and physical agitation. As the anger gets very intense, it exerts a force upon S to perform an act of retribution. Because acts of retribution are dangerous and/or socially unacceptable, S has a responsibility to control his anger. Moreover, loss of control is damaging to S's own wellbeing, which is another motivation for controlling anger. Stage 3: Attempt at Control S attempts to control his anger. Stage 4: Loss of control. Each person has a certain tolerance for controlling anger. That tolerance can be viewed as the limit point on the anger scale. When the intensity of anger goes beyond that limit, S can no longer control his anger. S exhibits angry behavior and his anger forces him to attempt an act of retribution. Since S is out of control and acting under coercion, he is not responsible for his actions. Stage 5: Act of Retribution S performs the act of retribution. The wrongdoer is the target of the act. The intensity of retribution roughly equals the intensity of the offense and the scales are balanced again. The intensity of anger drops to zero. The course of anger depicted in the prototype scenario is by no means the only course anger can take. In claiming that the scenario is prototypical we are claiming that according to our cultural folk theory of anger, this is a normal course for anger to take. Deviations of many kinds are both recognized as existing and recognized as being noteworthy and not the norm. Let us take some examples: - Someone who turns the other cheek, that is, who does not get angry or seek retribution. In this culture, such a person is considered virtually saintly. - Someone who has no difficulty controlling his anger is especially praiseworthy. - A hothead is someone who considers more events offensive than most people, who has a lower threshold for anger than the norm, who cannot control his anger, and whose acts of retribution are considered out of proportion to the offense. Someone who is extremely hotheaded is considered emotionally 'unbalanced'.

37 30 METAPHORS OF ANGER, PRIDE, AND LOVE On the other hand, someone who acts in the manner described in the prototypical scenario would not be considered abnormal at all. Before turning to the nonprototypical cases, it will be useful for us to make a rough sketch of the ontology of anger: the entities, predicates and events required. This will serve two purposes. First, it will allow us to show in detail how the nonprototypical cases are related to the prototypical model. Second, it will allow us to investigate the nature of this ontology. We will include only the detail required for our purposes. It is part of our folk concept of a person that he can temporally lose control of his body or his emotions. Implicit in this concept is a separation of the body and the emotions from the Self. This separation is especially important in the ontology of anger. Anger, as a separable entity, can overcome someone, take control, and cause him to act in ways he would not normally act. In such cases, the Self is no longer in control of the body (B). A fuller treatment would probably also require viewing the mind as a separate entity, but that is beyond our present purposes. Since anger has a quantitative aspect, the ontology must include a scale of anger, including an intensity (I(A)), a zero point (Z) and a limit point (L). The basic anger scenario also includes an offending event (0) and a retributive act (R). Each of these has a quantitative aspect, and must also include an intensity, a zero point and a limit. In the prototypical case, the offending event is an action on the part of a wrongdoer (W) against a victim (V). The retribution takes the form of an act by an agent (A) against some target (T). The ontology of anger also includes a number of predicates: displeasing (D), at fault (AF), exert force on (F), cause (C), exist (E) control (CL), dangerous (DR), damaging (DG), balance (B), and outweigh (OW). There are also some other kinds of events: the physiological events (PE); the angry behaviors (AB); and the immediate cause of anger (IC), in case it is not the same as the offending event. Summary of the ontology of anger Aspects of the person: Self: S Body: Β Anger: A Offense and retribution: Offending event: 0 Retributive act: R

38 THE CONCEPT OF ANGER 31 Scales of intensity: Intensity of Anger: I(A) Intensity of Offense: I(0) Intensity of Retribution: I(R) End points: Zero: Ζ Limit: L Predicates: Displease: D At Fault: AF Cause: C Exist: Ε Exert force on: F Control: CL Dangerous: DR Damaging: DG Balance: Β Outweigh: OW Other events: Physiological Reactions: PE Angry Behaviors: AB Immediate cause: IC 2.6. Restatement of the prototypical scenario Given the above ontology and principles of the folk model, we can restate the prototypical anger scenario in terms that will facilitate showing the relationships among the wide variety of anger scenarios. We will first restate the prototypical scenario and then go on to the nonprototypical scenarios. Prototypical anger scenario: Constraints: V = S: Victim = Self A = S: Agent of Retribution = Self Τ = W: Target of Anger = Wrongdoer IC = 0: Immediate cause of Anger = Offending event AB = R: Angry behavior = Retribution

39 32 METAPHORS OF ANGER, PRIDE, AND LOVE Stage 1: Offending Event 0(W,S): Wrongdoer offends Self AF(W): Wrongdoer is at fault D(0,S): The offending event displeases Self OW(I(0),I(R)): The intensity of the offense outweighs the intensity of the retribution (which equals zero at this point), thus creating an imbalance. C(0,E(A)): The offense causes anger to come into existence. Stage 2: Anger E(A): Anger exists. PE(S): S experiences physiological effects (heat, pressure, agitation). F(A,S) SO THAT ATTEMPT(S,R): Anger exerts force on the Self to attempt an act of retribution. Stage 3: Attempt to control anger F(S,A): S exerts a counterforce in an attempt to control anger. Stage 4: Loss of control I(A)>L: The intensity of anger goes above the limit. CL(A,S): Anger takes control of S. AB(S): S exhibits angry behavior (loss of judgment, aggressive actions). DG(S): There is damage to S. DR(W): There is a danger to the target of anger, in this case, the wrongdoer. Stage 5: Retribution R(S,W): S performs retributive act against W (this is usually angry behavior directed at W). B(I(R),I(0)): The intensity of retribution balances the intensity of offense. I(A) = Z: The intensity of anger drops to zero. NOT(E(A)): Anger ceases to exist The non-prototypical cases We are now in a position to show how a large range of instances of anger cluster about the above prototype. The examples are in the following form: a nonprototypical anger scenario with its name, followed by an informal description; an account of the minimal difference between the given scenario and the prototype scenario, first in English, then in approximate formal notation; finallly, an example sentence. Insatiable Anger: You perform the act of retribution and the anger just doesn't go away. In stage 5, the intensity of anger stays above zero and the anger continues to

40 THE CONCEPT OF ANGER 33 exist. Stage 5:I(A)>Z and E(A). Example: His anger lingered on. Frustrated Anger. You just can't get back at the wrongdoer and you get frustrated. It is not possible to gain retribution for the offensive act. S engages in frustrated behavior. Option: S directs his anger at himself. Stage 5: NOT POSSIBLE(R(S,W)). AB(S): ACTS OF FRUSTRATION. Option: T=S, R(S,S). Examples: He was climbing the walls. She was tearing her hair out. He was banging his head against the wall. He's taking it out on himself. Redirected Anger. Instead of directing your anger at the person who made you angry, you direct it at someone or something else. The target of anger is not the wrongdoer. Stage 5: NOT (T = W). Examples: When I lose my temper, I kick the cat. When you get angry, punch a pillow until your anger goes away. When something bad happened at the office, he would take it out on his wife. Exaggerated Response: Your reaction is way out of proportion to the offense. The intensity of retribution outweighs the intensity of offense. Stage 5:OW(I(R),I(0)). Examples: Why jump down my throat? You have a right to get angry, but not to go that far. Controlled Response: You get angry, but retain control and consciously direct your anger at the wrongdoer. S remains in control. Everything else remains the same. Stage 4:CL(S,A). Example: He vented his anger on her. Constructive use: Instead of attempting an act of retribution, you put your anger to a constructive use. S remains in control and performs a constructive act instead of a retributive act. The scales remain unbalanced, but the anger disappears. Stage 4: CL(S,A). Stage 5: CONSTRUCTIVE ACT in place of R(S,T). OW(I(0),I(R)). Example: Try to channel your anger into something constructive.

41 34 METAPHORS OF ANGER, PRIDE, AND LOVE Terminating event: Before you have a chance to lose control, some unrelated event happens to make your anger disappear. Anger doesn't take control of S. Some event causes the anger to go out of existence. Stage 4: NOT(CL(A,S)). There is an event such that NOT(e = R) and C(e,(NOT(E(A))). Example: When his daughter smiled at him, his anger disappeared. Spontaneous cessation: Before you lose control, your anger just goes away. Anger doesn't take control of S and the intensity of anger goes to zero. Stage 4: NOT(CL(A,S)) and I(A) = Z. Example: His anger just went away by itself. Successful suppression: You successfully suppress your anger. S keeps control and the intensity of anger is not near the limit. Stage 4: CL(S,A) and I(A) IS NOT NEAR L. Example: He suppressed his anger. Controlled reduction: Before you lose control, you engage in angry behavior and the intensity of anger goes down. S does not lose control, S engages in angry behavior and the intensity of anger goes down. Stage 4: NOT (CL(A,S)) and AB(S) and I(A) GOES DOWN. Example: He's just letting off steam. Immediate explosion: You get angry and lose control all at once. No Stage 3. Stages 2 and 4 combine into a single event. Example: I said "Hi Roundeyes!" and he blew up. Slow burn: Anger continues for a long time. Stage 2 lasts a long time. Example: He was doing a slow burn. Nursing a grudge: S maintains his anger for a long period waiting for a chance at a retributive act. Maintaining that level of anger takes special effort. Stage 2 lasts a long time and requires effort. The retributive act does not equal angry behavior. Don't get mad, get even: This is advice (rarely followed) about the pointlessness of getting angry. It suggests avoiding stages 2,3, and 4, and instead going directly to stage 5. This advice is defined as an alternative to the prototypical scenario.

42 THE CONCEPT OF ANGER 35 Indirect Cause: It is some result of the wrongdoer's action, not the action itself, that causes anger. The offense is not the immediate cause of anger, but rather the cause of the immediate cause. Stage 1: NOT (O = IC) and CAUSE(0,IC). Example: Your secretary forgets to fill out a form that results in your not getting a deserved promotion. O = secretary forgets to fill out form. IC = you don't get promotion. You are angry about not getting the promotion. You are angry at the secretary for not filling out the form. In general, about marks the immediate cause, at marks the target, and for marks the offense. Cool anger: There are no physiological effects and S remains in control. Anger with: To be angry with someone, S has to have a positive relationship with the wrongdoer W, W must be answerable to S, the intensity is above the threshhold but not near the limit. Perhaps the best example is a parent-child relationship, where the parent is angry with the child. Righteous indignation: O is a moral offense and the victim is not S. The intensity of anger is not near the limit. Wrath: The intensity of the offense is very great and many acts of retribution are required in order to create a balance. The intensity of the anger is well above the limit and the anger lasts a long time. There appears to be a recognizable form of anger for which there are no conventional linguistic expressions, so far as we can tell. We will call this a manipulative use of anger. It is a case where a person cultivates his anger and does not attempt to control it, with the effect that he intimidates those around him into following his wishes in order to keep him from getting angry. This can work either by fear or by guilt. The people manipulated can either be afraid of his anger or may feel guilty about what anger does to him. This form of anger is fairly distant from the prototype and it is no surprise that we have no name for it. Interestingly enough, there is a linguistic test that can be used to verify that what we have called the prototypical scenario is indeed prototypical. It involves the use of the word but. Consider the following examples (where the asterisk indicates a semantic aberration): - Max got angry, but he didn't blow his top. - *Max got angry, but he blew his top.

43 36 METAPHORS OF ANGER, PRIDE, AND LOVE - Max blew up at his boss, but the anger didn't go away. - *Max blew up at his boss, but the anger went away. - Sam got me angry, but it wasn't him that I took my anger out on. - *Sam got me angry, but it was him that I took my anger out on. The word but marks a situation counter to expectation. In these examples, the prototypical scenario defines what is to be expected. The acceptable sentences with but run counter to the prototypical scenario, and thus fit the conditions for the use of but. The unacceptable sentences fit the prototypical scenario, and define expected situations. This is incompatible with the use of but. Thus we have a linguistic test that accords with our intuitions about what is or isn't prototypical. Each of the nonprototypical cases cited above is a case involving anger. There appear to be no necessary and sufficient conditions that will fit all these cases. However, they can all be seen as variants of the prototypical anger scenario. Part of what makes the prototypical scenario prototypical is that it is sufficiently rich so that variations on it can account for nonprototypical cases. The point is that there is no single unified cognitive model of anger. Instead there is a category of cognitive models with a prototypical model in the center. This suggests that it is a mistake to try to find a single cognitive model for all instances of a concept. Kinds of anger are not all instances of the same model; instead, they are variants on a prototypical model. There is no common core that all kinds of anger have in common. Instead, the kinds of anger bear family resemblances to one another Conclusions We have shown that the expressions that indicate anger in American English are not a random collection but rather are structured in terms of an elaborate cognitive model that is implicit in the semantics of the language. This indicates that anger is not just an amorphous feeling, but rather that it has an elaborate cognitive structure. Moreover, if Rosaldo's (1980) account of anger among the Ilongot is correct, it would follow that the cognitive model of anger implicit in English is anything but universal. However, very significant problems and questions remain. - First, there are aspects of our understanding of anger that our methodology cannot shed any light on. Take, for example, the range of offenses that cause anger and the corresponding range of appropriate responses. Our methodol-

44 THE CONCEPT OF ANGER 3 7 ogy reveals nothing in this area. - Second, study of the language as a whole gives us no guide to individual variation. We have no idea how close any individual comes to the model we have uncovered, and we have no idea how people differ from one another. - Third, our methodology does not enable us to say much about the exact psychological status of the model we have uncovered. How much of it do people really use in comprehending anger? Do people base their actions on this model: Are people aware of the model? How much of it, if any, do people consciously believe? And most intriguingly, does the model have any effect on what people feel? Certain things, however, do seem to be clear. Most speakers of American English seem to use the expressions we have described consistently and make inferences that appear, so far as we can tell, to be consistent with our model. We make this claim on the basis of our own intuitive observations, though to really establish it, one would have to do empirical studies. If we are right, our model has considerable psychological reality, but how much and what kind remains to be determined.

45

46 3. THE CONCEPT OF PRIDE 3.1. Some additional issues In this chapter two issues will receive particular emphasis. The first issue has to do with the causes of pride. In the discussion of anger, the 'causal aspect' of the concept has not been dealt with in detail. All I have said was that, typically, there is an offence which gives rise to anger. However, the detailed study of this causal aspect of concepts (that have one) can be important. Not only does it make the conceptual analysis of the concept more complete, but it also has some significance for the lexical behavior of the word corresponding to the concept. This issue will be discussed in the chapter that deals with how this lexical approach can contribute to a better understanding of some lexical phenomena. In this chapter, among other things, the most typical causes that lead to pride will be examined. The second issue can be presented in the following way. The conceptual domain of PRIDE contains a variety of concepts. These include CONCEIT, SELF-ESTEEM, JUSTIFIED PRIDE, FALSE PRIDE, VANITY, SELF- ASSURANCE, and more. Our intuition that all of these concepts have something to do with each other and PRIDE is supported by (or arises because of) the fact that they can all be referred to by the word pride (if the context requires, with some qualification like justified, too much, etc.). Dictionary definitions also make abundant use of this. They tend to define the various senses of pride by means of the other concepts which, in turn, are defined by means of pride. Thus our conceptual system seems to operate with a superordinate category PRIDE that defines a conceptual domain with concepts like PROPER PRIDE, CONCEIT, VANITY, etc. in it. The question arises: how is this conceptual domain organized? And this question entails such further questions as: what is the exact relationship between these concepts, what conceptual organization each of these concepts has, and whether there is a concept in this category of concepts which serves as a 'cognitive reference point' (i.e. as a prototype) for the understanding of the other concepts that are related to it? My answer will be that there is one and that it is what can be termed as 'ba-

47 40 METAPHORS OF ANGER, PRIDE, AND LOVE lanced pride as an immediate response'; and my primary concern will be to characterize this kind of pride and to give an account of how other forms of pride can be systematically defined in relation to it. Indeed, my main argument for the existence of such a reference point is that while it is possible to derive the other forms of pride from 'balanced pride as an immediate response', the other forms could not serve this purpose in any natural, intuitively satisfying way Some metonymies for pride Let us begin with the folk model of the physiological effects of pride. This part of the folk model maintains that THE PHYSIOLOGICAL EFFECTS OF PRIDE ARE REDNESS IN THE FACE, INCREASED HEART RATE, INTERFERENCE WITH ACCURATE PERCEPTION, AND INTERFERENCE WITH NORMAL MENTAL FUNCTIONING. Given the general metonymic principle, THE PHYSIOLOGICAL EFFECTS OF AN EMOTION STAND FOR THE EMOTION we get the following system of metonymies for pride: REDNESS IN THE FACE: He flushed with pride. INCREASED HEART RATE: His heart was throbbing with pride. Her heart fluttered with pride. INTERFERENCE WITH ACCURATE PERCEPTION: He was blinded by conceit. His conceit prevented him from seeing clearly. She was blinded by her own glory. Winning the class election turned his head. INTERFERENCE WITH NORMAL MENTAL FUNCTIONING: Success went into his head. It is important to see that REDNESS IN THE FACE and INCREASED HEART RATE apply to what is called justified or proper pride, whereas INTERFERENCE WITH ACCURATE PERCEPTION and INTERFER ENCE WITH NORMAL MENTAL FUNCTIONING apply to conceit, and that the latter two are more intense physiological effects than the former

48 THE CONCEPT OF PRIDE 41 ones. Furthermore, INTERFERENCE WITH ACCURATE PERCEP TION and INTERFERENCE WITH NORMAL MENTAL FUNCTION ING are physiological effects that impair one's normal functioning. These observations lead to further principles in the folk model of the physiological effects of pride: AS PRIDE INCREASES, ITS PHYSIOLOGICAL EFFECTS INCREASE. THERE IS A LIMIT BEYOND WHICH THE PHYSIOLOGICAL EFFECTS OF PRIDE IMPAIR NORMAL FUNCTIONING. The linguistically-based folk model of the behavioral reactions of pride maintains: THE BEHAVIORAL REACTIONS OF PRIDE ARE ERECT POST URE, CHEST OUT, BRIGHTNESS OF THE EYES, SMILING, TELLING PEOPLE ABOUT ONE'S ACHIEVEMENTS, etc., HEAD HELD UNNATURALLY HIGH, CHEST UNNATURALLY THRUST OUT, FORMS OF WALKING, OSTENTATIOUS BEHAVIOR, THINKING ONE IS UNIQUE, AND BOASTING. Given the general principle THE BEHAVIORAL REACTIONS OF AN EMOTION STAND FOR THE EMOTION we get the following system of metonymies for pride: ERECT POSTURE: After winning the race, he walked to the rostrum with his head held high. CHEST OUT: After winning the race, he swelled with pride. BRIGHTNESS OF THE EYES: She was beaming with pride. He was glowing with pride. There was pride in his eyes as he looked at his son. SMILING: He was smiling proudly after winning the race. TELLING PEOPLE ABOUT ONE'S ACHIEVEMENTS: He ran home to tell his mother about his success.

49 42 METAPHORS OF ANGER, PRIDE, AND LOVE HEAD HELD UNNATURALLY HIGH: She's going around with her nose in the air. Don't high-hat me again! She's looking down her nose at everyone. She said "No!", bridling up. CHEST UNNATURALLY THRUST OUT: He was bloated with pride. He was going around with his chest out. Why is he so puffed up? John's so swelled up with pride that I think he'll burst. He's inflated with pride. You shouldn't be so chesty. FORMS OF WALKING: The manager strutted along the hall. He swaggered down the street after winning the fight. She's going around like a peacock. He's walking around as if he owned the place. OSTENTATIOUS/THEATRICAL BEHAVIOR: She's giving herself airs. The child was putting on airs. You don't need to put on the dog. He's just a show-off. She's flaunting her new fur coat. THINKING ONE IS UNIQUE: He thinks he is it. You needn't think your shit doesn't stink. She thinks the sun shines out of her asshole. BOASTING: He's always singing his own praises. Stop bragging. Here he is blowing his own horn again. He's always broadcasting his own achievements. You like talking big, don't you? Here, again, we should notice that ERECT POSTURE, CHEST OUT, BRIGHTNESS OF THE EYES, SMILING, TELLING PEOPLE ABOUT ONE'S ACHIEVEMENTS are behavioral reactions which characterize jus-

50 THE CONCEPT OF PRIDE 43 tified pride, and that the rest of the reactions are typical of conceit. Also, the reactions that characterize justified pride are less intense and less salient reactions than the ones that typically go together with conceit. This is particularly obvious in such comparable pairs as ERECT POSTURE vs. HEAD HELD UNNATURALLY HIGH, CHEST OUT vs. CHEST UNNATURALLY THRUST OUT, and TELLING PEOPLE vs. BOASTING. This gives us another principle in the folk model of the behavioral reactions of pride: AS PRIDE INCREASES, ITS BEHAVIORAL REACTIONS INCREASE Some metaphors After the metonymies let us now turn to some of the metaphors for pride. The discussion of the metaphors will enable us to see the different kinds of pride, the aspects of pride that they address, and the overall cognitive organization of the different kinds of pride. Pride also has its version of THE EMOTIONS ARE FLUIDS IN A CONTAINER metaphor. The PRIDE IS A FLUID IN A CONTAINER metaphor is partially motivated by the THE BODY IS A CONTAINER FOR THE EMOTIONS metaphor. The experiential basis for the metaphor PRIDE IS A FLUID IN A CONTAINER seems to be provided by the physiological effects INCREASED HEART RATE (involving the heart as a container with blood in it) and the behavioral reaction CHEST OUT (involving the chest as a container). PRIDE IS A FLUID IN A CONTAINER His good performance filled him with pride. Pride welled up inside him at the sight of his garden. He was full of pride after beating the former champion. Her pride rose as she watched her children perform. In this metaphor the container is the body. There is another metaphor, where the container is the heart: PRIDE IS (A FLUID) IN THE HEART Her heart swelled with pride. The news filled her heart with pride. Pride swelled his heart. These metaphors address a variety of issues concerning PRIDE. In particular, they highlight the aspect of intensity (full of, swell with), the change in its

51 44 METAPHORS OF ANGER, PRIDE, AND LOVE intensity (rise), that it has a cause (good performance, sight of one's garden, beating the champion, etc.), and that it has a beginning (well up, fill). The FLUID IN A CONTAINER and HEART metaphors are constituted by the following ontological correspondences: - the container is the body - the fluid in the container is pride - for the container to be full of the fluid is for pride to be intense - for the fluid to move upwards is for the pride to become more intense (cf. MORE IS UP) - for the fluid to appear is for pride to begin to exist 3.4. Causes of pride Now let us make a brief inventory of the causes that make people proud. I will ignore many details and deal with the issues only in the depth that is necessary for my purposes in this study. If we look at the kinds of things that people typically mention as the causes for their pride, we get the following groups: 1. ACHIEVEMENTS: winning the race, getting a good grade in a difficult exam, writing a good book, solving a problem, etc. 2. POSSESSIONS: owing a Mercedes, having a lot of money, having a diamond necklace, etc. 3. BELONGING TO A GROUP: the team you play for, the university where you study, your hometown, your nation, etc. 4. APPEARANCES: good looks, having a pretty face, having a shapely body, long dark hair, green eyes, etc. 5. PHYSICAL/MENTAL CAPABILITIES/SKILLS/PROPERTIES: how high you can jump, how strong you are, how fast you can read, how many things you remember, your health, etc. 6. MORAL QUALITIES: your honesty, that you cannot be bribed, that you have never lied in your life, etc. 7. SOCIAL POSITION/STATUS/CLASS: you are the manager, you are a well known singer, you are an aristocrat, etc. And the list could no doubt be continued. But this much will do for the present purposes. What makes the list relevant to our purposes is that the various kinds of pride seem to go together with particular causes. For example, JUS TIFIED PRIDE seems to go together typically with ACHIEVEMENTS, VANITY with APPEARANCES, HAUGHTINESS with SOCIAL

52 THE CONCEPT OF PRIDE 45 STATUS/CLASS, etc. Of course, this does not mean that the various kinds of pride are restricted to one particular cause. The claim is that a given form of pride favors a particular kind of cause, but less typically can also take others. In order for a person S (short for the Self) to be appropriately proud, S must be involved in some way in one of the states of affairs given in the above taxonomy. That is, S must be the 'agent' in ACHIEVEMENTS, the 'owner' in the case of POSSESSIONS, a 'member' of the GROUP, etc. in order to be able to be proud of an ACHIEVEMENT, POSSESSION and GROUP, respectively. In other words, I cannot be proud of a valuable book as such. I can only be proud of it if I am the owner of the book. However, as we will see below, this condition can be weakened in less prototypical cases of pride. This is, however, not sufficient for S to be proud. For S to be appropriately proud of X (=something), X must have some value. The things that can be the causes of pride either have built-in social values or the proud person assigns individual values to them. Thus we have a scale of social values and a scale of individual values. Things with built-in social values are, for example, achievements, expensive goods, things considered beautiful in a culture, honesty, being the manager of a company, etc. (Many of these things are oriented UP: VIRTUE, SOCIAL STATUS, etc). The individual rates things highly on the scale of individual values if in a given context it is difficult for him to achieve X (something). In this case he can be justifiably proud of X. Thus one can only be proud of X if X has a built-in social value or X can be assigned a value by the individual (=S). The difference between built-in social values and assigned individual values can be illustrated by a simple example. Under normal circumstances, it is not an achievement to lift (in the sense of 'raise') an ordinary pen. That is, lifting a pen does not have a built-in social value. If, however, S is a paralytic and after a long period of exercising he is eventually able to lift a pen, he can be justifiably proud of lifting the pen. For S, lifting the pen can be assigned a value on the scale of individual values, and thus S will consider lifting the pen as an achievement that forms a proper basis for his pride. In general, it can be maintained that individual value scales are accepted to base one's pride on when there are socially accepted extenuating circumstances for ignoring social value scales. One such extenuating circumstance that is socially accepted is one's illness. We said that S can be proud of X if he has some direct role ('agent', 'owner', 'member', etc.) in one of the possible causes for pride in the above

