Emotions in Crosslinguistic Perspective [Reprint 2010 ed.] 9783110880168, 9783110170641

Citation preview

Emotions in Crosslinguistic Perspective

W DE

G

Cognitive Linguistics Research 17

Editors Rene Dirven Ronald W. Langacker John R. Taylor

Mouton de Gruyter Berlin · New York

Emotions in Crosslinguistic Perspective

Edited by Jean Harkins Anna Wierzbicka

w DE

G Mouton de Gruyter Berlin-New York 2001

Mouton de Gruyter (formerly Mouton, The Hague) is a Division of Walter de Gruyter GmbH & Co. KG, Berlin

© Printed on acid-free paper which falls within the guidelines of the ANSI to ensure permanence and durability.

Library of Congress Cataloging-in-Publication Data Emotions in crosslinguistic perspective / edited by Jean Harkins, Anna Wierzbicka. p. cm. — (Cognitive linguistics research ; 17) Includes bibliographical references and indexes, ISBN 3110170647 (he. : alk. paper) 1. Language and languages. 2. Emotions. 3. Psycholinguistics. I. Harkins, Jean II. Wierzbicka, Anna. III. Series P107.E46 2001 410-dc21 2001055803

Die Deutsche Bibliothek — Cataloging-in-Publication Data Emotions in crosslinguistic perspective / ed. by Jean Harkins ; Anna Wierzbicka. — Berlin ; New York : Mouton de Gruyter, 2001 (Cognitive linguistics research ; 17) ISBN 3-11-017064-7

© Copyright 2001 by Walter de Gruyter GmbH & Co. KG, D-10785 Berlin All rights reserved, including those of translation into foreign languages. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Printing: WB-Druck, Rieden/Allgau Binding: Lüderitz & Bauer, Berlin Printed in Germany

Contents Introduction Anna Wierzbicka and Jean Harkins

1

Testing emotional universals in Amharic Mengistu Amberber

35

Emotions and the nature of persons in Mbula Robert D. Bugenhagen

69

Why Germans don't feel"anger" Uwe Durst Linguistic evidence for a Lao perspective on facial expression of emotion N. J. Enfield

115

149

Hati: A key word in the Malay vocabulary of emotion Cliff Goddard

167

Talking about anger in Central Australia Jean Harkins

197

Meanings of Japanese sound-symbolic emotion words Rie Hasada

217

Concepts of anger in Chinese Pawei Kornacki

255

Human emotions viewed through the Russian language Irina B. Levontina, Anna A. Zalizniak

291

A culturally salient Polish emotion: Przykro (pron. pshickro) Anna Wierzbicka

. . . . 337

vi

Contents

An inquiry into "sadness" in Chinese Zhengdao Ye

359

Subject and name index

405

Words and phrases index

413

Introduction Anna Wierzbicka and Jean Harkins

The study of human emotions needs input from the study of languages, now more than ever before. Advances in the science of brain physiology are identifying in ever greater detail the specific wave patterns and locations of brain activity that correspond to different affective or emotional states. Studies of this kind require high technology, and most of them are conducted in predominantly English speaking research environments. It is expected that their findings will apply to human brains generally, not just those of a particular language or cultural group (cf. e.g. Davidson and Ekman, eds. 1994; Ekman and Rosenberg, eds. 1997). Further research may confirm or challenge this expectation, but the research itself and the interpretation of its results hinge at certain crucial points upon questions of language. These questions have to be approached seriously, for the validity of scientific studies of emotional response may depend upon whether the issue of language is addressed in an informed way or by default. Along with increased technical understanding of the workings of human brains, there have also been major advances in the study of emotions from the perspectives of cognitive and crosscultural psychology, psychological anthropology, and sociology. Some of this work has recognised the role of language as central to the study of emotions, particularly when examining instances where the cultural life of one group seems to focus attention on emotional states for which other groups don't even have names. Detailed descriptions of the meanings and manifestations of emotional states in different linguistic and cultural groups have added much to the understanding of emotions in cultural context. The purpose of the crosslinguistic studies presented in this volume is to demonstrate how the tools of linguistic analysis can be applied to produce more accurate descriptions of the meanings of emotion words and, more generally, ways of speaking about emotions in different languages. Such analyses of linguistic meaning not only complement findings from other approaches to the study of emotions, but help to resolve methodological problems that arise when these other approaches have to deal with data

2

Anna Wierzbicka and Jean Harkins

from different languages. Before proceeding to the language-specific studies, we draw readers' attention to the relevance of language in the study of human emotions, and give some background to the approaches to analysing language data that are used in these studies.

1. Emotions and languages Emotions are among the least tangible aspects of human experience, yet they exert powerful influences upon our thoughts and actions, and even upon our physical appearance and physiological processes occurring within our bodies. Examining the outward manifestations of facial expression and measurable physiological responses is one approach to studying these inner states and events. But an emotional feeling is so internal to the experience of the person who has it, that it has rightly been questioned whether it is even reasonable to think that there is necessarily, or demonstrably, very much in common between one person's experience of, for example, anger in a particular situation, and a different person's experience of anger in the same or some other situation. Moreover, people's emotional responses obviously differ, so that a context in which one person feels anger may evoke a different kind of emotional response in another person. Even if there is a general social or cultural expectation that a person would or should feel anger in this situation, not everyone reacts in the same way. Neither do most people respond the same way all the time; something that makes me angry today may prompt a different reaction tomorrow. Hence, emotions cannot be defined purely in terms of situation, context or eliciting conditions. The extent to which emotional states can be detected through measurable or observable reactions is, as mentioned above, a focus of much current scientific research. But we know that observable displays of emotion vary according to many factors. One may turn red with anger, glower and shout in one situation and appear white-faced, expressionless and icily polite in another. Both social and cultural pressures may encourage people to adjust their outward appearance to please others even while inwardly seething. One may not even be aware of feeling angry about something that happened, until talking or thinking about it afterwards. Whatever the conditions that produce an emotion like anger, whether or not it is visibly expressed, and whatever physiological responses accom-

Introduction

3

pany it, it is only through language (if at all) that we can know that what is experienced is anger: that is, if the experience! says so, or says other things by which we know that the person feels anger. Even if an identical physiological pattern is observed on two occasions, only language enables us to connect these two events with the emotion of anger. Most, though of course not all, of what we know about people's inner feelings comes to us via language, and it is only language that enables people to use the same word, anger or angry, to designate two different instances of an emotional experience. While theories of word meaning and use differ, the point here is that it is language that provides a conceptual connection between two disparate emotion experiences, by giving them the same label. The same label, however, may not apply across languages. Bilingual dictionaries usually equate English anger with French colere, for example, and deal in varying ways (if at all) with the fact that they cannot correctly be substituted for each other in every situation. The further observation that the adjective angry is often, though not invariably, better translated as/wrieux (which therefore in turn cannot be regarded as an exact equivalent of furious) is only part of the story (cf. Wierzbicka 1988). So, even if we think we can understand what anger means for two people who use this English word to label their experience, how are we to understand the supposed equivalent or equivalents in another language and culture? Even in the case of these two fairly closely related European languages, it is widely perceived that the differences in usage of emotion words are connected in some way with cultural attitudes and cultural identity. The different "feel" of the words angry, furious, furieux has much to do with the kinds of things English and French speakers do, how they look, sound and behave, when they feel these emotions. Studies of societies in other parts of the world have explored how certain patterns of emotional response seem to be centrally involved in the social behaviour of a cultural group. For example, Briggs (1970) examined an Eskimo group where very little attention was devoted to feelings comparable to those described in English as angry, while more socially cohesive emotions received greater cultural emphasis. The passionate feeling of liget described by Rosaldo (1980) in the Philippines had some similarities to anger, but was seen by Ilongot speakers in a more positive light, as a constructive force motivating much worthwhile human activity. Links between cultural identity and emotions identified by particular words, often held to be untranslatable, have long been recognised. They

4

Anna Wierzbicka and Jean Harkins

have often been seen as an important key to understanding cultural patterns and motivations, as for example in Benedict's (1946) discussion of the Japanese emotion haji in comparison with English shame and guilt. A Japanese psychiatrist has devoted a whole book to analysing "a particularly Japanese emotion" (Doi 1981: 169), amae, which he views as central to personality structure and social relations in his own culture. This is a good example of a practically untranslatable emotion concept, discussed by various writers in terms of affection, love and dependency, but clearly not equivalent to any of these English words. More recent crosscultural studies have explicitly challenged prevailing "Western" views of emotion and its place in relation to personal and social life. As Russell (1991) has pointed out, "emotion" is itself a cultural category, and there are many groups of people in the world who neither recognise such a category nor have a word for it (cf. also Wierzbicka 1995, 1999). Lutz (1985, 1988) has deliberately avoided importing categories from outside, aiming instead to describe emotional meanings as they emerge through social and interpersonal processes in Ifaluk society. The thorough and insightful account that results from this approach includes extensive discussion of many Ifaluk words that involve emotional meanings, and of how these words differ in meaning from various English words for emotions. The concept of emotion is also being reexamined in the light of poststructuralist critiques of essentialist approaches to personality and to culture. If selves are understood as shifting and plural, it may become difficult to identify anything like an emotion or an experiencer of it. Kondo (1990) presents a view of Japanese identities as fluid and unbounded, constantly created and recreated through a variety of discourses within the society. At the same time, her account shows that the discourses through which these selves are crafted rely upon many linguistic elements, and that for example the interdependency associated with the emotion amae mentioned above plays an important part in the fluidity of boundaries among the selves she describes. Thus, even approaches that call into question central assumptions about emotion and self are still having to deal with language data: with the ways that people talk about what they are feeling, and with the words that they use in whatever language they speak. The studies presented here are intended to illustrate that examining these words and the contexts in which they are spoken can lead to a better understanding of the meanings associated with them.

Introduction

5

2. Linguistic evidence in studying emotions The discussion so far has touched upon several matters in which caution is necessary in approaching linguistic data concerning emotions. All of the papers in this volume challenge the kind of ethnocentric view that assumes that an English word such as anger (or a French or Ilongot word like colere or ligei) denotes the same experience or concept as its supposed equivalents in other languages or in other cultural contexts. Neither do these studies assume that any society is so homogeneous as to have a single, unitary understanding or ideology of, for example, anger, colere or liget that is shared by all its members or implied by every use of a corresponding word. What we do assume is that a careful examination of linguistic data can provide clues to what people mean when they use such words or expressions. Without some such empirical evidence, any assertions about similarities or differences in emotional meanings, either within or across cultures, would be hard to maintain. By looking closely at how words are used, we can observe both similarities and differences in the use of an English word like anger and its equivalents or near-equivalents in other languages. This emotion is of particular interest here because "anger" is often assumed to be basic to human psychology, along with a small group of other emotions including "fear", "surprise", "sadness", "joy", "disgust" (see Izard 1977; Ekman 1992, 1993; Moore et al. 1999). Anger is indeed intuitively simpler than related emotion concepts like outrage or indignation (in English), which are often explained in terms of it. For example, outrage is considered to be a "stronger version" of anger, and indignation to be anger arising from some specific offence or injustice. Thus it may appear that "anger" is basic to a number of other emotions, and in lexicography as well as psychology there is a well established tradition of defining complex emotions in terms of ones that appear to be simpler. While this may be a reasonable or at least practical approach within the English language, the crosslinguistic evidence raises some questions about whether "anger" really is a basic human experience. The Yankunytjatjara language of Central Australia, for example, has three main translation equivalents for the English anger or angry (as in This person is angry). These three Yankunytjatjara words are examined in detail by Goddard (1991, 1998), who shows that none of them is sufficiently general to serve as a basic equivalent for angry, as each is appropriate in some contexts but not others. The three words focus on different as-

6

Anna Wierzbicka and Jean Harkins

pects of angry feelings and behaviour: pikaringanyi is associated with active hostility, mirpanarinyi with a sense of grievance, and kuyaringanyi with resentfulness. None of the three is basic or general enough that the others can be defined in terms of it. In the light of evidence like this, it would surely be ethnocentric to assert that Yankunytjatjara speakers are really experiencing a basic emotion "anger" but that they call it different things on different occasions, suggesting that they do not or cannot generalise by giving it a single label the way English speakers do. Another language, Malay, has a single word, marah, that is usually translated as angry. But on closer examination of its use in Malay, Goddard (1996, 1998) finds evidence that the concept of marah is not identical with the English concept of anger. For example, the word marah cannot be used to describe a protester's feelings about the destruction of rainforests; it is a more personally oriented feeling of having been treated badly. In this way it is somewhat closer to the English offended, and both the attitudes and behaviour associated with Malay marah are consistent with this. But in most contexts there is no better translation for angry in Malay than marah. Here again, it would be ethnocentric to suggest that Malay speakers have no word for the supposedly basic emotion of "anger", but only for a more complex feeling of personal offence. Obviously these words from other languages share some elements of meaning with the English anger. Many of them involve "feeling something bad because someone did something", many involve "wanting to do something", and often, "wanting to do something (bad) to the person who did this". Semantic elements like these may indeed be common to words for anger-like feelings in many, perhaps even most languages, and in this sense there may after all be a basic human experience of something like "anger" (cf. Wierzbicka 1998 a, 1999). But in the light of crosslinguistic evidence, it would be wrong to say that an emotion concept precisely equivalent to the English anger, or indeed to the Malay marah, or any other language-specific emotion term, is common to all humans. Whatever does emerge as a possible "emotional universal" in this area will need to be identified in universal terms, not in terms of the English emotion lexicon. The papers in this volume focus on words and other linguistic expressions related to emotions from a diverse selection of languages, and examine in depth the linguistic evidence concerning their meanings. Attention is devoted not only to words for emotions, but also to a range of other

Introduction

7

lexical and grammatical devices from interjections to reflexives that are deployed in various languages to express emotional meanings. In analysing the language data, we show that it is possible to identify what elements of meaning are specific to the language under examination, and what elements are shared with similar words and expressions from other languages. While this book is written in English, we explicitly recognise the limitations imposed by the language of inquiry. We distinguish in a principled way between words belonging to particular languages, and elements of meaning that may be applicable across languages and cultures. Throughout the volume we indicate typographically whether we are speaking of a language-specific word or concept, such as English anger or Malay marah (in italics), an English gloss such as 'anger' (in single quotation marks), or elements of meaning involved in such a concept, such as "feeling something bad". If a more general reference is made without typographic marking, for example to an angry expression or an anger-like feeling, it is recognised that these are English words and cannot be assumed to apply across languages and cultures.

3. Emotion concepts as cultural artefacts The semantic differences among words for anger-like feelings in the languages mentioned above are fairly subtle, and the similarities in meaning may seem to outweigh the differences. It could be argued that the emotion denoted by these various words is essentially the same, and that any differences in meaning are due to the social context, that is, to when and how these feelings emerge and are expressed in social interaction. From this perspective, the fact that people can be said to feel anger but not marah about environmental damage, or that one has to choose between describing someone as actively pikaringanyi or broodingly kuyaringanyi seem to be cultural and individual matters. However, several different lines of recent research in the cognitive sciences have independently arrived at a remarkably similar view of what constitutes an emotion, suggesting that socially accepted ways of thinking about the kinds of events that provoke anger, marah or kuyaringanyi and the kinds of behaviour that result from these feelings are integral parts of the emotion itself. Analyses of emotions by cognitive psychologists such

8

Anna Wierzbicka and Jean Harkins

as Frijda (1986), Ortony et al. (1988), Johnson-Laird and Oatley (1989), social constructionists like Harre (1986), and linguistic semanticists like Goddard (1995 a, b, 1996,1998); Harkins (1990,1996); Wierzbicka (1986, 1992 a, b, 1999); Ye (In press) and Zalizniak (1992) all indicate that to produce an adequate description of an emotion one must include both situational or social and cognitive elements in addition to the obvious affective elements. If this is correct, then we must indeed regard anger, marah, pikaringanyi and so on as different concepts, though clearly closely related ones; and we cannot treat English emotion words such as anger as neutral, self-explanatory, and culture-independent terms by means of which human emotional experience in all cultures can be validly and meaningfully described. There are of course many scholars who claim that nothing is truly selfexplanatory and everything is culture-dependent, and who appear to accept and even to rejoice in the idea that there is no way out of a "hermeneutical circle". Charles Taylor (1979 [1971]: 34) applied this idea specifically to emotions when he wrote that "our understanding of these terms moves inescapably in a hermeneutical circle. An emotion term like 'shame', for instance, essentially refers us to a certain kind of situation, the 'shameful', or 'humiliating'... But this situation in its turn can only be identified in relation to the feelings which it provokes... We have to be within the circle." There is an important truth in what Taylor is saying here, but it is a partial truth, as discussed more fully in Wierzbicka (1999). It is true that there are "communities of meaning", sharing a familiarity with certain common meanings, such as, for example, the meaning of the Russian words toska (roughly, 'melancholy-cum-yeaming') or zalet' (roughly, 'to lovingly pity someone'; for detailed semantic analyses see Wierzbicka 1992 a; Zalizniak 1992), or the Ifaluk concept fago (roughly, 'sadness/compassion/love', cf. Lutz 1995). It is also true that verbal explanations of such concepts cannot replace experiential familiarity with them and with their functioning within the local "stream of life" (to use Wittgenstein's phrase, cf. Malcolm 1966: 93). But it is not true that no verbal explanations illuminating to outsiders are possible at all. The crucial point is that while most concepts (including toska, zalet', fago, shame, emotion, implication) are complex (decomposable) and culture-specific, others are simple (non-decomposable) and universal (e.g. FEEL, WANT, KNOW, THINK, SAY, DO, HAPPEN, IF); and that the former can

Introduction

9

be explained in terms of the latter. For example, while there is no word in English matching the Russian word toska, one can still explain to a native speaker of English what toska means, relying on concepts shared by these two languages (as well as all other languages of the world): it is how one feels when one wants some things to happen and knows that they cannot happen (see Wierzbicka 1992 a). Crucially, this (simplified) definition can be translated word for word into Russian, and tested against the intuitions of native speakers. Shared, universal concepts such as FEEL, WANT, KNOW, THINK, SAY, DO, HAPPEN and IF (in Russian, CUVSTVOVAT', XOTET', ZNAT', DUMAT', SKAZAT', SDELAT', SLUCiT'sjA, ESLi) constitute the bedrock of intercultural understanding. And these concepts are the stepping stones by which we can escape the "hermeneutical circle". Since the cognitive scenarios linked with guilty and toska can be stated in the same, universal human concepts (such as FEEL, WANT, BAD, DO, and so on), these scenarios can be understood by cultural outsiders, and the kinds of feeling associated with them can be identified, explained, and compared; and both the similarities and differences between scenarios lexicalised in different languages can be pinpointed. But the very possibility of comparisons rests on the availability of a universal tertium comparationis, provided by universal concepts like FEEL, WANT, BAD, GOOD or DO, that is, simple concepts which are maximally self-explanatory and at the same time maximally culture-independent.

4. Natural Semantic Metalanguage (NSM) as a tool for crosscultural analysis The basic idea underlying this book is that language is a key issue in "emotion research" and that progress in the understanding of "emotions" requires that this issue be squarely addressed. "Human emotions" vary a great deal across languages and cultures, but they also share a great deal. Neither the diversity nor the universal aspects of "emotions", however, can be studied without an appropriate metalanguage. All attempts to study "human emotions" in terms of ordinary English (or any other natural language) are bound to lead to distortions, because every natural language contains its own "naive picture of the world" (cf. Apresjan 1992 [1974], 2000), including its own "ethnopsychology". By relying, uncritically, on

10

Anna Wierzbicka and Jean Harkins

ordinary English words we unwittingly fall prey to the "naive picture" that is reflected in them. Nor can we transcend this "naive picture" by adopting the conventional technical language of traditional psychology, because this conventional language has developed on the basis of ordinary language and is coloured by the naive picture embedded in it. The reliance of many psychologists on English folk-distinctions such as that between emotion and mood is a good case in point (cf. Mandler 1975). Despite insights from, inter alia, psychologists like Mandler, the faith in technical English as a suitable metalanguage for describing human emotions across languages and cultures is still deeply entrenched in emotion research in general, and the psychological literature on emotion in particular. For example, Van Geert (1995: 260) writes (with reference to one of the present authors): "Anna Wierzbicka argues (correctly, I think) that a theory of emotion universals cannot be stated in terms such as 'sadness is a universal emotion' because the word 'sadness' has a culture- and languagespecific meaning, and can therefore not be used to refer to universal meanings". But the solution that Van Geert proposes is a retreat into a technical language (that is, in practice, technical English): It is very possible that the number of [emotion-related] physiological patterns is limited and universal, but that there is no universality in the corresponding subjective experience. That is, the universals are of an entirely technical nature, they require particular technical instruments and theories and do not correspond with intuitive experiential categories accessible to the layperson. [... ] The emotions are experiential universals, that is, all people are in principle able to entertain a set of similar subjective experiences called the universal emotions. But the only way to refer to such shared emotions is by means of a technical language, (pp. 261-262)

The claim that ordinary people's subjective experience can only be validly discussed in a technical language (which means, in effect, technical English), is, in our view, a form of ethnocentrism, as well as misguided scientism. Van Geert suggests that only an expert, armed with technical language, can know what an ordinary person (non-expert) "really" experiences: "what a person experiences is not a separable set of such emotion components, but an inseparable subjective unity. It requires technical expertise and a technical language in order to successfully entangle [sic] and identify the components". (Cf. also Kupperbusch et al. 1999). But the goal of devising a technical language for describing other people's emotions is very different from that of getting insight into other

Introduction

11

people's subjective experience and their own way of thinking about it. Here, technical English will help us even less than ordinary English. In ordinary English, we can single out some words (and grammatical constructions) which can be rendered in the language of the speakers whose experience we are trying to understand - words like "good" and "bad", "feel", "want" and "think", "people", "body", "part", "inside" and so on. Technical English, on the other hand, is totally unrelated to the experience of people in other cultures - or, for that matter, that of most people in English-speaking cultures. "Technical universals" of emotion (whatever they are) cannot throw light on ordinary human experience and conceptualisation of experience. To understand this ordinary experience, and the way "ordinary" people think about it, we must pay attention to the way "ordinary" people talk; and to try to understand their talk we must rely on concepts "they" (the "laypeople") and "we" (the "experts") share. In this task, simple and universal words such as "good" and "bad", "feel", "want" and "think", "people", "body", "part" and "inside" will be our most reliable tools. (For further discussion, cf. Enfield and Wierzbicka, In press.) The approach to "emotions" adopted in most chapters of this book (and in earlier publications by the editors and colleagues) seeks to break the dependence on any one natural language as the source of "common sense insights" by anchoring the analysis in universal human concepts and their "universal grammar", according to the Natural Semantic Metalanguage methods developed by Wierzbicka and colleagues. Both the universal concepts and their universal rules of combination (i.e. their "grammar") have been arrived at by empirical crosslinguistic investigations carried out by several linguists over many years, working with typologically diverse and genetically unrelated languages, including Chinese (Chappell 1986, 1991, 1994), Japanese (Hasada 1997, 1998; Onishi 1994, 1997; Travis 1997), Malay (Goddard 1995 a, 1996), Lao (Enfield In press), Mbula (Bugenhagen 1994), Ewe (Ameka 1990 a, 1990b, 1994), French (Peeters 1994, 1997), several Australian Aboriginal languages (Goddard 1991b; Harkins 1995, 1996a; Harkins and Wilkins 1994; Wilkins 1986), among others. Most words in any language are specific to this particular language or to a group of languages, and are not universal. For example, neither English nor Spanish nor Malay has a word with a meaning corresponding exactly to the meaning of the German word Angst (cf. Wierzbicka 1999). At the same time, there is evidence that all languages have words with meanings corre-

12

Anna Wierzbicka and Jean Harkins

spending exactly to the meanings of the English words good and bad, or know and want. This suggests that the concepts of "good" and "bad", "know" and "want" are universal, and can therefore be used as elements of a culture-independent semantic metalanguage. Furthermore, the use of conceptual primitives allows us to explore human emotions (or any other conceptual domain) from a universal, language-independent perspective. Since every language imposes its own classification upon human emotional experience, English words such as anger or sadness are cultural artefacts of the English language, not culturefree analytical tools. On the other hand, conceptual primitives such as 'good' and 'bad', or 'want', 'know', 'say' and 'think' are not cultural artefacts of the English language but belong to the universal "alphabet of human thoughts"; and they do appear to have their semantic equivalents in all languages of the world. Basing our analysis on lexical universals we can free ourselves from the bias of our own language and reach a universal, culture-independent perspective on human cognition in general and on human emotions in particular. The latest empirical evidence (cf. Goddard and Wierzbicka (eds.) 1994, In press; Wierzbicka 1996; Goddard 1998) suggests that the full set of hypothetical conceptual primitives, lexicalised in all languages of the world, looks something like this: Substantives

I,

YOU,

SOMEONE

(PERSON),

SOMETHING

(THING), PEOPLE, BODY Determiners Quantifiers Attributes Mental predicates Speech Actions, events, movements Existence and possession Life and death Logical concepts Time

THIS, THE SAME, OTHER ONE, TWO, SOME, MANY/MUCH, ALL GOOD, BAD, BIG, SMALL THINK, KNOW, WANT, FEEL, SEE, HEAR SAY, WORD, TRUE DO, HAPPEN, MOVE THERE IS, HAVE LIVE, DIE NOT, MAYBE, CAN, BECAUSE, IF WHEN (TIME), NOW, AFTER, BEFORE, A LONG TIME, A SHORT TIME, FOR SOME TIME

Space

WHERE (PLACE), HERE, UNDER, ABOVE, TOUCH (CONTACT), BELOW, FAR, NEAR; SIDE, INSIDE

Intensifier, Augmentor Taxonomy, partonomy Similarity

VERY, MORE KIND OF, PART OF LIKE

Introduction

13

To reflect the special status of such words as exponents of universal human concepts, we can render them in capital letters, as GOOD and BAD, or BUENO and MALO (Spanish), or BAIK and BURUK (Malay), thus indicating that they are being used as elements of a special semantic metalanguage. At the same time we can identify them with the meanings of ordinary English, Spanish and Malay words (good and bad, bueno and malo, baik and buruk), and require that semantic formulae including these words be testable via natural language. Since the words of ordinary language are often polysemous, we need to identify the meanings in question by means of specified "canonical" sentences such as, for example, this person did something bad (English), esta persona ha hecho algo malo (Spanish), and orang ini buat sesuatu yang buruk (Malay), or something good happened to me (English), algo bueno me ha sucedido (Spanish), sesuatu yang baik terjadi kepada aku (Malay). Proceeding in this way, we can overcome both the incomprehensibility and unverifiability of a technical language relying on "experience-distant" concepts, and the ethnocentrism of descriptions using a full natural language such as ordinary English, in all its culture-specific richness. Thus, a configuration of conceptual primes such as "I feel (something) good now" appears to be universally possible and can therefore be plausibly proposed as the meaning of a smile, in preference to culture-specific English words like enjoyment or happy. Similarly, configurations such as "I want to do something", "I know I can't do anything", or "I know: something bad happened" also appear to be universally present and can be assigned as plausible semantic components to emotion words such as, for example, the English sad and the Russian grust', helping to map the similarities and differences between them.