53 46 METAPHORS OF ANGER. PRIDE, AND LOVE taxonomy. However, in less typical cases of pride, this is not always the case. We often hear people proudly say things like: My son's got straight A's in school. I know a man who can eat fifty eggs in one sitting. Someone has an '86 Mercedes on our street. In reporting these utterances, we could say: He's proud of his son. He's proud of someone he knows. He's proud of someone living on his street. What this suggests is that S (=he, in the example) can be proud of X not only if he (S) has some direct role in the cause, but also if S is a relative of someone (my son) who plays a direct role ('agent') in the cause; or if S knows someone (I know a man) who plays a direct role ('agent') in the cause; or if S lives in the vicinity of someone (someone on our street) who has a direct role ('owner') in the cause of pride. Thus it seems that, less typically though, S can also be proud of X if S 'is a relative of Ρ ('someone other than S'), 'knows' P, and 'lives nearby' P, provided Ρ plays a direct role ('agent', 'owner', etc.) in X. Of course, we sense a lessening in the degree of acceptability as a cause for pride between: S 'plays a direct role' ('agent', 'owner', 'member', etc.) in X. S 'is a relative of' Ρ who plays a direct role in X. S 'knows' Ρ who has a direct role in X. S 'lives nearby' Ρ who has a direct role in X. It is very likely that S's playing a direct role in X produces a much better example of pride than S's living nearby Ρ who has a direct role in X Scales, related concepts and the prototype So far I have talked about only one scale: A value scale for actions, possessions, appearances, social positions, etc. This scale can be imagined as a scale that is oriented UPWARDS, and that has a threshold associated with it. The actions, possessions, appearances, social positions, etc. that are above the threshold have (social or individual) value and therefore are proper causes for pride, given that the other conditions are also met. The idea of this scale gives us an explanation for the expression justified pride. One's pride is justified if the cause of one's pride is above the threshold on the value scale. And we have to allow for the possibility that opinions in judging what's above

54 THE CONCEPT OF PRIDE 47 and below the threshold may differ. Now consider a sentence like: He's more proud of it than he should be. This sentence implies not just one, but two scales. The two scales are a value scale and a pride scale. In effect, what the speaker says is that S has more pride than is justified by X (the cause of his pride); i.e. X occupies a certain position above the threshold on the value scale and S's pride is at a point on the pride scale that is higher than the point that corresponds to the point on the value scale. Notice that this is not the same case as justified pride. In justified pride the issue is whether the cause of pride is above or below the threshold on the value scale. If it is above, the pride is justified. In the case under discussion, the issue is the measure of pride; i.e. whether S has or does not have the amount of pride that he is entitled to, given the amount of value on the value scale. That is, whether the pride is justified or not is not called into question; what is suggested is that the pride is excessive relative to its cause. It is important to realize that the entire discussion of value in relation to actions, states and properties is couched in metaphor. States of affairs (actions, states, and properties) are not commodities, they don't have values literally. Thus our thinking about states of affairs seems to be metaphorical to a large extent, as the following metaphor indicates this: STATES OF AFFAIRS ARE COMMODITIES That was a valuable victory. He had to pay a high price for his dishonesty. Give her some roses in return for her kindness. His paper isn't worth looking at. Did your plan pay off? The value of his work is tremendous. Just watch out! I'll pay you back for this. The analysis of the kind of pride under consideration ('pride as immediate response') has to be augmented by what can be called 'related concepts' (for a detailed explanation of this notion, see the chapter on love). It can be argued that pride as characterized so far has JOY as an inherent concept. This would seem to follow from at least the following linguistic evidence: First, there are certain behavioral reactions (correspondingly, certain metonymies) that are shared by both PRIDE and JOY. These include BRIGHTNESS OF THE EYES and SMILING. Second, linguistic usage seems to indicate that, at least in the

55 48 METAPHORS OF ANGER, PRIDE, AND LOVE kind of pride under discussion, one cannot be proud of something and not feel joy:?he was proud of winning the race, but he was not happy. *He was proud of winning the race, but he was happy. It has to be noticed that the cases where PRIDE entails JOY in the most clear way are those where pride arises from achievements, rather than the other causes. The reason perhaps is that, unlike achievements, social status, appearance, possession, etc. are permanent states or properties, which do not give rise to such immediate events as joy. It is partly this immediacy which makes 'pride resulting from an achievement' a more prototypical form of pride than the forms of pride that result from other causes. (Some other reasons are discussed at the end of the chapter). And, thirdly, PRIDE and JOY seem to appear together in such conventionalized expressions as one's pride and joy, which can be taken as an indication that the two concepts are closely related. One can only feel joy if one is satisfied with what one has done, and an achievement also presupposes that one is satisfied with what one has done. Thus SATISFACTION is another inherent concept of PRIDE (of the kind under investigation). Linguistically, in addition to the but-test, this is shown by he fact that the following two sentences entail each other: He looked at his son with pride. He looked at his son with satisfaction. In the light of what has been done so far, we can see the following prototypical model of PRIDE emerge: 1. S does X. S is directly involved in X. S values X on a social value scale; X's value is high. S perceives X as an achievement. 2. S is proud of X; the pride is intense (but not inappropriately so) The pride scale is the same height as the value scale. S experiences physiological effects: Redness of the face, increased heart rate. S exhibits behavioral reactions: Erect posture, chest out, brightness of the eyes, smiling, telling people about the achievement. Inherent concepts: Joy, satisfaction. This model reveals pride not only as 'immediate response' but also as 'ba-

56 THE CONCEPT OF PRIDE 49 lanced'. Many of the concepts within the domain of pride can be seen as deriving from this notion of pride Self-esteem As some of the examples we have seen so far suggest, there is a form of pride that comes about as an immediate response to a state of affairs that S perceives as an achievement (cf. Winning the race filled him with pride). However, not all forms of pride are like this. Consider the following metaphor: PRIDE IS A PERSON His criticism hurt her pride. They humbled his pride. His pride revolted against the treatment he received. His national pride was roused/awakened by the stupid comments. Don't say anything that may wound his pride. Her pride was deeply injured. The pride in this metaphor is not an immediate response to a situation. Rather, it is something that S has had for some time and that, on a given occasion, can revolt, can be hurt, can awaken, etc. This seems to be the kind of pride that one takes in one's possessions, one's country, one's success in a field (but not in just one achievement, which is the pride as immediate response), etc. By and large, in these cases self-esteem is a synonym for pride. Most of the examples in the PERSON metaphor could be used with the word self-esteem in place of pride (e.g. His self-esteem was injured). The PRIDE IS A PERSON metaphor is constituted by the following ontological correspondences: - the person is pride - the physical harm done to the person is emotional harm to S - physical revolt against something is S's emotional revolt - humbling the person is reducing S's pride - for the person to awaken is for pride to begin or continue to function The word self-esteem can be used as a synonym for the word pride in the following metaphor as well: PRIDE IS A SUPERIOR; THE SELF IS AN INFERIOR His pride prevents him from doing anything dishonorable. His pride did not allow him to accept the aid. Pride did not let her do what she wanted to do the most.

57 50 METAPHORS OF ANGER, PRIDE, AND LOVE The main focus of this metaphor seems to be how pride has an influence on what we should, and in particular, on what we should not do. The issue of how pride can determine our actions shows up in other metaphorical expressions as well: swallow one's pride, pocket one's pride, and set one's pride aside. These are expressions that are not instances of the PRIDE IS A SUPERIOR conceptual metaphor. Nevertheless, the diversity of these expressions indicates that this is an important aspect of pride for us. Although in many of the examples self-esteem could replace pride, selfesteem is, on the whole, different from pride. It has to do with one's overall evaluation of oneself. Unlike the kind of pride exemplified by the CON TAINER and HEART metaphors (which have their particular causes), the kind of pride represented by the following metaphor does not have particular causes. SELF-ESTEEM IS AN ECONOMIC VALUE (WHICH A PERSON ESTIMATES HIMSELF TO HAVE) He has low self-esteem. She values herself highly. Her achievements boosted her self-esteem. Don't overestimate yourself. She underestimates herself. Don't let him lower your self-esteem. This metaphor enables us to see self-esteem as something that can be represented on a scale: It can be low, it can be boosted, and it can be high. When self-esteem is high, the person is proud of himself. That this is so can be seen from the following two sentences that entail each other under normal circumstances: She has high self-esteem. She is proud of herself. This close relationship between pride and self-esteem is likely to be the result of the way we structure ourselves and our actions, possessions, appearances, etc. (i.e. the causes pride) in metaphorical terms. Since we see our actions, possessions, etc. and ourselves as having VALUES, we expect people with actions, possessions, etc. that have HIGH VALUES to see themselves as having a HIGH VALUE, that is, as having high self-esteem. Of course, implicit here is a principle we use for understanding ourselves: A PERSON'S SELF-ESTEEM IS DETERMINED BY HIS

58 THE CONCEPT OF PRIDE 51 ACTIONS, POSSESSIONS, ETC. But this takes place only indirectly, through pride. The particular actions, states and properties that have HIGH VALUES result in pride, and the particular instances of pride give rise to self-esteem. Self-esteem is thus a generalized form of pride that is not directly linked to particular causes of pride: S is just proud of himself (i.e. values himself highly), and not of some actions, state or property involving him. To complete the survey of this form of pride, mention must be made of another metaphor. The metaphor is PRIDE IS AN OBJECT. PRIDE IS AN OBJECT His self-esteem was shattered when he had found out the result. He lost his pride and began to implore her not to leave him. She managed to destroy my self-esteem. He tore her pride to shreds. He kept his pride in the face of all adversities. The main focus of this metaphor is the existence of pride and how pride can play a role in determining our actions. In the discussion of the PRIDE IS A SUPERIOR metaphor, I said that our pride can have an influence on what we do. Now we can see how this works. If S has his pride (i.e. has kept it), then it does not allow him to do certain things, but if he does not have it (i.e. has lost it), then S does certain things he would not do otherwise. The ontological correspondences are: - the object is pride - the intact object is existing pride - the destroyed object is pride out of existence - having the object is having pride The epistemic correspondence that accounts for keeping and losing one's pride is: Source: An object can only work for S if S has the object. Target: Pride can only determine S's actions if S has the pride. The word dignity could be a natural replacement for pride in the expressions keeping pride and losing pride. This would seem to suggest that dignity is that aspect of pride (or self-esteem) that plays a role in determining our actions. This seems to have some basis in the fact that the most prevalent metaphor for DIGNITY addresses just this issue:

59 52 METAPHORS OF ANGER, PRIDE, AND LOVE DIGNITY IS AN OBJECT He cast off his dignity. He preserved his dignity. She lost all her dignity. His grey hair and solemn manner lent him dignity. He set aside his dignity and accepted a menial job. She descarded her dignity. She kept her dignity in the face of all adversities. He disposed of his dignity. Similarly to a portion of the PRIDE IS AN OBJECT metahor, DIGNITY is seen here as something which, if S has it, prevents S from doing certain things and if he does not, then S does things he would not do otherwise. As can be seen, we have quite a few expressions at our disposal to denote loss of pride, and hence giving up the controlling effect of pride (i.e. dignity). These expressions in the metaphor are: lose, set aside, cast off, discard and dispose of dignity. However, there are some subtle differences between them. Cast off, discard and dispose of carry with them a blaming of S by an outside observer for S's not keeping his dignity. This is shown by the unacceptability of sentences involving cases of justified self-esteem (pride). *He cast off his pride and accepted a menial job, and I think he was right. That this sentence would be acceptable when the kind of pride involved is unjustified pride is another matter. What is at issue here is that it is inconsistent for someone to express blaming (cast off) in one part of the sentence and agreement (I think he was right) in another. Compare: He set aside his dignity and accepted a menial job, and I think he was right. The implication of blame in cast off, discard and dispose of is possible because they seem to draw on a special case of the OBJECT metaphor: DIGNITY IS A WORTHLESS OBJECT. Since DIGNITY is considered as something DESIRABLE, or VALUABLE, in our folk model of human qualities (we have not justified this claim here), it is the idea of DIGNITY being a WORTHLESS OBJECT that allows the expressions to function as expressions of blame. Set aside does not have the implication of blame. Instead, the setting aside of dignity is seen as justified because it is done for a larger purpose. This is why the speaker can approve of it. And losing dignity is different from setting aside dignity in that it is unintentional. It is seen as a consequence

60 THE CONCEPT OF PRIDE 53 of other events, and not as something that is done for a purpose. What has been said about SELF-ESTEEM yields the following prototypical model: 1. S is directly involved in many X's. S sees the X's as having high value. 2. As a result, S values himself highly. S's self-esteem is high; S is proud (of himself). The value scale and the pride scale are even. S has dignity: It determines what S should not do Conceit So far we have only considered cases where the value scale and the pride scale are balanced. Now I will turn to cases where the pride scale is higher than the corresponding value scale. First we will discuss CONCEIT, and then VANITY. Some of the behavioral reactions and physiological effects that are assumed to accompany CONCEIT seem to form the basis of some of the major metaphors for CONCEIT. In particular, the behavioral reaction HEAD HELD UNNATURALLY HIGH seems to provide the motivation for the CONCEITED PERSON IS UP/HIGH metaphor: A CONCEITED PERSON IS UP/HIGH He's been on his high horse ever since he was elected captain. You don't have to be so stuck-up. Don't you think he's a little bit uppish? She became high and mighty when she won the prize. It was another lofty speech. Look at that uppity bastard. It's time you got off your high horse. And also, historically, superior and haughty. And the behavioral reaction CHEST UNNATURALLY THRUST OUT seems to motivate a related metaphor for CONCEIT: A CONCEITED PERSON IS BIG When the teacher made John the monitor, he became too big for his boots. He's big as you please.

61 54 METAPHORS OF ANGER, PRIDE, AND LOVE He got a big head after winning the race. She has a swollen opinion of herself. She's had a swelled head ever since she got her Ph.D. Both of these metaphors are concerned with greater than normal physical size: Greater than normal height and greater than normal bigness, respectively. This defines a scale with two points on it: A point indicating normal size and a point indicating greater than normal size. There seems to be an obvious link between this scale and the pride scale I have mentioned before. The point indicating normal size would seem to correspond to an appropriate amount of pride and the point indicating greater than normal size to more than an appropriate amount of pride. An appropriate amount of pride is here defined as a point on the pride scale that is on the same level as the corresponding point on the value scale. We have seen how the folk model of behavioral reactions captures the increase of pride to a level which is inappropriate: AS PRIDE INCREASES, ITS BEHAVIORAL REACTIONS INCREASE. Thus ERECT POSTURE becomes HEAD HELD UNNATURALLY HIGH and CHEST OUT becomes CHEST UNNATURALLY THRUST OUT. This same aspect of pride is also captured by the metaphors we use. Thus cases of pride with an appropriate amount of pride seem to be expressed by the FLUID IN A CONTAINER and HEART metaphors that confine pride to a level within the body as a CONTAINER, while cases of inappropriate amount of pride are captured by metaphors (greater than normal HEIGHT and BIGNESS) that define a level of pride as being, as it were, outside the normal body. The metaphor A CONCEITED PERSON IS UP/HIGH indicates more than just 'a person with an inappropriate amount of pride'. It also indicates that this person considers himself as superior to other people. It is important to note, however, that it is often only in the judgement of an outside observer that S considers himself superior. This attribution of superiority goes hand in hand with the attribution, also by an outside observer, of more pride than is appropriate. The concept of SUPERIORITY may arise as a result of the way our emotion concepts (and some related concepts) are structured metaphorically. Since in this system THE OBJECT OF CONTEMPT IS DOWN/ LOW, it is easy for the metaphor A CONCEITED PERSON IS UP/HIGH

62 THE CONCEPT OF PRIDE 55 to have the implication of SUPERIORITY. SUPERIORITY is thus an inherent concept of CONCEIT. People displaying this attitude to others are called haughty. As has been mentioned at the beginning, the principal causes associated with HAUGHTINESS are SOCIAL CLASS/STATUS/POSI TION. This is not surprising in view of the fact that these are concepts which are all oriented UP, just like CONCEIT. Thus the UP orientation of CON CEIT is maximally coherent with other parts of the metaphorical system. There is another metaphor in our metaphorical system that is relevant to our understanding of PRIDE. This metaphor is SIGNIFICANT IS BIG (taken from Lakoff and Johnson 1980): SIGNIFICANT IS BIG He's a big man in the garment industry. He's a giant among writers. It's no big deal. I was astounded by the enormity of the crime. His accomplishments tower over those of lesser men. He's head and shoulders above everyone in the industry. Whereas with other forms of pride the proud person attaches importance to the cause of his pride, in the case of CONCEIT he attaches importance to his own person. This makes SELF-IMPORTANCE inherent in CONCEIT. THE CONCEITED PERSON IS BIG metaphor can be viewed as a special case of the SIGNIFICANT IS BIG metapor. Thus we have a case where a metaphor is doubly motivated. Part of its motivation comes from the behavioral reaction CHEST UNNATURALLY THRUST OUT and another part from a more general metaphor in the system: SIGNIFICANT IS BIG. A third concept inherent in CONCEIT is SELF-LOVE. This is suggested by the following expressions that can mean either CONCEIT or SELF-LOVE or both, depending on the context: He is full of himself. She is entirely wrapped up in herself. He is stuck on himself. In the physiological reactions of CONCEIT, we have seen that there is INTERFERENCE WITH ACCURATE PERCEPTION and INTERFER ENCE WITH NORMAL FUNCTIONING that are assumed to accompany CONCEIT. That is, these physiological effects emphasize inability to function normally. The same idea is brought into focus by the following two

63 56 METAPHORS OF ANGER, PRIDE, AND LOVE metaphors. CONCEIT IS AN OPPONENT/SUPERIOR He was completely overcome by conceit. Conceit robbed him of the ability to judge things properly. Conceit did not allow her to see things as they are. The metaphor CONCEIT IS A DANGEROUS ANIMAL can be regarded as a special case of the OPPONENT metaphor: CONCEIT IS A DANGEROUS ANIMAL She was swallowed up by conceit. He was devoured by conceit. These metaphors make us see conceited people as being unable to function normally. This discussion yields the following prototypical model for CONCEIT: 1. S is directly involved in S (or several X's) S perceives X as having high value on the value scale. 2. S's pride scale is high; S is proud of X (or of himself). In the opinion of an outside observer, the pride scale is higher than the value scale. S is unable to function normally. S experiences physiological effects: Interference with accurate perception, interference with normal mental functioning. S exhibits behavioral reactions: Head held unnaturally high, chest unnaturally thrust out, forms of walking, ostentatious behavior, thinking one is unique, boasting. Inherent concepts: Superiority, self-importance, self-love Vanity Let us now turn to VANITY. The central (i.e. the most productive) metaphor for this concept appears to be the metaphor VANITY IS A (SENSUAL) PERSON. VANITY IS A (SENSUAL) PERSON He is really wallowing in the praises of his new book. What you said wounded his vanity. Her vanity is just insatiable. Her praises soothed his vanity.

64 THE CONCEPT OF PRIDE 57 They tickled his vanity by praising his work to the skies. It only fed her vanity. Your words of appreciation pandered to his vanity. Her vanity was eventually gratified. His vanity was offended by your critical comments. He tried desperately to cater to her vanity, but nothing satisfied her. She indulged her vanity by looking at her hair in the mirror for hours. We noted at the beginning that the typical causes for VANITY have to do with APPEARANCES. ACHIEVEMENTS, POSSESSIONS, MORAL QUALITIES, BELONGING TO A GROUP, and SOCIAL POSITION/ STATUS/CLASS are not proper causes for VANITY. However, as some of the examples suggest, INTELLECTUAL CAPABILITY can also be a cause for VANITY. These (unfortunately unexplained) facts would then explain why some of the sentences below are acceptable and why some of them are not: She is vain of her long dark hair. He is vain of his cleverness. *He is vain of winning the race. *He is vain of being an aristocrat. In addition to this feature of VANITY, there are some other important aspects of it that are focused on by this metaphor. Perhaps the most important of these is the aspect which is captured by the expression insatiable vanity. Here the question is what it is that VANITY has an appetite for. Let us call this aspect the purposive aspect of the concept. The related aspect of VAN ITY is captured by such expressions as feed, cater to and, to some extent, also pander to. The expressions suggest that it is mostly some FOOD that can satisfy the person's VANITY. More generally, it can be claimed that there is a DESIRE in VANITY that needs to be satisfied. The satisfaction of VAN ITY'S DESIRE can take various forms, of which FEEDING is one. Some other forms are soothing and tickling, and the general forms gratify and satisfy. But what, in turn, corresponds to the food in feed or cater to, the cool(ing) fluid or air in soothe, and the gentle touch in tickle in the concept of VANITY? As the examples suggest, the corresponding entity is praise or flattery. An obvious question to ask at this point is: Why does the vain person need praises or flattery? One might say that he needs it because APPEAR ANCE, the typical cause for VANITY, and maybe also ABILITIES we are

65 58 METAPHORS OF ANGER, PRIDE, AND LOVE born with, are things that we are not responsible for, cannot take credit for and thus need confirmation of. These things are the opposites of ACHIEVE MENTS that are the typical things a person can take credit for, and thus be proud of. A further aspect of VANITY is that the vain person enjoys the praises and flattery (wallow in praises). And not only does he enjoy them but actually invites them. This would seem to follow from the use of such expressions as He's always fishing for compliments. And since the vain person is out for compliments and praises, it is only natural that he does not tolerate criticism. The expression fish for compliments deserves some attention, for it reveals a great deal about VANITY. Fishing implies trying to catch fish. Catching is an act the target of which (the fish) does not readily lend itself to the act. Correspondingly, the person whose praises must be sought is not a natural and/or sincere source of praise. Praises and flattery are forms of public acknowledgement. Fishing for compliments can then be seen as seeking unwillingly given public acknowledgement. So far we have not seen any explicit indication of the presence of PRIDE in VANITY. However, there is a minor but important metaphor whose instances clearly indicate that PRIDE is present in VANITY. VANITY IS AN INFLATED OBJECT His sarcastic remarks pricked her vanity. His vanity has been punctured. Here the inflated object can only correspond to bigness in the CONCEITED PERSON IS BIG metaphor, which in turn is motivated by the CHEST UNNATURALLY THRUST OUT behavioral reaction. We have seen in the VANITY IS A (SENSUAL) PERSON metaphor that VANITY has a DESIRE. The same aspect of vanity can be found in the OPPONENT metaphor. VANITY IS AN OPPONENT She yielded to her vanity. Her vanity has been appeased. She overcame her vanity. He gave in to his pride. In this metaphor it is the expression appease that shows that VANITY has a DESIRE. Appease presupposes that the person (the opponent) has demands that have to be met. Similarly to CONCEIT, SELF-LOVE is inherent in VANITY. This is

66 THE CONCEPT OF PRIDE 59 indicated by such expressions as the following: He loves the sound of his own voice. She is completely wrapped up in her own beauty. SELF-LOVE brings with it undue preoccupation with oneself. The cause of vanity becomes the center of the vain person's attention. This attitude can be exemplified by: She is always looking at herself in the mirror. We can conceive of this type of behavior as a behavioral reaction that is assumed to accompany vanity. Our treatment of VANITY gives rise to the following prototypical model for the concept: 1. S has/is X (where X has to do with appearances). S perceives X as having high value. 2. S is proud of X. The scale of pride is higher than the scale of value. Behavioral reaction: Undue preoccupation with oneself. Inherent concept: Self-love. S wants more public recognition than is due to him or her Conclusion We have seen that 'balanced pride as immediate response' can be regarded as a cognitive reference point in the system of concepts that we ordinarily designate with the term pride. Pride of this kind can serve as a prototype with respect to which other related concepts, as characterized by their own respective prototypes, can be defined. The term pride can be used of this prototypical pride as well as of self-esteem, dignity, conceit, and maybe even some cases of vanity. However, the terms self-esteem, dignity, conceit and vanity could not be used to refer to what has been called prototypical pride. This is because these forms of pride derive from 'balanced pride as an immediate response', and not the other way around. There is also some non-linguistic evidence which points to the same conclusion. As Lakoff (1987) has pointed out, prototypes often take the form of ideals. In the case of two related scales, such as the pride and value scales, we tend to take the state of balance between the two as the ideal. This phenomenon is not limited to pride but is pervasive in our thinking. Another example from the domain of emotion will be discussed in the chapter on love. 'Ba-

67 60 METAPHORS OF ANGER, PRIDE, AND LOVE lanced pride' is thus a more ideal form of pride than such 'non-balanced' forms as conceit and vanity. The other piece of evidence comes from emotion research. W. Lyons (1980) argues that what he calls 'ongoing' emotional states are more paradigmatic than dispositions or traits. Thus, 'pride as an immediate response' would be a more paradigmatic form, or better example, of pride than conceit and vanity, which can be viewed as dispositions and/or traits. The point is that given these pieces of non-linguistic evidence it would really be surprising if linguistic use did not reflect these cognitive facts in some way. (More evidence of another kind can be found in chapter 6.1.).

68 4. THE CONCEPT OF ROMANTIC LOVE 4.1. Some further aspects of a concept Romantic love is commonly thought of as a mysterious emotion which it is notoriously difficult to pin down. This idea is shared not only by laymen. Various authors in various disciplines have expressed the view that this is a concept that is difficult to grapple with and define. Consider, for example, what Bloomfield (1933) had to say about this: "[...] and we can define the names of plants or animals by means of the technical terms of botany and zoology, but we have no precise way of defining words like love or hate, which concern situations that have not been accurately classified and these latter are in the great majority". (Bloomfield 1933: 139) Or let us take as another illustration Smith and Medin's (1981) work. They study the nature of concepts and the role of concepts in categorization from a psychological perspective. At the beginning of their study, the authors discuss their reasons for dealing with concepts corresponding to animals, plants, etc. rather than with those corresponding to geometric objects like a square or to abstract entities like love or brilliance. They write the following in connection with the latter two: "Similarly, had we chosen as our domain abstract concepts, such as 'love' or 'brilliance', we might have prejudiced the case against the classical view; for no mathematician or metaphysician has come even close to constructing a classical-view description of such concepts. Thus natural objects and human artifacts offer an in-between case between concepts that any schoolboy can define and concepts that no scholar can grapple with". (Smith and Medin 1981:5) The point that I would like to make here simply is that, contrary to these claims, the lexical approach that is employed in this study enables us to come up with a fairly clear definition of the concept of LOVE. In this chapter I will attempt to show that through the detailed examination of the conventionalized language about love a great deal can be uncovered about the constituents and structure of the concept of ROMANTIC LOVE.

69 62 METAPHORS OF ANGER, PRIDE, AND LOVE In addition to the conceptual tools utilized so far (metaphors and metomymies), some further aspects of concepts will also be discussed. In particular, I will look in some detail at what have been called 'related concepts'. In the chapter on anger, it was implicitly assumed that DISPLEASURE is a concept inherent in ANGER. In the chapter on pride, explicit mention was made of concepts like JOY and SATISFACTION, which characterize the concept of PRIDE. A large portion of the present chapter will be concerned with the issue of inherent concepts and concepts that are otherwise related to ROMANTIC LOVE. Finally, another aspect of love that I wish to discuss has to do with the issue of how many prototypical models are necessary to characterize a concept in its entirety. Lakoff (1987) has shown that there are several kinds of prototype. In the case of love, I will try to show that we need at least a typical model and an ideal model if we wish to claim that at least a relatively full description of the concept has been given. Some of the implications of this distinction for theories of conceptual structure will be discussed in the next chapter. A more complete description of the concept of ROMANTIC LOVE can be found in Kövecses (n.d.) The central metaphor Amongst the various ways of conceptualizing love the model according to which love is a UNITY OF TWO COMPLEMENTARY PARTS is perhaps central. LOVE IS A UNITY (OF TWO COMPLEMENTARY PARTS) We were made for each other. We are one. She is my better half. Theirs is a perfect match. We function as a unit. They are inseparable. What makes the UNITY metaphor a model of love that has held for more than two thousand years? Why is it that the metaphor has pervaded our thinking so much that we hardly notice that it is a metaphor at all? There seem to be essentially two reasons for the naturalness and obviousness of this metaphor. One is that the metaphor entails several metaphorical consequences which allow us to understand a number of our love experiences, the understanding of which would be difficult without the UNITY metaphor.

70 THE CONCEPT OF ROMANTIC LOVE 63 Another reason is that we see a similarity of sorts between certain love experiences and the unity of two complementary physical, chemical, etc. parts, and because of this the linguistic examples of the metaphor appear natural and obvious to us. As a result the language describing love does not strike us as metaphorical at all. The UNITY metaphor suggests perfect harmony, an idyllic state. What gives rise to this is that by virtue of their perfect fit, or match the two parts form an ideal unity in which the two parts maximally complement each other. In the biological version of unity, the two parts live in symbiosis, one part being dependent upon the other. One part is incomplete and cannot really function without the other part. In a love relationship this is experienced as the lover being only a half, the other half being made up by the beloved. It is important to notice however that this view of love is mainly a result of the metaphor and it is not something that we should take as inevitable. There is nothing in the view that true love is perfect harmony that could not be otherwise. We see and experience true love as perfect harmony because we have made it so by virtue of the concept of UNITY OF TWO COMPLEMEN TARY PARTS. The UNITY metaphor has other consequences as well. Once we conceive of love as the unity of two parts, then we can also conceive of love as the bond, or attachment, between the two parts. That we indeed think of love in this way is shown by the following examples: She has an attachment to him. There are romantic ties between them. There is something between them. The words attachment, ties and bond correspond to love. These are the things that hold the two persons together. And the word something also indicates the presence of such a bond or tie. Bonds or ties can be strong or weak. We make use of this knowledge about physical things in the understanding of the stability aspect of our love relationships. Thus we speak of indissoluble bonds and unbreakable ties between two people. But to speak of love in this way reflects a deeply embedded conceptual metaphor in this culture. According to this metaphor, the STABILITY OF LOVE is understood in terms of PHYSICAL, CHEMICAL, etc. STABILITY. That is, the conceptual metaphor that would account for these examples would be the following: THE STABILITY OF ALOVE RELATIONSHIP IS PHYSICAL (CHEM ICAL, etc.) STABILITY.

71 64 METAPHORS OF ANGER, PRIDE, AND LOVE The UNITY metaphor, or rather its metaphorical consequences enable us to capture some further aspects of the concept LOVE. The unity of the two parts 'can be dissolved'. Underlying sentences like Sally and John broke up there is the knowledge that the cohesive force between two physical objects may weaken as time goes by and the unity may cease to exist altogether. This is the knowledge that we seem to carry over to the domain of love. The 'breaking up' in the source domain (UNITY OF TWO COMPLEMEN TARY PARTS) corresponds to the end of the love relationship in the target domain (LOVE). Also, the two physical objects can form a unity, they can become a single unit. Corresponding to this in our love experiences is the beginning of a love relationship. An expression which is based on this idea is get together and it occurs in sentences like They got together last year. The purpose of this detailed (though not complete) analysis of metaphorical consequences is to try to give an account of why we feel that the UNITY metaphor is so particularly appropriate to our thinking (and hence talking) about love. I believe that one source of the metaphor's appropriateness is that it captures a large number of our love experiences. Thus we see the success of a particular metaphor mainly as a result of the extent to which the metaphorical entailments of a source domain (like UNITY OF TWO COMPLEMENTARY PARTS) can be utilized in the target domain (like LOVE). As we have just seen, we carry over a great deal of our knowledge about the unity of two physical objects to the domain of love and this enables us to comprehend, or even create, such aspects of love as the perfect harmony associated with love, the stability of love, the beginning and end of a love relationship, etc. However, this is not the complete picture. The account of the intuitive appeal of love as the unity of two physical objects has to be supplemented by another possible explanation. The core of our argument will be a phenomenon which most people consider the most characteristic feature of love, namely physical closeness. The distinguished role of physical closeness in love is illustrated by the abundance of linguistic expressions that have to do with physical closeness: I want to be with you all my life. We're always together. He follows her everywhere. They walked along the Danube holding hands. "I wanna hold your hand". Please don't ever let me go.

72 THE CONCEPT OF ROMANTIC LOVE 65 I want to hold you in my arms forever. You are so far away, I wish you were here. I hasten to remark that these linguistic expressions are not metaphorical. They simply describe something that we commonly observe; namely, that the lovers can often be seen together, and want to be in each other's company. Yet, on hearing these expressions one of the things we can reasonably infer is that the people mentioned are, in all probability, in love. But how do we know this when all we are literally told is that the two persons are, or want to be, close to each other physically? Well, this is a reasonable guess because we know that people in love spend, or want to spend, a lot of time together. In other words reference to the physical closeness of the two people may serve to indicate for us that the two people are in love. Linguistic expressions of this kind are called metonymies. When we use a linguistic expression metonymically we make reference to X which is related in some way to Y and mean or imply Y. With respect to love this means that reference to physical closeness (X) may indicate love (Y). This is possible because there seems to be a causal relationship between the concept of LOVE and the concept of PHYSICAL CLOSENESS. Cases like this can be called, on the analogy of the term conceptual metaphor, conceptual metonymies. A more precise formulation of this causal relationship in the domain of emotions would be to say that according to our folk model of how the emotions work, a given emotion (like love) is accompanied by, or manifests itself in certain behavioral reactions (like physical closeness). It is often the case with the emotions that the most typical behavioral reactions corresponding to an emotion serve as the basis for the central metaphor of the emotion (or rather, emotional concept). Thus it seems likely that it is the behavioral reaction PHYSICAL CLOSENESS that forms the experiential basis of the LOVE IS A UNITY conceptual metaphor, which plays a central role in the structuring of the concept of LOVE. This may be another explanation of the intuitively felt appropriateness of the UNITY metaphor. The experiential basis of the metaphor is provided by the common everyday observation that the lovers tend to be together as much as they can. Thus the concept of PHYS ICAL CLOSENESS seems to be present in both the metonymy (the physical closeness of the two people) and the metaphor (the physical closeness of the two physical objects that form the unity). It is important to notice however that the concept of PHYSICAL CLOSENESS in the metonymy is very different from PHYSICAL CLOSENESS in the metaphor. The physical closeness of the metonymy is to be understood literally between two people, whereas

73 66 METAPHORS OF ANGER, PRIDE, AND LOVE the physical closeness of the metaphor is to be understood metaphorically between two physical objects in unity. Nevertheless, it is this shared concept of PHYSICAL CLOSENESS that makes the UNITY metaphor a maximally motivated metaphor for us. The UNITY metaphor is felt to be appropriate because a feature associated with love (PHYSICAL CLOSENESS) is also found in metaphorical form in the (UNITY) metaphor. So far I have been concerned with trying to give an account of the central role and naturalness of the UNITY metaphor. I have said that the naturalness of the metaphor is due to the fact that the UNITY metaphor has a very strong experiential basis in one of the most typical behavioral reactions associated with love, PHYSICAL CLOSENESS. The account of the central role of this metaphor is based on the metaphorical entailments of the metaphor. It has been shown that a large number, and perhaps also the most significant, of our love experiences can be captured by the network of metaphorical entailments associated with love as a unity of two complementary parts. However, I have not yet brought to light all of these important metaphorical consequences. So let us explore these a little further. In conversations about love we often hear sentences like the following: I just can't go on without her, She will always remain my one and only love, I'll never find a love like that, and This love will last forever. As we shall shortly see, these sentences represent different but closely related attitudes towards love. They can be viewed as further consequences of the UNITY metaphor. Their examination is all the more important since these attitudes seem to have a great deal of influence on the way we experience and evaluate our love relationships. A consequence of the conception of love as the UNITY OF TWO COM PLEMENTARY PARTS is that we see love as some kind of need. This attitude is formulated in the sentence / can't go on without her. The view of love as a need arises from the metaphor because if I think of love as a unity of two complementary parts, then one part alone will be viewed as dysfunctional, unable to work normally until it is completed by its complementary other part. One part is seen as incomplete, indeed a 'half as in my other half and my better half. Another consequence of the UNITY metaphor is the notion of true love. We wait for Mr. or Mrs. Right or the one and only love to come along. This also follows in part from what we know about the UNITY OF TWO COM PLEMENTARY PARTS. If we accept that a part can only be completed by

74 THE CONCEPT OF ROMANTIC LOVE 67 one and only one other part that matches it perfectly, then it follows that there is only a single perfect match for everyone. That is, there is one and only one true love. This way of thinking shows up, for example, in the sentence He is a perfect match for me. That is, true love corresponds to the perfect match. The notion of true love brings a false dichotomy into our thinking. Namely, it maintains that there is a single true love and there are one or more loves that are not true. Loves that are not true are interchangeable and replaceable, while one's true love naturally isn't. Thus we have come to the idea of the irreplaceability of true love. However, irreplaceability can also be derived from the UNITY metaphor. For if there is only one part that matches the other part perfectly, then this part is irreplaceable and non-interchangeable. What this implies for love is that true love is irreplaceable. It is of course impossible for love to last forever and yet we often come across and catch ourselves saying sentences like I'll love you forever, Our love will last forever, My love will never die, etc. In comparison, Only death can part us seems to represent a more realistic version. The main issue is: What does the idea that love lasts forever have to do with the UNITY metaphor? I would be inclined to think that some of the metaphorical entailments of the metaphor jointly produce this idea. Once we see love as a need and think that there is only one true love which is irreplaceable, then the desire that love should last forever becomes perfectly understandable The object of love The object of love is conceptualized by means of a variety of conceptual metaphors. The study of these metaphors is important because the various metaphors bring to light other emotions and attitudes that, in the idealized version of love, we find inseparable from it. Let us begin with the metaphor that seems to be maximally coherent with the idea that love is a need. One of our most important needs is food. And so it is not surprising that we see the ob ject of love in terms of appetizing food. THE OBJECT OF LOVE IS (APPETIZING) FOOD Hi, sweetheart. She's my sweet and sugar. Hi, sugar! Honey, you look great today! She's the cream in my coffee. Hello, sweetie-pie.

75 68 METAPHORS OF ANGER, PRIDE, AND LOVE The fact that we conceptualize the OBJECT OF LOVE as APPETIZING FOOD does not only link love with needs but also with liking and sexual desire. How is this possible? Let us take the concept of LIKING first. We do not eat appetizing food, especially sweet things, only to satisfy hunger, that is, to satisfy a need. We eat it because we enjoy it, like it. For something to be pleasant to the taste is for us to like it. But to be pleasant to the taste is only one kind of liking. Another kind of liking is when something is pleasant to the sight. And what is pleasant to the sight is usually something that we find beautiful. This gives us perhaps the most pervasive belief about love: THE OBJECT OF LOVE IS BEAUTIFUL Let's go, beautiful. Hi, cutie! Well, gorgeous? Shall we go, angel-face? It is important to note that the object of love is always considered beautiful. For it is always the lover who defines who is beautiful for him or her. The object of love may be the ugliest person in the world (for others), and yet the lover will see him/her as beautiful. In other words, once we are in love, we do not seem to be very choosy about whom we call my beautiful or cutie. And if we think twice about whom we call my beautiful we are so much less in love. The concept of BEAUTY plays a very important role in the conceptualization of love. It is important because it helps us understand more about the nature of the concept of LIKING, and secondly, because it seems to form the basis of an entire system of metaphors which, as we shall see in a later section, will turn out to be one of the decisive factors in the way we comprehend, experience and live our love relationships. Consider the following metaphor for BEAUTY: BEAUTY IS A FORCE (PHYSICAL and PSYCHOLOGICAL); LIKING IS A REACTION TO THAT FORCE She bowled me over. Who's that attractive man over there? She's a dazzling beauty. I was hypnotized by her beauty. What a bombshell! I was knocked off my feet. She's enchanting. Look at all these glamour girls here! She's dressed to kill.

76 THE CONCEPT OF ROMANTIC LOVE 69 In this metaphor the force takes a variety of forms. It can be magnetic force {attractive), a mechanical force {bowl over, bombshell, knock off), a strong light {dazzling, glamour), and a magical force {hypnotize, enchanting). What is common to them is that beauty is viewed as a force which produces some effect on the self (the person who observes it). This effect, or reaction, corresponds, in this metaphor, to liking. There are two things to note about this particular way of conceptualizing liking. The first is that our reaction to beauty, as portrayed by the metaphor, is essentially passive. That is, according to the metaphor beauty is something that we experience in such a way that it is happening to us, that it affects us, without the self being actively involved as an agent in the process. The second observation is that our reaction to the force is such that as a result the self loses control (cf. bowl over, knock off, hypnotize, enchant, kill). As we shall see in section 6, not only beauty but also love is conceptualized in these ways. However, liking does not necessarily involve the liking of such physical attributes as beauty. If we ask someone How did you like her?, there can be more than one acceptable answer and not just She's beautiful. The answer may legitimately refer to some non-physical characteristic: Her personality, her nature, etc. Thus the concept of LIKING which is present in LOVE is broader in focus than the concept of LIKING as merely indicating physical appearance. But since we use the same concept (LIKING) to refer to, or focus on, several human characteristics, it can perhaps be asssumed that we have to do, in this case also, with the mechanism that seems to pervade our entire conceptual system: The mechanism on the basis of which we try to understand, or indeed construct, the non-physical aspects of human beings (like love) in terms of certain aspects of the physical world (like taste and vision). But what does sexual desire have to do with the APPETIZING FOOD metaphor? What establishes a connection between the two is the fact that the object of sexual desire is also conceptualized as (APPETIZING) FOOD. And at the same time we come to view sexual desire as HUNGER for (AP PETIZING) FOOD. Let us see some examples (taken from Lakoff 1987): THE OBJECT OF SEXUAL DESIRE IS (APPETIZING) FOOD; SEXUAL DESIRE IS HUNGER She had him drooling. He's sex-starved. You have a remarkable sexual appetite. You look luscious.

77 70 METAPHORS OF ANGER, PRIDE, AND LOVE She's quite a dish. She had kisses sweeter than wine. Let's see some cheesecake. Look at those buns! What a piece of meat! I hunger for your touch. He's a real hunk. I thirst for your kisses. Only some of these examples indicate more or less clearly that love and SEX UAL DESIRE are related concepts in our conceptual system. She had kisses sweeter than wine, I hunger for your touch and I thirst for your kisses may be used to indicate the presence of love in addition to sexual desire. The user of She's quite a dish or He's a real hunk does not necessarily want to say that he or she is in love with the person referred to. The person using these sentences is more likely to imply a liking of the other person and a willingness to have sexual intercourse with him or her. This would mean then that there is only a remote and indirect conceptual relationship between LOVE and SEXUAL DESIRE which is mediated by the concept of LIKING. However, as I will show later, there is a clear and direct relationship between love and sexual desire. The best indicators of the conceptual link between the two concepts in the SEXUALITY IS HUNGER metaphor are perhaps those that have to do with kissing and touching. I believe that this is not so by accident. The reference to love in these examples in a more straightforward way is possible because the examples draw heavily on a concept that is closely associated with LOVE: INTIMATE SEXUAL BEHAVIOR. The relationship between LOVE and INTIMATE SEXUAL BEHAVIOR is metonymical: INTIMATE SEXUAL BEHAVIOR STANDS FOR LOVE She showered him with kisses. It was a fond embrace. He caressed her gently. She held him to her bosom. He embraced her tenderly. The above linguistic expressions represent various forms of intimate sexual behavior. Thus we find instances of kissing, touching, and embracing. The expressions corresponding to them also indicate the presence of romantic love. That is, INTIMATE SEXUAL BEHAVIOR is metonymically related

78 THE CONCEPT OF ROMANTIC LOVE 71 to LOVE in the same sense that, as we have seen, PHYSICAL CLOSENESS is metonymically related to it. In other words, INTIMATE SEXUAL BEHAVIOR is one of the behavioral reactions associated with LOVE. But the concept of INTIMATE SEXUAL BEHAVIOR does not coincide with the concept SEX. The relationship between intimate sexual behavior and romantic love is established by the concept of INTIMACY that figures importantly in our concept of LOVE. Intimate sexual behavior is a manifestation of intimacy in the same way as physical closeness is a manifestation of intimacy. Sex as sexual intercourse does not presuppose intimacy, or, as a matter of fact, it does not even presuppose the presence of intimate sexual behavior. This can explain that a sentence like Sally went to bed with John does not imply that Sally is in love with John. This does not mean, however, that the relationship between sex and love does not hold from the other direction; that is, the concept of LOVE, at least in the ideal cases, presupposes the concept of SEX. However, this link between romantic love and sex is by no means independent of space and time. More than two thousand years ago Plato separated love and sex. Today this 'emasculated' version of love is known as 'platonic love'. But the metonymy INTIMATE SEXUAL BEHAVIOR STANDS FOR LOVE gives us occasion to mention some additional emotions accompanying love. It would be wrong to claim that one can only kiss, caress or embrace someone that one loves romantically. We can also kiss, caress and embrace our children, our parents, or our friends. In these cases we would think of what we do as expressions of our affection, or fondness. Thus the additional emotional concept that this metonymy brings to light is AFFEC TION/FONDNESS. If we love someone romantically, the chances are that we also feel affection for, or are fond of that person. And this is in turn the reason why sentences like She's fond of him can be used as euphemisms to indicate, or talk about, romantic love. Romantic love also presupposes kindness. And kindness is also presupposed by affection, or fondness. As some of the examples indicate (gently, tenderly), the attitude of the lover to the beloved involves gentleness, tenderness. TENDERNESS is one of the major ways in which KINDNESS is conceptualized (cf. soft-hearted, tender-hearted, etc.). Thus KINDNESS seems to be a concept closely associated with both ROMANTIC LOVE and AFFECTION. It is for this reason that the sentence You are not kind to me anymore can be uttered to convey a complaint that one is no longer loved romantically.

79 72 METAPHORS OF ANGER, PRIDE, AND LOVE A metaphor in terms of which we understand the LOVERS is the LOV ERS ARE DOVES metaphor. Consider the following examples: THE LOVERS ARE DOVES They sat there billing and cooing till after midnight It was all lovey-dovey. Look at those two lovebirds on the bench over there! Here come the love doves again. Their love nest has been discovered. This metaphor also emphasizes the presence of AFFECTION and KIND NESS in the conceptual network of (ROMANTIC) LOVE. Doves are viewed as symbols of peace, love, and gentleness. The gentleness of the doves is the gentleness of the lovers. Gentleness invites care, or caring. This gives us the further related concept: CARE. If we love each other, we want to take care of each other. The link between (ROMANTIC) LOVE and CARE/ CARING enables us to use expressions from the domain of CARE/CAR ING to indicate ROMANTIC LOVE: Sally cares a lot about John. A view of the object of love that is very different from what we have seen so far is portrayed by the metaphor which makes us see the beloved as some kind of deity. Let us see some examples: THE OBJECT OF LOVE IS A DEITY I adore you. She loves the air he breathes. He worships the ground she walks on. He put her on a pedestal. She devoted herself to him entirely. He fell on his knees before her. She prayed him not to leave her. She idolizes him. He is forever singing her praises. She has sacrificed her whole life for the love of her husband. The special significance of the DEITY metaphor lies in the fact that it brings to light several of the emotional concepts associated with love; in particular, the concepts of RESPECT, ADMIRATION, DEVOTION, SACRIFICE and ENTHUSIASM. There is an obvious link between the concepts of DEITY and RESPECT. (Note such expressions as the praise of God). It seems very likely that the original form of respect was the cultic worship of some deity. The

80 THE CONCEPT OF ROMANTIC LOVE 73 conceptualization of the object of love as a deity brings in respect by virtue of the fact that the object of respect is also conceptualized as a deity. A clear indication of this is that in the appropriate context some of the expressions in the above metaphor can be used to denote the emotion of respect. Such expressions are worship and idolize. To be sure, the meaning of these words contains an element of undeserved, or undue respect. And this also applies to love. When applied to love, these expressions indicate that the object of love receives more love than he or she deserves. It is important to note that the words worship and idolize are not used by the lovers to describe their own experiences but carry the judgement of an outside observer. RESPECT, maybe undue RESPECT, is then an integral part of our conception of love. But what is undue respect for an outsider is admiration for those on the inside, the admiration of the beloved. The outsider says that the lover is singing the beloved's praises, while the lover whispers that he or she adores the other and thinks that he or she is the most wonderful person in the world. What is condemned by the outsider is a very natural thing for the lover. Admiration and worship are just two faces of the same coin. What is important here is that ADMIRATION is a part of LOVE. If we look up the etymology of the word devote, we find that it goes back to a form that meant something like 'from a vow'. The reference is possibly to the vow that one makes when one dedicates, or devotees oneself to the service of God. The conceptual link between LOVE and DEVOTION is that the lovers see themselves as being devoted to each other in much the same way as a priest sees himself as being dedicated to the service of God. Thus the notion of devotion in the domain of love seems to enrich our idea of love with two elements. First, in our everyday conception of romantic love we tend to think of love as an emotion that can be directed at only one person. (Just as a priest can only serve one God). Second, people in love see themselves as 'servants' to the beloved. Most of us are familiar with phrases like I'll be your slave forever. The concepts of DEVOTION and SACRIFICE are difficult to distinguish. But whatever the differences between them, the important issue for our purposes is that our idea of love seems to incorporate the idea of (SELF-)SACRIFICE. If you are not willing to make sacrifices for your love, you are not considered as really being in love either. ADMIRATION, WORSHIP and DEVOTION all presuppose the idea of ENTHUSIASM. If I admire you, adore you and am devoted to you, I am also enthusiastic about you. I will have more to say about enthusiasm later. What makes it especially appropriate to mention enthusiasm in connection

81 74 METAPHORS OF ANGER, PRIDE, AND LOVE with the DEITY metaphor is that the word can be traced back to Greek and Latin words meaning 'possessed by the God'. Current usage of the concept suggests that there is indeed a close link between LOVE and ENTHUSIASM. If I have occasion to say that He enthuses over that pretty girl in his class, you would feel justified to conclude from my statement that he may be in love with her. Finally, I must mention yet another metaphor that we use for conceptualizing the object of love. THE OBJECT OF LOVE IS A VALUABLE OBJECT metaphor enriches the network of concepts related to love with two emotion concepts. THE OBJECT OF LOVE IS A VALUABLE OBJECT Hello, my precious! We have to leave now, my dear. You're my treasure! and on etymological grounds: Darling, give me a kiss! In the same way we treasure and want to keep possession of a valuable object, we are attached to the person we love. Thus the related concept this metaphor brings into focus is ATTACHMENT. That attachment is really a concept related to love is shown by the fact that we can use sentences like She's deeply attached to him to indicate love. The VALUABLE OBJECT metaphor makes it appropriate to mention another related concept. This is PRIDE. If I am in love, I am not only attached to the object of love but I am also proud of her because I value her highly. This is possible because, as we have seen, one of the typical sources of pride is a possessed valuable object Related concepts As we have seen so far, LIKING, SEX, SEXUAL DESIRE, RESPECT, DEVOTION, SELF-SACRIFICE, ADMIRATION, ENTHUSIASM, KINDNESS, AFFECTION, ATTACHMENT, CARING are all concepts that in the ideal case are closely related to the concept of ROMANTIC LOVE. In this section what I would like to do is to complete the list of emotions and emotional attitudes associated with love. Moreover, I will try to show that these related concepts are not a part of the conceptual network constituting LOVE to the same degree, and that there can also be

82 THE CONCEPT OF ROMANTIC LOVE 75 differences among the concepts in the nature of their relationships to LOVE. Finally, in this section I also wish to provide further linguistic evidence for the claim that the above concepts are indeed related to the concept LOVE. We have noted that physical closeness, intimate sexual behavior and sex are behavioral reactions associated with love. Both the concepts of INTIMATE SEXUAL BEHAVIOR and SEX involve the concept of PHYSICAL CLOSENESS. Both intimate sexual behavior and sex presuppose the physical closeness of the participants. The behavioral reaction of physical closeness plays a very significant role in the conceptualization of love. It appears to be the case that one of the most important concepts related to LOVE, INTIMACY, is nothing but a metaphorical counterpart of PHYSICAL CLOSENESS. I have already mentioned that the UNITY metaphor has in it the concept PHYSICAL CLOSENESS that has to be understood metaphorically (since the two parts can be found in each others physical vicinity). When we say things like They are very close or (on etymological grounds) They are very intimate, we in fact talk about this metaphorically understood notion of PHYSICAL CLOSENESS. That is, the claim is that INTIMACY is a concept that was metaphorically created on the analogy of the concept of PHYS ICAL CLOSENESS. Such synonyms of the word intimacy as psychological or spiritual closeness suggest that there must indeed be a conceptual metaphor according to which PSYCHOLOGICAL/SPIRITUAL CLOSE NESS is indeed understood, or perhaps has been created, on the analogy of PHYSICAL CLOSENESS. But so far I have only been concerned with the nature of the relationship between the concepts of PHYSICAL CLOSE NESS and INTIMACY and have not produced any linguistic evidence to show the relationship between LOVE and INTIMACY. The following may be considered such evidence. In certain contexts the sentence They are close to each other may also mean that the two people are in love. But perhaps more convincing examples are those that have to do with the lessening of love in a romantic relationship: The distance has grown between them or We have drifted apart lately. In the discussion of PHYSICAL CLOSENESS we have seen examples like the following: I want to be with you all my life, Dont ever let me go and You're so far away, I wish you were here. What is common to all of these examples is that, in a more or less straightforward way, they all contain the idea that lovers desire to be physically close to each other, they want to be together. Let us call this feeling of desiring to be together 'longing'. This gives us LONGING as another concept related to LOVE.