5. Describing feelings through prototypes In literature, feelings are frequently described by means of comparisons: the hero felt as a person might feel in the following situation... (which the author then describes). To give just one typical example from Tolstoy's novel Anna Karenina, when Anna has finally left her husband: He felt like a man who has just had a tooth drawn which has been hurting him a long time. (Tolstoy 1970[1918]: 254)

14

Anna Wierzbicka and Jean Harkins

This is quoted from the translation by Louise and Aylmer Maude; for more detailed discussion of this literary practice, see Wierzbicka (1973). The same mode of description is also often used in everyday discourse, as well as in popular songs and other similar texts. A simple example is seen in the blues song, "Sometimes I feel like a motherless child". Much could of course be written about what it means to "feel like a motherless child", but the expression "I feel like..." itself cannot be defined or explained any further: it is as simple and clear as anything can be. There is no point in trying to define or explain the meaning of "I", "feel", "like", or the combination "I feel like (this)". The understanding of the whole line depends not only on the assumption that one knows (or can imagine) how "motherless children" feel, but also that the meaning of the expression "I feel like this" is intuitively clear. Although "feeling" cannot be defined, "ordinary people" generally assume that the way one feels can be described and that one can tell other people how one feels. There are many ways of describing to other people how one feels, but most of them can be reduced to three basic modes: (1) one can tell other people that one "feels good" or that one "feels bad"; and (2) one can tell other people that one feels like a person feels in a certain situation and then identify, in one way or another, that prototypical situation, (3) one can tell people what seems to be happening inside one's body. If I tell someone that I "feel wonderful" or that I "feel awful", I am following the first mode of describing feelings. If I tell them that I "feel like a motherless child", or that I "feel lost", or that I "feel abandoned", I am following the second mode. If I tell them that "my heart is breaking" or that "my heart is heavy", I am following the third. For example, we can say that to "feel pleased" means, roughly, "to feel something good, like a person does who thinks: something good happened, I wanted this to happen"; that to "feel delighted" means, roughly, "to feel something very good, like a person does who thinks: I know now that something very good happened, I didn't know before that this would happen"; that to "feel hope" means, roughly, "to feel something good, like a person does who thinks: I don't know what will happen, some good things can happen, I want these things to happen"; and that "to feel fear" means, roughly, "to feel something bad, like a person does who thinks: I don't know what will happen, some bad things can happen, I don't want these things to happen". Using a standardised mode of semantic description we can represent these meanings as follows:

Introduction

15

Pleased (X was pleased) (a) (b) (c) (d) (e) (f) (g)

X felt something because X thought something sometimes a person thinks: "something good happened I wanted this to happen" when this person thinks this, this person feels something good X felt something like this because X thought something like this

Delighted (X was delighted) (a) X felt something because X thought something (b) sometimes a person thinks: (c) "I know now: something very good happened (d) I didn't know that this would happen" (e) when this person thinks this, this person feels something very good (f) X felt something like this (g) because X thought something like this Hope (Xfelt hope) (a) (b) (c) (d) (e) (f) (g) (h)

X felt something because X thought something sometimes a person thinks: "I don't know what will happen some good things can happen I want these things to happen" when this person thinks this, this person feels something good X felt something like this because X thought something like this

Fear

(Xfeltfear)

(a) (b) (c) (d) (e) (f) (g) (h)

X felt something because X thought something sometimes a person thinks: "I don't know what will happen some bad things can happen I don't want these things to happen" when this person thinks this, this person feels something bad X felt something like this because X felt something like this

16

Anna Wierzbicka and Jean Harkins

In these semantic formulae, the symbol "X" stands quite simply for "this person". This symbol is used for convenience, because in reading the formulae quickly, confusion might arise between the "this person" of the utterance scenario (this person (X) was pleased) and the "this person" of the prototype scenario (sometimes a person thinks ...)· The two are in fact quite distinct on careful reading, as the prototype scenario is a self-contained subset of the utterance meaning. The composition of semantic representations of emotion words and expressions is more fully explained in Wierzbicka (1999).

6. Emotion and culture The literature on "emotions" often contrasts "biology" with "culture", as if the two were mutually exclusive. It is worth remembering, therefore, that it was actually a psychologist (William James), not an anthropologist or a linguist, who said that the categorisation of feelings depends on "the introspective vocabulary of the seeker", which in turn depends on his or her language and culture (James 1890: 485). James held that feelings represented the subjective experience of biological (physiological) events, but he recognised that feelings can be categorised in a variety of ways, and that they are differently categorised in different cultures. This is not to say that there are no common threads. But the diversity is very considerable indeed. The meaning of English "emotion words" has actually changed a great deal in the course of history. Had Shakespeare been interested in proposing a basic "emotional keyboard" (cf. Shweder 1985: 200) it would have been different from that proposed by twentieth-century psychologists - even if it contained some of the same words, for example angry. The view of anger as something that can be manipulated - "controlled", "vented", "released", left "unresolved", "directed" at this or that target, "stirred up", "repressed", "expressed", "suppressed" and so on (for examples see, e.g., Pendergrast 1998: 23, 24, 219, 242, 243, 364) - is entirely modern and goes far beyond the semantic range of the Shakespearian anger (cf. Logan 1998; cf. also Stearns and Stearns 1986). What this example of historical change shows is that an apparently "basic and fundamental human concept" like anger is in fact linked with a certain cultural model, and so cannot be taken for granted as a "culture-free" analytical tool or as a universal standard for describing "human emotions".

Introduction

17

Examples of this kind provide an answer for those who, like Paul Harris (1995), ask "whether we can draw any conclusions - other than lexical conclusions - about the emotional universe of a culture by examining its emotion lexicon." The answer is that, by examining the meaning and the use of words like anger and angry in contemporary English, we can indeed learn a great deal about the "emotional universe" of the speakers of contemporary English. "Emotion words" such as anger reflect, and pass on, certain cultural models; and these models, in turn, reflect and pass on values, preoccupations and frames of reference of the society (or speech community) within which they have evolved. They reflect its "habits of the heart" (Bellah et al. 1985) and the concomitant "habits of the mind". The English anger and angry (with their current range of meanings and use) both reflect and reinforce what Bellah et al. (1985) call the "therapeutic culture" in modern Anglo society; and the shift from the Shakespearian wrath to modern anger both reflects, and constitutes an aspect of, the democratisation of society and the passing of the feudal order (cf. de Tocqueville 1953 [1835-40]; Stearns and Stearns 1986; Logan 1998). Naturally, it is not only the lexicon that provides clues to the "emotional universe of a culture". Grammar does too, as do phraseology, discourse structure, gestures, intonation, interjections, swearwords, forms of address, culture-specific facial expressions and bodily postures, and so on. For example, it is clearly significant that "active" verbs like rejoice have all but disappeared from modern English usage, giving way to "passive" adjectives like happy or pleased; and that those "emotion verbs" which remain tend to have pejorative or humorous connotations (cf. e.g. fume, fret, sulk, pine, enthuse, rage, and so on; for detailed discussion see Wierzbicka 1988, 1995; cf. also Hübler 1998, Dirven 1997, Osmond 1997 on various emotion-related grammatical patterns). Similarly, the Russian "emotional universe" is reflected in Russian expressive derivation, including notably the numerous "diminutive" suffixes with different emotional shadings encoded in each of them (for detailed discussion see Wierzbicka 1992b; see also Friedrich 1997). It is also reflected in the culture-specific phraseology centred on the human body and expressive bodily (and facial) behaviour (cf. lordanskaya and Paperno 1996; Wierzbicka 1999). The semantic explorations in this volume go far beyond the lexical semantics of emotion words and their equivalence or otherwise in different languages. As several of these authors point out, the "emotional world" or "emotional universe" of the speakers of Chinese, Russian, Mbula or Malay

18

Anna Wierzbicka and Jean Harkins

is much richer than just their set of lexical labels for particular feelings. "Emotion talk" or talk about people's feelings - one's own or other people's feelings or both - may receive greater or lesser emphasis in each cultural and social setting, but in every language examined thus far, people don't just name feelings: they have recourse to a whole realm of discourse about human feelings and their manifestations. (Cf. also Palmer 1996; Palmer, Bennett and Stacey 1999, among others.) The domains of emotion discourse and phraseology in each language and culture are represented in these studies in a wealth of different ways. Semantic analysis of particular emotion concepts has to pay close attention to evidence from their distinctive collocations and associated phraseology. To give just one example from this volume, an understanding of the nature of the Malay hau or seat of emotion relies substantially on the discourse concerning it: the hau can "speak" to the person, though it apparently doesn't "move" (unlike the seat of emotion in many other languages); it can be sakit 'sick',/7«os 'content', kecil 'small' and many more - and each of these feeling-states has its own semantic content that cannot be derived by simply adding together the meanings of the individual words. Such phraseology can involve both metaphorical and non-metaphorical cultural models (cf. Emanatian 1999; Gibbs 1999; Yu 1998), which in turn pose challenges for how to represent these meanings in semantic explication. Even the non-verbal facial and bodily signals of emotional feeling are in turn given linguistic representations - and these too have their distinctive interpretations in each language and culture, so that the Mbula mata- sigsig (lit. 'eye red') hasn't anything to do with crying, but rather with eagerness; and the Japanese hiya-hiya 'shivery feeling' isn't so much a "frisson" as a helpless apprehensiveness. It is this wealth of associated discourse that makes the domain of feelings such a rich field for linguistic, anthropological, psychological and social science investigation. Some evidence is seen in these studies for possible universals of emotion phraseology, for example, description of emotional feelings via bodily events, sensations, images (Wierzbicka 1999), and culture-specific syntactic and pragmatic patterns (Goddard Forthcoming a; Enfield In press). All these aspects of both verbal and non-verbal communication deserve careful attention and analysis both within their specific linguistic and cultural contexts, and across cultures. None of them, however, can be studied with precision and without an ethnocentric bias if attention is not given to the vital matter of a suitable metalanguage. Whether one wishes to focus

Introduction

19

on the study of the emotion lexicon, collocations, grammatical devices, common metaphors, interjections, gestures, facial expressions, or whatever, one cannot escape the issues of methodology; and the question of an appropriate metalanguage is a key methodological issue.

7. Cultural semantics in cognitive linguistics Each of the language-specific studies included in this volume demonstrates the vital contributions that can be made by means of detailed analysis of emotional meanings based on empirical linguistic evidence. This places the work firmly within the broader cognitive linguistics enterprise, although we recognise that the pluralism of approaches valued by cognitive linguists like, for example, Niemeyer (1997), Athanasiadou and Tabakowska (1998) and Langacker (1999) is criticised by some who would prefer to narrow the field. For example, Athanasiadou and Tabakowska (1998: xxi), in the same series as this volume, remark that their book "represents a wide spectrum of cognitive trends, thereby testifying to pluralism within the cognitive linguistic paradigm: the metaphorical-metonymical Lakoffian approach (Kovecses), the semantic-primitives approach (Wierzbicka), and the semasiological-structure approach (Geeraerts/Grondelaers)." On the other hand, other scholars have strong ideological objections to an analysis of linguistic and cultural meanings based on a theory of conceptual primes. Goddard (Forthcoming b), contrasts the view quoted above with that of Geeraerts (1997, cf. 1999), who, he says, "characterises present-day cognitive linguistics as having two methodological extremes. In the 'good corner' there are the data-driven, empirically-minded linguists doing psycholinguistics, neurophysiological modelling and corpus analysis. In the 'bad corner' there is the 'idealistic tendency' represented by Wierzbicka and her colleagues, with their appeals to intuition and platonistic views about conceptual primes." Another apparent conflict of emphasis here concerns the issues of visual (diagrammatic) versus verbal (paraphrastic) representations of meaning, and the emphasis on either the experiential or the conceptual aspects of communication and cognition. Our own view is, like Goddard's (Forthcoming b), that the two approaches should be seen as complementary rather than inimical:

20

Anna Wierzbicka and Jean Harkins Just as it is easy to fall into "either-or" thinking in relation to verbal and diagrammatic representation, so too with the contrast between symbolic (conceptual) meaning and its experiential (pre-conceptual) underpinnings. However, one may very well grant that embodied, pre-conceptual experiential Schemas (kinaesthetic image-schemas) underlie, constrain and support the emergence of conceptual meaning, without accepting that conceptual meaning is reducible to experiential Schemas.... [B]oth conceptualist and experientialist perspectives are needed if we are to get anything approaching a full picture of language as a human phenomenon (just as we need both universalism and relativism...)

Conceptual analysis of meaning carried out in universal semantic primes is, in our view, essential to the understanding of human communication, cognition, and emotion; and we see it as a necessary part of cognitive linguistics, alongside other models and approaches. The usefulness of this approach in the area of culture-laden aspects of language, such as the language of values and emotions, seems to us particularly clear in the studies presented here.1 With our strongly semantic orientation and our concern for precise yet culture-free specifications of linguistic and cultural meanings, we aim both to complement the work of other cognitive scientists by contributing analyses based on rich linguistic data; and to challenge them to strive also for maximally clear and culturally unbiassed analyses and explanations within their own research frameworks. This volume joins the valuable body of crosslinguistic and crosscultural studies assembled in recent years by Niemeier and Dirven (1997), Athanasiadou and Tabakowska (1998), Palmer and Occhi (1999), Gibbs and Steen (1999), but seeks to extend and refine the ways we deal with both the universal and the culture-specific dimensions of human emotional communication and meaning.

8. Cognitive and semantic analysis in practice Above all, our hope is that these papers will encourage others to experiment with the methods demonstrated in them, as useful research tools for the investigation of emotions and related aspects of culture. This is an eminently practical way of exploring cultures. Of the increasing number of linguists now actively using aspects of the NSM approach in their work, it would probably be fair to say that few have adopted it through an a priori theoretical attraction to semantic universals. People use this method first and foremost because it works, particularly for those who wish to explain cul-

Introduction

21

tural meanings without imposing external interpretive frameworks. These papers demonstrate its applications in linguistic fieldwork, corpus-based studies, literary analysis, and analysis of contemporary cultural material. The tools of linguistic analysis, and in particular, the use of a languageindependent analytical method provided by the Natural Semantic Metalanguage, can point the way out of the hermeneutical circle inherent in culture-dependent frameworks. In so doing we can, as so clearly shown in these studies, resolve much of the confusion surrounding the dual character of emotions in human experience. On the one hand, we can pinpoint, explain and validate in terms of linguistic evidence the culture-specific and culture-dependent nature of the set of categories of emotional experience identified by each cultural group through its language. And on the other, we can observe certain tendencies that do indeed seem to be universal, including the existence in all languages of words for cognitively-based feelings, and substantial overlaps between certain elements in the semantic composition of some of these, overlaps that can be specified in languageindependent terms, thus explaining the extremely common crosslinguistic phenomenon of partial but never perfect intertranslatability of at least some emotion words. The studies presented here demonstrate the usefulness of the proposed approach to studying a variety of different emotions and feelings expressed in a wide range of languages from diverse cultural groups around the world. Each of these papers examines a set of words or expressions used to talk about particular feelings in one language, giving an in-depth analysis of their semantic content, and thus showing what elements of meaning they may have in common with words for similar feelings in other languages, and what elements of meaning may be specific to the language under examination. The papers are not grouped thematically or by language groupings because the various themes explored in each of them link and resonate with several of the others, cutting across genetic, areal and thematic boundaries. While the authors share a commitment to seeking language-independent, non-ethnocentric ways of analysing culturally situated linguistic meanings, grounded in accurate linguistic description and tested against natural language data, readers will notice also a diversity of perspectives and practices in applying this semantic approach. As a research tool, the NSM metalanguage is undergoing vigorous development, and each application of it yields theoretical insights into patterns of linguistic meaning as well as

22

Anna Wierzbicka and Jean Harkins

into the structures of the languages under investigation. These papers demonstrate a greater convergence than previously found between different researchers' NSM formulations, but important variations in explicatory style and practice remain. Alternations in phrasing between, for example, "thinks something like:..." vs "thinks something like this:...", or "I didn't think before now: this will happen" vs "I didn't think before that this would happen", or "sometimes a person thinks" vs "sometimes people think" are more than trivial stylistic variations, in that they represent different positions with regard to the semantic valency (or combinatorial properties) of the primes "like" and "think", or the association of prototypical cognitive scenarios with individuals or groups. Such theoretical issues, which are only resolved through ongoing empirical work of the kind seen here, are not generally pursued in this volume, but are explored in Goddard and Wierzbicka (In press). Researchers should not be deterred from using this method to describe cultural meanings of interest to them, by feeling that they must master a rigidly formalised system of representation. Although the emerging formal properties of the metalanguage are indeed worthy of close examination (see Goddard and Wierzbicka In press), it can readily be employed as a practical tool for investigating meaning. Statements of meaning can be proposed, tested against more data, and tested for intelligibility and translatability as shown in several of these studies. It takes many successive revisions and much analytical work to arrive at optimally clear and accurate explications, but many researchers are now finding this process immensely rewarding in terms of the cultural insight gained through it, in addition to the usefulness of the outcomes for language teaching and intercultural communication. In the first paper, Mengistu Amberber demonstrates that the meanings of both lexical and grammaticised expressions of emotion in Amharic can be fully described in terms of semantic universale, allowing far greater accuracy than approaches based on partial translation equivalents. The polysemy of the verb (t3)s9tnma 'feel; hear' is resolved on formal, languageinternal criteria; and it is found that the Amharic data support the main proposals that have been put forward about universal trends in the linguistic expression of cognitively-based feelings. The anger-like verb tak'ot't'a reveals an interesting link between emotion and speech act, containing a semantic element of wanting to rebuke the person who aroused the angry feeling.

Introduction

23

The Mbula language of Papua New Guinea is one of many in its region that make extensive use of body image expressions, as the most productive and semantically rich means of talking about emotional experiences. Robert Bugenhagen's detailed explications of expressions based on mata'eye', lele- 'insides', kete- 'chest, liver', ni- 'being', kuli- 'skin', kopo'stomach' show how a language that reflects a view of the human person markedly different from "Western" concepts (cf. also McVeigh 1996), draws nonetheless on the same kinds of semantic components (THiNKing something, FEELing something GOOD/BAD, WANTing to do something) that other languages do in emotion expressions. These explications occasionally introduce elements from outside the basic NSM set of primes. Although such elements could themselves be reduced to combinations of primes, it sometimes works better to retain them as "semantic molecules" to enhance the readability of long definitions, and explicate these molecules separately elsewhere. Turning to a European language, a corpus-based analysis by Uwe Durst of Ärger, Wut and Zorn in German illustrates how even a very closely related language does not have an exact match for the supposedly basic English anger, but that the differences in distribution of these words can be explained in terms of subtle differences in their semantic structure. He also demonstrates some of the ways in which historical changes in cultural models, as suggested above, may be reflected in semantic changes and shifts in usage of a group such as the anger-like emotions. Several genetically unrelated languages of Asia provide rich sources of data and analysis. Ways of talking about facial expressions in Lao, studied by Nick Enfield, reveal not only a culture-specific set of associations between emotions and facial gestures, but also a distinctive ethnolinguistic perspective on what are the significant features or components of facial expression. At the same time, the Lao evidence provides support for proposals concerning the universality of "cry" and "smile" as facial expressions of feeling (cf. Wierzbicka 1999). While this paper does not propose explications in the same style as most of the others, it nonetheless demonstrates the same kind of empirical and non-linguocentric approach to the identification and discussion of meanings. Cliff Goddard gives a thorough overview of the Malay cultural keyword hati, referring to what is seen as the seat of human feelings and a highly productive source of linguistic expressions of emotion. Semantic analysis of this word provides a key to Malay ethnopsychology, both in its under-

24

Anna Wierzbicka and Jean Harkins

standing of human nature, and in the role of interpersonal feelings in Malay social and emotional life. Goddard points to some developments in NSM metalanguage syntax, such as the avoidance of semantically complex "and" and "feel something towards someone" in favour of combinations of syntactically simpler elements. Like some of the other contributors, he employs a simplified version of the prototype framework explained above. He also provides a table of the NSM primes in Malay, though he does not actually give full explications in that language. These latter themes are taken up in Jean Harkins' exploration of processes and issues involved in developing language-internal explications with speakers of Arrernte. A striking phenomenon in Japanese is explored by Rie Hasada, who shows how the semantic complexity of the psychomimetic words reflects unique cultural perspectives. These words are among the most difficult for learners of Japanese to grasp, despite their frequent use and importance to effective communication, and an analysis of them in terms of universal semantic elements can help to overcome this problem. The strongly embodied nature of these concepts raises issues of how to represent auditory elements and physical prototypes in explications (I could hear something like: "dokiQ"; this person thinks: "if I am in a cold place I will feel something like this" [hiya-hiya]). (Cf. also Occhi's 1999: 157, 167 observation about their power to invoke intuitive recall of an image.) The theme of anger-like emotions is revisited in Pawel Kornacki's study of a family of five Chinese words: nu, shenglqi, nao(huo),fen, and taoyan. Elements of these emotional states are compared with Western proposals concerning symptoms and stimuli related to anger-like feelings (cf. also Yu 1998), and some overlapping components of meaning are found, although they also differ in ways that are specified in Kornacki's definitions. As with anger-like emotions in many other languages, impulses to action (I want to do something...) are often involved, but in each case the role and phrasing of elements of this kind differs in important particulars. The rich emotional vocabulary of Russian is addressed by Irina Levontina and Anna Zalizniak, who cover more than a dozen emotion concepts that have not received in-depth treatment in the already extensive literature on Russian emotions. By focussing on aspects of the system of culturally significant stereotypes of experience reflected in the language (which could also be understood in terms of cultural models or imageschemas), they are able to explain some of the most frequently observed characteristics of Russian emotional life - and literary reflections there-

Introduction

25

of - without making unwarranted generalisations about "national character". A single highly salient Polish emotion, przykro, is analysed in depth by Anna Wierzbicka, who places it in context of the set of interpretive categories provided by the Polish language. She shows how the different set of categories provided by English would link similar feelings with several different emotion concepts including hurt, offended, sorry and feeling bad, and how przykro is linked also to Polish cultural values and expectations of interpersonal "good feelings", the perceived lack of which produces this painful emotion. Finally, Zhengdao Ye illustrates another area of Chinese ethnopsychology with a study of three of the most salient words in Chinese literature and poetry: bei, ai and chou, often glossed interchangeably as 'sadness', 'sorrow' and 'grief. She demonstrates conclusively that they are by no means interchangeable, nor are they equivalent to Western characterisations of "sadness". Moreover, she provides detailed semantic explications of these three emotion words in both English and Chinese metalanguage formulae - an important demonstration of the principles of intertranslatability of NSM formulae, and of language-internal definition (that each language has the capacity to provide its own definitional metalanguage). By presenting detailed semantic descriptions of culturally-situated meanings of culturally salient words used in the "emotion talk" in different cultures, we can offer glimpses into other people's emotional lives - without imposing on those lives a perspective derived from the vocabulary and other resources of our own native language. Since the descriptions presented here are phrased in universal, that is, shared, concepts, they can be both faithful to the perspective of the speaker whose emotions we purport to be talking about, and intelligible to others. (These others include scholars, who often don't seem to realise that they too are speakers of another language, with their own spectacles, tinted by their own native language.) We can combine the insiders' point of view with intelligibility to outsiders.