83 76 METAPHORS OF ANGER, PRIDE, AND LOVE Physical closeness serves as a model not only for intimacy. There is at least one other concept that is also understood in terms of PHYSICAL CLOSENESS. This is the concept FRIENDSHIP. We also say of our friends that they are close, and people we do not know very well are often referred to as distant acquaintances. Of course, the reason why I bother to mention these linguistic trivia is that FRIENDSHIP also belongs to the conceptual repertory of LOVE. Perhaps the clearest indicators of this conceptual relatedness are the words girlfriend and boyfriend when they are used to refer to one's sweetheart. It might be objected that these words only represent some euphemistic usage. I would argue, however, that what makes this euphemistic usage possible at all is that there exists a conceptual link between LOVE and FRIENDSHIP. Finally, mention must be made of a concept that is undoubtedly a part of the rich conceptual network associated with LOVE. This is the concept INTEREST. The presence of interest among the concepts related to love is shown by such examples as She is not interested in him anymore, which is almost equal to saying She is not in love with him anymore. The relatedness of LOVE to INTEREST could also be shown through the concept of ENTHUSIASM, which is a more intense form of INTEREST and thus presupposes its presence. So far I have treated the concepts associated with LOVE, such as LIK ING, SEXUAL DESIRE, RESPECT, ADMIRATION, ENTHUSIASM, INTIMACY, FRIENDSHIP, as if the relationship between LOVE and these other concepts were equally strong for each of the related concepts. But this is not so. The relationship between romantic love on the one hand and the various emotions and emotional attitudes on the other is strong in some cases and weak in others. But the question is how we can decide in which case it is strong and in which case it is weaker. We can employ the following procedure. As an illustration, let us take two concepts: POOR and UNHAPPY. In our everyday conceptual system these are related in such a way that we think of a poor man also as an unhappy man. What evidence is there to show this? Well, evidence can be found in the use of the word but. The word but is used appropriately (at least in one of its major uses) when it is followed by something contrary to our expectations. That is, if we say Peter is poor, but..., then the appropriate continuation of the sentence can only be he is happy. This is because in this case it is being happy that is contrary to our expectations. The sentence Peter is poor, but unhappy sounds odd because but is followed by a state of affairs which is not contrary to our expectations set up by 'being poor' ;

84 THE CONCEPT OF ROMANTIC LOVE 77 so the word but is not used appropriately. Of course, there may be people for whom being poor is not associated with being unhappy but with being happy. For these people the sentence Peter is poor but unhappy would be perfectly acceptable. At this point the link between being poor and being unhappy or being poor and being happy becomes an empirical issue. Although I have not carried out any sociological investigations into this conceptual link, I would say that we would find more people who associate being poor with being unhappy than people who associate being poor with being happy. And consequently, I would expect to come across more people who find Peter is poor but happy more acceptable than Peter is poor but unhappy. Thus we could say on the basis of such (probable) majority opinion that such concepts as ENTHUSIASM, AFFECTION, SACRIFICE, SEXUAL DESIRE, INTIMACY, etc. are related to our idea of ROMANTIC LOVE. Linguistic evidence seems to bear this out. Sentences like I am in love with her but I feel affection toward her, I am in love with her but I admire her, I am in love with her but I would make any sacrifice for her, We are in love but we are close to each other, and I am in love with her but I feel sexual desire for her etc. are not really acceptable sentences. If however we negate the emotion after but, we must get acceptable sentences in each case. This is true of some of the cases like I am in love with her, but I don't feel sexual desire for her, We are in love with each other, but we are not really friends, I am in love with her, but I don't respect her, I am in love with her, but I am not kind to her. In the case of some other emotions we would hesitate to consider the sentence as clearly acceptable (the question marks before the sentences indicate this uncertainty of judgement):? I am in love with her, but I don't admire her,? I am in love with her, but I would not make any sacrifice f or her,? I am in love with her, but I am not devoted to her,? I am in love with her, but I don't care about her. And again there are negated versions where one would expect more agreement. It is likely that most people would judge the following sentences unacceptable: I am in love with her, but I don't feel affection for her, I am in love with her, but I do not long for her, I am in love with her, but I am not interested in her, We are in love with each other, but we are not close to each other. Maybe in some of these cases other people's judgement would differ from mine. But this is not at issue now. What is really significant is that the sensitivity of these concepts to the but-test seems to show at least three degrees of variation. As the above tentative results indicate, we can imagine non-ideal cases of love which lack, say, respect or sexual desire or kindness

85 78 METAPHORS OF ANGER, PRIDE, AND LOVE or friendship. Secondly, I believe that my examples show that people tend to think that romantic love is somewhat less likely to occur without admiration or sacrifice, for example. Finally, the chances are that we would consider someone a fool if he said that he was in love with a girl but at the same time he did not feel affection for her or was not interested in her or did not long for her. What all of this suggests is that some emotion concepts are indeed related to LOVE in stronger ways (like AFFECTION, LONGING) than others (like KINDNESS, SEXUAL DESIRE). Furthermore, there seem to be cases between the two (like DEVOTION, SACRIFICE), which is an indication that the 'strong-weak link' is to be conceived of as a gradient rather than a rigid dichotomy. Of course, this does not mean that the emotion concepts that are linked with love in only looser ways do not form an integral part of our everyday idea of romantic love. Our conceptual network associated with the ideal cases of romantic love incorporates respect, kindness, sexual desire, caring, etc., just as it does affection, longing, enthusiasm, emotional closeness. In addition to the linguistic evidence cited so far, this is what the but-test also seems to suggest. Insofar as the but-test reveals anything real about the relationship of two concepts and the strength of the link between them, and insofar as I am in agreement with other people (or at least with a significantly large number of people) concerning the acceptability of the above sentences, we can say that the conceptual core of ROMANTIC LOVE is constituted by AFFECTION, ENTHUSIASM, INTEREST, INTIMACY, and LONGING. These are concepts that are most closely related to LOVE. Another way of putting this is to say that these concepts are inherent in LOVE. A definition of romantic love along these lines would be something like this: Romantic love is a feeling which is characterized by affection, enthusiasm, interest, longing and intimacy. This is the kind of definition which resembles definitions as given by dictionaries. Dictionaries try to give the necessary and sufficient conditions for the correct application of a word. For example, the Heritage Dictionary defines love in the following way: "An intense affectionate concern for another person". That is, this dictionary emphasizes, among other things, affection and concern. As we have seen, affection is also present in the definition that I have come up with. However, I have not found concern an inherent part of love. Perhaps concern can be equated with what I have termed caring, and in that case it would be a concept which is related to, though not inherent in, love. But all this creates a problem. Why doesn't the dictionary

86 THE CONCEPT OF ROMANTIC LOVE 79 define the concept of LOVE with the other concepts that I have found inherent in it (except for AFFECTION)? Several answers are possible. One is that my linguistic intuitions do not coincide with the intuitions of the editors of the dictionary. Another answer is that the but-test does not reveal what my examples would seem to indicate, since concern, or caring, has received a status in the dictionary different from the one I have given to it. And thirdly, we should not exclude the possibility that the dictionary may be wrong in that it does not faithfully represent the conceptual organization of love. However, our problems do not end here. If we look up the word love in other dictionaries, we find that the dictionaries are not in agreement as to which concepts can be used to define love. The Random House College Dictionary finds tenderness and affection important. Their definition is: "A profoundly tender, passionate affection for a person of the opposite sex". But Funk and Wagnalls stress devotion and affection. Their version goes like this: "A deep devotion or affection for another person". Now what are we to make of all this? It seems that I am not in agreement with Heritage as to the necessary ingredients of love and Heritage is not in agreement with Random House, and Random House is not in agreement with Funk and Wagnalls, etc. The only concept that we perhaps all take as relevant is AFFECTION. But of course no one would want to say that romantic love is exclusively constituted by affection, since this would amount to saying that the two are the same. Perhaps the best thing we can do is to admit that LOVE is a concept which it is difficult to define in terms of inherent features, and whose definition might depend on such factors as one's general view of the world or one's subculture or method of definition. In other words, we may not have defining features for love, or as a matter of fact, for the majority of our categories, which have the same validity forever and for everyone. If the but-test is a reliable method of arriving at what the defining features of a category are, then it can provide us with these features insofar as people agree in their acceptability judgments. So far we have looked at emotions and attitudes that are related to love: Liking, longing, intimacy, sexual desire, admiration, etc. Another emotion concept that is also related to love is HAPPINESS. But the relationship between love and happiness is different from the relationship between love and the above emotions. As we have seen, the concept of LOVE presupposes, or takes for granted the presence of LIKING, LONGING, INTI MACY, SEXUAL DESIRE, ADMIRATION, etc. Another way to express the same idea is to say that if I like, long for, am emotionally close to, feel sex-

87 80 METAPHORS OF ANGER, PRIDE, AND LOVE ual desire for, admire, etc. someone, then I am in love with that person. (Obviously, in the everyday ideal model of love some other conditions will also have to be met: For example, the lovers would have to be of different sexes. Clearly, then, I am describing the ideas of mainstream culture in this study.) However, the relationship between love and happiness seems to be different. Instead of saying that romantic love presupposes, or takes for granted happiness, we should rather say that if I am in love (and my love is returned), then I am happy. That is, in this case love results in happiness, (cf. This love has made me happy). To put the difference briefly: If I like, long for, am emotionally close to, etc. someone, then I am in love, and if I am in love, then I am happy. But this way of putting things blurs important differences in the meanings of the two constructions (the constructions being: 'If liking, longing for, etc. someone, then love' on the one hand and 'if love, then happiness' on the other). The presence of liking, longing, admiration, etc. allows us to describe someone (including ourselves) as being in love, whereas in the case of happiness being in love leads to, or brings about happiness. All this seems to suggest that while there is a causal relationship between the concepts of LOVE and HAPPINESS, there does not seem to be a similar relationship between LOVE on the one hand and LIKING, LONGING, ADMIRA TION, etc. on the other. Perhaps the relationship between love and these emotions can be best put by saying that in order for me to describe someone as being in love, these emotions must be present in the situation. It seems then that liking, longing, admiration, etc. specify and/or constitute the rules of the applicability of the concept of ROMANTIC LOVE. If they are present I can appropriately say of a person that he or she is in love. However, this should not be understood as implying that all of the concepts we have discussed so far must be present in order for me to correctly describe someone as being in love. We have already seen in the discussion of the but-test that the majority of the concepts associated with love need not each be present. I can be truly in love with someone even if I don't feel sexual desire for the person, provided that enough of the other concepts are present. That is, almost any one of these related emotions can be absent (except maybe affection, intimacy, etc., which yielded unacceptable sentences on the but-test. But even this is an open empirical issue.) Of course, I would still want to maintain that the ideal model of romantic love in our everyday thinking is very much characterized by the complete repertory of emotions I have mentioned in this and the previous chapters. There is at least one ('scientific') study in which an attempt is made to define love by means of what I have called related concepts (Newton-Smith

88 THE CONCEPT OF ROMANTIC LOVE ). In this study the following concepts, or "love-comprising relations" as the author calls them, are found to comprise the concept of LOVE: (1) A knows Β (or at least knows something of B) (2) A cares (is concerned) about Β A likes Β (3) A respects Β A is attracted to Β A feels affection for Β (4) A is committed to Β A wishes to see B's welfare promoted (Newton-Smith 1973: ) With the exception of (1) and (4), I have mentioned all of the other concepts. (1) is intended to be a precondition for love, and I will also incorporate it in the characterization of the everyday model of love in the form 'A comes along' and 'A finds B'. (4) can perhaps be accommodated by claiming that it roughly corresponds to what I have termed SELF-SACRIFICE. In the light of what has been done in this and the previous chapters, we can see two weaknesses in this approach. First, the conceptual features given appear to be a fairly randomly selected choice of concepts. For example, where are concepts like INTEREST, INTIMACY, and LONGING? And if RESPECT is included, why not include FRIENDSHIP, ADMIRATION, and DEVO TION? The second weakness that needs to be pointed out has to do with the inadequacy of any approach that tries to define a concept in terms of necessary and sufficient conditions alone. We have seen that different dictionaries find different concepts important in their definitions of love. This is an indication of the difficulty involved in the task of coming up with the necessary and sufficient conditions for the application of the concept LOVE. But let us suppose for a moment that this is possible. The concept defined only in terms of these features would be extremely impoverished, given what we know and can reveal about love on the basis of everyday language use. We have already discussed in detail the extent to which the UNITY metaphor and the various conceptualizations of the object of love contribute to the concept of ROMANTIC LOVE. Without these metaphors our understanding of the various aspects of the concept would virtually be reduced to zero. It is therefore important to move on and consider other aspects of love as these are captured by other metaphors and metonymies.

89 82 METAPHORS OF ANGER, PRIDE, AND LOVE 4.5. Intensity In this section I will be looking at the intensity aspect of love. INTEN SITY is an abstract concept that has evolved from more concrete and conceptually simpler concepts. Such simple concepts are DEPTH, QUANTITY, EFFECT, CLOSENESS, etc. We can conceive of INTENSITY as a scale which represents the amount of a substance, the degree of depth, the degree of an effect, the degree of closeness, etc. Thus INTENSITY seems to be a concept that is characterized by a scale indicating lower and higher degrees. As we shall see, love is an emotion whose intensity, at least in the prototypical cases, is always very high. The conceptual relationship between INTENSITY and DEPTH is easy to see. When we say of a person that he is deeply wrapped up in his thoughts what we usually mean is that the person is thinking intensely. Similarly, when we speak of deeply felt emotions, one of the things that we can mean by this is that one experiences one's emotions with intensity. One of the metaphors for LOVE (that I will discuss in the next section) is the LOVE IS A NATU RAL FORCE (FLOOD, etc.) metaphor. Examples of this metaphor are the sentences: She is deeply immersed in love and Waves of passion came over him. Both the waves coming over someone and immersion in something presuppose depth. Thus because of the conceptual link between INTENSITY and DEPTH, one of the implications of the sentences will be that the person in question has intensive love experiences. This is even more transparent is examples like She is deeply in love and Jimmy is up to his eyes in love, where the immersion indicates the intensity of the love experience. The intensity of love is often expressed by the amount of substance in a container. If there is a lot of substance in the container, then the intensity of love is high and if there is only a small amount, then it is low. The relationship between intensity and quantity is captured most clearly by the LOVE IS A FLUID IN A CONTAINER metaphor. This metaphor is primarily concerned with the issue of intensity. Let us see some examples for this conceptual metaphor: LOVE IS A FLUID IN A CONTAINER She was filled with love. Warm feelings welled up inside him. She overflowed with love. He poured out his affections on her. He was full of love for her. She couldn't hold in her love for him any longer.

90 THE CONCEPT OF ROMANTIC LOVE 83 In our conceptual system our body is often seen as a container for our emotions and the emotions as fluids inside the container (cf. He was boiling with anger, She was filled with jealousy, etc.). THE BODY IS A CONTAINER FOR THE EMOTIONS metaphor enables us to conceptualize INTENSITY as AMOUNT, or QUANTITY (INSIDE A CONTAINER). In love, this means that the more fluid there is in the container the greater is the intensity of love. Thus She couldn't hold in her love for him any longer indicates more love, and therefore more intensive love, then Warm feelings welled up inside him. This is because the metaphor makes use of a height scale measuring quantity, or amount. The height scale corresponds to the height of the container. It has two end points. The lower end point (the threshold on the intensity scale) corresponds to the bottom of the container and the upper end point (the limit point on the intensity scale) to the top of the container. These two end points define three degrees of the intensity of love. When the amount of love is above the threshold (love wells up from the bottom), love comes into existence; that is, the intensity is relatively low. When the amount of love reaches the limit (love fills the container), it indicates a high degree of love. And when the amount goes beyond the limit (love overflows), it indicates the highest degree. The conceptualization of intensity as quantity is not limited to the metapor LOVE IS A FLUID IN A CONTAINER, which is based on the more general metaphor THE BODY IS A CONTAINER FOR THE EMO TIONS. In addition to the body, the heart can also serve as a container for this purpose. This is exemplified by expressions like these: She filled my heart with love, I love you with all my heart, etc. These examples come from the metaphor LOVE IS IN THE HEART. What makes this metaphor particularly appropriate is the fact that the heart is a container with blood (a fluid) pumping through it. Or, indeed, it may well be that the HEART metaphor is in a sense more basic than the BODY AS A CONTAINER metaphor. It seems more basic in the sense that, as we shall see, the physiological effect of increased heart rate is one of the most important bodily responses associated with love, and also with many other emotions. This would then seem to account for the fact that we have conceptualized our emotions, including love, in terms of the heart and thus consider the heart as the seat of the emotions. Before we go on to the major way of conceptualizing love's intensity, which is heat, let us see some minor ways that we use for this purpose; namely, closeness and effect. We have already seen in the discussion of the

91 84 METAPHORS OF ANGER, PRIDE, AND LOVE UNITY metaphor and the PHYSICAL CLOSENESS behavioral reaction that the concept of CLOSENESS plays an important part in the way we conceptualize love. What we need to point out in this connection is that we seem to make use of a closeness scale too. Thus we often use sentences like You seem to be so distant now and We are getting real close to each other, where the expressions distant and close indicate the two end points of the closeness scale. When two people are close to each other on this scale, their love is intense. (They can even be thick with each other). And when they occupy the end points, there is no love at all. Thus the closeness scale appears to be a fitting device in terms of which a number of variations in the dynamics of human relationships can be interpreted. Another concept in terms of which love's intensity is understood is EFFECT. Intensity is often captured by the impact something has on us. So in the same way as we can describe the intensity of some noise as piercing, we can describe the intensity of somebody's love with such sentences as He was blinded by love. The more extreme and deadly the effect is, the higher is the intensity (cf. He was dead set on it, I am dead tired, etc.). I will say more about examples like blinded by love towards the end of this section. However, the intensity of love, and several other emotions, is for the most part conceptualized in terms of the concept HEAT. HEAT manifests itself primarily in the metaphor LOVE IS FIRE. Of course, intensity is not the only aspect of love that this metaphor brings into focus. But let us see some examples: LOVE IS FIRE My heart's on fire. He was burning with love. The old-time fire is gone. She set my heart on fire. There were sparks. She is his latest flame. The fire slowly went out. That kindled love in his heart. I don't want to get burned again. He was consumed by love. I just melted when she looked at me. She carries a torch for him. This metaphor highlights the intensity of love (fire, flame, sparks, consume),

92 THE CONCEPT OF ROMANTIC LOVE 85 the existence of love (on fire), the coming into and going out of existence of love (kindle, go out), its duration (flame), the cause of love (kindle, set on fire), the frustration caused by love (get burned), and how it can render a person unable to function normally (consumed). Before we discuss intensity, let us see the correspondences between the source and target domains that constitute the LOVE IS FIRE conceptual metaphor. Source: FIRE Target: LOVE - the fire corresponds to love - the thing burning is the person in love - the cause of the fire is the cause of love - being burned by the fire is the frustration caused by love - the burning of the fire is the existence of love - the intensity of the fire is the intensity of love - the inability of the thing burning to function normally is the ability of the person in love to function normally Some of these correspondences are possible because we seem to carry over our knowledge of the source domain to the target domain. For example, we know that a thing burned by fire is unable to function normally. Now when we carry this over to love we get the (ontological) correspondence according to which the inability of the thing burned to function normally and the inability of a person in love to function normally are the same. That is, just as the thing consumed by fire is unable to function normally, so a person in love is unable to function normally. A further bit of knowledge we have of fires is that they can burn us and can cause pain. If we carry this over to love, what we get is that love can also burn us and can cause pain. It is this metaphorical (but nonetheless metaphorically real) pain that we feel at the time of a frustrating love experience. The conceptualization of EMOTIONAL PAIN as PHYS ICAL PAIN is again a very general conceptual metaphor in our conceptual system. Just as a fire can start, stop and exist, so we can conceptualize love as having a beginning, an end and being in existence. When the fire is kindled love begins. While the fire is burning, love exists. When it goes out, love ends. Something which is burning can suddenly burst into flames. Flames can last for a very short time and then return to their original intensity or go out completely. Some of our love experiences are such that we love a person very intensely for a short time and then our feelings abruptly come to an end. These cases are called infatuation, and this is when we talk about our latest

93 86 METAPHORS OF ANGER, PRIDE, AND LOVE flame. However, real fires are not such short-lived flames. They last and go on burning for a long time. We imagine real love to be a fire that burns a long time also. But a fire represents the top of a heat scale ranging from cold through warm to hot. And as a result of the correspondence between LOVE and FIRE, real love is always hot. In fact, it is so hot that it can even melt us. And a still higher degree of love is represented by the case when we are consumed by love (corresponding to fire). We have already seen that the intensity of love can be expressed by the effect of love on the subject of love. Taking this into account, we can see that the particular appropriateness of the words melt and consume to express love's intensity is motivated by two concepts that have to do with intensity: HEAT and EFFECT. In order for love to melt and consume us, which are effects, it needs the heat of fire. In these cases heat functions as cause which leads to an effect. Like the closeness scale, the heat scale provides us with a conceptual device in terms of which a great number of our love experiences seem to be constructed. These scales represent various degrees and shades of love or the lack of love. What is common to them, however, is the fact that real love is always seen as the highest grade on the scales: Depth as opposed to shallowness, closeness as opposed to distance, utmost effect as opposed to lack of effect, and intense heat as opposed to cold. The FIRE metaphor is perhaps the most important metaphor concerning the intensity of love. And since intensity plays a very marked role in the way we think of LOVE, the FIRE metaphor has a central role in the concept of LOVE as a whole. In the same way as the UNITY metaphor, the FIRE metaphor has several metaphorical consequences, which we use to bring to light various aspects of love. The FIRE metaphor also resembles the UNITY metaphor in that it is also grounded in experience in a clear way. As with other emotions (e.g. anger, lust, etc.), we find the physiological effect of body heat also in love. It is this bodily reaction that serves as an experiential basis for the FIRE metaphor. However, body heat is not the only physiological effect characterizing love. Our use of conventionalized language suggests that love is also characterized by the presence of other physiological effects in our conceptual model of love. Now let us see these effects as illustrated by the corresponding linguistic expressions: BODY HEAT I felt hot all over when I saw her.

94 THE CONCEPT OF ROMANTIC LOVE 87 You really have the hots for her, don't you? "I love you", she whispered in the heat of passion. It was a torrid love story. INCREASE IN HEART RATE She had palpitations. He's a heartthrob. His heart was throbbing with love. Her heart began to pound when she saw him. Increased body heat and/or heart function is assumed to cause redness in the face. BLUSHING There was a glow of love in her face. She blushed when she saw him. And a physiological effect which has already been mentioned: INTERFERENCE WITH ACCURATE PERCEPTION He was blinded by love. He saw nothing but her. I only have eyes for her. To summarize, the physiological effects that are assumed to accompany love are INCREASED BODY HEAT, INCREASED HEART RATE, BLUSHING, and INTERFERENCE WITH ACCURATE PERCEP TION. Of these, INCREASE IN BODY HEAT, and also indirectly INCREASE IN HEART RATE and BLUSHING, serve as the experiential base for the FIRE metaphor. The felt appropriateness of the metaphor arises as a result of these physiological effects. We have already seen in the discussion of the FIRE metaphor that a person in love is not able to function normally. Perhaps this idea becomes clearer if we look at the physiological effect of INTERFERENCE WITH ACCU RATE PERCEPTION. This provides us with more examples of how a person in love is indeed not able to function normally. (Of course, this applies only to the most intense forms of love). Similarly to anger and pride, we seem to have in our folk model of love two principles: AS LOVE INCREASES, ITS PHYSIOLOGICAL EFFECTS INCREASE and THERE IS A LIMIT BEYOND WHICH THE PHYSIOLOGICAL

95 88 METAPHORS OF ANGER, PRIDE, AND LOVE EFFECTS OF LOVE IMPAIR NORMAL FUNCTIONING. This second principle in fact tells us that if we are really in love we are unable to function normally. Inability to function normally in love involves a lack of control over love. This conceptualization of love's intensity brings about a serious conflict in our conception of love. Since the maximal degree associated with 'real' love involves a lack of control over love and since we have the responsibility to control our emotions, we have to make a choice. We have to decide whether we want love (together with a lack of control) or whether we want control over love (in which case, however, there is no room for 'real' love as it is defined by our culture). I will return to this and related issues in the next section and the section on typical love. To sum up, then, we have seen that to a large extent we conceptualize the intensity aspect of love in terms of HEAT, which is a concept realized in the FIRE metaphor. The heat scale consists of various degrees of heat, ranging from cold to hot. On this scale romantic love is conceptualized as an emotion with the highest degree of intensity. Furthermore, the combination of HEAT as cause with EFFECT yields an even higher degree for love's intensity. The case when love's heat drops to 'mere' warmth will be discussed in the section on typical love Passivity, lack of control, pleasantness We have seen that as a result of the UNITY metaphor the lovers desire to form a unity, and if they are successful, they live in the state of perfect harmony. The primary focus of the FIRE metaphor is that it gives us a sense of the intensity of love. In the discussion of the intensity aspect of love I have mentioned the examples She is deeply immersed in love and Waves of passion came over him which, through the concept of DEPTH, indicate the degree of love's intensity. But these two examples are not isolated linguistic expressions we use to talk about love, but instances of a system of metaphorical expressions which group around the concept of NATURAL FORCES. As we shall see, the LOVE IS A NATURAL FORCE metaphor addresses a number of issues that I have not discussed so far. But first let us see some further examples of this metaphor: LOVE IS A NATURAL FORCE (FLOOD, WIND, STORM, etc.) She swept me off my feet. Waves of passion came over him. She was carried away by love.

96 THE CONCEPT OF ROMANTIC LOVE 89 It was a whirlwind romance. She let herself go. I was swept away by love. We were riding the passions. It was a surging love. She was deeply immersed in love. We were engulfed by love. This metaphor brings into focus three aspects of love that have been mentioned already but which have not yet been examined in detail. They include passivity, lack of control, and pleasantness. In the metaphor above, the NATURAL FORCE embodies love itself and, consequently, the person carried away by the flood or wind corresponds to the person in love. The person who is carried away plays a passive role with respect to the force in the sense that he can't help being transported. The same idea appears in love in the following form: The person who falls in love can't help falling in love either. That is, according to the metaphor love is a force independent from us which can affect us without our active participation. The metaphor also contains the idea of lack of control. The person who is carried away by the flood or wind has no control over the situation, it is the force that has the upper hand. Corresponding to this in love is the idea that the person in love is unable to function normally. Passivity and lack of control are not easy to distinguish. We will need further conceptual metaphors to see clearly how these two notions differ from each other with respect to love. I have mentioned pleasantness as the third aspect captured by the NATURAL FORCE metaphor. This can be present in the metaphor because we think of being carried away by a force outside us as something pleasurable. Expressions like letting oneself go and riding the passions indicate that we readily take part in this. As a matter of fact, there are in our culture several pastimes and sporting activities which are based on the idea that, in them, force or energy outside the human body is used for change of location. Sailing, skiing, gliding, etc. can be mentioned as examples. Let us now consider these three aspects of love one by one. We can begin with passivity. Perhaps the clearest indication that we indeed conceive of ourselves as being passive in relation to love is that we talk about its coming: My true love will come along one day. That is, LOVE is thought of as a MOVING OBJECT. According to the LOVE IS A MOVING OBJECT metaphor love comes along, while we are standing still, waiting passively for it to happen to us. There are many other common linguistic expressions that also suggest our

97 90 METAPHORS OF ANGER, PRIDE, AND LOVE passive attitude towards love: we fall in love, we fall for another person, and we may even fall for him or her hook, line and sinker. Falling is something that happens to us and not something that we do. A similar idea comes through in one of the most widely used metaphors for love: LOVE IS A PHYSICAL FORCE. According to this metaphor a person in love is like a physical object that obeys a larger physical force. Here are some examples to illustrate this metaphor (from Lakoff and Johnson 1980): LOVE IS A PHYSICAL FORCE (MAGNETIC, CHEMICAL, GRAVITATIONAL, etc.) I could feel the electricity between us. There were sparks. I was magnetically drawn to her. They are uncontrollably attracted to each other. They gravitated to each other immediately. His whole life revolves around her. The atmosphere around them is always charged. There is incredible energy in their relationship. They lost their momentum. The aspect of passivity is reflected, in this metaphor, by expressions like magnetically drawn, attracted to, gravitate to, revolve around. What all of these examples indicate is that the object of the physical force (i.e. the person in love) obeys a stronger, larger physical force (i.e. love), it merely undergoes the effect of the force without having any active role in the interaction. The object of the force cannot do anything but obey the physical force. This way of conceptualizing love has the consequence that love is viewed as something for which we are not responsible and in which we have no choice, just as the object of the force cannot be responsible for what happens to it (since it is only an object affected by a larger force) and, for the same reason, has no alternative in its reactions to the force. We have seen that the PHYSICAL FORCE metaphor is at the root of the idea that the lover can't help being in love, since all he or she does is obey a force larger than him or her. The MAGIC metaphor, which is another metaphor for love, also contains this idea, but at the same time we can recognize in it the notion that love is something that has control over us. Lakoff and Johnson (1980) provide the following examples for this metaphor: LOVE IS MAGIC She cast her spell over me.

98 THE CONCEPT OF ROMANTIC LOVE 91 The magic is gone. I was spellbound. She had me hypnotized. He has me in a trance. I was entranced by him. I am charmed by her. She is bewitching. Similarly to the PHYSICAL FORCE metaphor, what we get here is that there is a psychological force (the magic) whose effect the person in love undergoes. In other words, he or she can't help being in the state he or she somehow got into. This is what we have called the passivity aspect of love. The idea of lack of control shows up perhaps most clearly in the examples spellbound, hypnotized, and in a trance. If I am spellbound, hypnotized or in a trance, I am unable to function normally. With the help of the metaphor, it is this aspect of magic that we seem to carry over to the domain of love. A person in love, we believe, is a person who lacks control over the state he or she happens to be in. A corollary of this is the belief that the more control we have over our emotion (in this case love), the less we are in love. Thus it seems that we find the idea that love goes together with our inability to function normally in both the NATURAL FORCE and MAGIC metaphors. This is what I have been referring to as lack of control in our everyday models of love. The claim that this is a part of our everyday model gains further substance from what was said about lack of control in connection with the FIRE metaphor. In that metaphor, lack of control took the form of one of the physiological effects of love: INTERFERENCE WITH ACCU RATE PERCEPTION. Lack of control, just like intensity, seems to have degrees. Love appears to be conceptualized as having the highest degree in this respect as well. Consider the examples of the following metaphor (from Lakoff and Johnson 1980): LOVE IS INSANITY I am crazy about her. She drives me out of my mind. He constantly raves about her. He's gone mad over her. I'm just wild about him. I'm insane about her. She's nuts about him.

99 92 METAPHORS OF ANGER, PRIDE, AND LOVE In this metaphor the insane person corresponds to the person in love. The person who makes us insane is the person that we are in love with. And the behavior of the insane person corresponds to that of the person in love. Insanity is the ultimate lack of control. The insane person is not responsible for his actions. If and when the INSANITY metaphor is applied to love, it follows that a person in love is not responsible for what he or she does either. In the discussion of the DEITY metaphor it was pointed out that enthusiasm is one of the concepts that are related to love. The INSANITY metaphor can give us further support for this claim. This is because, similarly to LOVE, ENTHUSIASM is a concept which is also comprehended in terms of INSANITY (cf. He's crazy about stamps). We can be enthusiastic about people in the same way as we can be about stamps, for example. The uncontrolled nature of the state of insanity is extended to both. But enthusiasm is just one characteristic of love. As we have already seen, many additional conditions have to be met for a relationship to be called true, or ideal love. Let us now briefly look at what I have called the pleasantness aspect of love. I have mentioned in connection with the NATURAL FORCE metaphor that according to this metaphor love is a pleasant journey. Its pleasantness comes from a special circumstance attaching to journeys of this kind; namely, there is an external source that provides the energy required for the change of location. We know, however, of many other sources of pleasure in this culture. Of these perhaps one of the most typical is what can be called 'rapture' (or more informally, a 'high'). It is not surprising then that we also conceive of love as rapture: LOVE IS A RAPTURE They were besotted with love. He is intoxicated with love. I am giddy with love. She was euphoric with love. I have been high on love for weeks. She is drunk with love. In this metaphor, corresponding to the person under the influence of the drug is the person who is in love, and to the drug, love itself. But what comes through most forcefully in this metaphor is the correspondence between the pleasantness of the rapture or high and the pleasantness of the love experience. The idea that love involves a lack of control is also present in the RAP-

100 THE CONCEPT OF ROMANTIC LOVE 93 TURE metaphor. A rapture, or high, similarly to insanity and being spellbound, is a state in which a person is not in possession of his or her faculties. In summary, we can say that LOVE is conceptualized as a FORCE which takes control over the person. The FORCE can be either physical (MAGNETIC, CHEMICAL, etc.) or psychological (MAGIC). In the ideal case there is ATTRACTION at the very beginning of the relationship (love at first sight), which quickly reaches the point on the intensity scale, and goes beyond this point, leading to the state of lack of control (I was spellbound, She had me hypnotized) The ideal model In the light of what has been said so far, we can now try to pull together the pieces and see what the ideal model of romantic love looks like. We have seen that on the basis of the UNITY metaphor the lovers view themselves as a unity composed of two complementary parts. This enables them to live the relationship as a state of perfect harmony. Furthermore, lovers will experience the relationship as one in which they need each other on almost a biological level. They also believe that their love is true love, that the beloved is irreplaceable, and that their love will last forever. Love is seen by the lovers as an unbreakable emotional bond or tie which guarantees the stability of the relationship. Ideal love is also characterized by a number of behavioral reactions such as physical closeness, intimate sexual behavior, and sex. Of these, physical closeness, intimate sexual behavior presuppose love, while sex does not. But, as was previously noted, love does not presuppose only the former two but also sex. With these behavioral reactions we have not yet exhausted the whole range of manifestations that are understood as accompanying romantic love in our ideal folk model. A further and very salient behavioral expression of love is what can be called loving visual behavior. Again, this can be gathered from our use of conventionalized language: Her eyes were beaming, Love showed in his eyes, She's starry-eyed, etc. Not only do the lovers exhibit certain behavioral reactions, but they also experience certain physiological effects. The ones we have arrived at by studying language include increase in body heat, increase in heart rate, blushing, and interference with accurate perception. Perhaps partly as a result of some of these physiological effects, the lovers experience their love as a state of pleasant giddiness, as some kind of rap-

101 94 METAPHORS OF ANGER, PRIDE, AND LOVE ture or high. Another possible source of this feeling may be sexual desire. It was mentioned that sexual desire and sex play a significant part in our contemporary conception of romantic love. Some of the physiological processes attaching to sexual desire may in part coincide with those physiological effects that characterize our folk model of love and thus may reinforce them. (One such shared effect can be body heat, as suggested by expressions like a hot-blooded woman). In addition to sexual desire, there are many other emotions and emotional attitudes that are related to love on a conceptual level. As has been seen, these include liking, respect, devotion, self-sacrifice, enthusiasm, admiration, kindness, affection, care, attachment, pride, intimacy, longing, friendship, interest and happiness. Moreover, it seems that these emotions and emotional attitudes are related to love with varying degrees of strength. Some of them appear to be merely associated with love, while others seem to be inherent in the concept of love itself. What is common to them, though, is that they are all present in the ideal case. A major characteristic of these concepts is that they in fact define, or constitute, the nature of the emotional relationship between the self and the beloved. Love, as we have seen, is a fire which burns everything, it is a flood which carries us away, it is a depth which swallows us up, it is a fluid which fills a container and then overflows. That is, love's intensity is always at a maximum high no matter which scale (like that of heat) we use to measure it. Our folk model also contains the principle that the more intense the love is, the more intense its physiological effects are. Thus love's intensity can increase to a point where it leads to interference with accurate perception. Interference with accurate perception is one of the ways that suggests that in love we are unable to function normally. And we have seen that there are several other physical or psychological states in which we are seriously impaired in terms of our ability to maintain control over what we do. The conceptualizations of love as insanity, magic, and rapture indicate in the clearest way that according to our folk model love is a state characterized by a lack of control. In ideal love, true love comes along, and all we have to do is wait patiently. In other words, love has an active part, while we are passive. If we are lucky, and it does come along, the other person begins to attract us immediately and irresistibly. The attraction (love) reaches the limit on the intensity scale at once and then goes beyond the limit. In ordinary language this is what we call love at first sight. There is a lack of control over love. The

102 THE CONCEPT OF ROMANTIC LOVE 95 intensity of love stays above the limit indefinitely, which in love means forever. However, the self's irresistible attraction to the other is seen not just as a result of an independent force outside us, but also as one which is capable of supplying us with extra energy. So far we have been concerned with the various aspects and features of love, but have not discussed a major feature in our conception of ideal love. This is the idea that love should be mutual. In the ideal version of romantic love, I love you and you love me to the same degree as I do. That is, love is not only mutual but also equal in degree. Where does the idea of love's mutuality and equality come from? For an answer we have to turn to yet another conceptual metaphor which is pervasive in our thinking: LOVE IS A VALUABLE COMMODITY (IN AN ECONOMIC EXCHANGE) I gave her all my love. I didn't get much in return. I've lost all my love for her. He received a lot of love from her. She rewarded his love by taking care of him. What am I getting out of this relationship anyway? I am putting more into this than you are. She's invested a lot in that relationship. This relationship isn't worth anything anymore. It should be clear from the examples above that a large part of the concept of LOVE is viewed as and comprehended in terms of commercial transactions. Thanks to the metaphor, love is seen as a situation in which love functions as a valuable commodity and the lovers as merchants exchanging goods. In a commercial transaction the giving of goods presupposes the taking of goods and also that the goods given must be of the same value as the goods received. Corresponding to the former in love is the idea that love is mutual and to the latter that the love given and received be of equal amount. The events, states, and properties I have listed above as characterizing the ideal case of romantic love do not of course occur all at the same time. There seems to be some temporal order in which the events, states and properties can be arranged. If we take into account the temporal dimension of the folk model, then ideal love looks something like the following: 1. True love comes along. The other attracts me irresistibly.

103 96 METAPHORS OF ANGER, PRIDE, AND LOVE The attraction reaches the limit point on the intensity scale at once. 2. The intensity of the attraction goes beyond the limit point. 3. I am in a state of lack of control. Love's intensity is maximal. I feel that my love gives me extra energy. I view myself and the other as forming a unity. I experience the relationship as a state of perfect harmony. I see love as something that guarantees the stability of the relationship. I believe that love is a need. that this love is my true love. that the object of love is irreplaceable. that love lasts forever. Love is mutual. I experience certain physiological effects: Increase in body heat, increase in heart rate, blushing, and interference with accurate perception. I exhibit certain behavioral reactions: Physical closeness, intimate sexual behavior, sex, loving visual behavior. I experience love as something pleasant. I define my attitude to the object of love through a number of emotions and emotional attitudes: Liking, sexual desire, respect, devotion, selfsacrifice, enthusiasm, admiration, kindness, affection, care, attachment, intimacy, pride, longing, friendship, and interest. I am happy. These features seem to be the ones in terms of which our language-based folk model of ideal love can be characterized. As can be seen, this model sets very high standards for anyone to be truly in love. But this is in the nature of ideal models. Ideal models are ideal because they represent goals that are virtually impossible to achieve. A slightly more realistic and practicable model of love is what we can call the 'typical model'. Let us now turn to this Towards the typical model The ideal model we saw in the previous section is not the only model of romantic love we have. Another model is what can be regarded as the typical model. Just as the ideal model has been arrived at by examining conventionalized linguistic expressions, so I will try to lay bare the typical model through the examination of language. There will be perhaps more people

104 THE CONCEPT OF ROMANTIC LOVE 97 who find the concept of love as characterized in this section more familiar and realistic than the one outlined in the previous section. To the extent that this is true, the typical model can be said to have more social reality than the ideal. Thus this is the kind of model one would expect sociologists to find in the investigation of what ordinary people mean by romantic love. The languagebased typical model is likely to be closer to the model people actually use in the course of their everyday lives. In contrast, the language-based ideal model is one, I surmise, which, in all probability, we reserve for our 'romantic moments'. Also, it seems more likely that the ideal model is something that we somehow grow out of, or mature out of, as we go through our love experiences. Thus, although there can be a great deal of individual variation in this respect, on the whole, I would expect the ideal, rather than the typical, model to be more characteristic of the way adolescents think about love. These are of course merely conjectures and thorough empirical research would be necessary to justify any of these claims. According to our model of ideal love, true love comes along and we passively fall in love. The typical model does not share the myth of passivity to the degree the ideal model would like to suggest. But the question arises: How do we know? For an answer let us turn to a further metaphor: LOVE IS A HIDDEN OBJECT He sought for love in the wrong places. His search for love wasn't successful. She's continually looking for true love. After many years of unsuccessful search, he eventually found love. You're lucky to have found her. The main message of the HIDDEN OBJECT metaphor is that love is not something that comes along but it is something that we must go and find. In this metaphor, the hidden object corresponds to love, or the object of love, which stands in a metonymic relation to it (i.e. love can mean either the emotion or the object of love). And the search for the hidden object corresponds to the search for love or the object of love. Thus according to the HIDDEN OBJECT metaphor it is not the self who is passive and who waits for the coming of love, but love is passive and we try to find it. But this is not the only difference between the ideal and typical models. We have seen that in the ideal model love's intensity increases to the point of losing control. This is also characteristic of the typical model. However, while in the ideal case this point is reached immediately and without any attempt to

105 98 METAPHORS OF ANGER, PRIDE, AND LOVE prevent it from happening, in the typical case we make an attempt to avoid getting into the state of lack of control. As we shall see, this accords maximally with some of our folk beliefs concerning the emotions. But let us now look at those linguistic expressions that suggest that we really try to avoid lack of control over love: LOVE IS AN OPPONENT He tried to fight off his feelings of love. Eventually he surrendered to his love. She was struggling with her feelings of love. He tried to suppress his feelings in vain. She was overcome by love. Love took complete control over him. She was seized by love. The OPPONENT metaphor is constituted by a number of correspondences between the source domain (OPPONENT) and the target domain (LOVE). These are the following: Source domain: OPPONENT Target domain: LOVE - the opponent is love - the fight against the opponent is an attempt to avoid loss of control over love - losing to the opponent is getting into a state of lack of control - winning is being successful in maintaining control over love - surrender is giving up the attempt to maintain control and accepting love's control over us Thus the OPPONENT metaphor's main focus seems to be the control related aspects of the concept; in particular, the attempt to avoid lack of control. At this point we are confronted with a very important question: Why is it so important for us to try to avoid a lack of control in love? Why do we try to fight off feelings of love and why do we struggle with love? The application of the OPPONENT metaphor would make sense and would be felt to be justified in the case of such 'negative' emotions as, for example, anger, but in the case of such paradigmatically 'positive' emotions as love it certainly calls for some explanation. According to the teachings of Western civilization, we have to keep under control such emotions as anger, desire, romantic love, etc. It seems to me that this idea stems from a particular conception, or view of the emotions in Western civilization. This view is based on metaphor and can be put in the following way: PASSIONS ARE BEASTS INSIDE A PERSON

106 THE CONCEPT OF ROMANTIC LOVE 99 (cf. You bring out the beast in me). A special case of this metaphor also applies to love: LOVE IS A CAPTIVE ANIMAL His love got out of hand. He unleashed his love. She gave way to her feelings of love. He couldn't hold back his love. She let go of her feelings. He lost his grip on his feelings of love. Her feelings of love broke loose. It appears that it is the PASSIONS ARE BEASTS INSIDE A PERSON metaphor which serves as a basis for the view that the emotions (very likely those which we consider as 'passions') constitute the 'animal' part of the human psyche. Possibly, it is this conception of the emotions which has led to the belief that we have to keep our emotions under control. Interestingly, this expectation is perhaps less severe with women. The reason for the belief that we have to control our emotions seems to be a metaphorical consequence of the metaphor itself. For if we cannot keep the animal inside us, it can cause harm to others and as a result to us as well. Therefore, we are responsible for keeping the animal inside, that is, controlling our emotions. The LOVE IS A CAPTIVE ANIMAL metaphor works a little differently. According to this metaphor, the animal is not necessarily inside us. Rather, it seems to be outside and we try to keep it back as it tries to get away from us. In this metaphor the animal trying to get away is love and the person trying to keep it back is the person who is in love. Corresponding to successfully preventing it from getting away is the successful avoidance of losing control, and failure to keep the animal back corresponds to a loss of control over love. To keep the animal back requires a lot of energy. Corresponding to this is the energy needed to maintain control over love. The CAPTIVE ANIMAL and the BEAST INSIDE A PERSON metaphors entail some very interesting consequences that we make use of in our thinking about love. Thus, for example, since the animal can cause harm to others and also to ourselves, it is not desirable for us to let the animal loose. What corresponds to this in love is that it is also undesirable for us to give up control over love. It is perhaps this metaphorical consequence which shows most clearly the conflict that exists between one's individual desires and the expectations imposed on the individual by society. The individual appears to

107 100 METAPHORS OF ANGER, PRIDE, AND LOVE want complete lack of control, whereas society wants him to behave in a socially responsible manner. The same idea can perhaps be put in the following way: The ideal model is the model of individual desires, while the typical model involves what is expected by and offered as a model by society. The same metaphors have, as we have seen, another metaphorical consequence as well. It is that a lot of energy is needed to keep the animal inside or stop it from getting away from us. Correspondingly, it also requires a lot of energy to prevent love from taking control of us. Those who cannot maintain control over their love are considered weak. Women are believed to be especially weak (i.e. unable to control their emotions), and the less they are capable of controlling their emotions, the more feminine they are taken to be. The stereotype of women being emotionally weak and men emotionally stronger is coherent with another stereotype according to which men are physically strong and women are physically weak. I am not interested here in the validity of these views. What is of much more importance for my purposes is to see how certain views in our everyday thinking about the world correlate with each other. It appears that a view of women which arises from a metaphor is maximally motivated by, or is based on another (non-metaphorical) view we have of women. The widespread belief that women are emotionally weak is supported by the belief that women are also physically weak. It is perhaps not wholly mistaken to see in this another example of how our emotional world is based on, or has been created on the analogy of how we see ourselves in less abstract realms. The related concepts that can be found in the ideal model are also present in the typical model. The list of these concepts has to be augmented by two additional concepts, jealousy and loyalty, or faithfulness, which do not have a place in the ideal model because they simply do not arise in the case of ideal love. The conceptual relatedness of LOVE and JEALOUSY gains linguistic expression in the proverb: "Love is never without jealousy" (The Penguin Dictionary of Proverbs, 1983). One possible explanation of this relationship can be that, as we have seen, we conceptualize the object of love as a valuable object and love as a valuable commodity. Jealousy, in this respect, can be defined as demanding, or claiming a person's love for oneself. LOYALTY and FAITHFULNESS are closely linked to JEALOUSY. Given the definition of jealousy above, we can say that loyalty is not giving oneself and one's love to a person outside the love relationship. Most people take loyalty to be a precondition for love. This view is often expressed in the slogan "If you are not faithful to a person, you don't really love him" or in

108 THE CONCEPT OF ROMANTIC LOVE 101 another proverb: "Faithfulness is a sister of love" (The Penguin Dictionary of Proverbs, 1983). We are not surprised, in fact we find it natural, if after a while love's intensity (its amount, heat, degree of closeness, etc.) goes down. There are a number of conventionalized linguistic expressions that are used to indicate this change. Thus we find expressions used for this purpose in the FIRE metaphor: The old-time fire is gone, The fiery passion died down and gave way to warm affection, etc. What both examples suggest is that the intensity of fire decreases and as a result, at least according to our folk knowledge of physics, the temperature goes down too. The scorching, all-consuming fire becomes a friendly fire that radiates warmth. This change corresponds to the process in love that we often put as love turns into affection. ROMANTIC LOVE is represented in our conceptual system as HEAT, while AFFECTION is conceptualized as WARMTH. We have already seen linguistic evidence for the claim that romantic love is viewed as hot. In addition to the FIRE metaphor, it was pointed out that, at least in our language-based folk model of love, love is understood as being accompanied by the physiological effect of body heat; that is, in a fiery or wild love affair our body temperature is assumed to become hot. However, in the case of affection, as the examples below indicate, we reach only warmth on the heat scale: There was a warm glow inside. It was warm affection. She feels warm all over when her husband comes home from work. They created a warm family home for themselves and their children. Hearth and home. She wasn't worth warming up to. These examples stand in a metonymic relationship with the concept of AFFECTION. This means that with the help of WARMTH we can make reference to AFFECTION. We have already seen a similar case in the discussion of the behavioral reaction PHYSICAL CLOSENESS. Now it appears that we find a general metonymic principle also in the case of physiological effects: THE PHYSIOLOGICAL EFFECTS OF AN EMOTION STAND FOR THE EMOTION. This principle holds both for love and affection. With respect to the physiological effect INCREASE IN BODY TEMPERATURE, love seems

109 102 METAPHORS OF ANGER, PRIDE, AND LOVE to go together with HEAT, and affection with WARMTH. Of course, to the extent that AFFECTION is a part of ROMANTIC LOVE, warmth can also be used to indicate romantic love. So far I have tried to argue that according to our language-based folk model of love, love's intensity goes down as a natural course of events. This is when we talk about wild passions turning into warm affection. It is important to stress that this change occurs of its own accord, and not as a result of something external (though, of course, this may also happen). This is why our reaction to this change is acceptance and resignation. Another point to emphasize is that this change does not in the least imply the end of love, but only a drop from heat to warmth. But of course love can cease altogether, and to conceptualize this we can also employ the heat scale. If love is viewed as hot and affection is associated with warmth, then it is only natural that the cessation of love is conceptualized as cold. Here are some examples: Their relationship has cooled recently. Why are you so cold to me? This relationship is getting lukewarm. This is not to say, however, that only HEAT can be used for comprehending this aspect of love. Another aspect that can be utilized for this purpose is CLOSENESS, as the example suggests: We are growing apart. In the examples above there are expressions like home, family, children, etc. These expressions point to the concept of marriage as it is conceived of in our culture. LOVE and MARRIAGE seem to be concepts that are related in interesting ways in the typical model of love. First of all, it appears that in the typical model love ends in marriage. Moreover, the conventionalized language we use to talk about this aspect of love indicates that love does not simply end in marriage but does so in a special way. The use of the words culminate and fulfill in the following sentences suggests that over and above ending in it, marriage is the highest, the most desired, point indeed the fulfillment of love: Love must culminate in marriage, Their love was fulfilled in marriage. In the sentences the use of the verbs culminate and fulfill can be regarded as completely conventionalized. Another piece of conventionalized language, a proverb, also seems to give credence to the view that the peak of love in everyday thinking is marriage: "Love is a flower which turns into fruit at marriage" (The Penguin Dictionary Proverbs, 1983). Curiously enough, however, we have to notice that there appears to be some positive correlation between the decrease in love's intensity and marriage. More specifically, it looks like love's intensity begins to go down at around the time when mar-

110 THE CONCEPT OF ROMANTIC LOVE 103 riage appears on the scene. In the examples we have seen, the concepts that have to do with marriage begin to show up when love's intensity drops from HOT to WARM. There is no knowing, at least using our methodology, exactly when this takes place; it can be the time of marriage, the end of the honeymoon, or maybe some time after this. What is important, however, is that marriage seems to give rise to the decrease of love's intensity. There is further linguistic evidence to show that this is really the case. Consider the sentence They have been married for ten years now, and yet they are still wildly in love with each other. The word yet could not be used in the sentence if it did not express a state of affairs which is contrary to our expectations. The expectation then is that a fiery and wild love affair is supposed to turn into quiet and warm affection after marriage The typical model In the previous section we have surveyed the most important linguistically accessible differences between the ideal and the typical models. We are now in a position to put together the pieces and describe the typical model in detail. It should by now be clear that this model has a much more clear-cut temporal structure than the ideal model. Taking this into account, our language-based typical model could be given as follows: 1. I search for true love. I find true love. The other attracts me irresistibly. The attraction soon reaches the limit on the intensity scale. 2. I try to keep control of my emotions (the attraction): That is, I make an effort to prevent love's intensity from going beyond the limit. 3. The effort is unsuccessful, I lose control over love: Love's intensity goes beyond the limit. 4. I am in a state of lack of control. Love's intensity is maximal. I feel that my love gives me extra energy. I view myself and the other as forming a unity. I experience the relationship as a state of perfect harmony. I see love as something that guarantees the stability of the relationship. I believe that love is a need. that this love is my true love. that the object of my love is irreplaceable. Love is mutual.