Notes 1. The work on the language of emotions carried out within the NSM framework includes Ameka (1990), Bugenhagen (1990), Goddard (1990, 1991, 1995 a, b, 1996,

26

Anna Wierzbicka and Jean Harkins 1997 a, b), Harkins (1990, 1995, 1996), Harkins and Wierzbicka (1997), Hasada (1997,2000, In press), Kornacki (1995), Mostovaja (1988), Priestley (In press), Travis (1998), Wierzbicka (1973, 1986, 1990a, b, 1992a, b, c, 1993, 1994a, b, 1995a, b, 1997, 1998a, b, c, 1999, 2000), Wilkins (1986), and Ye (2000, In press).

References Ameka, Felix 1990

The grammatical packaging of experiences in Ewe: a study in the semantics of syntax. Australian Journal of Linguistics 10: 139-182. 1994 Ewe. In Goddard and Wierzbicka (eds.), 57-86. Apresjan, Jurij D. 1992 Lexical Semantics. Ann Arbor: Karoma. [first pub. 1974, in Russian]. 2000 Systematic Lexicography. Translated by Kevin Windle. Oxford: Oxford University Press. Athanasiadou, Angeliki and Elzbieta Tabakowska (eds.) 1998 Speaking of Emotions: Conceptualisation and Expression. Berlin: Mouton de Gruyter. Bellah, Robert N., Richard Madsen, William M. Sullivan, Ann Swidler, and Steven M. Tipton (eds.) 1985 Habits of the Heart: Individualism and Commitment in American Life. Berkeley: University of California Press. Benedict, Ruth 1946 The Chrysanthemum and the Sword. Boston: Houghton Mifflin. Briggs, Jean 1970 Never in Anger: Portrait of an Eskimo Family. Cambridge, MA: Harvard University Press. Bugenhagen, Robert D. 1990 Experiential constructions in Mangap-Mbula. Australian Journal of Linguistics 10(2): 183-215. 1994 The exponents of semantic primitives in Mangap-Mbula. In Goddard & Wierzbicka (eds.), 87-108. Chappell, Hilary M. 1986 Formal and colloquial adversity passives in standard Chinese. Linguistics 24: 1025-1052. 1991 Strategies for the assertion of obviousness and disagreement in Mandarin: a semantic study of the particle me. Australian Journal of Linguistics 11: 39-65. 1994 Mandarin semantic primitives. In Goddard & Wierzbicka (eds.), 109-147.

Introduction

27

Davidson, Richard J. and Paul Ekman (eds.) 1994 The Nature of Emotion: Fundamental Questions. New York: Oxford University Press. de Tocqueville, Alexis 1953 Democracy in America. Translated by H. Reeve, edited by P. Bradley. New York: Alfred A. Knopf. [1835] Dirven, Rene 1997 Emotions as cause and the cause of emotions. In Niemeier & Dirven 1997: 55-83. Doi, Takeo 1981 The Anatomy of Dependence. Tokyo: Kodansha. Ekman, Paul 1992 An argument for basic emotions. Cognition and Emotion 6: 169-200. 1993 Facial expression and emotion. American Psychologist 48: 384-392. Ekman, Paul and Erika Rosenberg (eds.) 1997 What the Face Reveals. New York: Oxford University Press. Enfield, Nicholas J. In press The syntax of Natural Semantic Metalanguage expressions in Lao. In Cliff Goddard & Anna Wierzbicka (eds.) Meaning and Universal Grammar: Theory and Empirical Findings. Enfield, Nicholas J. (ed.) Forthcoming Ethnosyntax. Oxford: Oxford University Press. Enfield, Nicholas J. and Anna Wierzbicka. In press Introduction. In: Nicholas J. Enfield and Anna Wierzbicka (eds.), Pragmatics and Cognition. (Special Issue on Emotions and the Body.) Friedrich, Paul 1997 Dialogue in lyric narrative. In Michael Macovski (ed.) Dialogue and Critical Discourse, 79-98. New York: Oxford University Press. Frijda, Nico 1986 The Emotions. Cambridge: Cambridge University Press. Geeraerts, Dirk 1997 Idealistic tendencies in cognitive linguistics. Plenary address delivered at Fifth International Cognitive Linguistics Conference, Amsterdam, 15 July 1997. 1999 Idealistic and empiricist tendencies in cognitive semantics. In T. Jansen and G. Redeker (eds.), Cognitive Linguistics: Foundations, Scope, and Methodology. Berlin: Mouton de Gruyter. 163-194. Geeraerts, Dirk & Stefan Grondelaers 1995 Looking back at anger: Cultural traditions and metaphorical patterns. In Taylor and MacLaury (eds.), 153-179. Gibbs, Raymond W. Jr. 1999 Taking metaphor out of our heads and putting it into the cultural world. In Gibbs and Steen (eds.), 145-166.

28

Anna Wierzbicka and Jean Harkins

Gibbs, Raymond W. Jr. and Gerard J. Steen (eds) 1999 Metaphor in Cognitive Linguistics. Amsterdam: John Benjamins. Goddard, Cliff 1990 The lexical semantics of "good feelings" in Yankunytjatjara. Australian Journal of Linguistics 10(2): 257-292. 1991 Anger in the Western Desert: a case study in the cross-cultural semantics of emotion. Man 26: 265-279. 1995 a Cognitive mapping or verbal explication? Understanding love in the Malay archipelago. Semiotica 106: 323-354. 1995 b Conceptual and cultural issues in emotion research. Culture and Psychology 1 (2): 289-298. 1996 The social emotions of Malay (Bahasa Melayu). Ethos 24:426-464. 1997 a Cultural values and "cultural scripts" of Malay (Bahasa Melayu). Journal of Pragmatics 27 (2): 183-201. 1997 b Contrastive semantics and cultural psychology: "surprise" in Malay and English. Culture and Psychology 3 (2): 153-181. 1998 Semantic Analysis: A Practical Introduction. Oxford: Oxford University Press. Forthcoming a Ethnosyntax, ethnopragmatics, indexicality, and culture. In Enfield (ed.). Forthcoming b In defence of paraphrase and the place of NSM semantics in cognitive linguistics. In: June Luchjenbroers (ed.) Cognitive Linguistics Investigations across Languages, Fields, and Philosophical Boundaries. Goddard, Cliff and Anna Wierzbicka (eds.) 1994 Semantic and Lexical Universals: Theory and Empirical Findings. Amsterdam: John Benjamins. In press Meaning and Universal Grammar: Theory and Empirical Findings. Amsterdam: John Benjamins. Harkins, Jean 1990 Shame and shyness in the Aboriginal classroom. Australian Journal of Linguistics 10: 293-306. 1995 Desire in language and thought: A study in crosscultural semantics. PhD dissertation, Department of Linguistics, Australian National University. 1996 Linguistic and cultural differences in concepts of shame. In David Parker, Rosamund Dalziell & Iain Wright (eds.), Shame and the Modern Self, 84-96. Melbourne: Australian Scholarly Publishing. Harkins, Jean and Anna Wierzbicka. 1997 Language: a key issue in emotion research. In: Angeliki Kofler (ed.), Innovation 10(4): 319-331. (Special Issue of Innovation: the European Journal of Social Sciences.)

Introduction

29

Harkins, Jean and David P. Wilkins 1994 Mparntwe Arrernte and the search for lexical universals. In Goddard and Wierzbicka (eds.), 285-310. Harre, Rom (ed.) 1986 The Social Construction of Emotions. Oxford: Basil Blackwell. Harris, Paul 1995 Developmental constraints on emotion categories. In Russell et al. (eds.), 353-372. Hasada, Rie 1997 Some aspects of Japanese cultural ethos embedded in nonverbal communicative behaviour. In Fernando Poyatos (ed.) Nonverbal Communication in Translation, 83-103. Amsterdam: John Benjamins. 1998 Sound symbolic emotion words in Japanese. In Athanasiadou & Tabakowska (eds.), 83-98. 2000 An exploratory study of expression of emotions in Japanese: Towards a semantic interpretation. PhD dissertation, Department of Linguistics, Australian National University. In press Body part terms and emotion in Japanese. In Enfield and Wierzbicka (eds.) Hübler, Axel (1998) The Expressivity of Grammar: Grammatical Devices Expressing Emotion Across Time. Berlin: Mouton de Gruyter. lordanskaya, Lidija N. & Slava Paperno 1996 The Russian-English Collocational Dictonary of the Human Body (RECDHB). Columbus, Ohio: Slavica. Izard, Carroll 1977 Human Emotions. New York: Plenum Press. James, William 1890 The Principles of Psychology. London: Macmillan. Johnson-Laird, Philip & Keith Oatley 1989 The language of emotions. Cognition and Emotion 6: 201-233. Kondo, Dorinne 1990 Crafting Selves: Power, Gender, and Discourses of Identity in a Japanese Workplace. Chicago: University of Chicago Press. Komacki, Pawel 1995 Heart and face: semantics of Chinese emotion concepts. PhD dissertation, Department of Linguistics, Australian National University. Kövecses, Zoltän 1995 Metaphor and the folk understanding of anger. In Russell et al. (eds.), 49-71.

30

Anna Wierzbicka and Jean Harkins

Kupperbush, C, Matsumoto, D., Kooken, K., Loewinger, S., Uchida, H., Wilson-Cohn, C. and Yrizarry, N. 1999 Cultural influences on non-verbal expression of emotion. In: P. Philippot, R. S. Feldman and E. J. Coats (eds), The Social Context of Nonverbal Behaviour, 17-44. Cambridge: Cambridge University Press. Langacker, Ronald W. 1999 Grammar and Conceptualization. Berlin: Mouton de Gruyter. Logan, Sarah 1998 The changing meaning of English emotion terms. BA Honours thesis, Department of Linguistics, Australian National University. Lutz, Catherine 1985 Ethnopsychology compared to what? In Geoffrey M. White & John Kirkpatrick (eds.) Person, Self and Experience, 35-79. Berkeley: University of California Press. 1988 Unnatural Emotions: Everyday Sentiments on a Micronesian Atoll and their Challenge to Western Theory. Chicago: University of Chicago Press. 1995 Need, nurturance and the emotions on a Pacific atoll. In J. Marks and R. T. Ames (eds.) Emotions in Asian Thought: A Dialogue in Comparative Philosophy, 235-252. Albany: State University of New York. Malcolm, Norman 1966 Ludwig Wittgenstein: A Memoir. London: Oxford University Press. Mandler, George 1975 Man and Emotion. New York: John Wiley & Sons. 1997 Foreword. In: James A. Russell and Jose Miguel Fernändez-Dols, The Psychology of Facial Expression, xii-x. Cambridge: Cambridge University Press. McVeigh, Brian 1996 Standing stomachs, clamoring chests and cooling livers: Metaphors in the psychological lexicon of Japanese. Journal of Pragmatics 26: 25-50. Moore, C. C., Romney, A. K., Hsia, T. and Rusch, C. D. 1999 "The universality of the semantic structure of emotion terms: methods for the study of inter- and intra-cultural variability". American Anthropologist 101 (3): 529-546. Mostovaja, Anna D. 1998 On emotions that one can "immerse into", "fall into" and "come to": the semantics of a few Russian prepositional constructions. In: Athanasiadou and Tabakowska (eds.), 295-329. Niemeier, Susanne 1997 Introduction. In: Susanne Niemeier and Rene Dirven (eds.), vii-xvii.

Introduction

31

Niemeier, Susanne & Rene Dirven (eds.) 1997 The Language of Emotions: Conceptualization, Expression, and Theoretical Foundation. Amsterdam: John Benjamins. Occhi, Debra J. 1999 Sounds of the heart and mind: Mimetic s of emotional states in Japanese. In Palmer and Occhi (eds.), 151-170. Onishi, Masayuki 1994 Semantic primitives in Japanese. In Goddard and Wierzbicka (eds.), 361-385. 1997 The grammar of mental predicates in Japanese. Language Sciences 19: 219-233. Ortony, Andrew, Gerald L. Clore and Anthony Collins 1988 The Cognitive Structure of Emotions. Cambridge: Cambridge University Press. Osmond, Meredith 1997 The prepositions we use in the construal of emotions: Why do we say fed up with but sick and tired of! In Niemeier & Dirven (eds.), 111-133. Palmer, Gary B. 1996 Toward a Theory of Cultural Linguistics. Austin: University of Texas Press. Palmer, Gary B., Heather Bennett & Les Stacey 1999 Bursting with grief, erupting with shame: A conceptual and grammatical analysis of emotion-tropes in Tagalog. In Palmer & Occhi (eds.), 171-200. Palmer, Gary B. & Debra J. Occhi, eds. 1999 Languages of Sentiment: Cultural Constructions of Emotional Substrates. Amsterdam: John Benjamins. Peeters, Bert 1994 Semantic and lexical universals in French. In: Goddard and Wierzbicka (eds.), 423-442. 1997 The syntax of time and space primitives in French. Language Sciences 19: 235-244. Pendergrast, Mark 1998 Victims of Memory: Incest Accusations and Shattered Lives. London: Harper Collins. Priestley, Carol In press Insides and emotions in Koromu. In Enfield and Wierzbicka (eds.). Rosaldo, Michelle 1980 Knowledge and Passion: I hngol Notions of Self and Social Life. Cambridge: Cambridge University Press.

32

Anna Wierzbicka and Jean Harkins

Russell, James A. 1991 Culture and the categorisation of emotions. Psychological Bulletin 110:426-450. Russell, James A., Jose-Miguel Fernändez-Dols, Antony S. R. Manstead and J. C. 1995 Wellenkamp (eds.), Everyday conceptions of emotion. Dordrecht: Kluwer. Shweder, Richard 1985 Menstrual pollution, soul loss, and the comparative study of emotions. In Arthur Kleinman & Byron Good (eds.) Culture and Depression: Studies in the Anthropology and Cross-cultural Psychiatry of Affect and Disorder, 182-215. Berkeley: University of California Press. Stearns, Carol Z. & Peter N. Stearns 1986 Anger: The Struggle for Emotional Control in America's History. Chicago: University of Chicago Press. Taylor, Charles 1979 Interpretation and the sciences of man. In Paul Rabinow & William M. Sullivan, eds. Interpretive Social Science: A Reader, 25-72. Berkeley: University of California Press. [1971] Taylor, John and Robert E. MacLaury (eds.) 1995 Language and the Cognitive Construal of the World. Berlin: Mouton de Gruyter. Tolstoy, Leo 1970 Anna Karenina. Translated by L. & A. Maude. New York: W. W. Norton. [1918] Travis, Catherine 1997 Kind, considerate, thoughtful: a semantic analysis. Lexikos 7: 130-152. (Afrilex Series 7). 1998 Omoiyari as a core Japanese value: Japanese-style empathy? In: Angeliki Athanasiadou and Elzbieta Tabakowska. Speaking of Emotions: Conceptualisation and Expression, 55-82. Berlin: Mouton de Gruyter. Van Geert, Paul 1995 Green, Red and Happiness: Towards a framework for understanding emotion in universals. Culture and Psychology 1: 259-268. Wierzbicka, Anna 1973 The semantic structure of words for emotions. In R. Jakobson, C. H. van Schooneveld & D. S. Worth, eds. Slavic Poetics: Essays in Honour ofKiril Taranovsky, 499-505. The Hague: Mouton. 1986 Human emotions: universal or culture-specific? American Anthropologist 88: 584-594. 1988 L'amour, la colere, la joie, 1'ennui: la semantique des emotions dans une perspective transculturelle. Langages 89: 97-107. 1990a DuSa (~ soul), toska (~ yearning), sud'ba (= fate): Three key concepts in Russian language and Russian culture. In: Zygmunt Saloni

Introduction

33

(ed). Metody formalne w opisie jgzykow stowianskich. [Formal methods in the description of Slavic languages], 13-32. Bialystok: Bialystok University Press. 1990b The semantics of emotions: Fear and its relatives in English. Australian Journal of Linguistics (Special issue on the semantics of emotions) 10(2): 359-375. Defining emotion concepts. Journal of Cognitive Science 16: 1992 a 539-581. 1992b Semantics, Culture, and Cognition. Oxford: Oxford University Press. 1992c Talking about emotions: semantics, culture and cognition. Cognition and Emotion (Special issue on basic emotions) 6(3/4): 285-319. 1993 Reading human faces: Emotion components and universal semantics. Pragmatics and Cognition 1(1): 1-23. 1994 a Cognitive domains and the structure of the lexicon: The case of emotions. In Lawrence A. Hirschfeld and Susan A. Gelman (eds.), Mapping the Mind: Domain Specificity in Cognition and Culture, 431-452. Cambridge: Cambridge University Press. 1994b Emotion, language, and "cultural scripts". In Shinobu Kitayama and Hazel Rose Markus (eds.), Emotion and Culture: Empirical Studies of Mutual Influence, 130-198. Washington: American Psychological Association. 1995 a Everyday conceptions of emotion: a semantic perspective. In: J. A. Russell et al. (eds.), 17-47. 1995 a Emotion and facial expression: a semantic perspective. Culture and Psychology 1: 227-258. 1997 Understanding Cultures through their Key Words: English, Russian, Polish, German, Japanese. New York: Oxford University Press. 1998 a 'Sadness' and 'anger' in Russian: the non-universality of the so-called 'basic human emotions'. In: Athanasiadou and Tabakowska (eds.), 3-28. 1998b Angst. Culture and Psychology 4 (2): 161-188. 1998c Russian Emotional Expression. Ethos 26(4): 456-483. 1999 Emotions across Languages and Cultures: Diversity and Universals. Cambridge: Cambridge University Press. 2000 The semantics of human facial expressions. Pragmatics and Cognition 8(1): 142-183. Wilkins, David P. 1986 Particle-clitics for criticism and complaint in Mparntwe Arrernte (Aranda). Journal of Pragmatics 10: 575-596. Ye, Zhengdao 2000 The language of emotions in Chinese: a study based on Hong Lou Meng. MA thesis, Department of Linguistics, Australian National University.

34

Anna Wierzbicka and Jean Harkins

In press

Yu, Ning 1998

Different modes of describing emotions in Chinese: Bodily changes, sensations, and bodily images. In: Enfield and Wierzbicka (eds.). The Contemporary Theory of Metaphor: A Perspective from Chinese. Amsterdam: John Benjamins.

Zalizniak, Anna 1992 Issledovanija po semantike predikatov vnutrennego sostojanija. Munich.

Testing emotional universals in Amharic Mengistu Amberber

1. Introduction In this paper I present a semantic description of emotion predicates in Amharic.1 I assume a distinct cognitive domain which specialises in the emotions, and that the basic architecture of this cognitive domain is innate and universal. I also assume that the best way of searching for universal emotion concepts is through the use of the conceptual primitive FEEL (see Wierzbicka 1990, 1994, 1996, 1997, 1999; Goddard & Wierzbicka 1994; Harkins & Wierzbicka 1997 among others) within the theoretical framework of Natural Semantic Metalanguage (NSM). Wierzbicka (1999) proposes the following set of working hypotheses in her study of emotional universals: (1)

a. b. c. d. e.

f. g. h. i.

All languages have a word for FEEL. In all languages, feelings can be described as "good" or "bad". All languages have "emotive" interjections (i.e. interjections expressing cognitively-based feelings). All languages have some "emotion terms" (i.e. terms for cognitively-based feelings). All languages have words overlapping (though not identical) in meaning with the English words 'angry', 'afraid', and 'ashamed'. All languages have words comparable (though not necessarily identical) in meaning to 'cry' and 'smile'. In all languages, people can describe cognitively-based feelings via observable bodily symptoms. In all languages, cognitively-based feelings can be described via bodily sensations, In all languages, cognitively-based feelings can be described via figurative "bodily images".

36

Mengistu Amberber

j.

In all languages, there are alternative grammatical constructions for describing (and interpreting) cognitively-based feelings.

The main purpose of the present study is to test the above set of hypotheses in Amharic. The paper is organised as follows. In § 2, a brief profile of Amharic grammar is presented in order to acquaint the reader with some of the basic grammatical features of the language. In § 3, a wide range of data from Amharic is investigated in terms of the set of hypotheses outlined

2. An overview of Amharic grammar Amharic belongs to the Ethiosemitic sub-branch of the Semitic language family. It is one of the major languages of Ethiopia.2 It is characterised by a typical Semitic morphology known as root-and-pattern morphology. This type of morphology involves roots consisting of a set of consonants, known as radicals, which bear the essential lexical meaning, and a set of vowels which occur with the radicals to form the basic stem. For example, the verb S9bb9r9 'he broke (tr)' consists of the triradical root Vsbr 'break' and the pattern C,3C2C23C39, encodes the perfect conjugation with the 3rd person masculine. The same three consonants occur in a whole range of stems. Thus, the infinitive of Vsbr 'break' is mosbdr'to break' which is formed by attaching the prefix mo- to the pattern -CIC23C3. Open lexical classes include Nouns, Verbs, Adjectives and Adverbs. However, most adjectives are open only by derivation: they are derived from other classes, notably nouns and verbs. The closed classes include Pronouns, Demonstratives, Adpositions, Conjunctions, Numerals and Interjections. The basic constituent order in the clause is S V (for intransitives) and A O V (for transitives). The verb is obligatorily marked for subject agreement as shown in (2): (2)

aster wods bet A. to home 'Aster went home.'

hedged go.PF-3F

Testing emotional universals in Amharic

37

Amharic is a nominative/accusative language and exhibits a mixture of dependent- and head-marking. A definite object NP (O) is obligatorily marked by the accusative suffix -n. The subject NP (S/A) is unmarked. Agreement with the object and indirect object is often optional. Whereas number distinction (singular and plural) is made in all persons (1st, 2nd, and 3rd), gender distinction (masculine and feminine) is restricted to the 2nd and 3rd person singular only.

3. The hypotheses 3.1. A word for FEEL The Amharic exponent for the conceptual primitive FEEL is found in the verb (t9-)s9mma. Consider the following examples:

(3)

a.

minim a-

yi-

ss^mma-

nfii-

nothing NEG- 3M.IMP- feelΊ feel nothing.'

b.

min

yi-

lo- NEG

sssmma- h-

what 3M.IMP- feel'How do you feel?'

mm

al

2M- IMP

The Amharic word is used to express concepts which are undifferentiated between "emotions" and "sensations", or feelings of "cognition" and feelings of "the body" respectively (cf. Wierzbicka 1997: 18)3. Thus, the following sentences can be felicitous replies to the question "how do you feel"?

(4)

a.

ddssita

yi-

happiness SM.IMPΊ feel happy.'

b.

ssamma-

fin-

al

feel-

i-

IMP

himam yi-

ssdmma- fin- al

pain 3M.IMPΊ feel pain.'

feel-

I-

IMP

The Amharic verb for 'feel' cannot be used to translate the English construction such as "I feel the chair" (say as in touching a chair while trying to find your way in darkness):

38

(5)

Mengistu Amber be r

*w9nb3rchair-

u t3DBF DT-

sammaPF.SM-

fin 10

One has to use a different construction which involves the verb nakka 'touch', in a syntactic frame such as: 'it seems that I touched the chair'. Neither can the Amharic verb sdmma 'feel' be used to translate constructions such as "the chair feels smooth". One has to use the adjective laslassa 'smooth', in a syntactic frame such as: 'the chair is smooth'. This is consistent with the NSM view that these uses are English-specific, and do not reflect aspects of the universal prime FEEL. The verb (t9-)s9mma which is the exponent of FEEL is polysemous with an exponent of another mental predicate, namely HEAR. Thus, the examples in (3) above can be glossed with the meaning of HEAR as in (6): (6)

a.

b.

minim aysamma- tint- mm nothing NEG- SM.IMP- feel10- NEG Ί hear nothing.' min yissamma- hal what 3M.IMP- feel2M- IMP 'What do you hear?'

As the following examples show dQssita 'happiness' and dims' 'noise' stand in a paradigmatic relationship: (7)

a.

b.

d9ssita yisssmma- fin- al happiness 3M.IMP- feelIIMP Ί feel happy.' dims' yissamma- nfi- al noise SM.IMP- feeli- IMP Ί hear a noise.'

Although both (7 a) and (7b) have formally identical predicates, they differ in meaning as the English translations show. I will argue that the verb (t9-)s9mma has two polysemous meanings and that it is possible to formally identify the environment of each meaning. First, the verb associated with the meaning 'feel' always occurs with a pronominal suffix that agrees with the argument that is undergoing the emotional experience (see § 3.10 below for further examples). This is not the case with the verb associated with the meaning 'hear':

Testing emotional universals in Amharic

(8)

a.

b.

39

ddssita tesammaw happiness DT- feel.PF.3M- SMO 'He felt happy.' * dassita te-samma

As can be seen in (8), when the pronominal suffix (-w '3Mo') is omitted from the verb samma 'feel', the sentence becomes ill-formed. This is in sharp contrast to the verb ssmma 'hear' as shown in (9): (9)

a.

b.

dims' te- ssmmanoise DT- hear.PF.3M'He heard a noise.' dims' t9S9mma noise DT- hear.PF.3M Ά noise was heard,'

w 3MO

Notice that in (9b), the sentence is still grammatical despite the absence of the pronominal suffix (-w '3Mo'). However, note that without the agreement suffix the verb now has a passive interpretation - 'something was heard by someone unspecified'. The second formal difference between the two verbs has to do with valency. The verb corresponding to the meaning 'feel' cannot occur in a canonical transitive syntactic frame, unlike the verb corresponding to the meaning 'hear': (10)

a. b.

*temma dassita samma Lemma happiness feel.PF.3M hmma dims' samma Lemma noise hear.PF.3M 'Lemma heard a noise.'

The essential difference between the two clauses in (10) is that in (lOb) the NP dims1 'noise' occurs in the Ο slot of the clause whereas the same possibility is not available to the NP dassita 'happiness'. The third difference between the two verbs is related to valency changing derivations: only samma 'hear' can take the causative derivation in which the causative prefix as- is attached to a verb stem: (11)

a.

*aster lamma- η d^ssita assamma- c i-w A. Lemma- ACC happiness CAUS- feel.PF- 3r- 3MO

40

Mengistu Amberber

b.

aster temma- n zafan asA. Lemma- ACC song CAUS(lit. 'Aster let Lemma hear a song.')

sdmma- ööihear. PF.

w SM

This does not mean that emotion concepts do not occur in the causative. The causative can be formed by attaching the causative prefix to the verb 'be happy' without the involvement of the predicate for 'feel' (see § 3.10 for discussion of the causative derivation). A fourth difference between sdmma 'hear' and S9mma 'feel' can be seen in terms of nominalisation patterns. The two verbs have different nominalised forms: (12)

a. b.

simmet 'feeling' mosmat 'hearing'

(cannot mean'hearing') (cannot mean 'feeling')

The nominal simmet 'feeling' is productively used in a possessive-attributive phrase to express various emotional concepts. (13)

a.

b.

yd- haz9n simmet of- sadness feeling 'feeling of sadness' ya- dassita simmet of- happiness feeling 'feeling of happiness'

As already pointed out, sdmma 'feel' can be used to express not only cognitively based feelings but also bodily ('corporeal') feelings including 'cold', 'hunger', 'pain': (14)

a.

b.

c.

aster bird t9A. cold DT'Aster felt cold.' aster rshab t9A. hunger DT'Aster felt hunger.' aster him9m t9A. pain DT'Aster felt pain.'

sammat feel.PF.3M- 3ro sammat feel.PF.3M- 3FO sammat feel.PF.3M- 3ro

Testing emotional universals in Amharic

41

It is important to note that physiological experiences such as 'be hungry', 'be cold', 'be in pain/be sick' can also be expressed without the use of the verb 'feel'. (15)

a.

b.

c.

aster (in) bdrr^dat A.(ACC) be.cold.PF.3M- 3ro 'Aster is cold.' aster (in) rabbat A.(ACC) be.hungry.PF.3M- 3ro 'Aster is hungry.' aster (in) amm^mat A.(ACC) be.ill.PF.3M- 3ro 'Aster is in pain/ Aster is ill.'

Therefore it is clear that Amharic speakers make an explicit distinction between FEEL and HEAR though the formal expression of the two concepts overlaps in certain morpho-syntactic environments.

3.2. "Feel good" and "feel bad" Emotional experiences can be expressed with the evaluative exponents "good" and "bad". Wierzbicka (1997) hypothesises that all languages have the capacity to describe feelings as "good" or "bad". The Amharic equivalent of "I feel good/bad" involves the use of the nominal simmet 'feeling' in the frame: "I feel a good/bad feeling" (16)

a.

b.

t'iru simmet te- sammagood feeling DT- feel.PF-3M'She feels good.' (lit. 'She feels good feeling.') mot'fo simmet te- ssmmabad feeling DT- feel.pF-3M'She feels bad.' (lit. 'She feels bad feeling.')

t 3FO

t 3ro

The above expressions refer to a bipolar opposition between positive emotions and negative emotions.

42

Mengistu Amberber

The predicate which is normally used to express the emotion 'happiness' can be used in a generic sense to express 'good' feeling, whereas its negation is used to express roughly 'bad' or 'unpleasant' feeling: (17)

a. b.

dass ya- mmil simmet happy RC- AUX. feeling 'good feeling' d9ss y9- mmayil simmet happy RC- AUX.NEG feeling 'bad feeling' (or 'unpleasant feeling')

The form glossed as Άυχ' in the above examples is a relativised form (RC = relative clause) of the verb aid which can be translated as 'say' or 'do'. While the pair of expressions in (16) and (17) generally refer to bipolar emotional experiences, it does not mean that all emotional concepts can be categorised in terms of a "good/bad" dichotomy (see Wierzbicka (1997: 23), for a similar point).

3.3. Emotive interjections Amharic is rich in its inventory of interjections. Most interjections are morphologically isolated words which do not belong to any word class. There is a small class of interjections which can refer to cognitively-based feelings. The following list is gleaned from Leslau (1995: 900ff): (18)

a. b. c. d.

ahe ahehe ih Hill

e. f. g. h. i. j.

9Γ3 issay awwoy ec' ' iffoy wa

'exclamation of surprise, sorrow' 'exclamation of sorrow, grief, or worry' 'sorrow, distress, or pain' 'joy, ululation' (trilling sound often made by women) 'expresses surprise at something unexpected' 'exclamation of joy, pleasure, happiness' 'exclamation of sorrow' ('woe is me!') 'expression of disgust, of impatience' 'sigh of relief 'exclamation of sorrow or sadness, expression of warning of threat ('watch out!')'

Testing emotional universals in Amharic

k.

wiy

1.

gud

43

'exclamation of (negative) surprise, pain, or sorrow' ('my goodness!, alas!') 'exclamation of (positive) surprise, of amazement, of awe, of wonder, of shock'

Most of the interjections listed above have been given multiple English glosses. This may seem to suggest that these interjections are, as it were, "polysemous" and can express different emotions. In fact, this is generally not the case, and the multiple glosses reflect an "Anglo" rather than an Amharic point of view. Looking at the use of these interjections from an Amharic point of view, one is more likely to conclude that each interjection has its own semantic invariant, which doesn't necessarily match that of an English emotion term. In some cases, however, it does seem justfiable on language-internal grounds to posit polysemy. I would suggest that this applies, in particular, to ahe, whose two meanings (ahe} and ahe^ could be represented as follows:

(19)

a/?e, I feel something sometimes a person thinks something like this: something is happening now I didn't think before that this would happen I want to know more about it because of this, this person feels something I feel something like this

The explication in (19) is similar to Wierzbicka's explication for the English surprise (1996: 179). Although the event is not evaluated by the speaker as something 'bad', it does not mean that the event is necessarily good. In other words, the interjection can be used in neutral contexts. When the event is evaluated as 'bad', we have a different meaning, perhaps roughly translatable into English as 'sorrow': (20)

ahe2 I feel something sometimes a person thinks something like this: something bad is happening now I didn't think before that this would happen because of this, this person feels something bad I feel something like this

44

Mengistu Amberber

The common key element in (19) and (20) is that the event that is taking place was rather unexpected or unforeseen. It is important to note that some of the interjections are used to vocalise both cognitively-based feelings and bodily feelings. For example, iffoy ((18 i) above) is an exclamation of 'relief either in the psychological or the physical domain. Hence, one can say iffoy\ after the aversion of some impeding disaster in which case the experiencer is vocalising a psychological relief- the anticlimax of a worrying state, so to speak. But one can also say iffoy\ when taking a break after a physically exhausting activity.

3.4. "Emotion" terms Amharic has a wide range of terms which refer to positive and negative emotions. In this section, we will sample some of the common emotion terms by focusing on the (intuitively) prototypical emotions for each category: roughly 'be happy' and 'be sad' respectively. For ease of exposition the emotion predicates are presented in the (unmarked) intransitive syntactic frame. Other possible grammatical strategies and syntactic frames for encoding emotions are presented in § 3.10.

3.4.1. Positive emotions One of the prototypical positive emotions is designated by the verb ta-dassate 'be happy/joyful' which has the component: "feel good": (21)

aster tadassdtaöö A. DT- be.happy.PF- 3F 'Aster is happy.'

The construction in (21) marks a change of state — a transition from not being in a state of happiness to being in a state of happiness. The semantic explication of the Amharic tdddssatd which roughly corresponds to 'be happy', 'be joyful' can be presented as follows:

(22) X feels something sometimes a person thinks something like this:

Testing emotional universals in Amharic

45

something good happened (to me) I wanted this to happen because of this, this person feels something good X feels like this The sub-component 'to me' in (22) is placed in parentheses to show that it is an optional component. That is, for X to be in a state of tsddssdta something good does not necessarily have to happen to X. X can be in a state of t3d3ss3t9 because something good happened to someone else. Note that teddssdt9 focuses on a change of state. There is a transition in X's state: there was a point in time where X was not in the state of t9d9ss9t^ and now X is in the state of t9dass3t3. On the other hand, a less dynamic state can be expressed by employing a different construction which involves the juxtaposition of the nominal dassitanna 'happy' with the copula verb n3w 'be': (23)

aster dassits nscc A. happy be.PF-3F 'Aster is a happy person.'

Thus, (23) expresses a relatively more stable state of affairs - a judgement regarding one's general temperament.

3.4.2. Negative emotions Perhaps one of the most common negative emotions is azzsns, comparable to 'be sad'. This emotion is expressed by a basic (i.e. non-derived) intransitive stem: (24)

aster azzanscc A. be.sad.PF- 3F 'Aster is sad.'

The verb azzanacan be used to translate English terms 'disappointment' and 'sympathy'. The former belongs to a family of emotions which express "events contrary to expectations" (cf. Wierzbicka (1996: 178ff)).4 The relationship between the Amharic expressions that correspond roughly to 'sadness', 'disappointment', and 'sympathy' is complex. Here I

46

Mengistu Amberber

can only present a brief general commentary. I will examine the relationship between the Amharic expressions which roughly correspond to 'sadness', 'disappointment' and 'sympathy'.

3A.2.1. azzand be sadlbe disappointed' English makes a distinction between 'be sad' and 'be disappointed'. In Amharic there is one concept as can be seen in the following examples: (25)

(26)

hmma bs- ihit- u fdtsna mo-wdsk' bat'am L. with- sister- 3POSS exam iNF-fail very azz9nd be.sad.PF.3M 'Lemma is very sad/disappointed because of his sister's failure in the exam.'

temma

ba-

ihit- u

mamot bat'am azzans

L. with- sister- SPOSS death very be.sad.PF.3M 'Lemma is very sad because of his sister's death.' Notice that in (25) the verb can be roughly translated into English by 'be sad' or 'be disappointed', in (26) it can only be translated as 'be sad'. The key difference between 'sadness' and 'disappointment' in English lies in the presence or absence of an (explicit) expectation regarding a particular goal. According to Wierzbicka's (1996: 179-180) explication of the two concepts in English, 'disappointment' has a component which spells out the idea that the experiencer had expected "something good will happen" which is absent in the explication of 'sad'. The emotional state of 'disappointment' sets in when that expectation is not fulfilled.5

3.4.2.2. Sadness and sympathy In Amharic the concept which roughly corresponds to the English 'sympathy' is expressed by grammatical means. Thus, when the verb azzand occurs with an argument marked by the preposition h 'for' (with or without a corresponding agreement marker on the verb) it has the approximate meaning 'sympathise (with)' or more literally 'be sad for someone':

Testing emotional universals in Amharic

(27)

temma

h-

hit-

u

azzana-

47

(ll-at)

L. for- sister- 3?oss be.sad.PF.3M- (FOR-BFO) 'Lemma is sorry for/sympathises with his sister.' (lit. 'Lemma is sad for his sister.') Notice that the verb is optionally marked by the prepositional suffix -11-, and an agreement suffix which agrees with the Object' of the emotion. When the verb is marked by the prepositional suffix -bb-, we get a different meaning which can be translated into English as 'be disappointed with'. Thus, the two complex verbs form a minimal pair, the only difference residing in the shape of the prepositional suffix, as shown in (28)-(29): (28)

azzdnd11- at be.sad.PF.3M- FOR- 3ro 'He feels sorry for/sympathises with her.' (lit. 'He is sad for her.')

(29)

azz9n^bbat be.sad.PF.3M- WITH- 3ro 'He is disappointed with her.' (lit. 'He is sad at her.')

The semantics of the two constructions can be represented by the following explications (-PER is an abbreviation used for the relevant personnumber-gender marking; it refers to the component 'that person' in the semantic explication):

(30) X feels something because X thinks something sometimes a person thinks something like this: something bad happened to that person I don't want things like this to happen to this person because of this, I want to do something because of this, this person feels something bad X feels something like this because X thinks something like this

(31)

azzdti9-bb-PER X feels something

48

Mengistu Amberber

because X thinks something sometimes a person thinks something like this: something bad happened (to that person) it happened because that person did something I don't want this person to do things like this because of this, I want to do something because of this, this person feels something bad X feels something like this because X thinks something like this Note that in (30) - the meaning that roughly corresponds to English 'sympathise' - the explication has the component "something bad happened to that person". In (31) on the other hand - the meaning that corresponds to the English 'disappointed with' - the component 'to that person' is shown to be optional. This is to show that something bad does not necessarily have to happen to someone in order to azzand-bb-PER 'be disappointed with' them. For example, if Υ had an expectation that X will execute (or not execute) some task and X failed to do so, Υ can say (31) - Ί am disappointed with X' though nothing bad happened to X. Note also that in (31) the component "it happened because that person did/didn't do something" clearly spells out that there is a specific party who is implicated in the negative outcome of the event.

3.5. "Fear-like" words, "anger-like" words, "shame-like" words "Fear-like", "anger-like", and "shame-like" words are hypothesised to constitute possible universal categories which appear "as reference points for emotion concepts" (Wierzbicka 1997: 27). In this section I will examine the Amharic equivalents of these emotion words to identify the languagespecific meanings.

3.5.1. terra 'fear' The verb fdira which can be translated into English as 'be frightened' or 'fear' is ambitransitive: it can occur either in an intransitive or transitive syntactic frame:

Testing emotional universals in Amharic

(32)

a.

b.

49

aster bat'am ferracc A. very be.frightened.PF- SF 'Aster is very frightened.' aster wissa- win farra -cc/- w A. dogDBF- ACC fear.PF- SF- SMO 'Aster feared the dog.'

The causative of the verb - 'make someone fear' or 'frighten (tr.)' - is derived by attaching the causative prefix as-: (33)

wissa- w aster- in dogDBF A.ACC The dog frightened Aster.'

ascaus-

fyrrafear-

t 3M-3FO

When expressing generic 'fear' — fear as a more or less permanent state of affairs - the verb occurs in the (habitual) imperfect aspect: (34)

aster wissa tiA. dog IMP'Aster fears dogs.'

farafear-

Itecc IMP- 3r

When somebody is overcome by fear, a special construction which involves the use of the nominal firhat 'fear' and the verb yazd 'hold, catch' is used: (35)

aster- in firhat yazat A.ace fear catch.pf.3M- 3ro 'Aster is overcome by fear.' (lit. 'Fear caught Aster.')

Extreme fear for something can be expressed by the reduplicative verb t9nbok'9bbok'9. The meaning of this verb is roughly equivalent to the English 'terrified' but often with a derogatory sense perhaps close to 'cowardly terrified': (36)

S9wiyy9w t9nbok'9bbok'9 man.DEF be.terrified.PF.3M The man is terrified.'

The verb fyrra can also be used in contexts where the target of fear is conceptualised as a powerful authority rather than a source of potential

50

Mengistu Amberber

danger. A typical example of a powerful (but non-adversative) authority can be represented by one's father or God:

(37)

aster abbat-wan ti- fara- lidöö A. father-DBF- ACC IMP- fear- IMP- 3F 'Aster fears her father.'

(38)

aster igziyabherA. God'Aster fears God.'

in ti- fsraACC IMP- fear-

Ihcc IMP- 3r

3.5.2. tenaddsdd 'be angry I distressed Wierzbicka (1997: 28) hypothesises that "all languages appear to have a word which shares two semantic components with the English word 'anger'" and the two components are identified as: "I don't want this" and "I want to do something because of this". In addition, many languages, including English, appear to combine the above two cognitive components with "a negative judgement" component - "someone did something bad". The Amharic verb tsnadddds appears to incorporate the two cognitive components. For a typical example, consider (39): (39)

setiyyawa gsnzab sil-t' affat tenaddsdaöd woman.DEF money as-lose.PF- 3F be angry.pp- 3F 'The woman is angry/distressed because she lost (her) money.'

Note that unlike the English 'anger' the Amharic tanaddada does not necessarily imply that the emotion is triggered by a negative judgement of someone else's action (or inaction). Thus, in (39) the experiencer is not necessarily identifying another person ('this person did something bad') as the culprit. In this sense, tsnadd9dd is perhaps closer to 'be distressed'. However, it is also possible that the negative judgement is made reflexively, i.e. the experiencer is taking full responsibility for the situation. In such cases, the relevant cognitive component is "I did something bad" instead of "this other person did something bad". This component can be explicitly stated by a prepositional phrase as in (40):

Testing emotional universals in Amharic

(40)

51

setiyy9wa b9- raswa tenaddada- cc woman.DEF with- herself be angry.PF- 3r 'The woman is angry with herself.'

Thus, in Amharic the verb t9nadd9d9 can be used when there is no negative-judgement component and also when there is a negative-judgement component. This can be summarised as follows: (41)

a. b.

with no negative-judgement component with a negative-judgement component predicated of either: (i) this other person (non-reflexive) or (ii) the experiencer (reflexive)

Incidentally, the verb t9nadd9d9 'be angry/distressed' appears to be related (albeit indirectly) to the verb n9dd9d9 'burn (intr.)' as in 'the burning of a candle'. Amharic has a family of "anger-like" verbs which appear to be metaphorically derived from verbs of 'burning': (42)

From the verb arr9r9 'be over-cooked' a. arr9r9 'to be extremely angry','furious' b. as-arr9r9 'cause someone to be extremely angry'

(43)

From the verb d9bb9n9 'be over-roasted' a. d9bb9n9 'infuriated' b. a-dabbana 'cause someone to be infuriated'

(44)

From the verb t9-k'at't'9l9 'burn (intr.)' (as in 'burning leaves') a. t9-k'at't'9l9 'be furiously angry','livid' b. a-k'at't'9l9 'cause someone to be furiously angry'

However, these 'anger-like' verbs often occur with the manner adverbial b9niddet (which consists of b9- 'with' and -niddet 'anger (noun)'): (45)

setiyy9wa b9-niddet arr9r9cc woman.DEF with-anger overcook.PF- 3F 'The woman is extremely angry.' (or 'fuming with anger')

These verbs are often used to express "suppressed" anger in a way which is perhaps comparable to the English 'fuming'. Furthermore, the verbs express the level of the emotional intensity and their function is similar to that of intensifiers such as 'very' and 'extremely'. This is supported by the fact

52

Mengistu Amberher

that some of the verbs, most notably te-k'at't'9te (lit. 'burn'), can be used in the context of other emotions such as 'jealousy' and even 'love' as in 'that man is burning with jealousy' (= i.e., 'that man is extremely jealous of someone'). The fact that "anger-like" verbs can be related to verbs which express non-emotional events associated with the implementation of fire is also common in English, as argued by Lakoff (1987: 383): The folk theory of physiological effects, especially the part that emphasises 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 CONTAINER. (...) When ANGER is HEAT is applied to solid, we get the version ANGER IS FIRE (...)

There is another "anger-like" verb, the verb ts-k'ot't'a, which is always used to express verbally vented anger: (46)

aster td- k'ot't'a cc A. DT- be.angry.PF- 3F 'Aster is angry.'

When this verb is used as a transitive verb - 'be angry at someone' - its meaning overlaps with that of the English verb 'rebuke' (or 'reprimand'). (47)

aster ihit- wa- n te- k'ot't'acd- at A. sister- her- ACC DT- be.angry.PF- 3r- 3ro 'Aster rebuked her sister.'

One way of analysing the transitive tsk'ot't'a is to assume that it has two predicates in its conceptual structure: the mental predicate FEEL and the speech predicate SAY. A first approximation of the explication looks like (48): (48)

Xt3k'ot't'a = X feels something because X thinks something sometimes a person thinks something like this about this other person: this other person did something bad I don't want things like this to happen

Testing emotional universals in Amharic

53

I want to do something because of this I want to say to this other person: "I don't want this to happen after this" because of this, this person feels something bad because of this, this person says something like this to this other person X feels something like this because X thinks something like this The crucial components in the explication of the transitive tek'ot't'a are those that relate to speaking: "I want to say to this other person: don't want this to happen after this'; ... because of this, this person says something like this to this other person". Note that unlike tanadddda, tak'ot't'a has the "negative judgement" component "this other person did something bad". The culprit is perceived to be non-reflexive, that is, it is a person other than the experiencer. Thus, tdk'ot't'a does not make sense if the negative judgement is made reflexively, i.e. if the judgement is about the experiencer (which is in marked contrast to the verb t^nadddd^ (see example (40)). (49)

*setiyy9wa ba-raswa tsk'ot't'a - öö woman.DEF at-herself be.angry.PF- 3F 'The woman is angry with (or rebuked) herself.'

3.5.3. affdrd 'be ashamed!embarrassed' 'Social' emotions (cf. Goddard 1996) comparable to those lexicalised by the words 'shame' and 'embarrassment' in English are expressed by the verb affars in Amharic. In other words, the concepts for something like 'shame' and something like 'embarrassment', which are lexicalised in English (see Wierzbicka 1997: 31), appear to be expressed by one word in Amharic. (50)

aster aff9racc A. be.embarrassed.PF- 3r 'Aster is embarrassed/ashamed.'

54

Mengistu Amberber

The basic semantic explication of words referring to 'social emotions' suggested by Wierzbicka (1997) captures the meaning of the verb affara 'be ashamed, embarrassed': (51)

someone thinks something like this: people can think something bad about me I don't want this because of this, this person feels something

The verb affara can also have a polysemous meaning comparable to the English 'be shy' (see Harkins 1990 for the expression of 'shame' and 'shyness' in Australian Aboriginal English): (52)

aster affordcc A. be.shy.pf- 3Έ 'Aster became shy.'

When people feel 'shy' it is not necessarily because they think that someone thinks bad about them. The feeling may be due to general 'self-consciousness' as pointed out by Darwin (1872/1965).6 Thus, the semantic explication of affara in the sense of (52) will not have the evaluator 'bad' in the second component: (53)

aff9r9 (something like 'shy') someone thinks something like this: people can think something about me I don't want this because of this, this person feels something

Incidentally, in Amharic being shy is typically regarded as a positive cultural attribute. It often incorporates evaluation in the visual domain - for example a shy person does not gaze or stare at people. In fact the noun referring to 'a shy person' is a compound derived from a nominal of affar9 and the noun for 'eye': (54)

aynaffar eye.shy 'a shy person'

Testing emotional universals in Amharic

55

This nominal is contrasted with another compound which involves the noun ay n 'eye' and the adjective d9rak' 'dry':

(55) 'shameless', 'impudent', 'cheeky' (lit. 'dry-eyed') The transitive of affara 'be embarrassed/ashamed' (and also that of affaro 'be shy') is formed by attaching the causative prefix as-. The transitive verb can be translated into English as 'make X be ashamed/embarrassed/shy' and also 'humiliate'. (56)

y9- iht- wa sira aster- in asPOSS- sister-poss.3p deed A.- ACC CAUSaffarat be.embarrassed.PF3r 'What her sister did embarrassed Aster.'

The verb affara also has a non-emotional meaning which can be roughly translated into English as 'going back on one's word' or 'back away from an agreement'. Perhaps this is not unlike the relationship between 'shy' and 'shy away' in English. Thus, it would be reasonable to assume that afford 'be embarrassed' (also 'be ashamed', 'be shy') is polysemous with affara 'going back on one's word'.

3.6. 'Smile' and 'cry' The Amharic words comparable in meaning with 'smile' and 'cry' are fygagateand atek'k '959 respectively. The former is typically expressed in a complex verb which involves a non-verbal stem and an auxiliary (see § 3.10 below for complex predicates). (57)

a. b.

aster fygsg ahco A. smile.vn aux.pf.- 3r 'Aster smiled.' aster alsk'k'dsa- co A. cry.pf.3r 'Aster cried.'