111 104 METAPHORS OF ANGER, PRIDE, AND LOVE I experience certain physiological effects: Increase in body heat, increase in heart rate, blushing, and interference with accurate perception. I exhibit certain behavioral reactions: Physical closeness, intimate sexual behavior, sex, and loving visual behavior. I experience love as something pleasant. I define my attitude to the object of love through a variety of emotions and emotional attitudes: liking, sexual desire, respect, devotion, selfsacrifice, enthusiasm, admiration, kindness, affection, care, attachment, intimacy, pride, longing, jealousy, faithfulness, friendship, and interest. I am happy. 5. Love is fulfilled in marriage. Love's intensity decreases, it goes below the limit: Love turns into affection. It seems then that the typical model retains many features of the ideal model, but at the same time there are also major differences. The two main points of divergence involve control and marriage. The typical model is supplemented by an attempt to control love and by love's culminating in marriage. In addition, our passivity in waiting for love is replaced by our search for love. In the typical model we don't find the (metaphorical) belief that love lasts forever either. It is replaced by the idea that wild romantic love turns into peaceful affection in a natural way. As a matter of fact, it is debatable to what extent, or whether at all, the other beliefs (those in stage 4) form a part of the typical model. Perhaps we may even go so far as to question the presence in the typical model of everything which results from the UNITY metaphor. It may well be that everything that this metaphor contributes to our idea of romantic love is only relevant to the ideal case. Unfortunately, at present it seems that our methodology does not help us decide these issues. Nevertheless, even if we leave out the products of the UNITY metaphor from the model above, the typical model will remain a sufficiently rich conceptualization of some of our love experiences. I feel that we cannot dismiss the possibility that there is in the typical model a feature that I have not talked about so far. Intuitively it perhaps makes sense to believe that in the typical case the beloved does not return the self's love at once or to the same degree as the self does. Maybe the typical case includes a situation in which one person conquers another, or one in

112 THE CONCEPT OF ROMANTIC LOVE 105 which the otherplays hard to get. Conquest corresponds to 'not falling in love with the other immediately' (this is why we have to conquer him/her) and hard to get to 'not falling in love to the same degree' (if the person fell in love to the same degree he or she would not play hard to get). In the case of conquest, the typical model would change in the following way: I fall in love, but the other person does not fall in love with me, and as a result I make attempts to have the other fall in love with me too. I succeed, and all the rest continues according to model as specified. This interpretation would be made possible by the expression conquer in particular and the LOVE IS WAR metaphor in general. This metaphor is discussed by Kövecses (to appear). The expression play hard to get, which comes from the metaphor LOVE IS A GAME (see also Kövecses, to appear), would necessitate a slightly different change. In this case I fall in love and the other person may also fall in love, but he or she does not show her emotions. Eventually, however, he or she may give in too and things go on as described. The possibility of these alternatives in the model raises a very significant methodological issue. Namely, the problem is that we have at our disposal linguistic material which may give rise to more than one distinct typical model (at least on the basis of my intuitions), and we have no linguistic test that we could use to decide which one of them is the typical model. Until such a linguistic test is found, the methodology suffers from a major weakness. Or else, it may turn out that we indeed have more than one typical model. Perhaps to decide on either of these issues, we will need the help of sociologists and social psychologists.

113

114 5. IMPLICATIONS FOR THEORIES OF CONCEPTUAL STRUC TURE 5.1. The structure of a concept The study of the three emotion concepts, ANGER, PRIDE and LOVE, suggests that at least some concepts have a structure with at least the following parts: a. a system of conceptual metonymies related to the concept in question b. a system of conceptual metaphors related to the concept in question a set of concepts linked with the concept in question d. a category of cognitive models, some of which, the prototypical ones, representing the concept in question as we commonly think of it Before I go into the various issues that emerge from this way of looking at the structure of a concept, let me briefly illustrate each of the four parts with examples from the foregoing analyses. Firstly, we have seen that related to the concepts of ANGER, PRIDE and LOVE is a number of conceptual metonymies: INCREASED BLOOD PRESSURE, ERECT POSTURE, and BODY HEAT, respectively. Secondly, there is a group of conceptual metaphors that seem to structure these concepts. For example, the concepts of FIRE, CAPTIVE ANIMAL, NATURAL FORCE, NUTRIENT, PHYS ICAL UNITY, VALUABLE COMMODITY, etc. each provide partial structuring for the concept of LOVE. Thirdly, in the case of LOVE again, we have found that a number of other concepts are related to LOVE. These include FRIENDSHIP, RESPECT, INTIMACY, AFFECTION, DESIRE, ENTHUSIASM, etc. And fourthly, we have seen that the concepts of ANGER, PRIDE and LOVE can each be partially characterized by one or (in the case of LOVE) two prototypical cognitive models and a large number of non-prototypical ones. Maybe the most important claim that I would like to make is that we need to take into account at least these elements if we wish to describe the

115 108 METAPHORS OF ANGER, PRIDE, AND LOVE structure of ANGER, PRIDE and LOVE in a way that represents the complexities of (psychological) reality. Most current theories of lexical meaning in linguistics see the structure of a concept as being definable by a set of sense components. These sense components are viewed as 'basic' concepts that recur in various parts of the lexicon, and the more often they show up as parts of other concepts the more universal and basic they are taken to be. We could, according to this analysis, build up the concept of ANGER, for example, out of such sense components as FEELING, INTENSE, DISPLEA SURE, yielding: 'a feeling of intense displeasure'. In the light of the three studies above, I believe that this way of analyzing the structure of a concept suffers from several weaknesses. Some of these weaknesses have been mentioned in the study of anger and love, and they repeat, implicitly or explicitly, criticisms by others. We are now, however, in a position to address the issues that were raised in chapter 1. These issues have to do with some additional weaknesses of componential analysis and the need for considering reference as just as important a part of lexical meaning as sense. To begin with, it can be claimed that, against the background of conceptual structure as given in a, b, c, and d, the structure of a concept as defined by sense components is grossly oversimplified. Intuitively, one feels that a definition in terms of some basic and atomic concepts does not do justice to what we usually mean by and know of a concept. The metonymies, metaphors and related concepts lend some of our categories a much richer texture than what is captured by sense components. One could counter this by saying that the goals of the two approaches are different, and consequently the means used to achieve these goals are different; and that componential analysis serves the purposes to which it is put in a satisfactory way. But this does not seem to be true. It appears that componential analysis is not a satisfactory way of handling some fundamental semantic phenomena. In the next chapter I will try to show that it does not enable us to account for polysemy, collocation and lexical fields. The issue of oversimplicity will be raised again in another connection in this chapter. The second point is related to the first. One source of our impression that componential analysis oversimplifies matters is its insistence on the smallest possible number of sense components in the characterization of meaning. I have suggested that the structure of a concept like LOVE consists of conceptual metonymies, conceptual metaphors, related concepts, and a category of cognitive models which includes some prototypical models and a number of

116 IMPLICATIONS FOR THEORIES OF CONCEPTUAL STRUCTURE 109 nonprototypical ones. Of these four parts what resembles most the structure of a concept as it is conceived of on the basis of componential analysis is what has been called prototypical cognitive models. Indeed, one could say that the models of ANGER, PRIDE and LOVE are simply extended feature (componential) analyses of these concepts. To some extent this is true. These models do employ features as well. However, there are also some major differences. Perhaps the most conspicuous difference is in the number of features employed in the two approaches. Componential analysis attempts to capture meaning in terms of a minimal number of components, whereas I have used extensive but structured lists of features in the characterization of ANGER, PRIDE and LOVE. In other words, in my analysis the features cohere into more or less complete but not infinitely long folk models. A third point has to do with folk models. The analysis of sense in componential analysis can be seen as an attempt to arrive at those conceptual features which are objectively real; i.e. features which the objects, states or events possess in and of themselves, independently of human beings. These are then features in terms of which an objectively-oriented scientific methodology would like to characterize and define objects, states or events. There are two kinds of evidence in the present study that considerably weaken the claim that sense components should be objectively real. First, they are not objectively real in the sense that, as Lakoff and Johnson (1980) argue, they often arise via conceptual metaphors, and hence they are the products of how the human mind works. I will take up this issue in section 2 in this chapter. The second kind of evidence is that the features that characterize ANGER, PRIDE and LOVE do not come from 'scientific' theories but arise from how ordinary people in everyday life think of and interpret their experiences. In this work I have called conceptual models and interpretations of this kind folk models. We can take a concrete example to illustrate the point. We have seen in the case of LOVE, for example, that one of the ingredients of love in our folk model is that we experience it as something that happens to us; i.e., we have a passive attitude towards it. However, some scientific theories of love may inform us that this not what is really going on. Solomon, for example, tells us that in reality we create love and do not undergo it (Solomon 1982). That is to say, according to this scientific model what is really the case is that we have an active, rather than passive, attitude towards love. But most people do not seem to be worried about what features scientific models of love find important and keep falling (!) in love in accordance with their folk models. That is, they regard passivity as one of the features of love. On the

117 110 METAPHORS OF ANGER, PRIDE, AND LOVE whole, in a large number of life situations folk models seem to play a more decisive role than objectively real scientific characterizations. Fourthly, another weakness of concentrating on sense to the exclusion of reference is that we cannot account for the following dilemma: If we accept componential definitions as a way of representing the structure of concepts, then we should have only a very limited range of experiences corresponding to the concepts in question. That is, a given conceptual structure as represented by a minimal number of sense components should make available for us a very narrow range of fairly simple and undifferentiated experiences. Thus there appears to be a big gap, on componential analysis, between thought and experience. For example, with respect to anger we should only be able to experience the 'feeling of intense displeasure' and with respect to love, depending on which definition we choose, say, 'a profoundly tender, passionate affection for a person of the opposite sex'. But, as we have seen, there is much more going on in both anger and romantic love. The richness of what is involved in experiencing anger is given in the prototypical model of anger and, as we have seen, the concept of LOVE involves even more complexities. The dilemma arises in componential analysis because componential analysis overemphasizes and takes for granted a distinction between sense and reference, and claims that the proper domain of the study of lexical meaning is sense. The notion of sense corresponds to the notion of concept, as defined by a minimal number of sense components and reference is a relationship between a given aspect of the world and a corresponding concept which codes most of our experiences as regards this aspect of the world. Thus the idea of 'concept as sense' cannot ensure a fit between thought (concept) and experience. If, however, we conceive of a concept as the 'concept of reference', then the match between thought and experience can be captured. This relationship between concept and experience is stated succinctly by Winch (1958): "The concepts we have settle for us the form of the experience we have of the world". (Winch 1958: 15). The fit between conceptual structure and experiential structure (i.e. what we experience) as it is conceived of in this work is made possible by allowing all the four parts metonymy, metaphor, related concepts and cognitive models to play a role in the structure of concepts. As we have seen throughout, they contribute to a large extent to our understanding and experience of anger, pride and love. Without taking all of them into account, we could not explain the wide range of experiences we have of anger, pride and love. Since componential analysis has a place for these parts in the struc-

118 IMPLICATIONS FOR THEORIES OF CONCEPTUAL STRUCTURE 111 ture of concepts only to a very limited degree, it cannot match conceptual structure with experiential structure. So far I may have created the impression that the four parts in the structure of concepts like ANGER, PRIDE and LOVE are unrelated. This is not so. First of all, there is global interconnectedness among the parts in the sense that the metonymies, metaphors, related concepts can, in some form, all be found in the prototypical cognitive models. We have seen how the various conceptual metonymies, conceptual metaphors and related concepts I have discussed all map onto a part of the prototypical scenario, and how they jointly converge on that scenario. This enables us to show exactly how the various metonymies, metaphors and related concepts are related to one another, and how they function together to help characterize a single concept like ANGER, PRIDE or LOVE. This is something that Lakoff and Johnson (1980) were unable to do. Secondly, any two parts of a concept as characterized here are related to each other in interesting ways. Of these perhaps the most important kind of relationship can be found between conceptual metonymies and conceptual metaphors. Before I discuss this issue and the issue of the metaphorical nature of the concepts under investigation, let me briefly survey and summarize the main points that have emerged in connection with the prototypical models. The most obvious, and perhaps the least interesting, feature of the models is that they display the events and properties that they are composed of in a temporal order. This temporal dimension seems to be more elaborate in some cases than in others. The model which is the most elaborate in this respect is that of anger and what I have called typical love (both were found to comprise 5 stages). Pride and the model of ideal love show less structuring in terms of time. Obviously, not all concepts have a temporal dimension. The cognitive models of objects (entities) like CHAIR and GUN or substances (mass entities) like WATER and SAND or properties like RED and HON EST will not have a time aspect built into them. Although, for example, anger is not a mass entity like water, it is conceptualized as such. Consequently, as we have seen, it takes the grammar of mass nouns. The study of LOVE has shown that there can be more than one prototypical cognitive model for a concept. It has been found necessary to distinguish the ideal model from the typical one. The idea that prototype effects may arise as a result of various types of models for categories comes from Lakoff (1987). Lakoff distinguishes between social stereotypes, typical examples,

119 112 METAPHORS OF ANGER, PRIDE, AND LOVE ideals, paragons, etc. for categories. These are all cases that give rise to prototype effects. The importance of this idea for the purposes of this study is that we do not need to provide an indefinitely long list of features for categories like LOVE. Instead, we will have one model for love which, among other things, is characterized by, for example, the idea that love lasts for ever (the ideal model) and another model which, among other things, is characterized by the idea that wild passions turn into warm affection (the typical model). For componential analysts one of the rationales for breaking down the meaning of a category into a minimal number of sense components (i.e. just enough components to be able to distinguish one meaning from another) is that the other solution involves ending up with an indefinitely long list of features, which is clearly undesirable. This is, it is claimed, because there is no natural stopping point in listing the characteristic features of objects, states and events and thus this undertaking becomes an attempt to describe the entirety of the world of experience. As Leech (1982: 13) writes in this connection "... any characteristic of the referent... may contribute to the connotative meaning of the expression which denotes it". But as we have seen, the typical model of anger includes only so much, the typical model of love can similarly be characterized by a 'human-sized' list of features, and ideal love is also describable in terms of a relatively short, though not necessarily minimal, list of components. In other words, our folk models like typical examples, ideals, stereotypes, etc. delimit to a considerable degree the range and number of features that, for us, comprise a category. Thus we will not need an indefinitely long list of features for the characterization of a category and we will not need to jam all the features of a category, like those of ROMANTIC LOVE for example, into a single componential formula. Instead, we can have an ideal model and/or a typical model and/or a stereotype, etc., and each will have, in additioin to the possible overlaps, its own distinctive features. By opting for this solution we can arrive at cognitive models for categories which do not have unduly long lists of features and are not composed of a minimal number of features either but which reflect faithfully what people mean by these categories. Another feature of the prototypical cognitive models of anger, pride and love is that they seem to be essentially composed of a number of entities and a number of predicates. The entities for love include the self, the beloved and love itself, and for anger they include the self, the target, the wrongdoer, etc. Part of what makes one model a model of love rather than that of anger, and vice versa, lies in the differences in the number and kinds of entities that make

120 IMPLICATIONS FOR THEORIES OF CONCEPTUAL STRUCTURE 113 up particular models. In this respect, anger seems to be more elaborate than love. It allows the participants in the anger scenario to play more than one role. Prototypes often involve clusters of conditions and the prototypical anger scenario is no exception. The clustering can be seen explicitly in identity conditions such as: Victim = Self, Target = Wrongdoer, Offence = Immediate Cause, etc. When these identity conditions do not hold, we get nonprototypical cases. For example, with righteous indignation, the victim does not have to equal the self. In the case of an indirect cause, the offence does not equal the immediate cause. Usually the act of retribution and the disappearance of anger go together, but in the case of spontaneous cessation and insatiable anger, that is not the case. And in the Don't-get-mad-get-even case, angry behavior is avoided, and is therefore not identical to the act of retribution. Thus part of what makes a prototypical scenario prototypical is that it has a conflation of conditions which are not conflated in nonprototypical cases. The predicates in the prototypical cognitive models can be factored into two types: Properties and relations. Properties characterize entities and relations hold between the entities. Examples of relations and properties are: An offence 'displeases' the self, the offence 'causes' anger to come into existence, the intensity of anger, pride or love is 'high', the beloved is 'beautiful', I 'value' myself highly, I 'form a unity' with the beloved, etc. 'Displease', 'cause', 'value' and 'form a unity' exemplify relations, while 'beautiful' and 'high' illustrate properties. Predicates like these come closest to what we think of as sense components, or features, in componential analysis. This means that the kind of analysis that I have been suggesting in this study is not a complete breakaway from componential analysis. To be sure, there are overlaps between the two approaches, but as we have seen throughout, there are also major differences. A tentative third category of elements that make up cognitive models is what I would like to call cognitive attitudes. Their presence in the models is especially noticeable in the case of love. I 'believe' that love lasts forever, I 'view' myself and the beloved as forming a unity, I 'feel' that love gives me extra energy, I 'experience' the relationship as a state of perfect harmony, etc. The various entities and predicates seem to form the content of these cognitive attitudes. For example, the entities 'self and 'beloved' plus the relation 'form a unity with' serve as the content of the cognitive attitude 'view'. If cognitive attitudes like these are indeed a part of prototypical cognitive models,

121 114 METAPHORS OF ANGER, PRIDE. AND LOVE then we can have an even more refined way, or means, of specifying the structure of a concept. We could say that cognitive models consist of three elements: Cognitive attitudes, entities and predicates. Differences between various categories could then be seen as resulting from not only differences in the number and kind of entities and predicates but also from the kinds of cognitive attitudes characterizing the categories and from their particular configurations in the models. Related concepts appear in the prototypical models as ways of defining certain relationships betweeen the entities; for example, in the case of love, between the self and the beloved, in the case of pride, between the self and the source of pride. Thus, in this respect, they are similar to relations, and can perhaps be taken as belonging to the set of predicates. As we have seen in the discussion of love, related concepts can be of two kinds: Inherent concepts and associated concepts (with perhaps a middle ground between them). Inherent concepts are the ones which must be minimally present in order for us to use a category appropriately, and associated concepts are those which may be absent but are presupposed (in a loose sense of 'presuppose') in the prototypical cases by the concept in question. Inherent concepts also come close to what a 'componentialisť would consider sense components in terms of which a category can be defined. Some objections have been raised to the analysis of meaning in terms of inherent concepts alone in the chapter on love and in this chapter. However, the fact that I have objections to the idea that meaning is defineable in terms of inherent concepts alone should not be taken as implying that I do not see an important place for inherent concepts in the analysis of meaning. The examination of love has shown that these concepts are very much a part of the structure of categories. As we have seen for all the three concepts under study, the conceptual metonymies appear in the prototypical cognitive models as behavioral reactions and physiological effects. I suspect that these 'outward' manifestations of 'internal' states must form a crucial and integral part of the cognitive models associated with internal states in general. The reasons for this, as can be determined in the light of the three studies, are twofold. In the discussion of polysemy it will be seen that it is not possible to give an adequate account of polysemy without taking into account these behavioral reactions and physiological effects. The other reason is that conceptual metonymies often serve to make concepts what Lakoff and Johnson (1980) call 'embodied'. We have seen that the conceptual metonymy BODY HEAT STANDS FOR ANGER provides the experiential basis for the ANGER IS HOT FLUID IN

122 IMPLICATIONS FOR THEORIES OF CONCEPTUAL STRUCTURE 115 A CONTAINER and ANGER IS FIRE conceptual metaphors, the metonymy PHYSICAL CLOSENESS STANDS FOR LOVE for the LOVE IS A UNITY metaphor, ERECT POSTURE STANDS FOR PRIDE for the PROUD PERSON IS UP metaphor, etc. These conceptual metonymies give us understandings of abstract concepts in terms of direct physical experience. Since the metonymies provide experiential motivation for the metaphors and the metaphors map onto a particular cognitive model which specifies the corresponding concept, the concept, no matter how abstract, will be anchored to direct physical experience via the metonymies. That is to say, the concept will be embodied rather than just a disembodied abstraction and thus will be available for experiential, and not just intellectual, understanding. This argument requires us to give a more detailed account of how metaphors map onto a cognitive model than has been given so far Metaphorical aspects of concepts If one looks only at the ontology (the entities and predicates) of the prototypical scenarios associated with anger, pride and love, it is tempting to think - that the concepts of ANGER, PRIDE and LOVE exist and are understood independently of any metaphors. - that the ontology of anger, pride and love and the category of scenarios represent the literal meaning of anger, pride and love. - that metaphors do no more than provide ways of talking about the ontology of anger, pride and love. These ideas entail the following: - The elements of the anger, pride and love ontologies really, literally exist, independent of any metaphors. A brief examination of the ontology of anger, pride and love reveals that this is simply not the case. In the anger ontology, for example, anger exists as an independent entity, capable of exerting force and controlling a person. This is what Lakoff and Johnson (1980) refer to as an 'ontological metaphor'. In this case it would be the ANGER IS AN ENTITY metaphor. A person's anger does not really, literally exist as an independent entity, though we do comprehend it metaphorically as such. In the ontology, there is an intensity scale for anger, which is understood as being oriented UP, by virtue of the MORE IS UP metaphor. This is what Lakoff and Johnson call an Orientational metaphor'. The intensity scale has a limit associated with it another ontological metaphor. As has been noted, anger is understood as being capa-

123 116 METAPHORS OF ANGER, PRIDE, AND LOVE ble of exerting force and capable of taking control of a person. The FORCE and CONTROL here are also metaphorical, based on physical force and physical control. The anger ontology also borrows certain elements from the ontology of retributive justice: Offence and retribution, with their scales of intensity and the concept of balance. These are also metaphorical, with metaphorical BALANCE based on physical balance. In short, the anger ontology is largely constituted by metaphor. As we have seen, the concept of CONTROL can also be found in the typical model of love, where it emerges via the LOVE IS AN OPPONENT and LOVE IS A CAPTIVE ANIMAL metaphors. The concept of INTENSITY is also present in both pride and love, and is also based on physical intensity. VALUE is a predicate in the prototypical model of pride (in 'S values himself highly'). A person does not really, literally have a value, but by virtue of the PEOPLE ARE VALU ABLE COMMODITIES metaphor we have integrated this predicate into our thinking about people. And finally, to take another example from love, two people who are in love do not really, literally form a unity, but they do comprehend their relationship metaphorically as such. In sum, in most, or perhaps all, of these examples we seem to understand at least certain aspects of ANGER, PRIDE and LOVE in terms of the physical world or, at any rate, in terms of those aspects of the world that can be viewed as more physical than anger, pride or love is. Or, in another formulation that I will clarify shortly below, it could be said that some aspects of the concepts of ANGER, PRIDE and LOVE have been created on the analogy of certain physical aspects of the world. These formulations suggest that the ontologies corresponding to ANGER, PRIDE and LOVE are to a large extent constituted by metaphorically understood/created concepts like ENTITY, INTENSITY, LIMIT, FORCE, CONTROL, VALUE, UNITY, etc. These concepts, together with the concepts of ANGER, PRIDE and LOVE, form what can be termed constitutive metaphors, which, if spelled out fully, can be stated in the following ways: ANGER is understood as being, or is created to be, an ENTITY, ANGER and LOVE are understood as having, or are created to have, a CONTROL aspect; the PROUD PER SON is understood as having, or is created to have, a VALUE; etc. Let us now examine these constitutive metaphors. Their source domains ENTITY, INTENSITY, LIMIT, FORCE, CONTROL, VALUE, UNITY, etc. all seem to be superordinate concepts, that is, concepts that are fairly abstract. By contrast, the principal metaphors that map onto the anger, pride and love ontologies HOT FLUID, FIRE, INSANITY, BURDEN,

124 IMPLICATIONS FOR THEORIES OF CONCEPTUAL STRUCTURE 117 STRUGGLE, ECONOMIC VALUE, PHYSICL UNITY (like a BOND), etc. appear to be basic-level concepts that are linked more directly to experience, concepts that are information-rich and rich in conventional mental imagery. Let us call the metaphors that are based on such concepts 'basiclevel metaphors'. I would like to suggest that most of our understanding of anger, pride and love comes via these basic-level metaphors. For example, the HOT FLUID and FIRE metaphors give us an understanding of what kind of entity anger is and the STRUGGLE and OPPONENT metaphors for anger and love, respectively, give us a sense of what is involved in controlling anger and love. Without these metaphors, our understanding of anger, pride and love would be extremely impoverished, to say the least. One is tempted to ask which is more primary: The constitutive metaphors or the basic-level ones. My intuitions tell me that the basic-level ones are more primary. It seems to me that we could not see ANGER as having a CONTROL aspect if we had not conceptualized it in terms of OPPONENT IN A STRUGGLE, or we could not see ANGER and LOVE as having an INTENSITY aspect if we had not conceptualized them in terms of HOT FLUID, FIRE, INSANITY, FLOOD, etc., or we could not see LOVE as being a UNITY if we had not conceptualized it in terms of PHYSICAL UNITY, etc. Of course, this is a historical perspective on the issue, and if we view it in a synchronic perspective the question loses its relevance. Historically, the basic-level concepts can, in my view, be said to have partially created the concepts of ANGER, PRIDE and LOVE. Looked at from a synchronic perspective, perhaps all we should say is that both the constitutive and basic-level metaphors exist, and have their separate functions: The basic-level metaphors allow us to comprehend and draw inferences about these concepts, using our knowledge of familiar, well-structured domains. The constitutive metaphors provide the bulk of the anger, pride and love ontologies. So far in this study I have only been concerned with cases in which a more abstract domain is understood in terms of a physical domain. Examples in the case of, say, anger are fire, insanity, burden, vicious animal, etc. The concepts of FIRE, INSANITY, BURDEN, etc. are familiar and relatively wellstructured concepts which are used to understand (and create) aspects of such abstract, and hence less familiar and well-structured, concepts as ANGER. But the opposite process can also take place. This is when we use an adjective like raging in such expressions as raging storm to describe storms. But this gives us a problem. If it is true that, on the whole, we use relatively physical domains for the comprehension of more abstract domains, and