56

Mengistu Amberber

The form fegag encodes a facial expression of a happy emotional state. However, it also has a related meaning comparable to the English 'brighten'. For example, the same complex verb can be predicated of the noun 'sky' with the meaning 'the sky brightened'. The event expressed by the verb atek'k'asa 'cry' may or may not be associated with noise. Thus, one can atek'k'asa 'silently' or 'loudly'. On the other hand, there is a separate word for something like 'laugh', as distinct from 'smile': (58)

aster sak'd cc A. laugh.PF.- 3F 'Aster laughed.'

It is important to note that in addition to expressing joy or happiness, laughter can express amusement.7 The fact that sounds are produced during laughing is an important element which may distinguish sak'd 'laugh' from feg9g ate 'smile'. The unmarked case is that sak'a is associated with (audible) sounds. Thus, in Amharic something like 'he laughed in his stomach' as in (59) below makes sense only in the context of inhibited laughter - i.e. when the event in question does not socially sanction laughter. (59)

ba- hod-u sak'3 in- stomach-3poss laugh.PF.3M 'He laughed quietly to himself.' (lit. 'He laughed in his stomach.')

In the case of fagag ah 'smile', one cannot have the equivalent of 'he smiled in his stomach'. This is due to the fact that fegagah 'smile' must be visually transparent.

3.7. External bodily symptoms Some emotions are described in Amharic through the use of externally observable bodily images or movements. For example, one can say something like 'his face became ashen' to express sudden shock or fright:

Testing emotional universals in Amharic

(60)

57

fit-u amsd face-3M.POSS ash be.alike.PF.3M 'His face became ashen (with fright).'

This sentence depicts a negative emotion, specifically that of being in a shock or gripped by fear. The semantic explication of such bodily symptoms of emotion is rather complex as it should incorporate the relevant emotional trigger for the externally observable expression. Suppose that the experiencer's facial expression is triggered by something like 'being frightened'. I assume that the semantic explication of 'frightened' is part of and embedded within the semantic explication of the bodily-symptom expression. A rough approximation of this is presented in (61): (61)

fit-u amdd something was happening to X's face people could see this because of this, people could know that X felt something because X thought something sometimes a person thinks something like this: something very bad is happening because of this, something very bad can happen to me now I don't want this to happen because of this I want to do something if I can I don't know what I can do because of this, this person feels something very bad because people could see X, they could know that X felt like this

Although the bodily symptom — of acquiring the color of ash — alone conveys negative emotion, often the bodily symptom is expressed together with the relevant emotion term: (62)

fit-wa bd-diniggat'e amad face-3p.poss with-shock ash be.alike.PF.3M 'her face became ashen with shock.'

(63)

S3winst-wa ba-firhat tenk'at'ak'k'at'a body-3p.poss with-fear tremble.PF.3M 'her body was trembling with fear.'

58

Mengistu Amberber

Notice that the emotion occurs as an oblique manner adverbial bo-diniggat'e 'with-shock', bs-firhat 'with-fear'.

3.8. Mapping from the senses In § 3.1. it is argued that the Amharic verb for the exponent FEEL has formal properties which overlap with the verb for HEAR. For example, the following interrogative sentence refers to either an auditory perception (HEAR) or an emotional state (FEEL): (64)

min yissdmma- hal what 3M.IMP- hear/feel- 2M- IMP 'What do you hear?'/'How (lit. What) do you feel?'

There is some evidence which suggests that the polysemy between sensation and emotional predicates is not restricted to the auditory domain. The following examples from the domains of taste and olfaction show similar polysemy. (65)

a. Taste msrrars b. FEEL t3m9rr3r3

(66)

a. b.

'taste bitter' 'be irritated', 'be embittered'

Smell kanakkans 'tickled (as in pepper in the nostrils)' FEEL kanakkana 'cause anxiety/uneasiness in one's mind'

The remaining two senses - touching and seeing - are also involved in the expression of emotions: (67)

a. b.

Touch t9n9kka FEEL tanakka

'be touched' 'be (emotionally) moved'

(68)

a. b.

See ayya FEEL ayy9

'see' 'have an emotional experience'

The last two occur in constructions such as (69)-(70): (69)

aster bahazan tanakkado A. with- sorrow (touch)feel.PF.- 3r 'Aster is touched with sorrow.'

Testing emotional universals in Amharic

(70)

aster mdkarawa-

n

59

ayya-öc

A. distress/misfortune (see)feel.PF- 3F 'Aster is distressed/miserable.'

3.9. Internal bodily images The internal parts of the body - especially the heart, the stomach, and the intestine - can be used to express various types of emotions in Amharic. In the following example, the heart's natural rhythm, expressed by the onomatopoeic tir tir, is manipulated to express something comparable to 'fear' or 'anxiety' with the cognitive components 'someone thinks like this'; 'something bad can happen to me now':

(71)

lib-

e

tirtir

ah

heart- POSS.I ONOM AUX (SAY).PF.3MS 'My heart was pounding (with fear or anxiety).' (lit. 'My heart was saying "tir tir'") In the following example, the stomach is projected as a seat for the feeling of sympathy: (72)

hod-e tambboc'abboc'a stomach-1.POSS move.PF.3M felt sorry for someone.' (comparable to 'sympathy/compassion') (lit. 'My stomach moved.')

The verb tembbcfc 'abbao 'a is normally used to express the movement of a liquid substance bounded within a container. (73)

wiha- w tambboö'abbsö'd water- DBF move.PF.3M 'The water moved.'

Following Wierzbicka's (1997: 37) prototype for this domain, I propose the following explication for the construction in (72):

(74)

My stomach tambboc 'abbac '9 I felt something because I thought something

60

Mengistu Amberber

sometimes a person thinks: something bad happened to this other person I don't want things like this to happen because of this, I want to do something because of this this person feels something bad I felt like this because I thought like this I say: something moved inside my stomach, because I felt like this I say this because I want someone to know how I felt Another negative emotion which can be expressed by an internal bodily image is something like 'anger': (75)

ddm-u Mia blood-I.POSS boil.PF.3M 'He is incensed.' (lit. 'His blood boiled.')

3.10. The grammar of Amharic emotion expressions As we saw in the previous section, Amharic emotions are expressed by the use of verbal predicates. The experiencer of the emotion can be realised either as Subject or Object depending on the lexical semantics of the verb. Consider the following examples: (76)

a.

b.

aster t9- dassataco A. DT- be.happy.PF- 3r 'Aster is happy/joyful.' temma aster- in asdassatat L. A.ACC CAUS- be.happy.PF.3M- SFO 'Lemma pleased Aster' ('made her happy/joyful')

In the first example, the experiencer of the emotion (something like 'happiness/joy') is realised in the S function, whereas in the second example the experiencer is realised in the Ο function. For ease of exposition we can refer to these patterns as SUBJECT EXPERIENCER (SubjExp) and OBJECT EXPERIENCER (ObjExp) respectively (cf. Pesetsky 1995).

Testing emotional universals in Amharic

61

In Amharic, SubjExp predicates exhibit two morphological patterns: those which are morphologically unmarked (Pattern A) and those which occur with the prefix t- (Pattern B). Consider the following examples: (77)

a.

Pattern A: aster hmma- n wsddada- cciA. LACC love.PF.- 3rs'Aster loved Lemma.'

b.

Pattern B: lamma taL. DT-worry.PF.3MS 'Lemma is worried.'

w 3MO

Syntactically, Pattern A verbs have two obligatory arguments which can be referred to as Experiencer of emotion and Target of emotion (adopting the term 'Target' from Pesetsky (1995)). Pattern B verbs have one obligatory argument, the Experiencer. Further examples of the two patterns are presented below: Table 1. Patterns of Experiencer Predicates

Pattern A

Pattern B

SubjExp

ObjExp

SubjExp

W9dd9d9 'love y'

as-w9dd9d9 'make χ love y'

f9rra 'fear y'

as-f9ira 'make χ fear y'

'be pleased'

'make χ pleased'

t'slla 'hate y'

as-t'alla 'make χ hate y'

t9-k'ot't'a 'be angry'

as-k'ot't'a 'make χ angry'

'worry'

ObjExp 'make χ worry'

t9-d9SS9t9

It is important to note that the ObjExp verbs are formed by the causative affix as-, and not by the causative affix a-. Thus, compare the sentences in (78) below : (78)

a.

hmma t9- d9ss9t9 L. DT. be.please.PF.3MS 'Lemma is pleased.'

62

Mengistu Amberber

b.

aster temma-n asdassataA. L.-ACC CAUS- be.please.pf.'Aster pleased Lemma.'

oc- iw SFS- SMO

Pattern B SubjExp verbs are of two types depending on whether or not they occur independently without the prefix t-. The verbs which cannot occur independently are referred to as Prefix requiring (P-verbs). Thus, the P-verbs are 'bound' in a morphologically transparent sense: they require overt affixation. Consider the following examples: (79)

a. b.

*ddss3t9 > te-dassata- 'be happy' *k'ot't'a > t9-k'ot't'a 'be angry'

Some Pattern B SubjExp verbs can occur with or without the prefix t-. Consider the examples in (80)-(81): (80)

a.

b.

(81)

a.

b.

aster- (in) c'snnsk'at A.- (ACC) worry.pf.3ms- 3ro 'Aster is worried.' aster to- o'anndk'd- oc A. Dt- worry.pf.- SFS 'Aster is worried.' aster- (in) ddnndk'at A.- (ACC) astonish.pf.3MS- 3ro 'Aster is astonished.' aster te- ddnnak'a- cc A. Dt- astonish.pf.- 3FS 'Aster is astonished.'

For ease of exposition, I refer to the constructions in the (a) examples as Type 1 and the constructions in the (b) examples as Type 2. What is interesting about the Type 1 construction is that (a) the Experiencer occurs with an (optional) accusative case, and (b) the Experiencer appears to control object agreement as can be seen in (80 a) repeated below: (82)

aster- (in) o'annsk'at A.(ACC) worry.pf.3ms- 3po 'Aster is worried.'

Testing emotional universals in Amharic

63

The presence (albeit optionally) of the accusative case marking on the subject noun phrase in (82) is unusual in Amharic, as only (definite) objects are marked by the accusative. Thus, for example object noun phrases with the "patient" role are marked by the accusative case as in (83): (83)

sawiyydw asterin mattat man.DEF A.ACC hit.pf.3M- 3ro 'The man hit Aster.'

However, recall that in § 3.1 (example 15) we saw that the subjects of certain physiological predicates (comparable to 'be hungry' and 'be cold') show exactly the same morphological marking. This raises the question: why is the experiencer of some SubjExp predicates (as exemplified in (80)-(81)) morphologically behaving as if it is the grammatical object of the clause? The fact that many Subject Experiencer predicates exhibit quirky case is by now cross linguistically well-documented (cf. Zaenen et al. 1985). Consider the following example from Icelandic (Zaenen et al. 1985): (84)

Calvini liki verkid Calvin-D like thejob-N 'Calvin likes the job.'

According to Zaenen et al. (1985), the experiencer argument is a grammatical subject as established by various tests for subjecthood. However, when the Experiencer is expected to be marked by the nominative, it is marked by the dative case. It is hypothesised that certain emotion predicates behave in this way due to their underlying conceptual structure. The subject of an emotional predicate such as 'worry' is an affected argument with less volition, thus aligning itself with prototypical objects (cf. Hopper and Thompson 1980) rather than prototypical subjects. In other words, the "experiencer" argument of the emotion predicates has much in common with typical "patient" arguments. The primary cognitive component connected to the "patient" role is expressed by the primitive HAPPEN as in 'something bad happened to X', or "something bad can happen to me" with the clear implication that X has little or no control over the event (see also Wierzbicka 1996: 123). Thus, the experiencer argument of certain emotion predicates behaves like the

64

Mengistu Amberber

"patient" argument of typical action predicates (e.g. 'hit') because the experiencer has little or no control over the onset of the event. There is some language-internal evidence which supports this claim. The two most common emotional verbs which do not allow the experiencer to occur in the accusative case are waddadd and t'9lla roughly corresponding to 'love' and 'hate' respectively. This fact can be accounted for by assuming that with emotions like 'love' and 'hate' the experiencer is relatively more volitional compared to the experiencer with emotions like 'anxiety' and 'shock' and thus not aligning itself with patient-like arguments. Some emotion predicates take a prepositional oblique argument to express the 'target' of emotion. Thus, for example with the intransitive verb k'9nna which roughly corresponds to 'be jealous', or 'be envious', the target of the emotion occurs in a prepositional phrase: (85)

aster ba- (i)htwa k'anna-öö A. on- sister.poss.3p be.jealous.PF-3ps 'Aster is jealous of her sister.'

But it is also possible to cast the target of emotion as an object with the accusative case and an obligatory agreement marker on the verb: (86)

aster ihitwa-n k'9nnacc- ibb- at A. Sister.poss.3p-Acc be.jealous.PF- 3p- ON- 3ro 'Aster is jealous of her sister.'

To summarise this section, the important point is that the experiencer argument of certain emotion predicates behaves morphologically as if it is the object of the clause rather than the subject. This is accounted for by proposing a semantic motivation: experiencers by virtue of having less control over the onset of the (emotional) event align themselves with prototypical "patient" arguments rather than with prototypical "agent" arguments. However, the alignment is partial, manifesting itself mainly in morphology: the occurrence of an accusative case morpheme which is otherwise restricted to definite object NPs of agentive verbs. As far as the overall clausal structure is concerned there is no evidence which suggests that experiencers are grammatical objects.

Testing emotional universals in Amharic

65

4. Conclusion This study has presented a general overview of the semantics of some emotional (linguistic) expressions in Amharic. The main purpose of the study has been to test the set of hypotheses proposed by Wierzbicka (1997,1999) regarding emotional universals. The description and analysis presented in the study shows that emotional universals are borne out by the Amharic data. At the methodological level, the study shows that a thorough description of emotional concepts can be accomplished only through the use of a "meta" language rather than via a language-specific vocabulary. Although English words are used in the study, the translations have been stated only as rough approximations of the Amharic concepts. When there is a mismatch between concepts expressed by the individual languages, the metalanguage can be used to clearly articulate in exactly what sense the words in question are different in the relevant languages. The study also shows that in investigating the expression of emotion concepts one should pay close attention to the semantics of the grammar as a whole. Thus, in Amharic the concept which is close to the English 'sympathy' is expressed by grammatical means, i.e. attaching a prepositional suffix on the verb which corresponds to 'sad/sorry'. It has been demonstrated that by simply using a different prepositional suffix the meaning comparable to 'be disappointed with' can be expressed. One of the crucial differences between the semantic explications of the Amharic expressions roughly corresponding to 'be sad/sorry for someone' and 'be disappointed with someone' is the presence of the component 'it happened because that person did something' in the latter. The problem of polysemy has figured rather prominently in the present study. The standard strategy employed in postulating polysemy is based on the discovery of language-internal formal evidence (cf. Goddard and Wierzbicka 1994: 446). Thus, it has been shown that the lexical exponent of the semantic primitive FEEL and that of the mental predicate HEAR in Amharic occur in a number of distinct morpho-syntactic environments. Of course, some aspects of the analysis presented in the study are tentative and only a detailed study of a wider range of data will provide a more in-depth insight into the cognitive domain of emotions.

66

Mengistu Amberber

Notes 1. I would like to thank Anna Wierzbicka for her detailed comments and suggestions on an earlier draft of this paper. 2. It is also spoken by a sizeable minority in Israel and Egypt (cf. Grimes 1996). 3. The material referred to here from Wierzbicka's (1997) paper is published in her subsequent book on emotions (Wierzbicka 1999). 4. It is interesting to note that according to Johnson-Laird & Oatley (1989: 120), the definition of sympathy incorporates the emotion sadness. They analyse sympathy as a complex emotion expressing "sadness for (and understanding of) someone as a result of imagining oneself in their situation". In their model, sadness is one of five basic emotion "modes" in terms of which a family of other complex emotions can be elaborated. 5. In the semantic-field analysis of Johnson-Laird and Oatley (1989), the two emotion concepts are treated as containing the basic emotion "sadness". They define "disappointment" as "sadness caused by failure to achieve goal" (Johnson-Laird & Oatley 1989: 112). 6. For Darwin (1872/1965) the important element in "shyness", "shame", and "modesty" is "self-attention". Although Darwin's discussion of these emotions is in the context of describing the biological properties of "blushing" what he has to say about the mental states that trigger the feelings in question is relevant to our purpose. He argues that the expectation of "praise" and "admiration" can trigger blushing in as much as the expectation of "blame" or "disapprobation". 7. The relationship between "smile" and "laughter" is by no means simple. Smiling can be an onset of laughing, or as Darwin (1872/1965: 196) put it, "smiling... graduates into laughter". But there are clear differences as well. For example, if you see a friend you are likely to smile but not necessarily laugh.

References Darwin, Charles 1872/1965 The Expression of the Emotions in Man and Animals. Chicago: The University of Chicago Press. Goddard, Cliff 1996 The "social emotions" of Malay (Bahasa Melayu). Ethos 24(3): 426-464. Goddard, Cliff and Anna Wierzbicka (eds.) 1994 Semantic and Lexical Universals: Theory and Empirical Findings. Amsterdam: John Benjamins. Grimes, Barbara 1996 Ethnologue. 13th edn. Dallas: Summer Institute of Linguistics.

Testing emotional universals in Amharic

67

Harkins, Jean and Anna Wierzbicka 1997 Language: A key issue in emotion research. Innovation 10(4): 319-331. Harkins, Jean 1990 Shame and shyness in the Aboriginal classroom. Australian Journal of Linguistics 10(2): 293-306. Hirschfeld, Lawrence A. and Susan A. Gelman (eds.) 1994 Mapping the Mind: Domain Specificity in Cognition and Culture. Cambridge: Cambridge University Press. Hopper, Paul and Sandra Thompson 1980 Transitivity in grammar and discourse. Language 56 (2): 251-99. Johnson-Laird, Philip N. and Keith Oatley 1989 The language of emotions: An analysis of a semantic field. Cognition and Emotion 3 (2): 81-123. Lakoff, George 1987 Women, Fire and Dangerous Things: What Categories Reveal about the Mind. Chicago: University of Chicago Press. Leslau, Wolf 1995 Reference Grammar of Amharic. Wiesbaden: Otto Harrassowitz. Pesetsky, David 1995 Zero Syntax. Cambridge, MA: MIT Press. Wierzbicka, Anna 1990 The semantics of emotion: Fear and its relatives in English. Australian Journal of Linguistics 10(2): 359-375. 1994 Cognitive Domains and the Structure of the Lexicon: The Case of Emotions. In Hirschfeld and Gelman 1994: 771-97. 1996 Semantics: Primes and Universals. Oxford: Oxford University Press. 1997 Emotional universals. ms. Australian National University. 1999 Emotions across Languages and Cultures: Diversity and Universals. Cambridge: Cambridge University Press. Zaenen, Annie, Joan Maling and Höskuldur Thräinsson. 1985 Case and grammatical functions: The Icelandic passive. Natural Language and Linguistic Theory 3: 441-483.

Emotions and the nature of persons in Mbula Robert D. Bugenhagen 1. Introduction Leavitt (1996), in his survey of anthropological treatments of emotion, states that historically, the analysis of emotions has wavered between two poles: 1) emotions as a matter of nature and body - culturally universal, biologically-based feelings in the body, and 2) emotions as a matter of nurture and mind, with different socio-cultural groups dividing up the emotional pie in different ways and having different conceptions of emotions. Advocates of a universal physical/physiological basis to emotions have a long history in the western scholarly tradition. The ancient Greeks believed that there were four characteristic emotional temperaments, each based on the dominance of a particular fluid in the body: sanguinary (blood), melancholic (black bile), choleric (yellow bile), and phlegmatic (phlegm) (Zimbardo and Ruch 1976: 225). Darwin (1872) held that emotional patterns are largely inherited, innate responses that have had biological utility in evolution. William James (1884) held that emotions are perceptions of physiological changes which are the response to some stimulus. Psychologists have observed correlations between subjects' reports of their emotions and physiological responses by the brain, viscera, endocrine system, and gastric acid, inter alia. Ekman, Sorenson, and Friesen (1969) report that emotions like happiness, sadness, fear, anger, and disgust are commonly displayed on the face with the same distinctive configuration of facial muscle movements. Leavitt (1996: 518) summarises: "Scholars in this tradition conclude that basic emotions are transcultural and that while they may be nuanced in different ways in different societies, at core they must be biologically determined and always the same." On the other hand, a purely physical analysis does not seem entirely plausible, since a given physical response can be consistent with a wide range of emotions and situations. For example, Zimbardo and Ruch (1976: 225) note that crying can be an expression of joy, pain in one's eyes, or relief after great anxiety. Babies cry when they are hungry, and young children cry when they throw temper tantrums. Some people cry at wed-

70

Robert D. Bugenhagen

dings. After a bad meal, a feeling of discomfort in the stomach is simply a physical fact. After a bad quarrel, the same kind of feeling can have an emotional interpretation of anger. And preceding a presentation before a large audience, the feeling can have an emotional interpretation of anxiety. So the immediate context has a very important disambiguating function on the emotional interpretation of physical responses. Leavitt (1996: 524) notes the importance of situation in emotion, stating that there has been a convergence of biological and sociocultural approaches "around something like a notion that socialised human bodies, bodies that normally exist as groups and in interaction rather than as isolated entities, have their being in recurrent situations that call forth the meaning/feeling responses we recognise as emotions." Similarly, a number of the articles in Harre (ed.) (1986) speak about stereotypical situations or scenarios which are expected to call forth emotions. Recent anthropological research like that of Lutz (1982, 1987) and Rosaldo (1980) has demonstrated the importance of an individual's socio-cultural background for their conceptualisation of emotions. Leavitt (1996: 516) says: "The flood of anthropological studies of the 1980s and 1990s has made it abundantly clear that different societies define something that westerners call emotion in different ways." Or, as Geertz (1973: 81) puts it: "Not only ideas but emotions too, are cultural artefacts in the mind." But if emotions are cultural artefacts in the mind, is it somewhat naive to hope that one could ever empathically understand what someone from a different culture is feeling? After all, establishing exactly what is going on in someone else's mind is not a trivial task. In addition to relying on speakers' self-reports of their emotional experiences, one fruitful investigative strategy has been a lexicographic one: "In the 1970s and early 1980s cultural anthropologists and like-minded philosophers elaborated the position that since we can't get inside other people's heads and hearts, we must instead work out other people's definitions of emotions. In accord with this interest a genre of analysis developed around the explication of emotional vocabulary and the interpretation of emotions as cultural categories." (Leavitt 1996: 521) The present paper is an instance of such a strategy. It seeks to precisely specify the meanings of a number of emotion expressions in the Mbula1 language of Papua New Guinea, focussing on those involving body part images. In doing so, use is made of a natural semantic metalanguage2, since:

Emotions and the nature of persons in Mbula

71

To compare meanings one has to be able to state them. To state the meaning of a word, an expression or a construction, one needs a semantic metalanguage. To compare meanings expressed in different languages and different cultures, one needs a semantic metalanguage independent, in essence, of any particular language or culture - and yet accessible and open to interpretation through any language. (Wierzbickal991:6) To study different cultures in their culture-specific features, we need a universal perspective; and we need a culture-independent analytical framework. We can find such a framework in universal human concepts, that is in concepts which are inherent in any human language ...We need to find shared terms; that is, universal concepts. I suggest we can find such concepts in the "universal alphabet of human thoughts" (Leibniz 1903: 430), that is, in the indefinable (i.e. semantically simple) words and morphemes of natural languages. (Wierzbicka 1991: 9-10)

For further defence and exemplification of such an approach, see Wierzbicka (1980: Introduction), and Goddard and Wierzbicka (1994). To see how it applies to Mbula, see Bugenhagen 1994 and In press.

2. Overview of Mbula constructions encoding experiences By way of orientation, there are seven different syntactic constructions expressing experiential notions in the Mbula language - and an eighth one, to be discussed below. Bugenhagen (1990) describes the syntax and semantics of most of these in some detail. 1. Uninflected experiential verbs with coreferential Experiencer Subjects and Objects. Semantically, these divide up into two groups: 1) those encoding physical sensations and experiences: petel 'be hungry', miri 'be thirsty', uze 'be sweaty', basma 'be hungry for meat', and 2) those encoding emotions: menmeen 'be happy' and kaipa 'selfishly rejoice'. These two semantic groups are distinguished syntactically by their ability to cooccur with the prohibitive adverbial pepe 'don't', with only the emotional terms having this possibility. (1)

Nio menmeen yo pa sua ku tana. IS.NOM be.happy IS.ACC OBL talk 2S.LOC that Ί am very happy about that talk of yours (so).'

(2)

Ni petel=i. SS.NOM be.hungry=3s.ACC 'He is hungry.'

72

Robert D. Bugenhagen

2. Inflected experiential verbs with Experiencer Subjects. These fall into the following groups: 1) cognition: -ute 'know', -kilaala 'recognise, perceive', 2) perception: -re 'see, look, watch out for, consider', -leg 'hear, listen to, obey', -yoozo 'smell', -yamaana 'feel', -kanamaala 'spiritually sense that someone or something is present', 3) semi-controllable physical states: -meete 'die, become unconscious', -mel 'fall down, stumble', -ggur) 'cough', -martu 'sneeze', -lulu 'vomit', 4) the 'live' sense of the verb -mbot (which also has the senses 'be located at', 'stay', and 'exist'), and 5) emotional responses: -morsop 'be startled', -murur 'be surprised', -twer 'long for something', -rju 'be offended by someone and make it known to them by keeping one's distance from them when in their presence', -mbur 'be jealous', -pok 'be uncomfortable when in someone's presence and therefore seek to avoid the person'. The experiential stimuli are encoded as either direct Objects (3) or as oblique prepositional phrases governed by the preposition/?« (4). (3)

Nio aq-yamaana yoyou-rja-na biibi. IS.NOM ls-feel hurt-NMZ-3s.GEN big.one Ί feel great pain.'

(4)

Ni i-yu piti. 3s.NOM 3s-be.offended IP.INC.OBL 'He is offended and keeping his distance from us (INC).'

3. Inflected experiential verbs with Experiencer Objects. Such verbs typically express physical stimuli and include: -namut 'be sweet (to)', -yes 'cause the sensation of burning pain (to)', and -tuntun 'cause to itch'. (5)

Marasin ti i-yes medicine this 3s-burn 'This medicine burns me.'

yo. IS.ACC

4. A construction involving the verb -kam 'do, get, receive', which nearly always encodes unpleasant experiences. The experiential stimulus is specified by Subjects like: tata 'busyness', mete 'disease', ggumtur 'embarrassment', sir) 'blood, bleeding', motoijana 'fear' - a nominalised form from the verb -moto 'fear', and dadaru 'doubt', and the Experiencer is encoded as the direct Object.

Emotions and the nature of persons in Mbula

(6)

Mete i-kam disease 3s-get Ί am sick.'

73

yo. IS.ACC

5. A construction in which the inalienable nouns le- and ka- are added immediately following the verb. Le- can be used for both pleasant and unpleasant experiences, whereas ka- is used only for unpleasant ones. (7)

Zin ti-we le-n. BP.NOM 3p-bathe RECX-3p.GEN They are cooling off for pleasure (in the water).'

6. An existential construction in which certain unpleasant experiences are expressed by nouns occurring as modifiers of ka-. (8)

Zin ka-n miai}. 3P.NOM PASS.RECX-3p.GEN shame/embarrassment They are ashamed/embarrassed.' (lit. '[There is] their shame/embarrassment.')

7. The expression -ur koipa 'put enemy to' which functions to express 'malice/hate' and the desire to have nothing to do with someone, as well as simple uncooperativeness, when it collocates with personal nouns. With event expressions like mbulu sananrjana/pakaamrjana/zigzikyana 'bad/false/destructive behaviour', it implies a negative evaluation of such behaviour and a refusal to do it. An explication of its meanings is given below: (9)

a.

Y -ur koi pa Z:personal noun sometimes people think like this about someone: this person is bad because of this, if this person wants something, I don't want the same I don't want to do good things for this person I don't want to be in the same place as this person I want to do something bad to this person when these people think like this they feel something bad when Υ thinks about Ζ, Υ feels something like this

74

Robert D. Bugenhagen

b.

Υ -ur koi pa Ζ':non-personal, event noun sometimes people think something like this about something: it is bad if someone does this because of this, I don't want to do this when these people think like this they feel something bad when Υ thinks about Ζ, Υ feels something like this

Regarding such explications, it will be observed that they express a prototypical situation resulting in some sort of feeling, this being embedded under the component "sometimes people think something like this". A person is then said to "feel something like this". The approximative "feels something like this" serves to indicate that there can be a range of emotions which are compatible with the lexeme, as long as they are basically similar to that invoked by the prototypical situation. When we predicate of two individuals that they are angry, we do not necessarily mean that their feelings are completely identical. The given formulation allows for a certain degree of flexibility in the extensions of a meaning. This is in accord with Wierzbicka (1990: 361): "the definition of an emotion concept takes the form of a prototypical scenario describing not so much an external situation as a highly abstract cognitive structure: roughly, to feel emotion Ε means to feel as a person does who has certain (specifiable) thoughts characteristic of that situation." 8. Body image expressions. McElhanon (1977) states that the use of body image expressions is quite common in both Austronesian and Papuan languages of Papua New Guinea. These are certainly the most productive and semantically richest means of talking about experiences, and are the focus of the remainder of this paper. (10)

Kopo-T) rru pu. stomach-ls.GEN seek + RED ZS.OBL Ί am worried/anxious about you (S).'

3. Body image expressions in Mbula Where do feelings 'happen'? A modern materialist might answer "In the brain and various glands of our bodies." An ancient Hebrew would have answered: In one's lev 'heart', or kilyowt 'kidneys', or nefesh 'throat, de-

Emotions and the nature of persons in Mbula

75

sire, life, soul', orayin 'eye' or mach 'wind, breath, spirit'. For example, in Psalm 73:21 the prosperity of the wicked leads to the following feelings in the psalmist : "My heart was soured/embittered and my kidneys were pierced." Earlier in the psalm, in verse 16, he talks about trying to understand this enigma and says it was "burdensome/laborious/oppressive to his eyes." In New Testament Greek emotions like compassion, longing, and refreshment are attributed to a person's splangxnon 'intestines', while lust, lack of generosity, and thinking and understanding or the lack thereof are attributed to a person's ophthalmos 'eye'. A Mbula speaker finds it relatively easy to identify with body image expressions like those in Hebrew and Greek, since his/her own language also abounds in such expressions. Of course, the English language, too, has its own collection of body image expressions like butterflies in the stomach and broken heart. But much greater reliance is placed upon nouns and adjectives encoding emotions like anxious, anxiety, hurt, offended, offence, happiness, happy, joy, joyful, peace, and peaceful, which do not require a mediating body part. One of the very important differences between English speakers and Mbula speakers is that when Mbula speakers want to talk about emotion, they do not generally have such lexical items to draw on. There are, relatively speaking, very few lexical items in the language which, on their own, encode emotional or physical experiences. In the preceding section, most of the listings of lexical items and constructions having an inherently experiential meaning are exhaustive. It will be noted that the range of experiences encoded by these lexical items with inherently experiential meanings is somewhat limited. For example, there are no specialised words for notions like 'anger', 'love', 'joy', or 'disappointment' . And yet, living among Mbula speakers over the past seventeen years, I can vouch for the fact that they certainly seem to experience feelings which strike me, albeit I am a cultural outsider, as bearing a considerable resemblance to what I understand such terms to mean. But for them, in order to talk about these notions, they must use body image expressions which localise the experience. There is no other option in their language. So for them, most experiences, whether physical sensations or emotions, are mediated via and localised in various parts of the body. When Mbula speakers want to talk about their emotions or the physical sensations they are experiencing, the body part terms which most frequently occur are: mata- 'eye', lele- 'inside(s), feelings', kete- 'liver', ni-

76

Robert D. Bugenhagen

'genitals, being', kuli- 'skin', and kopo- 'stomach'. Less commonly used are: kwo- 'mouth', taiga- 'ear', rjgure- 'throat', and zuru- 'lips'. (There are more specific terms for male and female genitals, but these are largely taboo words. Instead, the more generic ni- is customarily used.) With regard to the mapping of arguments and syntax in such constructions, the Experiencer is encoded as the ostensible genitive of the body part noun, and the experiential stimulus is encoded as the object of the Oblique preposition pa. However, as Bugenhagen (1995: 256-257) notes, the syntactic structure of such constructions is not: [NPGENiTiVE

Body PartN]suBJECTNP

[Verb [pa NP]]

Instead, possibilities of adverb placement indicate the structure to be: [NPsuBJECT]

[[BodyPartN Verb] [pa NP]]

with the body part being syntactically incorporated with the verb into a kind of complex predicate. This is evidenced by the fact that certain adverbs which normally occur between the Subject and the predicate do not occur between the body part noun and the verb, but instead occur between the NP encoding the "possessor" of the body part and the body part noun itself. In the remainder of this paper, the principal categories of experience associated with different body parts will be delineated. In many instances, there is no clear English equivalent for the emotion expressed, so fuller semantic explications using the natural semantic metalanguage are given. Due to limitations of space, examples given below in the various listings are representative only, and not completely exhaustive.

3.1. Mata- 'eye' The most productive body part of all in experiential expressions is mata-, a polysemous form whose various senses include: 'eye, face, front, presence, edge of a sharp object, first, chief, brightness'. (Examples of some of these non-experiential senses are: wooggo mataana 'the bow/front part of a canoe/boat/ship', tibeede mataana 'they wrote on/decorated his/her face', buza mataana 'knife's edge', tikam sua ila biibi mataana 'they talked in the presence/view of the big one/leader', tomtom mataana kana 'the first

Emotions and the nature of persons in Mbula

77

person', zog mataana men 'the sun is very bright'.) The least extended experiential meanings are observed in expressions like mata- yenrjeene 'eye sleepy - to be tired and find it difficult to keep one's eyes open', mata- ikam pak 'eye does opening - to wake up and open one's eyes', mata- iyoyou 'eye hurts - feel pain in one's eye', and in expressions having to do with seeing), like those below: (11)

Ymata-rikrik (lit. eye move.back.and forth) 'look around'

(12)

mata- ito (lit. eye follows) 'look at something which is moving away'

(13)

mata- ila I imar lisu I ise I imiili (lit. eye goes/comes/descends /ascends /returns) 'look ahead at something'/'look at the down' /' look up' /' look back'

speaker'/'look

(14)

Υ mata-ikam Z (lit. eye do/get/receive) 'involuntarily happen to see something' something happens in someone Y's eyes because of this, Υ sees something Ζ not because Υ wants to see it

(15)

Υ mata- mbeij (lit. eye night) 'to not see something which one should be able to see' (used to scold someone) someone Υ does not see something. I think: if Υ wanted to see this, Υ could see this because of this, I think Υ does something bad I say Υ is like someone who cannot see things because it is night3

(16)

mata-pisana (lit. eye-closed.one) 'blind'

78

Robert D. Bugenhagen

The next class of experiential examples with mata- have to do with life and consciousness, the capacity to think and know. (17)

Υ mata- iyaryaara (lit. Υ eye is shining) 'be alive/be awake with one's eyes open' someone/something Υ is alive (now Υ can see things)

(18)

mata- koukou (lit. eye white) 'unconscious'

(19)

mata- zugutgut (lit. eye becoming.dark) 'be on the verge of fainting and becoming unconscious'

(20)

mata- ipit se (lit. eye jump.up ascend) 'come back to consciousness after having been unconscious'

(21)

Ymata-imbol (lit. Υ eye is.strong) 'to be morally aware' (used of children from around the age of 8, who are then viewed as being responsible for doing wrong) some people know that if a person does something it is bad because of this, if someone does something like this people think: "this person did something bad" Υ is like this

(22)

Υ mata- kankaana pa Z (lit. Υ eye confused Oblique Z) Ύ is confused and as a result accidentally makes a mistake having to do with Z'

(23)

Υ mata- mburri (lit. Υ eye [?]) sometimes people think like this: many things are happening in this place because of this, I don't know what I want to do

Emotions and the nature of persons in Mbula

79

when these people think like this they feel something (bad) Υ feels something like this (There is another sense for this expression, which appears to be essentially synonymous with that ofmata- zugutgut: 'be on the verge of fainting and becoming unconscious'.) (24)

mata- iur pa (lit. eye puts Oblique) 'to know how to do something'

(25)

Υ mata- molo (lit. Υ eye long) when things happen in places far away from here, some people know these things happen Υ is like this ('omniscient' when used of God)

(26)

a.

Y mata- iygal Ζ (lit. Υ eye pierces Z) before now, Υ did not think about something Ζ now Υ thinks about this thing

b.

Υ mata- iygal Z:personal noun (pa W) (lit. Υ eye pierces Ζ Oblique W) before now Υ did not think about person Ζ now Υ thinks about this person Υ wants to do good things for this person because of this, Υ does something good for this person

(27)

Υ mata- iggal itu(lit. Υ eye pierces self) Ύ is careful and takes precautions' 1.

sometimes people think like this: something bad could happen to me I don't want this thing to happen to me if I do some other thing, this bad thing cannot happen to me because of this, these people do this other thing Υ does something like this

80

Robert D. Bugenhagen

2.

sometimes people think like this: I could do something bad I don't want this to happen if I do some other thing, I will not do this bad thing because of this, these people do this other thing Υ does something like this

(The expression mata- ise kor 'eye ascends above' has a similar meaning of being careful or wary.) (28)

Υ -ρο ZseY mata(lit. Υ ties Ζ ascend Υ eye) 'memorises' Υ thinks: after now, I want to be able to think about something Ζ because of this, Υ does something because of this, afterwards, Υ can think about Ζ

(29)

Υ mata- imiili pa Ζ (lit. Υ eye returns Oblique Z) something Ζ happened before now Υ thinks about this thing now

(This expression has another sense: 'look behind oneself.) (30)

mata- ila pa (lit. eye goes Oblique) 'think about something'

(This expression has another sense: 'look ahead'.) (31)

Υ mata- imbeleele Ζ (lit. Υ eye shrivels Z) Ύ forgets Z'

(32)

Υ mata- imunlmata munrjaana (lit. Υ eye be.closed/eyes closed.ones) 'ignorant' sometimes a person does not know something other people know this thing

Emotions and the nature of persons in Mbula

81

I think it is bad if a person does not know this thing person Υ is like this (33)

Υ mata- leyaleija (lit. Υ eye in.different.places/directions) 'unable to concentrate, scatterbrained' it is good if a person can think about one thing at one time sometimes when a person thinks about something, this person thinks about many other things at the same time because of this, this person cannot think about this one thing I think it is bad if a person is like this person Υ is like this

(34)

Υ mata- ise ki Z (lit. Υ eye ascend Locative Z) 'look up to, respect, admire' sometimes people think like this: some people do not do bad things these people do good things because of this, if a person of this kind says something, I want to hear it if a person of this kind says to me: I want you to do something, I want to do this thing when these people think like this, they feel something good when Υ thinks like this about someone Z Υ feels something like this

(In the Mbula culture, to show respect one is obedient/responsive to the desires of people who are admired and pays attention to what they say. Conversely, if one is not obedient/responsive to the desires of someone and doesn't pay attention to them, this indicates a lack of respect.) (35)

Υ mata- pasom Ζ (lit. Υ eye cause.not Z) Ύ despises Z as being worthless/insignificant'

The third class of experiential expressions involving mata- are more clearly emotional in nature.

82

(36)

Robert D. Eugenhagen

Υ mata- putput (lit. Y eye bulging) 'furious and about to act' sometimes people think like this: something very bad happened I want to do something very bad to someone now when some people think like this they feel something very bad their eyes are like they are bulging Y's eyes are like this now because of this, I think Υ feels something like this now

(37)

Υ mata- pambaara (lit. Υ eye crossed) sometimes people think like this: something bad happened I want to do something bad to someone after now when some people think like this they feel something bad their eyes are like they are crossed Y's eyes are like this because of this, I think Υ feels something like this now

In the Mbula translation of the Bible, Cain is said to have 'crossed eyes' before he kills his brother Abel. (38)

Υ mata- berber pa Ζ (lit. Υ eyes be.coming.into.view Oblique Z) 'covet' sometimes if people see something, they think like this: this thing is good I want to have this thing when these people think like this they feel something when Υ sees something Ζ, Υ feels something like this

(39)

Υ mata- imapalpaala pa Ζ (lit. Υ eye be.breaking Oblique Z) 'intense coveting to the point of being unable to concentrate, obsessed' sometimes if people see something, they think like this: this thing is very good

Emotions and the nature of persons in Mbula

83

I very much want to have this because of this, these people cannot think about other things when these people think like this, they feel something when Υ thinks about Ζ, Υ feels something like this (40)

Υ mata- kinigana (lit. Υ eye food.NMz) some people think about food all the time people like this want food all the time I think it is bad if a person is like this Υ is like this

Other items occurring in this mata- -...-gana construction are: moni 'money', pat 'stone/money', and moon 'woman'. (41)

Υ mata- mburmbur (pa Z) (lit. Υ eye jealous Oblique Z) 1.

2.

'jealous of another's good fortune' sometimes people think like this: if something good happens to someone else, this thing does not happen to me because of this, I don't want good things to happen to other people something good happened to someone else I want to do something bad to this person when these people think like this, they feel something bad when Υ thinks about Ζ, Υ feels something like this 'jealous of one's relationship with someone else', sometimes people think like this: another person is doing something with this person I want to do this thing with this person if this other person does this thing with this person, I cannot do this thing with this person because of this, I don't want anyone to do this thing with this person I want to do something bad to this other person when these people think like this they feel something bad when Υ thinks about Ζ, Υ feels something like this

84

Robert D. Bugenhagen

Although -mbur is a verb meaning 'jealous', it nearly always co-occurs with mata- rather than a simple pronominal subject. Inherent in the notion of jealousy is the idea of 'limited good', that more for someone else means less for me. This was graphically demonstrated a number of years back when the wife of someone in the village where we reside found a fibreglass boat washed up on the shore. At the time, there were no other boats in the village, and transportation of goods to and from the village was a considerable hardship. The family who found the boat took it and put it under their house for a number of months because they felt people were so mata- mburmbur that they were likely to take an axe and break up the boat, even though it had the potential to help the whole village. Another eye-opening incident was the time a neighbour had planted some watermelons which were nearly ripe and one morning noticed that they had all been smashed. When I talked with one language helper and asked why someone would smash the melons instead of stealing them and eating them themselves, he replied along the lines of, "The person who smashed the melons probably thought like this: 'Should they eat melons and others not eat them? No'." In other words, it was better that no one eat the melons than that some people should be able to eat them while others could not. (42)

Υ mata- iyoyou pa Ζ (lit. Υ eye hurts Oblique Z) 'not want to see something anymore' sometimes people think like this: if I see something/someone, I feel something bad before now, I saw this thing/person many times because of this, I don't want to see this thing/person now when these people think like this they feel something bad when Υ thinks about Ζ, Υ feels something like this

(43)

Υ mata- pot pa Z:event NP (lit. Υ eye float Z) 'do something reluctantly and poorly' sometimes people think like this: I don't want to do something because of this, I don't want to do this thing well I think it is bad if a person thinks like this when Υ thinks about Ζ, Υ thinks like this

Emotions and the nature of persons in Mbula

(44)

85

Υ mata- sirjsirj pa Z (lit. Υ eye red Oblique Z) 'eager, enthusiastic' sometimes people think like this: something Z is good I very much want to do this because of this, these people do this thing well I think it is good if a person does this person Υ does something like this

(45)

Υ iur mata- (pa Z) (lit. Υ puts eye Oblique Z) 'hope for, expectantly wait for something good' sometimes people think something like this: something/someone Z is good some time after now, this thing (person) will happen (be here) if Z happens (is here), I want to see Z because of this, I will do something because of this, after now if Z happens (is in this place), I can see Z when these people think like this they feel something because of this when Υ thinks about Ζ, Υ feels something like this

One further experiential expression with mata- is mata- sirjiini isu 'eye's blood descends — experience great physical suffering'. From the examples given above, it will be seen that mata- can be used to express thinking, knowing, wanting, and feeling. Its most common experiential function, however, is to express thinking and knowing. No other body part is so frequently associated with expressions indicating thinking and knowing. Many times, the experiences expressed using mata- are the result of seeing something, as in mata- iyoyou and mata- herber. With -ur matapa, someone desires to see something. In other instances, like mata- putput and mata- pambaara, the eyes have a signalling function. One recognises the person's emotional state from the appearance of their eyes/face.

86

Robert D. Bugenhagen

3.2. Lele- 'insides' The second most productive item in experiential body image expressions is the noun lele- 'part or space located inside of something, insides/feelings'. Its spatial sense can be seen in the following examples. (46)

Ni i-mbot ruumu lele-ene. SS.NOM 3s-be.at house inside-Ss.GEN 'He is inside the house.'

(47)

Ruumu lele-ene ru. house inside-3s.GEN two 'The house has two rooms/two spaces inside it.'

(48)

Ti-sap woorjgo lele-ene. 3p-hew canoe inside-3s.GEN 'They hollowed out the canoe.'

Our concern, however, is not with these spatial senses of lele- but with its use in sentences like the following: (49)

Lele-g i-saana pizin. feeling-ls.GEN 3s-deteriorate 3P.OBL Ί feel bad/sorry for them.'

(50)

Zin ti-ler)=i som, tabe lele-ene 3P.NOM 3p-listen.to=3s.ACC not and.so feeling-3s.GEN i-saana. 3s-deteriorate 'They didn't listen to him, and so he became upset.'

(51)

Mo IS.NOM le-T)

lele-rj ambai paso, insides-ls.GEN be.good because pata-rja-na sa

RECX-lS.GEN

heavy-NMZ-3s.GEN

NON.REF

som. NEG

Ί am happy/content because I don't have any troubles/problems.' (52)

Sua ku tana i-kam yo ma lele-rj ambai. talk 2s.LOG DEM 3s-do IS.ACC CSB insides-ls.GEN be.good 'What you (S) said made me happy./I am pleased with what you (S) said.'

Emotions and the nature of persons in Mbula

87

Unlike mata- , lele- is rarely used to express any sort of physical sensation or experience. The only instance I know of is lele- ibayou 'insides be.hot - to feel hot (and thirsty)' (because of being out in the sun and exerting oneself. Having a temperature/fever is expressed differently as kuliibayou 'skin be.hot'. There is also a handful of body image expressions where lele- appears to have a cognitive function. These are given below: (53)

a. b.

Y lele- iurpa Z.NP (lit. Υ insides puts Oblique Z) Υ lele- iur be Z:S (lit. Υ insides puts NAF Z)

1.

'someone Υ decides to do Z' before now, someone Υ did not want to do something Z now Υ wants to do this thing

2.

Ύ chooses Z and is determined to do good things for Z' sometimes people think like this about someone: before now, I did not want to do good things for this person now I want to do good things for this person after now, I will do good things for this person nothing can cause me to not want to do these things for this person4 Υ thinks like this about someone Z

This "choosing" is more along the lines of the biblical concept of election; i.e. Υ chooses to have a special relationship with Z in which Υ is committed to Z and is determined to do good things for Z. (54)

Υ lele- iwe ru pa Z (lit. Υ insides become two Oblique Z) 'doubt' 1.

sometimes people think like this: I want someone to do something this person might not do it because of something else because of this, I don't know if this person will do this when Υ thinks about something Ζ, Υ thinks like this

88

Robert D. Bugenhagen

2.

'doubt about something whether it would be good to do it or not' sometimes people think like this about something: I don't know if this is good I think it could be bad because of this, I don't know if I want to do this thing Υ thinks like this about Ζ

(55)

Υ lele- iur pa Z sorok (lit. Υ insides put Oblique Z without.basis) 'to suspect that someone has done something wrong' sometimes people think like this: I think someone did something bad I don't know if this person did this thing. Υ thinks like this about something Ζ

(56)

Υ lele- iurur (lit. Υ insides be.putting) 'perplexed, not knowing what one wants to do'

Apart from the few physical and cognitive uses given above, lele- is used to express desire and a wide range of emotional states, the more common of which are listed below. (57)

a.

Y lele- pa Z:N non-personal (lit. Υ insides Oblique Z) sometimes people feel something inside because of this, they want something Υ feels something like this because of this, Υ wants something Ζ

b.

Υ lele- pa Z:N personal 'love, like someone' sometimes people think about someone like this: when I think about this person, I feel something good inside I want to be where this person is I want to do good things for this person when these people think like this about someone, they feel something

Emotions and the nature of persons in Mbula

89

when Υ thinks about Ζ, Υ feels something like this (58)

Υ lele- be Z:S (lit. Υ insides NAF Z when Υ thinks about Ζ, Υ feels something inside because of this, Υ wants Ζ to happen

(59)

a.

Y lele- ambai (pa Z:S) (lit. Υ insides be.good Oblique Z) 'happy, contented, at peace, approving, grateful, free from worry, care, anger, and sorrow' Υ does not feel anything bad inside Υ feels something good inside Υ (because of Z)

b.

Υ lele ambai pa Z:NP (lit. Υ insides be.good Oblique Z) 'be pleased with someone, approve' Υ does not feel anything bad towards someone Z when Υ thinks about Ζ, Υ feels something good inside

(60)

Υ lele- ipuru Υ (lit. Υ insides uproot Y) = Y lele- imagga (lit. Y insides get up) before now, Y did not want to do something now Y feels something inside because of this, Y wants to do this thing

(61)

Ylele-imbol (lit. Y insides be.strong/firm) 'not easily persuaded to do things' when someone Y does not want to do something, if someone else wants Y to do it, Y will not do it

This expression is most frequently used in reference to people who are stingy, but it can also be used to refer to people of integrity who will not violate their good principles.