125 118 METAPHORS OF ANGER, PRIDE, AND LOVE this is the primary function of conceptual metaphors, then this 'reverse' metaphorization should not occur at all. In other words, the metaphorical process that goes from the more abstract to the more physical should not take place since it does not perform any important (cognitive) function that would motivate its application. (The question of what 'non-primary' functions conceptual metaphors can have will be discussed shortly). We can go about solving this apparent problem along the following lines. The example cited above is not an isolated metaphorical expression but forms a part of a system of metaphorical expressions. Consider the following conceptual metaphor: A STORM IS AN ANGRY PERSON It was a violent storm. The sea was lashing the rocks. The fierce wind caused a lot of damage. The angry waves calmed down as the dawn came. We had a raging storm last night. The savage storm whipped up the waves. There were roaring winds. The problem that this metaphor should not exist but it does ceases to be a problem if we consider the following. Primary metaphorization (in which the source domain is more physical than the target domain) is motivated by the cognitive function of understanding the more abstract in terms of the more concrete. This is possible because physical domains usually have, for us, a clearer structure than abstract domains. However, by virtue of structuring abstract domains in terms of metonymies and metaphors (which involve physical domains), we also get a fairly clear understanding of the abstract domains. The abstract domains structured and understood in this way can in turn be used to structure and understand further the physical domains that were originally employed to structure and understand the abstract ones. Let us call this process 'secondary metaphorization'. The processes of primary and secondary metaphorization can be seen as mutually enriching our understanding of the world. Domain A is used to understand domain and consequently can be used to understand A. The conceptual metaphor above is an instance of secondary metaphorization. Historically again, it must have been preceded, at least on this analysis, by instances of primary metaphorization. In this case the corresponding primary metaphor is ANGER IS A STORM:

126 IMPLICATIONS FOR THEORIES OF CONCEPTUAL STRUCTURE 119 ANGER IS A STORM After the row she stormed out of the room Waves of anger swept over him. We had a tempes tous meeting. He thundered with rage. His face clouded over with anger. After that stormy conversation, he wouldn't even talk to me. You could see rage surging up in him. The experimental basis of this metaphor is ANGRY BEHAVIOR, which stands metonymically for ANGER. Angry behavior includes INSANE BEHAVIOR, AGGRESSIVE BEHAVIOR and PHYSICAL AGITA TION, this last one also characterizing the first two. What is common to these forms of behavior is UNCONTROLLED MOVEMENT. It is this which serves as a basis for applying the concept of STORM to ANGER and that of ANGER to STORM. It should be noted that experiential bases for metaphors are not to be taken as similarities, in the ordinary sense, between a source and a target domain. Possibly we would not see any similarity between a storm and anger if we did not have the STORMS ARE ANGRY PERSONS metaphor. However, the metonymies corresponding to the experiential bases provide enough motivation for us to accept the conceptual metaphors as natural. It can be noticed that there is a difference in the style of the examples given for primary and secondary metaphors. Examples of secondary metaphors usually have a distinctly literary flavor: The expressions in the secondary metaphor A STORM IS AN ANGRY PERSON involving anger sound what we would normally call poetic, or imaginative, or fanciful in comparison to the examples in the ANGER IS A STORM primary metaphor. To account for this, we can bring in the notion of 'rhetorical function' for metaphors. As we have seen, this rhetorical function does not play a very important part in primary metaphorization whose crucial function is understanding. In secondary metaphorization, however, we can see a strengthening of the rhetorical function though the purpose of (further) understanding an already well-structured domain is not lost completely either. Since the target domains of secondary metaphors are typically already well-structured and familiar domains, in these metaphors there is less emphasis on the cognitive function of understanding. The result is that the linguistic examples of secondary metaphors will have a more literary or poetic character than those

127 120 METAPHORS OF ANGER, PRIDE, AND LOVE of primary metaphors. The issue of how the process of secondary metaphorization can be relevant to accounting for the collocational range of words like angry and proud will be taken up in the next chapter. To conclude what I have said about the structure of concepts in this chapter, an important point has to be made. I have tried to show that the structure of at least some concepts can be seen as having four parts: A system of conceptual metonymies, a system of conceptual metaphors, a set of related concepts and one or maybe more prototypical cognitive models with their several variants. However, no claim is made that all concepts have this level of complexity. In particular, it seems to me that it is mostly abstract, less clearly delineated concepts that are elaborately structured in terms of metaphors and metonymies, and that familiar and relatively clearly structured concepts, which typically denote physical phenomena, would lack these (metaphorical and metonymical) aspects of conceptual structure though they would remain characterizable in terms of prototypical cognitive models and related concepts (especially what have been called associated concepts). This difference in the structure of abstract and concrete concepts follows naturally from the cognitive functions of metaphor and metonymy as explained by Lakoff and Johnson (1980). Since metaphors and to a lesser degree also metonymies are primarily used for the purposes of understanding less clearly-structured domains in terms of more clearly-structured domains, we do not so much need metaphors and metonymies in the comprehension of more clearly-structured areas of experience.

128 6. IMPLICATIONS FOR THEORIES OF LEXICAL STRUCTURE In this chapter I will try to show how the approach presented in the investigation of ANGER, PRIDE and LOVE can be put to use in the study of some lexical issues. These issues include polysemy, collocation and semantic fields Polysemy The simplest view of polysemy maintains that polysemy involves a word that has a number related senses with at least one sense component shared by all the senses. In this section an attempt will be made to show that this is not how polysemy works in the case of the word proud. Linguists have come up with more subtle models for explaining polysemy, relying on the notion of literal, metonymical, and metaphorical extensions of meaning. These extensions of meaning are generally taken as being based on the notion of similarity (except, of course, metonymical extension). However, no account of the exact nature of similarity is provided. Moreover, little is said of the way in which the various senses of a word are related: Thus we do not get a picture of the internal structure of a category. Finally, senses are given in terms of minimal sense components, like 'feeling intense displeasure' in the case of anger and 'feeling satisfaction with oneself for pride. As a result, only some of the senses of a particular word can be accounted for and some of the subtler relationships between the senses cannot be made explicit. What is needed instead, I suggest, are prototypical cognitive models of the kind we have seen for anger, pride, and love. These are rich enough to enable us to derive a large number of senses in a systematic way and to recognize how the senses are related. Consider the various senses of proud extracted from several standard dictionaries: ARROGANT, CONCEITED, HAUGHTY asin He's too proud to talk to us ELATED, EXULTANT as in John is proud of winning the race

129 122 METAPHORS OF ANGER, PRIDE, AND LOVE HONORED, DELIGHTED, GLAD as in I am proud to introduce the new president METTLESOME, HIGH-SPIRITED as in aproud steed DIGNIFIED, STATELY as 'mproud cliffs SPLENDID, MAGNIFICENT as in those proud old castles CAUSING PRIDE, GLORIOUS as in It was a proud victory SELF-RESPECTING, DIGNIFIED as in They are poor but proud INDICATING PRIDE as in There was a proud look on his face PROTRUDING as 'mproud joint OVERGROWN as 'mproud flesh To claim that these senses are somehow similar is to claim that there is at least one sense component which is present in all of the senses above. This claim is easy to refute. What feature, or component, could possibly be shared by CONCEITED, ELATED, SELF-RESPECTING, METTLESOME, PROTRUDING, etc. each? In fact some of the senses suggest sharply opposing features: DIGNIFIED suggests calmness, slowness and METTLESOME suggests vigor. And in some cases it seems impossible to see any connection between the senses. One wonders what could possibly link PROTRUDING with, say, SELF-RESPECTING. This is not to say, however, that the idea of similarity does not play an important role in accounting for the various senses of proud. To be sure, there are some similarities between some of the senses on the list, but these similarities do not extend beyond two, three, or, maybe, four cases. For example, there is something in common between John's pride as a result of winning the race and self-respect. However, the point is that whatever similarities there are between the senses, these similarities are not shared by all the items on the list of senses forproud. If the similarity account does not take us too far, what can we offer in its place? Perhaps the intuitively simplest account of these relationships between the various senses could begin along the following lines. There is a prototypical, or central, or primary nuclear sense of proud. This sense is spelled out in the prototypical cognitive model, or prototype scenario, as described in the chapter on pride. Given this model, the wordproud denotes the following: A person has done something, he perceives what he has done as an achievement, and values it highly on a social value scale. As a result, he also values himself highly but proportionately to the achievement. Given this central sense of proud, we can begin to see how the various other senses can be defined in relation to this central sense. The sense which

130 IMPLICATIONS FOR THEORIES OF LEXICAL STRUCTURE 123 is related to the central sense in the most obvious way is that of INDICA TIVE OF PRIDE; as in a proud look. We have seen that looking in a certain way is one of the behavioral reactions that accompany pride in our folk model of pride. That is, the relationship between pride and the various behavioral reactions associated with it is that, in the folk model, the behavioral reactions result from the emotion. Because of the result from relationship, a behavioral reaction can indicate the existence of the emotion. Thus when the adjective proud is used to describe a behavioral reaction of pride (as in proud look) it will have INDICATING PRIDE as its sense. The relationship between INDICATING PRIDE and PRIDE/PROUD is that the former contains the latter as a part. That is, the two meanings are related metonymically. Another set of behavioral reactions associated with pride is erect posture and chest out. The further senses OVERGROWN and PROTRUDING (the latter found only in British English) can be derived from these manifestations of pride. Just as proud people 'Stand out' physically in relation to people who are not proud (i.e. who do not exhibit these behavioral reactions), so the flesh around a healing wound and a piece of wood in a joint can stick out physically in relation to the surface of the flesh or wood surrounding them. This metaphorical similarity accounts for the senses OVERGROWN and PROT RUDING forproud. When proud is used of castles, buildings, towers, etc., it acquires the sense SPLENDID, MAGNIFICENT. Splendor is associated in our conceptual system with properties like sumptuousness, tallness, bigness, etc. For our purposes it is tallness and bigness that are important. Tallness corresponds to the behavioral reactions of erect posture and head held unnaturally high, and bigness to those of chest out and chest unnaturally thrust out. The conceptual link between SPLENDOR on the one hand and TALLNESS and BIGNESS on the other can be seen if we try to decide which of the two expressions sounds more natural: a tall and splendid building or a small and splendid building. I believe that most people agree with me that we need less specialized contexts for accepting the former, rather than the latter, as more natural. Thus whenproud is used metaphorically of castles, buildings, etc., it will activate this conceptual link between TALLNESS and BIGNESS on the one hand and SPLENDID, MAGNIFICENT on the other. It is this conceptual link which will motivate and make natural the use of proud in the sense of SPLENDID, MAGNIFICENT. We have seen how pride is viewed as resulting in a number of behavioral reactions. Now let us consider cases where it is pride that resultsfrom a cer-

131 124 METAPHORS OF ANGER, PRIDE, AND LOVE tain state of affairs, this state of affairs being an achievement in the prototypical case. The sense that proud acquires when it is applied to achievements is termed GLORIOUS, or CAUSING PRIDE, or PRODUCTIVE OF PRIDE. The sense GLORIOUS arises because there is a state of affairs (an achievement) which is productive of pride. And here again we have a case of metonymy. But proud can be used in conjunction with several words which do not denote achievements (or other causes of pride) and yet it has the meaning PRODUCTIVE OF PRIDE. Some examples for this case are proud day, a proud sight, etc. I will consider these examples in the next section in which the possible collocations ofproud will be discussed. As has been noted in the chapter on pride, one of the behavioral reactions of pride is smiling. This is possible because JOY and SATISFACTION are concepts that are related to prototypical PRIDE. Smiling as a behavioral reaction is primarily associated with JOY, and since JOY is closely related to PRIDE, smiling becomes a manifestation of pride as well. Dictionaries tend to equate the central, or prototypical, sense of proud with the concept of JOY, defining this sense of proud as ELATED, EXULTANT. This is of course mistaken because a sense of a word cannot be given by substituting another concept for it as though somehow this concept could fully exhaust the sense in question. As the description of the prototype of PRIDE suggests, this concept involves a great deal more than just the idea of being happy as a result of an achievement. Nevertheless, this practice of lexicography is to some extent justified since the prototype of PRIDE involves as one of its crucial ingredients the concept of JOY. The word proud is employed in its prototypical sense when the person who achieved something equals the person who is proud. Butproud can also be used of a person who is not the 'achiever'. This is the case when, for example, I introduce to an audience a person who achieved something. In this situation I can say: I am proud to introduce X. This sense oí proud is given by dictionaries as FEELING HONORED, or GLAD, or DELIGHTED. These latter two definitions are possible for the same reason that has been given above for ELATED. Incidentally, this account also explains why, in the same situation, the sentences Iam glad to introduce X, Iam pleased to introduce X, and I am honored to introduce X can be used with no significant difference in communicative effect. Another sense that heavily exploits the concept of JOY as related to PRIDE is METTLESOME, or HIGH-SPIRITED, as it is applied to horses. One of the behavioral reactions that is understood as accompanying joy is intense (uncontrolled) movement (as in She was leaping for joy). Horses that

132 IMPLICATIONS FOR THEORIES OF LEXICAL STRUCTURE 125 are characterized by intense (uncontrolled) movement are called mettlesome, or high-spirited. It is this metaphorical similarity between intense movement in horses and in man from which the sense METTLESOME, HIGH-SPIRITED arises for the wordproud. Let us now return to the prototypical sense. The prototype for pride involves, as we have seen, a value scale and a pride scale that are balanced. In the case oí proud meaning CONCEITED, the pride scale is higher than the value scale. That is, there is more pride than would be appropriate. What justifies the extension of proud to this sense is that the model of conceit contains essentially the same elements (entities and predicates) as prototypical pride: A pride scale (entity), a value scale (entity), and balance (predicate). But in conceit the relation (balance) between the two entities is negated. This can be thought of as similarity of some sorts; but the similarity is of a very peculiar kind. Proud can also be used in the sense of SELF-RESPECTING. Selfrespect can be defined as an AWARENESS OF ONE'S WORTH. If in prototypical pride one values oneself highly for a particular achievement, in selfrespect one values oneself highly for no particular cause. In other words, in both pride and self-respect we find the idea that the self values oneself highly. Thus it seems thatproud is extended to the sense of SELF-RESPECTING on the basis of this shared feature, which, as we have seen, is metaphorically constituted. But there may be another motivation for this. There is, at least in the lexical domain of emotions, a productive semantic process by which new meanings are generated. This is a process whereby emotions that prototypically have their particular causes (like achievement for pride) become 'causeless'. The causeless version will typically get a new name but the name of the primary, or prototypical, version, out of which the new one developed, will also be used to express the new sense. Some examples that illustrate the process are: Pride vs. self-respect, anger vs. resentment, fear vs. anxiety, sadness vs. depression, etc. In all these pairs, the first word can be used to cover the meaning of the second, but not the other way around. Closely related to SELF-RESPECTING is the sense DIGNIFIED. In the chapter on pride, we have seen that DIGNIFIED is a concept inherent in SELF-RESPECT or SELF-ESTEEM. This allows proud to be used, via SELF-RESPECT or SELF-ESTEEM, of people in the sense of DIG NIFIED: A proud man can mean very much the same as a man with dignity. (I did not use the expression dignified man because dignified is primarily, though not exclusively, employed to describe the manifestations of dignity: Calmness, solemnity, ceremonial slowness, etc. And this is not the intended

133 126 METAPHORS OF ANGER, PRIDE, AND LOVE sense here.) What makes this further extension of proud even more understandable is that DIGNIFIED can also be defined in terms of WORTH. Webster gives the following definition for dignity: CONFIDENCE IN ONE'S WORTH, and, on the basis of other dictionaries, we may add: IN RELATION TO ONE'S CHARACTER. Thus the key concepts that seem to play a decisive role in the definition of dignity are: WORTH and CHARACTER. However, this still leaves us with the question: How can proud mean DIGNIFIED when it (proud) is used of mountains, cliffs, big rivers, etc.? The answer to this question requires us to look at how CHARACTER (in the sense of INTEGRITY) is conceptualized. A large part of our understanding of the concept of CHARACTER seems to arise from two metaphors. Let us take the metaphor CHARACTER IS STRAIGHT MOVEMENT first: CHARACTER IS STRAIGHT MOVEMENT They're OK. They're on the straight and level. He's a man of unswerving integrity. You're just a crook. I want a straight answer. She would never deviate from the truth. A particularly forceful, though archaic, example of this metaphor comes from a book on character (Smiles 1878): Without principles, a man is like a ship without a rudder or compass, left to drift hither and thither with every wind that blows (Smiles 1878: 6) The other metaphor that is relevant for the comprehension of CHARACTER is the following: CHARACTER IS PHYSICAL STRENGTH She's morally strong. Moral strength is above all. He has afirm character. He's a bit unstable, as far as character goes. She has weak character. You will find some loose women in that bar. A combination of both of these metaphors seems to be present in the word bent when it is used in the sense CORRUPT (as in Some policemen are bent). A hard physical object resists bending, i.e. it is strong. But if a lot of force is applied, it bends; i.e. it becomes not straight. A consequence of the PHYSI-

134 IMPLICATIONS FOR THEORIES OF LEXICAL STRUCTURE 127 CAL STRENGTH and STRAIGHT MOVEMENT metaphors is that we conceive of people with character as unyielding to such metaphorically construed forces as temptation (cf. I could not resist the temptation). The feature UNYIELDING appears in the STRAIGHT MOVEMENT metaphor as not deviating from a straight line, or course. Natural formations of large physical size like mountains, cliffs, big rivers, etc. are unyielding in a physical sense. Mountains and cliffs are immobile, stable and withstand the forces of nature affectig them. Big rivers follow a straight course and are unstoppable. It seems that it is metaphorical similarity of this kind between certain natural formations and character that is responsible for proud having the meaning DIGNIFIED when it is used of these natural formations. On the basis of the foregoing analysis, we can perhaps safely conclude that we need a much subtler account of polysemy than the one which is based on the undifferentiated notion of similarity between the senses. It is just not true to say that polysemy involves senses that all share at least one component. This view of polysemy can be represented in the following way (W stands for a word: s 1, s2, s 3 etc. stand for the various senses of W; d, e, f, etc. stand for various sense components. W:s 1 (c d,c e,c f...c),s 2 (c d, g, h,...c) s 3 ( d, i, j,... ) It can be noticed that the sense component c d is common to all of the senses. This view of polysemy may apply to some cases, but it certainly does not do as a general account. In the discussion of the various senses of proud, no single component has been found to be present in all the senses. Perhaps a subtler account of polysemy can be imagined along the following lines, using the same abbreviations as above: W: s 1 (c d,c e,c f...c),s 2 (c d, g, h,...c) s 3 ( d, i, j,... ) The difference between this account and the previous one is that in this case the first and the second senses share one component (c f ) and the second and third senses share another (c ), etc., but there is no single feature that the three senses have in common. This is of course Wittgenstein's idea of family resemblance, as he aplied it to concepts like GAME (Wittgenstein 1963). A similar idea has been around for quite a while in lexicographic practice. Consider the entry hop up in the Dictionary of American Slang: Hop up v.i., v.t. 1 To take narcotic drugs, as an addict; to drug a person... 2 To give a race horse a stimulating drug so that he will run faster

135 128 METAPHORS OF ANGER, PRIDE, AND LOVE than he normally would... 3 To increase the maximum speed,... of an automobile... The arrows indicate that the sense after the arrow developed out of the sense preceding it. Here we could say that senses 1 and 2 share the component STIMULATE WITH A DRUG and senses 2 and 3 have in common the feature IN ORDER TO INCREASE SPEED. But there is no component shared by all the senses. The analysis of proud reveals an even more sophisticated structure for polysemy. We have seen that proud (W) involves a prototypical sense (s ) defined by a number of components (c d, c e, c f,...). The various other senses ofproud (s 2, s 3, s 4,...) all seem to emerge from and cluster around particular components. Thus we have the following situation: W: s (c p,c e, c f,... n ) uster 1: s 2 (c d, c g, c h,...), s 3 (c h, c.,...). etc. Cluster 2: s 5 (c e, c k,...), s 6 (c R, c ļ5...), etc. Cluster 3: s 8 (c f, c m,...), s 9 (c m, c n,...), etc. etc. The prototypical sense ofproud has been characterized by such sense components as inherent concepts like JOY, behavioral manifestations like tallness and bigness, the idea of worth, the idea of balance between the pride and value scales, etc. The various senses can be seen as forming several clusters of senses based on different components. Thus tallness and bigness in the prototype give rise to the senses PROTRUDING, OVERGROWN and SPLENDID. The idea of worth is the grounds for the extension ofproud to mean SELF-RESPECTING, which in turn leads to DIGNIFIED. The sense METTLESOME emerges via the concept of JOY, which is also present in FEELING HONORED or GLAD. CONCEITED is based on the balance between the two scales. That is, all of these clusters of senses draw on differents aspects of the prototype scenario. In sum, the account of polysemy in the case of the wordproud required us to rely upon all the parts, or elements, of conceptual structure as given in the previous chapter. The conceptual metonymies, conceptual metaphors, related concepts, and the prototypical cognitive model of pride all seem to contribute to our understanding of what senses this word has and how they are related. It is not clear how this could be done in an approach which defines proud in terms of a minimal number of sense components like WORTH DUE TO AN ACHIEVEMENT, for example.

136 IMPLICATIONS FOR THEORIES OF LEXICAL STRUCTURE Collocation In the previous section the focus of the discussion was on the questions: How does a particular word acquire a given sense and how are these senses related? It was assumed that a word (like proud) combines with other words (like look, victory, building, etc.) to form larger syntactic units and that the various senses arise in these combinations. In this section I would like to look at some of the regularities that seem to govern with which other words a given word can combine to form acceptable larger units. For illustration, I will continue to use mostly the word proud. The problem of which words can combine with a given word is well known in linguistics. It goes under various names: In one school of British linguistics represented by Firth (1957) it is called collocation, or collocability and in more recent approaches it is usually discussed under the name of selection restrictions (Leech 1982; Lehrer 1974). Interestingly enough, the issue has been given some attention by philosophers as well. Austin, perhaps the most eminent philosopher to deal with the problem, couched the issue of collocability in a broader philosophical perspective and asked in a famous paper (Austin 1961): "why do we call different things by the same name?" The traditional answer to the question, according to Austin, is that we call different things by the same name because they are similar. Indeed, when we think about why the word head can apply to such different things as the human head, head of department, bridgehead, the head of a hammer, the head of beer, etc., we tend to think that there is something common in all of them; i.e. that they are all similar. In this case, maybe we could find a feature shared by all: Perhaps something like 'the uppermost part' or 'the most extreme position', although we would have difficulties accommodating examples like bridgehead or head of department. And there are obviously cases where the similarity position clearly does not work. Let us take Austin's examples: healthy body, healthy exercise, and healthy complexion. Austin's claim is that there is a primary nuclear sense of healthy in which it is used of healthy bodies. The other two senses are given by him as PRODUCTIVE OF HEALTHY BODIES (in healthy exercise) and RESULTING FROM HEALTHY BODIES (in healthy complexion). Austin's conclusion is that the body, the exercise and the complexion are not called healthy because they are in any way similar, but because they are related to each other in these particular ways. In his discussion Austin does not distinguish between the issue of sense (polysemy) and that of combinability of words (collocation, selection

137 130 METAPHORS OF ANGER, PRIDE, AND LOVE restrictions). He talks about senses in one place and possible combinations of words in another. For his purposes Austin did not need to make this distinction. However, my purposes require the two (obviously related) issues to be kept apart in a clear way. Polysemy involves the different senses of a word, whereas collocation involves the set of different words that a particular word can combine with. With this in mind, the basic issue concerning collocation can be put in the following way: How can we characterize the mechanism by which a certain word selects a set of other words as 'collocational partners'? The answer to the question is given according to which school of linguistics one belongs (Lehrer 1974: 173). The Firthian school maintains that collocation is a matter of lexical items and not a matter of the meanings of lexical items, although the set of collocations a word has forms a part of, or is one level of the meaning ofthat item. That is, this school correctly recognizes that one item collocates with a number of other items and that this is an important aspect of meaning, but offers no explanation of the mechanism that may govern the selection of some items rather than others. The other approach to the problem at hand is represented by the work of Lehrer (1974), Leech (1974) and Weinreich (1966). Their accounts are improved versions of proposals made by linguists like Katz, Fodor and Postal. Authors in this approach maintain that word A can collocate with word if a feature 'c' found in the meaning of word A is also present in the meaning of word or the feature 'c' can be transferred from A to without causing a clash of features in (Weinreich 1966). Leech (1974) claims to achieve the same result as Weinreich by what he calls 'contextual redundancy rules'. These rules work in the following way. Consider the sentence Gold owns the man (Leech 1982: 140). The semantic anomaly of the sentence, according to Leech, is due to a clash of the features +ANIMATE and -ANIMATE in the initial argument of gold. The verb own requires an initial argument which has the feature +PERSON. +PERSON adds the feature + ANIMATE to the initial argument by a redundancy rule. And another redundancy rule 'Gold requires -ANIMATE' produces the contradictory feature -ANIMATE in the same argument. The result is a clash of features within the initial argument and so the sentence becomes anomolous. Leech also notes that "many metaphorical uses of language involve a violation of selection resrictions" (Leech 1982: 141). He observes that a sentence like Gold owns the man or My gasmeter eats up 10ps" can be, and are, interpreted figuratively and thus are not thought of as anomolous. Eat requires the feature +ANIMATE in its initial argument and since meter has the feature ANIMATE, the result is a clash of features. However, we can interpret eat in

138 IMPLICATIONS FOR THEORIES OF LEXICAL STRUCTURE 131 this sentence metaphorically as 'uses up 10ps like an animal eating something' (p.41) and thus avoid the emergence of contradictory features in the initial argument. This seems to me to be an inadequate and ad hoc way of handling collocation in a large number of cases that involve metaphors. To say simply that a metaphor can be characterized by the statement 'X is like Y' is not to explain too much. In the first place, this is an ad hoc account because is like' is a catch-all phrase which is intended to account for all the metaphors in language. However, we have seen that if we work with conceptual metaphors we can get a much more precise understanding of how the human mind processes knowledge about the world. Second, this traditional view of metaphor based on the idea 'X is like Y' can only provide an after-the-fact analysis of metaphorical processes. On encountering a linguistic metaphor, all we can do, so this view maintains, is to say that the metaphor is possible because X is like Y. In other words, this view has no predictive power as to which metaphors will occur. In order to see how these accounts of selection restrictions can be improved, let us now return to the word proud. In the chapter on pride it was observed that in its prototypical use proud co-occurs with PERSONS as in John is proud of winning the race or John is a proud man. The selection restrictions for proud would state that proud requires the feature PERSON in the argument and thus we would get the correct results. However, we noticed in the previous section that proud can also collocate with victory, look, horse, building, etc. which are not persons. Perhaps we could, at the expense of losing important information, accommodate horse by saying that proud requires ANIMATE rather than PERSON. But what about the other items and a host of other examples? There is much more to these unaccounted-for cases than simply saying that, for example, in the case oí proud buildings the buildings in question resemble proud persons (metaphorical transfer of meaning) or in the case of proud victory victories produce proud persons (metonymical transfer of meaning). At this point we have to return to the issue of primary and secondary metaphorization in order to see how our conceptual system can have a direct influence on what at first glance appeared to be a purely lexical issue. What I would like to claim is that the collocational range of a word is in part determined by which other concept(s) the word (more precisely, the corresponding concept) is used to conceptualize. For example, we have seen that