90

(62)

Robert D. Bugenhagen

a.

Y Me- isaana pa Z:personal noun (lit. Υ insides deteriorate Oblique Z:N) 'feel sorry for someone and want to help them, compassion' sometimes people think like this: something happened to someone this is bad for this person when these people think like this, they feel something bad inside these people want to do something good for this person these people think: if I do this, this person will not feel bad after this when Υ thinks about Ζ, Υ feels something like this

A common situation where the expression lele- isaana pa is heard, is when someone is in need of food or money due to no fault of their own. People feel bad for such a person and want to do something to help them, such as to give them money or food. b.

Υ lele- isaana (pa Ζ .-non-personal noun) 'feel bad about, be upset by' sometimes people think like this: someone did something bad to something Ζ this person did something bad to me when people think like this, they feel something bad inside when Υ thinks about Ζ, Υ feels something like this

c.

Υ lele- isaana (pa Ζ:non-personal event noun) sometimes people think like this: someone did something bad this person did something bad to me when these people think like this, they feel something bad inside when Υ thinks about Ζ, Υ feels something like this

d.

Υ lele- isaana paso Z:S (lit. Υ insides deteriorate because Z:S) 'feel bad about, be upset' sometimes people think like this: someone did something

Emotions and the nature of persons in Mbula

91

this person did something bad to me when these people think like this, they feel something bad inside when Υ thinks about Ζ, Υ feels something like this The expression lele- iygis pa 'insides be.slanted Oblique' appears to be essentially identical in both meaning and syntax to the latter two senses of lele- isaana. An example of the first would be Ni leleene isaana pa mbu kini ta tikem na 'He is upset because of his areca nuts which they stole'. Event nouns are items like sua 'talk', kuumbu 'theft', zoorotjana 'rebellion/stubbornness'. An example with an event noun (62 c) would be: Ni leleene isaana pa sua kizin 'He is upset because of their talk'. (63)

Υ lele- ipata pa Ζ (lit. Υ insides be.heavy Oblique Z) 'very sad or troubled and helpless to do anything about the situation' sometimes people think like this: something bad happened when these people think like this, they feel something very bad inside they don't want to feel like this they want to do something because of this they cannot do anything when Υ thinks about someone/something Z, Υ feels something like this

Lele- isaana and lele- ipata differ in: 1) the intensity of the feeling (with lele- ipata expressing greater intensity), 2) the Experiencer's assessment of whether or not someone is culpable for having caused the experience, and 3) whether or not the Experiencer has the possibility of doing anything about the situation. If someone of lesser status steals a person's areca nuts or publicly disagrees with him/her, both of which are viewed as personal slights, the expression lele- isaana would be more favoured. If a relative dies, lele- ipata would be used, because the situation is something very bad and not really any particular person's fault, and there is nothing one can do to undo the situation.

92

(64)

Robert D. Bugenhagen

Υ lele- imbukmbuk pa Z (lit. Υ insides be.swelling Z) 'anxiety' sometimes people think something like this: something bad could happen to/because of something now when these people think like this, they feel something bad inside when Υ thinks about Ζ, Υ feels something like this

(65)

Υ lele- sa pa Ζ (lit. Υ insides non.referential Oblique Z) = Υ ni- sa pa Z (lit. Υ being non.referential Oblique Z) sometimes people think like this: someone did something bad to me when these people think like this, they feel something bad inside because of this, if these people and this person are in the same place, these people do some things because they do these things, other people know they think that this person did something bad to them I don't know what this person did to these people when Υ thinks about someone Ζ, Υ feels something like this

Lele- sapa and ni- sapa appear to be completely synonymous. The kinds of things done in response to this feeling are: 1) Υ does not look Z directly in the eye, 2) Υ says as little to Z as possible, and 3) Υ does not smile. Lelesapa is the expression used when one sees from another's actions that they are upset with someone, but does not know the reason why. (66)

Υ lele- ipasiksik Z (lit. Υ insides compel Z) 'unable to refrain from doing something'

(67)

Υ lele- ikam uraata (lit. Υ insides do work) 'be emotionally moved' someone Υ feels something

Emotions and the nature of persons in Mbula

93

This expression is the most generic way of all of expressing that someone is feeling emotion. It indicates nothing whatsoever about the kind of emotion being experienced. (68)

Υ lele imbai pa Ζ (lit. Υ insides be.poor Oblique Z) sometimes people think like this: someone did something bad because of this, I feel something bad inside I want this person to know this because of this, if this person wants me to do something, I will not do it when these people think like this, they feel something bad when Υ thinks about someone Ζ, Υ feels something like this

This expression is most commonly heard in households when the wife is upset with her husband and expresses this by refusing to do things like cook for him or wash his clothes. But it also is used when people in the village are upset with someone and therefore refuse to cooperate with him in traditional ceremonies or large work projects. (69)

Υ lele- iluumu (pa Z) (lit. Υ insides be.cool Oblique Z) before now, someone Υ felt something bad inside (because of Z) now Υ does not feel like this any more

Lele- iluumu is a generic expression indicating the cessation of unpleasant emotional states. When one is no longer angry or sad, their insides are said to be 'cool'. (70)

Υ lele- imiili pa Z:personal noun (lit. Υ insides return Oblique Z) 'reconciled and no longer angry' sometimes people think like this about someone: before now, when I thought about this person I felt something bad inside because of this, I did not want to do good things for this person

94

Robert D. Bugenhagen

I wanted to do bad things to this person now I don't feel like this any more Υ thinks like this about Z (71)

YiraoZlele(lit. Υ be.adequate Z's insides) 'something Υ satisfies someone Z'

(72)

a. b.

Y lele- imet (kat) pa Z:event NP (lit. Υ insides snap.in.two (really) Oblique Z) Υ lele- imet (kat) be Z:S (lit. Υ insides snap.in.two (really) NAF Z) 'someone Υ becomes determined to do something Z, regardless of any possible negative consequences doing it might have for Y' sometimes people think like this: I want to do something I know: if I do this, something bad can happen to me I don't want not to do it because of this when these people think like this, they feel something bad inside because of this when Y thinks about Ζ, Υ feels something like this

With regard to the component "Y feels something (good/bad) inside", it seems that the explications should make explicit the localisation of the experience implied by the use of the noun lele-. Explications for the other body image expressions will also frequently make use of such a component, with "inside" replaced by the corresponding body part term. Such explications make no claim about the validity according to western science of associating an emotion with a particular body part. There may or may not be a scientific basis for such associations. That is not the issue here. What we are seeking to do is to reflect the Mbula conceptualisation of emotion, as revealed by the way they talk about it. The five expressions lele- pa, lele- ambai, lele- isaana, lele- ipata, and lele- iluumu are certainly among the most frequently occurring body image emotional expressions in the language and correspond to a wide variety of emotion terms in English. Of note is the fact that lele- ambai spans a range of emotions which English sub-divides much more: 'happy, pleased, approving, grateful, content, at peace, free from worry, care, sorrow, and

Emotions and the nature of persons in Mbula

95

anger'. Its semantic invariant seems to simply be the absence of any negative emotion, with an optional component that one feels something good. It differs, in this respect, from the more positive, uninflected verb menmeen 'happy', which has an obligatory component of feeling good. Along similar lines, the expression lele- iluumu expresses a return to this neutral emotional state from any kind of bad one: sorrow, anger, jealousy, etc. It seems to be the case that the Mbula lexicon is much more precise in delineating negative emotions than positive ones. One wonders if there are cultural reasons for this. People who are emotionally troubled constitute a severe threat in the Mbula culture, since they are very likely to act on their feelings by physically attacking someone, or working sorcery on them, or destroying something, or, at the very least, ceasing to provide needed cooperation. Being able to precisely talk about bad feelings can be viewed as having considerable adaptive value, since it allows one to diagnose efficiently the causes and current or envisioned consequences of an individual's bad emotional state, in preparation for the process of restoring them to at least a neutral emotional state where they no longer constitute a threat. The relative paucity of lexical expressions for positive emotional terms also reflects an uneasiness in the Mbula culture with attributing positive emotions to oneself. Apart from religious contexts, where expressing one's happiness or contentment with life because of what God has done is acceptable, it is relatively rare for people to announce that they are happy. Perhaps, this is due to the large amount of jealousy in the culture. Saying that you are happy makes you more likely to be a target of other people's jealousy. It is somewhat more common for people to attribute happiness to others.

3.3. ^^-'chest/liver' With these comments about lele-, we now turn our attention to another extremely productive body-part term: kete- 'liver, chest, flat-shaped part of something'. (The semantic invariant of the physical referents of kete- is that they are parts of something which are flat in shape. For example, nama- keteene is the flat part of one's hand or palm, kumbu- keteene is the flat part of one's foot or sole, kar keteene is a flat area in the centre of a village, lele keteenerjana is a flat plain.) Unlike mata- and lele-, body image expressions containing kete- never express pure cognition. The term is

96

Robert D. Bugenhagen

used to express some physical states like hunger and different ways of breathing: kete- bilbil 'chest/liver empty -be very hungry', kete- imaraaza 'chest/liver tear - be so hungry as to feel sharp physical pain in one's stomach', kete- isalaklaaga 'be nauseated', kete- kutkut 'chest/liver be.pulsating - be breathing rapidly (due to exhaustion, pain, or fear)', kete- ipas 'chest/liver is removed - be out of breath', kete- ikam kug kek 'has ceased to breathe - dead', kete- nono 'chest/liver be.hunting - heart beats rapidly'. The preeminent emotional function of kete- is to express rash, impetuous responses which are not well thought through and strong emotions like anger. (73)

Υ kete- guruklpitpit (pa Z) (lit. Υ chest/liver move.a.bit Oblique Z) 'do something too quickly without thinking of the possible bad consequences of one's action' sometimes people feel something in the part of them where their liver is because of this, these people want to do something now these people do this thing now in a short time because of this, something bad could happen I think it is bad if a person does something like this I think if these people do not do this thing now, afterwards these people will think: if I do this, something bad could happen because of this, I will not do this thing because of this, these people will not do this thing because of something Z, someone Υ feels something like this

(74)

Υ kete- pitpit pa Z (lit. chest/liver be.jumping Oblique Z) 'respond to something too quickly without thinking of the potential bad consequences of one's action' sometimes when something happens, people feel something in the part of them where their liver is because of this, these people want to do something now in a short time these people do this thing now because of this, something bad could happen

Emotions and the nature of persons in Mbula

97

I think it is bad if a person does something like this I think if these people do not do this thing now, afterwards these people will think: if I do this, something bad could happen because of this, I will not do this thing because of this, these people will not do this thing because something Z happens, someone Υ feels something like this (75)

kete- ikam keglwek (lit. chest/liver does snapping/quickly moves vertically) 'be greatly startled by something unexpected happening and have one's breath taken away'

The unexpected event can be either something good or something bad. A good example comes from the Mbula translation of the book of Genesis. When Jacob hears that his long-lost son Joseph is alive and well in Egypt, this expression is used. It is, however, more frequently used to talk about responses to startling bad things. The same expression is also used to talk about steadying oneself when one loses one's balance and almost falls. (76)

Υ kete- ise (pa Z) (lit. Υ chest/liver ascends Oblique Z) 'become enthusiastic about doing something' sometimes people think like this about something: I want to do this now when these people think like this, they feel something in the part of them where their liver is before this, these people did not want to do something now these people want to do this thing very much when Υ thinks about something Ζ, Υ feels something like this

(77)

Υ kete- malmal (pa Z:personal noun) (pa W:event noun) (lit. Υ chest/liver fights Oblique Z Oblique W) Ύ is angry atZ' sometimes people think like this: someone did something bad when these people think something like this, they feel something bad in the part of them where their liver is

98

Robert D. Bugenhagen

these people want to do something bad to this person when Υ thinks about someone Z (and something W), Y feels something like this (78)

Υ kete- ibayou (pa Z:personal noun) (pa W:non-personal noun) (lit. Υ chestAiver be.hot Oblique Z Oblique W) = Υ kete- kembeita lai (lit. Υ chest/liver like wild yam) = Υ kete- ibeleu (pa Z) (pa W) (lit. Υ chest/liver swirls Oblique Ζ Oblique W) = Υ kete- ipakpak (pa Z) (pa W) (lit. Υ chest/liver be sour Oblique Z Oblique W) ' Υ is furious at someone Z (because of what Z did to something W)' sometimes people think like this: someone did something very bad when these people think like this, they feel something very bad in the part of them where their liver is these people want to do something very bad to this person when Υ thinks about someone Z (and something W), Υ feels something like this

All four of these expressions appear to be essentially synonymous. They express a greater intensity of anger which is likely to lead to a person doing worse things to someone else. (79)

Z ipas Υ kete(lit. Υ removes Z's chest/liver) 'Z enrages Y' sometimes people think like this: someone did something/ something Υ happened now I think this is very bad when these people think like this, they feel something very bad in the part of them where their liver is before this happened, these people didn't feel anything bad these people want to do something very bad to someone/something when Υ thinks about Ζ, Υ feels something like this

Emotions and the nature of persons in Mbula

99

The expression -pas kete- is used to talk about someone causing someone else to become angry and focuses upon the transition from a state of not being angry to one of being angry. (80)

Υ kete- imap (lit. Υ chest/liver ends) 'be astonished/dumbfounded because of something unusual happening and consequently have one's breath taken away'

3.4. Ni- 'being' The term ni- is somewhat difficult to define. It is used to talk about physical sensations which are not localised in any one particular part of the body, as in the following examples: (81)

ni- imbeeze ni- itekteege ni- ambai ni- iyoyou ni- isaana ni- ipeele

being is.smooth being is.holding being is.good being hurts being deteriorates being opens

'feeling fit or fresh' 'shiver with cold' 'healthy' 'physically suffer, ache all over' 'be exhausted, worn out' 'feel fully awake'

(A near equivalent of ni- isaana would be mbura- imap 'strength be finished'.) Ni- is not, however, equivalent to body in English. For example, a dead body is not referred to as someone's ni-. When talking about burying people, the person being buried is referred to by name, kin-term, or via a pronoun. A person's ni- is never said to be buried. If one wants to talk about a person's body going into the ground and decaying, it is their kuli- (also 'skin') which is said to undergo the process of decay. A second sense for ni- is that it functions as a euphemism for male and female sexual organs. And a third sense is that of a part of something which is long and narrow: ndapndap niini 'pestle', yok niini 'watercourse', lolo niini 'lightning bolt', ndeete niini 'step/rung of a ladder'. In the expression ni- ikilaalalikanamaala 'being recognises/senses', it functions to encode perception via some means other than the five senses. In emotional expressions, it often resembles a negative counterpart to lele-, functioning to encode dislike. But there is also one very important ex-

100

Robert D. Bugenhagen

pression encoding a positive emotion: ni- ise. Since ni- is specifically used to indicate feelings that are non-localised, it will be observed that in the following explications the component "feels something in a part of F" is absent.

(82)

F ni- pa Z som (lit. Υ being Oblique Z not) Ύ does not want Z'

(83)

F ni- gesges pa Ζ (lit. Υ being tedious Oblique Z) Ύ finds Z tedious, irksome' sometimes people think like this: something happened for a long time/I did something for a long time before now, if this thing happened/I did this thing, I did not feel something bad now, if this thing happens, I will feel something bad because of this, I don't want this thing to happen after now when these people think like this, they feel something bad when Υ thinks about something Ζ,Υ feels something like this

(84)

Ύηϊ-sapaZ (lit. Υ being non-referential Oblique Z) See explication for lele- sa pa.

(85)

F ni- sanaana pa Z (lit. Υ being sin Oblique Z) = F ni- puliizi pa Z (lit. Υ being [?] Oblique Z) 'upset with' sometimes people think like this: someone did something bad to me because of this I feel something bad I want this person to know: when I think about this person, I feel something bad when these people think like this, they do something bad because of this, these people do some things when Υ thinks about someone Ζ, Υ feels something like this

Emotions and the nature of persons in Mbula

101

The kinds of things done are: 1) Υ does not look Z directly in the eye, 2) Υ says as little to Z as possible, 3) Υ does not smile, and 4) when Υ is in Z's presence, Υ does not come close to Z.

(86)

Υ ni- ise (pa Z) (lit. Υ being ascends Oblique Z) 'be visibly exultant over one's good fortune' sometimes people think like this: something good happened to me this thing did not happen to other people because of this, I am not like other people now when these people think like this, they feel something very good these people do some things because of this, if other people see these people, they know that these people feel like this when Υ thinks about something Ζ, Υ feels something like this

The last expression, ni- ise, is a difficult one for which to find an English equivalent. It occasionally borders on pride, feeling that one is superior to others. But it is also used when someone just feels happy because of something good happening to them which has not happened to other people. The key concept is that the person somehow feels "special" and others can see it from the way the person behaves, because he smiles, is louder and more forceful in speech, walks more erectly and generally acts more vigorously than people normally do. The triggering events can be something as simple as wearing a new outfit of clothes, or getting a new lawn mower, or being on a winning soccer team. But the term can also apply to Christians who feel grateful and happy to have been touched by God's grace.

3.5. Kuli- 'skin' The next term to be considered is kuli- 'skin, bark, surface'. It is used to express various physical experiences in expressions like: kuli- isu 'skin descends — lose weight, skin becomes wrinkled due to aging', kuli- imalai 'skin wilt - feel tired', kuli- ila muriini 'skin goes.to its.place - feel rested and refreshed', kuli- ire yoyourjana 'skin sees pain - physically suffer', kuli- iswidit 'skin be.burnt - be sunburned', and kuli- isarakrak 'skin

102

Robert D. Bugenhagen

shivers - have shivers up one's spine (in response to being cold or encountering the supernatural)'. Kuli- is probably the nearest equivalent to the English concept 'body'. When someone is buried and their physical body decays, it is their 'skin' which is said to rot. But unlike body, in Mbula one's skin is not said to be buried; it is the person who is buried. The physical resurrection which, according to Christian belief, will occur on the last day is described as tamarjga kulindi munrjaana 'We (INC) get up, our skins whole'. And food that keeps one healthy is characterised as ikis kulindi 'it holds our (INC) skin'. It is possible for the skin to function as an organ of perception, as in the following example: (87)

Kuli-rj i-kilaalali-kam la rjgar kizin skin-ls.GEN 3s-sense/3s-get go thinking SP.LOC Ί somehow sensed/felt what they were thinking.'

kek. PERF

Emotions akin to fear, shame, and embarrassment are typically localised in the skin, as the following body image expressions illustrate: (88)

a.

Y kuli- ka- miarjpa Z:personal noun (lit. skin PASS-RECX shame Oblique Z) ' shy/embarrassed/ashamed' sometimes people think like this when someone sees them: this person is thinking about me I don't want this when these people think like this, they feel something bad in their skin these people do not want to see this person these people do not want to be in the same place as this person these people want to be in some other place when Υ thinks about someone Ζ, Υ feels something like this

The term miarj subsumes a whole range of English terms: shame, embarrassment, shyness. The experience is an uncomfortable one of being the focus of someone's attention when one does not want to be under the "spotlight", and is usually triggered either by having done something bad or being the focus of attention of someone whom one does not know well. The typical responses to such a feeling are: 1) averting one's eyes from the person and 2) seeking to avoid them.

Emotions and the nature of persons in Mbula

b.

(89)

103

Υ kuli- ka- miar) pa Z:non-personal noun 'shame' sometimes people think like this: before now, I did something bad if other people see me, they will think about this bad thing I don't want this when these people think like this, they feel something bad in their skin these people do not want these other people to see them these people do not want to be in the same place as these other people these people want to be in some other place when Υ thinks about something Ζ, Υ feels something like this

Υkuli- puliizi (lit. Υ skin embarrassed) 'be embarrassed because of being talked about' sometimes people think like this: other people are saying things about me I don't want this when these people think like this, they feel something bad in their skin these people do not want to be where these other people are these people want to be in some other place

(90)

Υ kuli- imoto (pa Z) (lit. Υ skin be.afraid Oblique Z) sometimes people think like this: something bad will happen to me because of something I don't want this when these people think like this, they feel something bad in their skin when Υ thinks about something Ζ, Υ feels something like this

Bugenhagen (1990: 211) states that this expression "is used to encode fears which are uncontrollable responses to having sensed something. The kinds of things which induce this sort of fear are reported to be: 1) feeling something in one's skin, 2) vaguely hearing or seeing something." Here,

104

Robert D. Bugenhagen

kuli- is being used in a wider sense than just 'skin', but 'skin' is included in the prototype proposed here. (91)

Υ kuli- mburaana pa Z som (lit. Υ skin its-strength Oblique Z NEG) Υ feels something in Υ's skin because of this, Υ does not want Z

(92)

a. b. 1.

Y kuli- imet (kat) pa Z:NP (lit. Υ skin snaps.in.two (really) Oblique Z) Υ kuli imet (kat) be Z:S (lit. Υ skin snaps.in.two (really) NAF Z) 'someone Υ becomes boldly determined to do something Z, regardless of any possible negative consequences doing it might have for Y'

In this first sense, this and the corresponding lele- imet (kat) appear to be, for all intents and purposes, synonymous. 2.