139 132 METAPHORS OF ANGER, PRIDE, AND LOVE ANGER is used to conceptualize the concept of STORM. This enables us to account for an expression like angry waves. Similarly, the concept PRIDE is employed to structure (further) the concepts of BUILDINGS, MOUN TAINS, etc. Let us examine the following secondary metaphor for PRIDE: ARTIFACTS (BUILDINGS, CITIES, etc.) ARE PROUD PERSONS NATURAL FORMATIONS (RIVERS, MOUNTAINS, etc.) ARE PROUD PERSONS And ahead of us emerged the proud city of London. Proud buildings lined the road on both sides. The Danube flows proudly underneath. Suddenly a proud tower emerged before us. A proud ship was slowly sailing by. Proud mountains surrounded the town on all sides. It is the metonymies ERECT POSTURE, HEAD HELD UNNATUR ALLY HIGH, CHEST OUT and CHEST UNNATURALLY THRUST OUT that form the basis of this metaphor. What this means in this case is that the ARTIFACTS and NATURAL FORMATIONS have to be of a particular kind; exactly what kind is specified by the metonymies above: i.e. they must be tall and/or big things. The importance of all this for our purposes is that the metaphors A STORM IS AN ANGRY PERSON and ARTIFACTS/NATURAL FOR MATIONS ARE PROUD PERSONS delimit, on a conceptual level, the range of possible words that angry and proud, and the other conceptually related terms, can combine with; namely, they tell us that these words can collocate with words that express, or are related to, the concept of STORM and certain kinds of ARTIFACTS and NATURAL FORMATIONS. However, it should not be thought that this statement is intended to apply to instances of secondary metaphorization alone. What type of metaphor is relevant for the delimitation of the collocational range of a word depends on the nature of the corresponding concept in question. Abstract concepts like ANGER and PRIDE can be used, as the metaphors above show, to conceptualize more physical domains. The delimiting effect of concepts associated with more physical domains is dependent upon which abstract concepts they are used to conceptualize. Thus the collocational possibilities of words that have to do with, for example, FIRE would seem to depend on which more abstract concepts it is used to structure. FIRE is present in the conceptualization of a large number of other concepts: Among them, as we have seen, in

140 IMPLICATIONS FOR THEORIES OF LEXICAL STRUCTURE 133 the conceptualization of LOVE. This fact would in turn enable us to account for, and on a conceptual level predict, the collocation fiery passions, for example. Thus it seems that the situation which is relevant for the issue of collocation is when the concept in question serves as a source, as opposed to a target domain, either in primary or in secondary metaphorization. Admittedly, the mechanism that has been used to delimit the possible collocations of proud and angry is not very precise in the sense that the delimitation has been done on a conceptual and not on a lexical level. That is to say, no attempt has been made to provide a complete list of items that can cooccur withproud and angry. That would mean that we simply state the collocations of words without trying to explain why the words have just those possible collocations that they have. Furthermore, listing all the items that can co-occur with another seems, in most cases, to be a futile job anyway since one can always add a new item to the list on the basis of the regularities provided by the conceptual metaphors and their experiential bases. It is this regularity and predictability that arises from the notion that we can call the 'collocability of concepts', rather than that of isolated words, which can be regarded as an advantage of the present lexical approach over the mechanisms proposed by Leech and Weinreich. The approach based on conceptual metaphors can make, however slight, predictions concerning possible collocations. It is this that the other approaches do not allow us to make because they perform after-the-fact analyses of isolated words. The germ of the idea of the 'collocability of concepts' can be found already in the work of Firth (1957) and Halliday (1966). Firth talked about 'formal scatters' by which he meant groups of related words. He suggested that collocation operates between scatters rather than words. Or, consider Halliday's examples (1966): a strong argument, the strength of the argument, he argued strongly, and his argument was strengthened. It is clear that it is possible to formulate a rule for the collocations of these items which would be on a more general level than the words that form a part of the collocations. In particular, we could say that one aspect of the concept of ARGUMENT (the strength of an argument) is comprehended in terms of the concept of PHYS ICAL STRENGTH and that items that express, or are related to, either of these concepts can combine with each other. Incidentally, the metaphor THE STRENGTH OF AN ARGUMENT IS PHYSICAL STRENGTH forms a part of a more general metaphor for arguments. The metaphor, as given by Lakoff and Johnson (1980), is ARGUMENTS ARE BUILDINGS. Let me now continue with cases that involve what is commonly referred

141 134 METAPHORS OF ANGER, PRIDE, AND LOVE to as transfer of meaning based on metonymy. To account for collocations like a proud achievement, a proud victory, a proud feat, a proud look, proud gait, etc., the following rule can be set up: The word proud can be used to modify a noun that denotes some of the prototypical causes or manifestations of pride. Here again, we are not concerned with particular cases. This statement helps us capture as possible collocations of proud a number of types of causes and manifestations which are specified in the prototypical cognitive model of pride. Within the types there can be a great deal of lexical variety. But things are more complicated than this. Consider the following examples: Proud gold medal, proud day, proud fields of Waterloo, proud sight, etc. The gold medal, the day, the fields of Waterloo, and the sight are not, strictly speaking, causes or manifestations of pride. The gold medal as such is not the achievement itself, the day as such is not something that can make one proud, etc. However, all these entities are related to achievements in certain specifiable ways: The gold medal is a token of an achievement, the day is the time when an achievement was accomplished, the fields are the place of an achievement, and the sight is a visual experience of whatever achievement was accomplished. We can call these cases instances of 'secondary metonymies'. These observations lead us to recognize a further principle of collocability forproud: The word proud can be used to modify a noun which denotes the possible tokens, the time, the place and the virtual experience of the prototypical cause of pride. These rules seem to be sufficiently general for the wordproud; i.e. they seem to allow us to capture the most important (metonymy-based) collocations of the word. But the question one can raise is to what extent these rules are applicable outside the domain of pride and in other languages. Their applicability, in their present form, to other domains and the same domain in other languages seems to be restricted in a number of ways. Languages may differ as to whether they observe the first principle or not. This principle covers both the causes and manifestations of pride. But, for example, Hungarian does not seem to make it possible to use the equivalent ofproud to modify causes of pride, only manifestations of it. Thus in Hungarian one cannot say the equivalents ofproud achievement, proud victory, etc. Second, the generality of the principles seems to be limited even within the same language. The first principle seems to hold for, say, fearful (as in a fearful animal and a fearful glance) but not for, say, angry; an angry insult does not mean an insult causing anger, although insults are fairly typical causes of anger. Third, since there are several kinds of causes that can produce pride, it has to be examined

142 IMPLICATIONS FOR THEORIES OF LEXICAL STRUCTURE 135 which of these do and which do not take proud as a modifier. Achievements do seem to take it. Possessions are less likely to do so:?α proud garden, Ία proud Mercedes,?α proud possession, etc. Appearances are even more doubtful:??her proud long black hair. And perhaps causes like positive moral qualities are even less likely to take it: cf.???his proud integrity. If there is a difference in the degree of acceptability between these cases, this may suggest an ordering from the most prototypical to the least prototypical causes. Fourth, we have even seen that the generality of the principles is limited even within the realm of emotional states. It remains to be seen whether and to what extent the first principle holds in the case of other states. It seems to work with physical states like health. But a mental property like intelligence appears to allow only manifestations (as in an intelligent look) but not causes. Fifth, another limitation involves what was called secondary metonymy. The set of secondary metonymies (token of cause, place of cause, time of cause, etc. for pride) that characterize a type of cause, if they have a cause component at all, may also be word-specific (or rather, conceptspecific). In other words, lexical items (concepts) may differ in the number and kinds of secondary metonymies they allow. For example, unlike proud, healthy does not seem to take places that are related to its prototypical causes. Thus although we have healthy exercise, we do not have healthy gym, in spite of the fact that gyms are places where exercise that produces healthy bodies can be performed. Are we to conclude, in view of these limitations, that the lexical approach I have presented has no role to play in the study of selection restrictions? Such a conclusion can be reached only by those who insist that collocations involve only semantic properties of individual words and have nothing to do with the way in which our conceptual system functions. However, as we have seen in this section, ignoring the study of conceptual structure in the attempt to account for collocation would mean the loss of certain significant generalizations (e.g. the role of conceptual metaphors and conceptual metonymies) for the collocability of the word proud. Instead of such a conclusion, what emerges from the foregoing analysis is that we need sufficiently general conceptually-based rules for particular words, where the rules may not be generalizable to (conceptual) domains outside the (conceptual) domain for which they have been established. The use of particular words in combination with other words would seem to depend largely on how the corresponding concepts are conceptualized, as we have seen for PRIDE and ANGER, although there may well be certain arbitrary limitations which have

143 136 METAPHORS OF ANGER, PRIDE, AND LOVE to be stated independently. When I say that the use of a word in collocations depends on how it is conceptualized, I do not intend to limit this statement to cover only the issue of which complex metaphorical and metonymical concepts the corresponding concept (i.e. corresponding to the word in question) in question forms a part of (as ANGER forms a part of the complex metaphorical concept A STORM IS AN ANGRY PERSON and PRIDE is a part of the complex metonymical concept THE CAUSES OF PRIDE STAND FOR PRIDE). In many cases, selection restrictions have nothing to do with metaphor or metonymy. These are cases where detailed prototypical cognitive models are needed to account for the use of the word in question in collocations. For example, whenproud is used in combination with a person, it is used non-metaphorically and nonmetonymically but the person in question must have enough of the properties as given by the prototype for PRIDE. Perhaps the majority of cases in the ordinary use of language are like this Semantic fields The area of linguistics which is explicitly concerned with the relationship between lexical structure and conceptual structure is what has come to be known as the theory of lexical, or semantic fields. In chapters 2 through 4 I have looked at a number of semantic fields: In particular, the semantic fields of anger, pride and love. Since theories of semantic fields can be interpreted as theories of how lexical structure and conceptual structure are related, it can be useful for us to see what and how the analyses of these fields can contribute to the theory of semantic fields in general. This is all the more important since the contribution that our lexical approach can make to the study of this area of semantics can be thought of as one measure of the validity of the approach. Trier (1931), one of the pioneers of field theory, made a distinction between conceptual and lexical fields. Furthermore, he assumed that a lexical field cuts up the corresponding conceptual space like a mosaic. This idea has been criticized by a number of authors, among them Lehrer (1974). Lehrer suggests that the 'mosaic' idea does not work because there are gaps and overlaps in most lexical, or semantic fields. She proposes another model for the analysis of semantic fields. Her model is based on Berlin and Kay's research on color as reported in their book Basic Color Terms (1969). Berlin and Kay found that, in the domain of color, there are certain focal points which people find easier to identify than nonfocal areas. The linguistic terms corresponding

144 IMPLICATIONS FOR THEORIES OF LEXICAL STRUCTURE 137 to the focal points are called 'basic color terms'. Lehrer generalizes this idea to the study of lexical fields. She proposes a distinction between basic and peripheral terms. Basic terms are clear instances of a given lexical field, whereas in the case of peripheral items people may differ in their judgements of whether the item does or does not belong to the field. The issue of which items are basic within a particular field can be decided on the following criteria (adapted from Lehrer): 1. Basic words are monolexemic. 2. The application of basic words is not restricted to a narrow class of objects (as it is in the case of auburn and blonde). 3. Basic words are psychologically salient for informants. 4. Recent foreign loan words are suspect as basic words. 5. The meaning of a basic term in not included in that of any other except the headword for the taxonomy (as it is in the case of crimson and scarlet which refer to types of red). According to Lehrer, on these criteria the following are basic words in the domain of cooking terms (Lehrer 1974: 30): cook, bake, boil, roast,fry, broil (or grill in British English) and steam. She then adds simmer, stew, poach, braise, sauté, French-fry, deep-fry, barbecue, grill (American English) and charcoal. The peripheral words in the same domain include parboil, plank, shirr, scallop, flamber, rissoler, and compounds like steam-bake and ovenfry. That some of the words that Lehrer calls basic do not in fact seem to be basic on the above criteria should not worry us. What is more important for my purposes at this point is to try to get a more or less complete picture of how the study of lexical fields is envisioned by previous researchers. This is necessary if we want to see the major differences between the approach represented by Lehrer and the approach that seems to emerge from the three case studies. As a next step, Lehrer gives us a diagram that shows how the various basic words are related to each other and to the two superordinate terms cook and bake. For example, we come to know thatfry is subordinate to cook, and sauté and French-fry are subordinate tofry, and that, in addition, sauté and French-fry are incompatible. It is significant to notice that all the terms listed by Lehrer belong to the same word class; they are all verbs. It is typical of studies of lexical fields to set up fields that include terms from only one word class. Lehrer also points out that in order for us to know that these items belong to the field of culinary terms and to know that they are related in these particular ways, we need to know which minimal sense components make up

145 138 METAPHORS OF ANGER, PRIDE, AND LOVE the meanings of particular items. In other words, what a theory of lexical fields must be capable of doing is the following: (1) specifying which grammatically homogeneous items belong to the field and which of these items are basic terms on the Berlin and Kay criteria, (2) showing the relationship between the superordinate term and the other items, (3) showing the relationships among the various items within the field, and (4) doing all this on the basis of minimal sense components as characterizations of meaning. It seems to me that one can be in agreement with this program as far as (2) and (3) are concerned without having to commit oneself to (1) and (4). And if we do not accept (1) and (4), our conception of semantic fields will be very different from the picture that the traditional views offer. To see the differences, we have to imagine what, for example, the field of anger would look like if we approached it along the traditional lines. The conceptual domain of anger would possibly be subdivided into four fields. One of them would correspond to the adjective angry and it would include words like choleric, enraged, incensed, indignant, irate, ireful, mad, seething, sore, wrathful. The second field could be thought of as being subordinated to the verb anger when it means CAUSE TO FEEL ANGER. The items would include burn up, enrage, incense, infuriate, madden, provoke, and steam up. The third field, corresponding to the verb anger when it means BE/BECOME ANGRY, would have such items as blow up, boil, bristle, explode, flare up, fume, rage and seethe. And finally, we could set up a field that corresponds to the noun anger as exemplified by rage, indignation, etc. (we shall see some more examples for this field in a moment). Let us now stop and try to evaluate this conception of the lexical field of anger. Several comments can be made. First, it does not seem to be intuitively right to have four distinct subfields for the domain. Our intuitions tell us that words like seethe and seething, enrage and enrage, etc. should go together because they express the same concept. One of the disadvantages of insisting on grammatically homogeneous word classes is that the presentation of the field will be very uneconomical. Items which are felt to be the same conceptually will have to be stated as many times as many word classes they belong to. The second issue has to do with the distinction between basic and peripheral items. It is claimed that it is sufficient to analyze the meanings of basic terms because it is these terms that carry all the important semantic contrasts within a field (Lehrer 1974: 10). This distinction, at least as it is made use of by Lehrer, makes it appear that there are important items within a field

146 IMPLICATIONS FOR THEORIES OF LEXICAL STRUCTURE 139 (i.e. the basic words) and there are unimportant ones (those that are peripheral). However, I find that the issue of importance has nothing to do with the organization of a semantic field. As we have seen in the study of anger, there is a large number of nonprototypical (peripheral) cases which are not in any way less important than the prototypical case. In fact, they seem to carry semantic features which cannot be found in the prototype. This, in a sense, makes these terms even more salient than the one which refers to the prototype. Thirdly, it can be objected furthermore that the lexical items in the four fields taken together do not provide a complete list of the linguistic expressions that are related to anger in some way. The analysis of anger in this work has been based on the study of more than 150 conventionalized expressions, and this is a much larger number than what we would get if we added up all the items in the four subfields given above. The reason is that in this study attention has been given to items, like various kinds of metaphors, metonymies, collocations, idioms, etc., which would be ignored in a traditional approach. However, as it has been pointed out, these expressions are just as important as the more 'basic' ones if we want to see the structure of the entire field. A final objection to the analysis so far is concerned with the number of aspects a particular concept has. The various subfields that correspond to the adjective angry, the noun anger and the verb anger make us believe that ANGER is a concept that has only a limited number of aspects. In particular, they suggest that we conceptualize anger as an emotional state (suggested by the adjective), an abstract mass entity (suggested by the noun), a state in which we are (suggested by one sense of the verb) and a state which we come to be in as a result of some cause (suggested by the other sense of the verb). But, as we have seen, there are several additional aspects which characterize anger: They include retribution, responsibility, controling anger, loss of control, physiological effects, cessation of anger, just to mention a few. By contrast, the aspects as based on grammatical classes feature the concept in only a very crude way. In order to see how a componential approach would handle the relationship between the superordinate term anger and its subordinates, let us now turn to the componential analysis of the fourth subfield, which corresponds to the noun anger. Some of the examples in this subfield are indignation, irritation, wrath, fury and rage. Anger has already been analyzed as STRONG FEELING OF DISPLEASURE (OVER A WRONG). The other items

147 140 METAPHORS OF ANGER, PRIDE, AND LOVE could be given the following componential definitions: Indignation can be defined as INTENSE DISPLEASURE DUE TO MORAL INJUSTICE, irritation as MILD DISPLEASURE, wrath as LONG-LASTING INTENSE DISPLEASURE, andfury and rage as EXTREMELY STRONG DISPLE ASURE. The way in which anger and the other terms are related is captured in a very simple way in componential analysis. This is possible because the sense components are designed to show minimal differences. In componential analysis each item receives a minimal conceptual definition, where minimal means that an item must differ from every other item (within the field) in at least one conceptual component. The exception to this is synonymy where two items are characterized by the same componential formula (like in the case offury and rage). The relationship between anger and the other terms is arrived at by pointing out that since anger is defined as INTENSE FEELING OF DISPLEASURE and that there is at least one component (DISPLEA SURE) which is present in the definition of all the other items in some (specified or modified) form, the word anger would be the super-ordinate term in this field. In the light of what has been said about anger in this study, some objections can be made to this analysis. Componential analysis takes the definition of concepts lightly. In the case of anger, all we are told is that it can be given as INTENSE FEELING OF DISPLEASURE. In contrast to this, we have seen that the conceptual definition of anger requires a lot more. It has been found important to set up five successive stages in characterizing the temporal dimension of anger. Moreover, a number of entities and predicates were used to describe (the various aspects of) each particular stage. The result was a large and intricate network of concepts that was called the prototypical scenario. This scenario is sufficiently rich so that a large number of non-prototypical cases can be defined in relation to it. Furthermore, the application of componential analysis to the other terms is deficient in the same way. As we have seen in the discussion of the nonprototypical cases of anger, wrath, for example, is not just LONG LASTING INTENSE DISPLEASURE. Rather, we should say that, in addition to lasting a long time and being intense, in wrath the intensity of the offense is well above the limit and many acts of retribution are required to create a balance between the offense, the anger and the retribution scales. In the case of indignation, it does not do to define it simply as INTENSE DISPLEASURE DUE TO MORAL INJUS TICE. It is just as important to say that the victim does not equal the self and that the intensity of anger is not near the limit. Rage andfury were defined as

148 IMPLICATIONS FOR THEORIES OF LEXICAL STRUCTURE 141 EXTREMELY INTENSE DISPLEASURE. Instead of this, it can be said that rage andfury primarily have to do with stage 4, that is, with loss of control and angry behavior. Perhaps these examples will suffice to make the general point that componential analysis singles out certain components as distinctive. These are the components that allow the analyst to distinguish one meaning from another in minimal terms. This in turn leads to the unfortunate consequence that some other features that may be just as important to define a term will be ignored. The lexical approach as described in this study offers us an alternative to this view. Namely, we can see semantic fields as categories of concepts with a prototype in the center. (This is what Lakoff (1987) calls 'radial categories'.) These categories of concepts, like those centering around PRIDE and ANGER, may emerge in two ways. A prototype may give rise to a variety of senses that deviate from the meaning of the prototype and these senses are thought of as the standard meanings of the term corresponding to the cognitive prototype. The senses are typically listed in the lexicon of the language. This is what is commonly referred to as polysemy and this is what we have in the case of the concept PRIDE. The various senses of the term pride and proud as listed by the dictionary can all be systematically derived from the prototype of pride as it is given in the prototypical cognitive model. Typically (though not always), these standardized meanings can also be conveyed by lexical items that are different from the one denoting the prototype. That is, we have words like conceited, mettlesome, splendid, etc. which are typically used to express the corresponding meanings. Another, and usually parallel, way in which a lexical field may arise is when the deviations from the prototype are not lexically standardized; i.e. are not thought of as standard lexical meanings of the word in question and consequently are not listed as such in the lexicon. This is what we have seen in the case of the concept ANGER. Most of the nonprototypical cases of anger are not standardized senses of anger: frustrated anger, exaggerated response, constructive use, indirect cause, etc. are not standard meanings of anger in the sense as CON CEITED, METTLESOME, etc. are standard meanings ofproud. Typically (though not always), these members of the lexical field of anger are not referred to by what can be called standard lexical items, or conventionalized expressions. Frustrated anger, exaggerated response, etc. are not standard lexical items in the sense, for example, conceited and mettlesome are; they are simply descriptions of certain nonprototypical cases. It is important to note that these generalizations concerning the emergence and lexical manifesta-

149 142 METAPHORS OF ANGER, PRIDE, AND LOVE tions of semantic fields should be taken to hold only typically. The two processes often intermingle, both as regards the standardization of meaning and the lexical manifestations of nonprototypical cases. In this study we have seen several examples for both. The question arises: If the meaning of anger is given as described in the prototypical cognitive model, then how can we characterize the meanings of nonprototypical cases? The answer depends on the ways in which something can deviate from the prototype, i.e. can be nonprototypical. There are cases where the meaning of an expression can be defined in terms of which part of the prototypical scenario it is linked up with. As we have seen above, rage and fury seem to derive their sense from a particular part of the prototype. Struggle with anger would have the meaning it has because it has to do with the aspect of control in the same model, and be boiling can be seen as deriving its meaning from the fact that it is connected with the intensity aspect of anger. The issue of which term in the field is linked up with which aspect of the model is determined by the mappings that have been discussed in the previous chapter. The mappings represent a process by which the various items capture, or translate into the terms of, the various aspects of the prototypical scenario. All the examples we have seen above are cases where a certain part of the prototypical scenario is focussed on. It is this partial focussing which makes these cases nonprototypical. But there is another type of nonprototypicality. This is when a term derives its meaning from negating a particular aspect of the prototype. These are cases where the differences between the prototype and the deviation from it have to be marked explicitly. It has been pointed out that, for example, in the case of indignation it is negated that the angry person is the same as the victim, that the intensity of anger goes beyond the limit, and that the cause of anger is an act intentionally performed to make someone angry. It seems then that there are various ways in which we can deviate from the prototype. A distinction has just been made between the deviation which works by focussing on a certain part of the prototype and one which works by negating a certain part of the prototype. As a consequence of this distinction, the meaning of the items belonging to the field can be specified in two distinct ways. In the former case, the meaning of an item does not need to be specified separately: It can be stated by referring the item to the suitable part of the prototypical model. For items that fall into the second category we can do

150 IMPLICATIONS FOR THEORIES OF LEXICAL STRUCTURE 143 either of two things: If there is a binary choice offered by the ontology of the prototype, we just need to refer the item to the entity or predicate that the item negates and mark it with a minus (for example, -INTENSE). Or, if there is no binary choice, we mark the difference with respect to the prototype in an explicit way (by saying, for example, that in the case ofindignation the typical cause for anger is to be replaced by some moral injustice). So far I have been concerned with the relationship between anger and the terms that are subordinated to it. But how can we handle the relationships among the various items related to anger? How do we know, for example, that someone who is boiling with anger hasn't hit the ceiling yet, that boiling involves a more intense form of anger than simmer, that someone who is struggling with his anger is ready to blow up, etc.? What a person who understands these expressions knows is which part, or segment, of the prototypical scenario the expression links up with. For example, since we know from the prototype that struggle with anger maps onto trying to control anger, that controlling anger presupposes that the intensity of anger is near the limit, and that blow up maps onto loss of control, which in turn takes place if the anger goes above the limit, we will know that the person who is struggling with his anger can blow up any minute. It is this kind of knowledge that speakers of English possess which enables them to see the various meaning relations among the terms. If it is true that this kind of knowledge is necessary for establishing certain meaning relations among various items in a field, then the conceptual structure that has been found important to characterize the concepts of ANGER, PRIDE and LOVE must play a very significant role in the description of meaning. It seems to me that without taking into account conceptual metaphors, prototypical scenarios, mappings onto these scenarios, the category of nonprototypical cases, etc., that is, is we solely rely on minimal sense components, we cannot faithfully represent the complexities of lexical meaning in natural language. Let us take the previous example again and see what componential analysis would say about the meaning of the words struggle and blow up when they apply to anger. Struggle would probably receive the meaning specification TRYING TO CONTROL ANGER and the meaning of blow up would in all probability be characterized as LOSS OF CONTROL OVER ANGER. All we would learn from these descriptions of meaning is that the two words have to do with anger and that they capture two different aspects of it, but we would not know exactly how these aspects are related within the model of anger. In other words, componential analysis could give

151 144 METAPHORS OF ANGER, PRIDE, AND LOVE no systematic account of how it is that people somehow do know why it is that a person who is struggling with his anger is ready to blow up. The major reason for this deficiency is that componential analysis sees word meanings as autonomous, isolated conceptual entities with a simple structure, rather than as overlapping segments in a rich conceptual texture which is organized in complex ways. It seems to me that lexical meaning can only be studied against this rich conceptual background.