'become accustomed to someone and no longer feel uncomfortable in their presence'

sometimes people think like this about someone: before now, I did not know this person because of this, when I was in the same place as this person, I felt something bad in my skin I did not want to be in the same place where this person is now I know this person because of this, if I am in the same place as this person now, I don't feel anything bad in my skin when these people think like this, they feel something when Y thinks about someone Ζ, Υ feels something like this

3.6. Kopo- 'stomach' The last body part term to be frequently used in experiential expressions is kopo- 'stomach, belly, cavity'. It is, of course, used in a number of expressions indicating physical sensations having to do with the stomach, like hunger, satiation and disease/pain: kopo- iggwor 'stomach make.noises - stomach is growling due to hunger', kopo- bok 'stomach

Emotions and the nature of persons in Mbula

105

be.full', kopo- isaana 'stomach deteriorate — feeling really stuffed with food', kopo- keke 'stomach be. stiff-constipated', kopo- imeete 'stomach dies - feel sick in one's stomach and lacking any appetite', kopo- suruunu 'stomach its.liquid - have diarrhea', kopo- iyoyou 'stomach hurts'. The preeminent emotional use of kopo-, however, is in encoding anxiety and sudden fright. (93)

Υ kopo- rru pa Ζ (lit. Υ stomach be.seeking Oblique Z) Ύ is anxious about Z' sometimes people think like this: after now something bad could happen to/because of something I don't want this when these people think like this, they feel something bad in their stomach when Υ thinks about something Ζ, Υ feels something like this

(94)

Kar kopoono isolooggo (lit. village stomach [?] ) 'because a whole village is anxious about something which is about to occur, people do not go anywhere and mill about restlessly in the village' sometimes people who live in the same place think something like this: after now, something bad could happen to us we do not want this when these people think like this, they feel something bad in their stomach these people think: if we all are in the same place, this bad thing cannot happen to us because of this, these people do not want to move to any other place

(95)

kopo- kutkut (lit. stomach be.beating) 'be anxious and have butterflies in one's stomach'

106

(96)

Robert D. Bugenhagen

kopo- ikam keu (lit. stomach does wrinkling) 'get a sudden fright'

3.7. Body parts less frequently used in body image expressions Due to limitations of space, just a representative example or two will be given of some of the remaining body parts used to express emotions. Like the preceding ones, all of these have obvious physical uses. (97)

Υ kwo- iyoyou pa Z (lit. Υ mouth hurts Oblique Z) 'tired of talking about something' sometimes people think like this: I want people to do something before now I said things about this thing many times because I thought people would do this thing because I did this people did not do this thing when these people think like this, they feel something bad now these people do not want to say anything more about this thing when Υ thinks about something Ζ, Υ feels something like this

(98)

7 kwo- itaanda pa Ζ (lit. Υ mouth be.wide.open Oblique Z) 'be astonished by something unusual with the result that one opens one's mouth widely'

(99)

Υ taiga- iyoyou pa Z (lit. Υ ear hurts Oblique Z) 'tired of listening to something' sometimes people think like this: before now, other people said things about something many times now if I hear other people say these things, I feel something bad because of this, I don't want to hear people say these things

Emotions and the nature of persons in Mbula

107

when these people think like this, they feel something bad when Υ thinks about something Ζ, Υ feels something like this In addition to these emotional uses, taiga- occurs in various expressions indicating the physical state of one's ears, such as taiga- imun 'ear be.closed - deaf, taiga- iyoyou 'ear hurts - ear ache', taiga- ikam 'ear do/ get - happen to overhear something', taiga- ikam kik 'ear does ringing - ringing in ears that hinders one's hearing'. (100)

Υ taiga- iluumu (lit. Υ ear be.cool) 'not have to hear something anymore' sometimes people think like this: before now, other people said things about something many times I don't want to hear these things any more when these people think like this, they feel something bad when they hear other people say these things now these people do not hear other people say these things because of this, now these people do not feel something bad Υ feels something like this

(The source of the annoyance could be an unpleasant, irritating noise, or it could be things people were saying which were unpleasant.) (101)

Υ taiga- ituntun pa Ζ (lit. Υ ear itches Oblique Z) 'always wanting to hear about something' if people say things about something Ζ, Υ wants to hear these things

(102)

Ζ ikam Υ ggure(lit. Ζ does/gets Υ throat) = Ζ izuuru Υ ggure(lit. Ζ pushes Υ throat) sometimes if people eat/drink something, these people feel something good in their throat because of this, these people want to eat/drink this thing Υ feels something like this when Υ thinks about Ζ

108

Robert D. Bugenhagen

Qgure- refers to a cylindrically shaped part connecting other parts. In addition to having the referent 'throat', it occurs in nama- rjgure- ' hand throat - wrist' and kumbu- rjgure- 'leg throat - calf of the leg'. It is also used to express purely physical sensations. Cf. rjgure- imamaaza 'throat is.dry - thirsty', rjgure- imun 'throat be.closed - what happens when a baby cries so hard that they run out of air in their lungs', rjgure- bulbulpa 'throat nauseated Oblique - be disgusted by the thought of eating something.' (103)

Υ rjgure- isaana pa Ζ (lit. Z neck deteriorates Oblique Y) 'longing to eat/drink something' sometimes if people eat/drink something, these people feel something very good in their throat these people did not eat/drink this thing for a long time because of this, these people very much want to eat/drink this thing now

(104)

Υ zuru- imbukmbuk pa Ζ (lit. Υ lip be.bulging Oblique Z) 'pouting and sticking out one's lip as an indication that one is upset with someone/about something'

(105)

Υ ikarrut zuru(lit. Υ is.biting lip) 'biting one's lips - a sign that someone is angry, has lost control of themselves, and is about to attack someone else'

4. Common semantic components of emotion expressions This now completes the survey of the most common body image expressions in Mbula. The explications for some of them, like kete- malmal 'anger', mata- berber 'covet', -ur mata- pa 'hope', lele- isaana 'feel bad', lele- imbukmbuk 'anxious', lele- ikam uraata 'be emotionally moved', ni-llele- sa pa 'be upset with someone' seem to map easily into corresponding English emotional terms. In other cases like ni- ise, as would be expected, it is difficult to find precise English equivalents, because they indicate a combination of semantic components unlike that of any English

Emotions and the nature of persons in Mbula

109

emotional expression. Perusing the different explications, we find the following sorts of components frequently recurring: 1. Situational experiential stimulus: something (good/bad) happens (to someone) (e.g. ni- gesges pa, ni- ise pa, kuli- imoto pa} 2. Experiential stimulus due to an intentional act: someone does something good/bad (to someone) (e.g. lele- isaanapa, kuli- ka miarjpa, kuli- puliizi pa) 3. Experiential stimulus due to the perception of something: someone sees/hears/feels something (e.g. mata- berber, mata- imapalpaala, mata- iyoyou, taiga- iyoyou) 4. Experiential stimulus due to cognitive activities: someone thinks something (e.g. mata- ise ki, iur mata- pa, lele- ipata, kopo- rru pa, kar kopoono soloorfgo) 5. Resultant feelings: because of this, someone feels something (good/bad) (in a part of his/ her body) [where 'part' is replaced by various body part terms] (e.g. lele- ambai, lele- isaana, kete- malmal, kopo- kutkuf) 6. External signals of the emotional state: these can be something in the appearance of the person (e.g. mata- putput, mata- pambaara), or their behaviour: Υ does things because Υ does these things, people know that Υ feels something (e.g. lele-lni- sapa, lele- imbaipa, ni- sanaanapa, ni-puliizipa, ni- ise pa, kwo- itaanda, zuru- imbukmbuk, ikarrut zuru-) 7. Intended responses/desire: because of this, someone wants/does not want to do something (good/ bad) (to someone) (e.g. mata- ise ki, mata- putput, lele-pa, kete- malmal pa, kete- ibayou pa, kuli- ka miai), Qgure- isaana pa) 8. Actual responses: because of this, someone does/does not do something (good/bad) (e.g. mata- irjgal, mata- sigsir), kete- guruk, kete- pitpit)

110

Robert D. Bugenhagen

9. Time of response: someone does something/wants to do something now/after now (e.g. mata- putput versus mata- pambaard) 10. Inceptive aspect: a) before now someone did not want to do something; now this person wants to do something b) before now someone did not feel something; now this person feels something (e.g. lele- iur pa/be, lele- ipuru, lele- imagga, kete- ise, ipas kete-, kuli-imet (kat) pal be) 11. Completive aspect: a) before now someone wanted to do something; now this person does not want to do this b) before now someone felt something; now this person does not feel this (e.g. lele- iluumu pa, lele- imiili. taiga- iluumu) 12. Differences in intensity: someone feels something (very) good/bad; someone does something (very) good/bad (e.g. mata- putput, kete- ise, kete- malmal, kete- ibayou) 13. Evaluation: I think this is good/bad (e.g. mata- pot, mata- sigsig, kete- guruk, kete- pitpit) By combining these components, one is able to specify the precise nature of the emotional state indicated by some expression.

5. The role of the body in Mbula emotion expressions The data presented in this paper amply illustrate the importance of the body in Mbula speakers' conception of emotion. In many cases it is impossible for them to talk about a particular emotion without making mention of some part of their body. Sometimes these parts are highly generic like lele'insides' and ni- 'being'. Other times, the parts are much more specific like

Emotions and the nature of persons in Mbula

111

mata- 'eye', kete- 'liver/chest', and kopo- 'stomach'. With regard to the debate over whether emotion is physiological or cultural and conceptual, the Mbula lexicographical evidence would suggest the answer: both. One interesting question is: What do these kinds of expressions imply about the Mbula concept of the person? People with a western background have variously viewed human beings as consisting of: 1) mind plus body, 2) body only (which includes the brain), 3) soul plus body, 4) spirit plus soul plus body, or 5) mind plus will plus emotions plus body (inter alia) - but how do Mbula speakers view the human person? The body image expressions presented in this paper do not support the neat divisions of people into various discrete components with clearly defined functions as the western traditions seem to do. For example, if matacan refer to one's physical eyes which see or experience pain, or experiences arising because of having seen something, or the part of one which thinks and even feels, what is the basis for a clear division between mind and body? It would be difficult to identify any one of the body parts described in this paper with the English terms mind, will, spirit, brain, or body. (There is a term in Mbula, rjgora, which corresponds to the physical brain organ. But it plays virtually no role in expressing thinking, feeling, or desire.) Along different lines, all of the body parts presented in this paper can be used to encode purely physical sensations, as well as feelings/desires with associated conceptualisations; i.e. emotions. And a wide range of body parts are used in body image expressions to indicate some sort of desire or the lack thereof. With regard to the division of a human being into a material and non-material part, at death the person is buried and their 'skin' decays, but their kunu- 'shadow, image, picture, ghost' survives and can appear to people, bringing them assistance or causing them mental or physical harm. (Kunu- is the form having an inalienable genitive; a related form kon has an alienable genitive and is restricted to the meaning 'ghost'.) Those Mbula speakers who are Christians believe that when they die it is their kunu- that goes to be with God in heaven. It is a striking fact, however, that unlike 'spirit' in Hebrew, Greek, and English, kunu- plays virtually no role in expressing emotion. These body image expressions are, furthermore, used to talk about the emotions of God, angels, and dead people's spirits, despite their lack of a physical body. The explications, as formulated here, allow for this, since they present a prototype of how people feel, and then attribute a feeling "something like this" to the individual experiencing the emotion. This

112

Robert D. Bugenhagen

allows body image terms to apply to noncorporeal beings by analogy. Thus when God, or an angel, or a ghost's liver is said to be fighting, he feels something like what people feel when their livers are fighting.

Notes 1. The Mbula language is an Austronesian language spoken by some 3500 people living on the eastern half of Umboi Island and on Sakar Island in the Morobe Province of Papua New Guinea. The full classification of the language according to Ross (1988: 20, 122) is: Austronesian, Malayo-Polynesian, Central-Eastern MalayoPolynesian, Eastern Malayo-Polynesian, Oceanic, Western-Oceanic, North New Guinea Cluster, Ngero-Vitiaz Family, Vitiaz linkage. The language exhibits Subject-Object-Verb word order and nominative-accusative morphological marking and is prepositional. Ordering of constituents within the noun phrase is basically: 1) inalienable genitive/topicalised alienable genitive, 2) Head Noun, 3) Attributive Modifiers, 4) Quantifiers, 5) Alienable Genitives, 6) Relative Clause 7) Demonstrative. Pronouns and verbal affixation distinguish: 1) first, second, and third person, 2) singular versus plural, and 3) (for first person non-singular) whether or not the hearer is included. The pronouns also exhibit dual forms, which the verbal morphology doesn't. There is no grammatical category of tense. Use is made of temporal adverbials to fix the temporal deictic centre whenever it changes, but most sentences have no indication whatsoever of the time of occurrence of the situation encoded. Most verbs are inflected with a set of subject prefixes. This is indicated in the paper by a preceding hyphen. A subset of the nouns is obligatorily inflected with a set of genitive suffixes. These are cited in this paper with a following hyphen. Nontransparent abbreviations used in this paper are: Accusative, CSB cosubordinating conjunction, DEMonstrative, EXC speaker excluded, GENitive, INC speaker included, LOCative, NMZ nominalisation, NOMinative, OBLique, PASSive, PERFect, RECX Recipient-Experiencer, REDuplication. I would like to very gratefully acknowledge the help of David Aibike, Lukas Aibul, and many other Mbula-speaking friends, who have spent many hours opening up for me a small window into their emotional world. 2. In the ensuing semantic explications, use is largely made of a restricted lexicon containing the following items: I, YOU, SOMEONE, PEOPLE/PERSON, SOMETHING/

THING, WORD, THINK, SAY, KNOW, FEEL, WANT, SEE, HEAR, THIS, THE SAME, OTHER, ONE, TWO, MANY/MUCH, SOME, ALL, DO, HAPPEN, MOVE, BE (THERE IS/ARE), LIVE, DIE, GOOD, BAD, BIG, SMALL, NOT, IF, IF... WOULD, MAYBE, CAN, BECAUSE, VERY, MORE, WHEN/TIME, WHERE/ PLACE, AFTER, BEFORE, A LONG TIME, A SHORT TIME, FOR SOME TIME, NOW, HERE, BELOW, ABOVE, FAR, NEAR, SIDE, INSIDE, PART OF, KIND OF, LIKE. However, occasionally some other items are used in order to enhance the readability of the explications.

Emotions and the nature of persons in Mbula

113

3. The phrasing of this component is not strictly NSM, but it is sufficient for the present purposes. 4. Again, the phrasing of this component is not strictly NSM, but it is sufficient for the present purposes.

References Bugenhagen, Robert D. 1990 Experiential constructions in Mangap-Mbula. Australian Journal of Linguistics 10(2): 183-215. 1994 The exponents of semantic primitives in Mangap-Mbula. In: Cliff Goddard and Anna Wierzbicka (eds.), Semantic and Lexical Universals, 87-108. Amsterdam: John Benjamins. 1995 A Grammar of Mangap-Mbula: An Austronesian Language of Papua New Guinea. (Pacific Linguistics C-101.) Canberra: Australian National University Press. In press. The syntax of universal semantic primes in Mbula. In: Cliff Goddard and Anna Wierzbicka (eds.), Meaning and Universal Grammar. Amsterdam: John Benjamins. Darwin, Charles 1872 The Expression of the Emotions in Man and Animals. London: Murray. Ekman, Paul, E. R. Sorenson and W. V. Friesen 1969 Pancultural elements in facial displays of emotion. Science 164: 86-88. Geertz, Clifford 1973 The Interpretation of Cultures: Selected Essays. New York: Basic Books. Goddard, Cliff and Anna Wierzbicka (eds.) 1994 Semantic and Lexical Universals. Amsterdam: John Benjamins. Harre, Rom (ed.) 1986 The Social Construction of Emotions. Oxford: Blackwell. James, William 1884 What is an emotion? Mind 9: 188-205. Leavitt, John 1996 Meaning and feeling in the anthropology of emotion. American Ethnologist 23(3): 514-539. Leibniz, Gottfried Wilhelm 1903 Opuscules et fragments inedits de Leibniz. Paris: Louis Couturat. [Reprinted 1961, Hildesheim: Georg Olms.]

114

Robert D. Bugenhagen

Lutz, Catherine 1982 The domain of emotion words on Ifaluk. American Ethnologist 9(1): 113-128. 1987 Goals, events, and understanding in Ifaluk emotion theory. In: Dorothy Holland and Naomi Quinn (eds.), Cultural Models in Language and Thought, 290-312. Cambridge: Cambridge University Press. McElhanon, Keith E. 1977 Body image idioms in Irianese and Papua New Guinean languages. Irian 6: 3-27. Rosaldo, Michelle 1980 Knowledge and Passion: Ilongot Notions of Self and Social Life. Cambridge: Cambridge University Press. Ross, Malcolm D. 1988 Proto Oceanic and the Austronesian Languages of Western Melanesia. (Pacific Linguistics C-98.) Canberra: Australian National University Press. Wierzbicka, Anna 1980 Lingua Mentalis: The Semantics of Natural Language. New York: Academic Press. 1990 The semantics of emotions: Fear and its relatives in English. Australian Journal of Linguistics 10: 359-375. 1991 Cross-Cultural Pragmatics. The Semantics of Human Interaction. Berlin: Mouton de Gruyter. Zimbardo, Philip and Floyd L. Ruch 1976 Psychology and Life. Glenview, Illinois: Scott, Foresman and Company.

Why Germans don't feel "anger

55

Uwe Durst

1. Introduction1 About two thousand years ago, Seneca (1970: 106-109) wrote: But you have only to behold the aspect of those possessed by anger to know that they are insane. For as the marks of a madman are unmistakable - a bold and threatening mien, a gloomy brow, a fierce expression, a hurried step, restless hands, an altered colour, a quick and more violent breathing - so likewise are the marks of the angry man; his eyes blaze and sparkle, his whole face is crimson with the blood that surges from the lowest depths of the heart, his lips quiver, his teeth are clenched, his hair bristles and stands on end, his breathing is forced and harsh, his joints crack from writhing, he groans and bellows, bursts out into speech with scarcely intelligible words, strikes his hands together continually, and stamps the ground with his feet; his whole body is excited and "performs great angry threats"; ...2

For centuries, if not millennia, people have attempted to explain the phenomenon of anger as well as other emotions, pointing to typical characteristics of behaviour, causal events, and physical incidents in a more or less scientific way. It could easily be shown that most of the symptoms described by Seneca (and other philosophers, doctors, and psychologists) would be attributed by people around the world to an angry or choleric person rather than a phlegmatic one. And, given a set of five or six "basic" emotion concepts, these symptoms would undoubtedly be attached to the category of "anger" rather than to categories such as "fear", "sadness", "happiness", "disgust", or "desire". However, to carry out such an experiment, this set of emotion words would have to be translated into the respective languages, so this test would really tell us nothing about the psychological reality of universal emotions or emotion concepts. It could only show the trivial fact that each of these languages has at least one word that can be used, among other possible applications, to name this kind of behaviour. But, provided that we could find such a word in all languages around the world, how could we be sure of what it really means?

116

Uwe Durst

In English translations, Seneca's musings De Ira are entitled On Anger, while German translations provide Vom Zorn or Über den Zorn? It is tempting, then, to assume that there is a semantic equivalence between z'ra, anger, and Zorn, and, following Seneca and his translators, that each of these words designates not just a single emotion but also a class of emotions. Although Seneca notes that Latin does not have the multiplicity of distinctive words for different kinds of anger that there is in Greek, he obviously uses the word ira as a hyperonym for various types of angry behaviour (1970: 116-117): ... and yet we call men bitter and harsh, and, just as often, choleric, rabid, clamorous, captious, and fierce - all of which designate different aspects of anger.4

Consequently, the German translation provides Zorn in this passage, and this can hardly be said to be wrong. But while Zorn has been well established in philosophical and theological writings throughout the centuries (cf. also der Zorn Gottes, 'the wrath of God'), recent books on psychological topics, on academic as well as popular levels, have included the word Ärger (roughly: 'annoyance', 'anger'), rather than Zorn, in their titles, whereas Harriet G. Lerner's popular book The Dance of Anger is translated as Wohin mit meiner Wut? (lit.: 'where.to with my anger/rage?'). So, what are they talking about? Is it the same emotion, feeling, psychological state, the same "referent" in the non-linguistic reality, so that the words Zorn, Wut, and Ärger could be seen as interchangeable synonyms? Or do we have to consider one of them as a hyperonym, labelling a category of what in English would be called anger! In other words: is there a hierarchy or taxonomy of emotion concepts, as suggested by the results of cluster analyses in psychological literature? Why is it that John Osborne's play Look Back in Anger could be translated into German literally as Blick zurück im Zorn, whereas the title of Theodore Isaac Rubin's The Angry Book presumably could not, so that one German translation takes the noun Ärger, another one the verb sich ärgern ('to be annoyed/angry'), instead of the adjective corresponding to Zorn, namely zornig? (These two translations of The Angry Book are: Mach deinem Ärger Luft, München: MvgVerlag 1989, and Ärgere dich gesund, Frankfurt: Ullstein 1971.) Thus, anger/angry seems to cover a wider range of use than each of the German words in question. Compare the following examples from two German novels with their English translations:5

Why Germans don't feel "anger"

117

(1)

Als Sabeth es endlich entdeckte, streckte sie die Zunge heraus; ich filmte sie mit der gestreckten Zunge, bis sie, zornig ohne Spaß, mich regelrecht anschnauzte. [MK1/LFH.OOOOO (Frisch, Homo Faber. 104)] When Sabeth finally noticed, she put her tongue out; I filmed her with her tongue out, until, really angry, she bawled me out properly. [Frisch 1974: 90]

(2)

Irgendeiner hatte ihn sogar zum Minister machen wollen, aber er war wütend geworden und hatte gesagt: "Ich bin Lehrer, und ich möchte wieder Lehrer sein". [MK1/LBC.OOOOO (Böll, Ansichten eines Clowns: 48)] Someone had even wanted to make him a Minister, but he had got very angry and said: "I am a teacher, and I want to be a teacher again." [Boll 1988: 31]

(3)

Ich hatte einmal eine ziemlich lange Nummer "der General" einstudiert, lange daran gearbeitet, und als ich sie aufführte, wurde es das, was man in unseren Kreisen einen Erfolg nennt: d.h. die richtigen Leute lachten, und die richtigen ärgerten sich. [MK1/ LBC.OOOOO (Böll, Ansichten eines Clowns: 264)] I had once spent a lot of time rehearsing a fairly long number called "The General," and it turned out to be what is known in our circles as a success: that is, the right people laughed, and the right people were angry. [Boll 1988: 215]

(4)

Ich machte die Augen auf und sah ihn an. Er war böse. [MK1/LBC.OOOOO (Böll, Ansichten eines Clowns: 197)] I opened my eyes and looked at him. He was angry. [Boll 1988: 159]

In (l)-(4) angry was used by the translators to convey four different German expressions: the adjectives zornig (derived from Zorn), wütend (related to Wut), and böse (lit.: 'bad'), and the reflexive verb form sich ärgern (which corresponds to Ärger). In fact, however, angry is by no means the only word that can be used to translate one of these four German expressions into English. Nor is anger the only noun that fits, for example, for the German word Zorn:

118

(5)

Uwe Durst

Ivy starrte mich an, dann schleuderte sie plötzlich ihren Mantel irgendwohin ins Zimmer, stampfend, außer sich vor Zorn ... [MK1/LFH.OOOOO (Frisch, Homo Faber. 73)] Ivy stared at me, then she suddenly threw her coat across the room and stamped her foot, beside herself with rage. [Frisch 1974: 65]

In (5) Zorn is conveyed by rage, which, in turn, can also be used in a context where the German author uses the noun Wut: (6)

Ich heulte vor Wut, zertrampelte die Schießscheiben und schrie den Jungen draußen, die mich bewachten, immer wieder zu: "Ihr Nazischweine". [MK1/LBC.OOOOO (Böll, Ansichten eines Clowns: 32)] I howled with rage, trampled on the targets, and kept on shouting at the boys outside who were standing guard over me: "You Nazi swine!" [Boll 1988: 19]

In sum, there is no German word that perfectly matches the English word anger, and none of the German words given in the examples above has a clear counterpart in English. Each of the German words has a meaning which is somewhat different from each of the English words, and there is no evidence for the "basicness" of one of these words. To grasp their meanings and to be able to compare them and to define them, we have to submit each word to detailed semantic analyses. In what follows, the lexical items Ärger, Wut, and Zorn, which constitute the most frequent and most common 'anger' words in German, will be subjected to semantic and comparative investigation. The analysis is given within the theoretical framework of the NSM approach to semantics, which has turned out to be a most useful way to gain suitable results for this task (see Wierzbicka 1996, 1999). But before we enter into details, we should have a brief look at some methodological problems in the work on German emotion terms and their semantics.

2. Methodological issues As emotions are, due to their nature, a matter of psychological research, it is not surprising that the first detailed studies on emotion words in German were done not by linguists but by psychologists.6

Why Germans don't feel "anger"

119

In a study on the verbal communication of emotions in German (Schmidt-Atzert 1980), 20 persons were given 60 emotion words written on small cards, which they were requested to arrange into piles according to their similaritiy regarding the dimensional features "angenehm" vs. "unangenehm" ("pleasant" vs. "unpleasant") and "körperliche Erregung" ("bodily excitation"). The results of the cluster analysis (1980: 57) suggest that Wut and Zorn are the most closely related concepts within a cluster of eight items, called "Aggressionslust" ("desire for aggression"), which means that nearly all test persons attached both of them to one pile. This tallies well with the fact that Wut and Zorn are interchangeable in most linguistic contexts and hence seem to be semantically identical, or almost identical. However, from a linguistic point of view there are two methodological problems, which raise some questions about the results presented by Schmidt-Atzert. First, the actions of the test participants are restricted by the given equipment (the previously selected emotion words), and by the sorting criteria (the dimensional features). In a two-dimensional diagram (1980: 60) Wut is located on a very high degree of "Erregung" ("excitation"), close to Zorn, both differing in this regard from the less "excited" states of Ärger and Aggressionslust. In another study presented by Wolfgang Marx (1982: 143), who investigated 20 German emotion words, the highest degree of "Aktivation" ("activation") is attached to Ärger, followed by Aggression (3rd position) and Wut (5th position), while Zorn is not found at all. Nor is it found in Schmidt-Atzert and Ströhm (1983: 135), where the dimensional approach was abandoned and the test participants had to deal with 56 words, which they were requested to classify intuitively into piles according to their similarity. Hence, the word most adjacent to Wut could no longer be Zorn, but now it was Groll ('rancour', 'resentment'), whereas in another study (Schmidt-Atzert 1987: 148, 150, 153) it was either Groll or Ärger, depending on whether Groll was part of the given set of emotion words or not. In the latter study (1987: 140) the participants had to "express given emotions in their own words into phrases", or to "find out which kind of emotion speakers... had expressed in their own words", respectively. This leads us to the second problem. Schmidt-Atzert claims that his investigations provide some information about the similarity of emotion words and their meanings: "The taxonomy proposed here is based on the similarity, or dissimilarity, of word meanings'.' (1983: 140; original em-

120

Uwe Durst

phasis) ("Die hier vorgeschlagene Taxonomie basiert auf der Ähnlichkeit bzw. Unähnlichkeit von Wortbedeutungen'.'} But all that can be yielded by this procedure are certain degrees of similarity between two or more emotion words, and thus a taxonomy that consists in a hierarchical order of hyponymy relations. In particular, we are not told what the words really mean, why they differ in their ranges of use, and which semantic components could be considered as being responsible for these different ranges of use. As Debus (1988: 125) puts it: "An emotion word is not used in view of the question whether a clearly distinguishable kind of experience is present or not, but whether the use of the word fits into the respective situation according to the linguistic conventions." ("Die Verwendung eines Gefühlswortes erfolgt nicht unter dem Gesichtspunkt, ob eine klar abgrenzbare Erlebnisqualität vorhanden ist, sondern ob die Verwendung des Wortes entsprechend den Sprachkonventionen in der jeweiligen Situation paßt.") Thus, to uncover the meanings of the elements of a lexical field we have to identify the linguistic conventions by observing the way people use these elements in actual spoken or written language, where they are built into certain verbal and non-verbal contexts that probably differ a great deal from those of test situations. As for the semantic representation, the idea that emotion concepts should be described in a prototypical framework has become widely accepted in the last decade. According to Russell (1991: 39), "the features that constitute emotion concepts describe the subevents that make up the emotion: causes, beliefs, feelings, physiological changes, desires, overt actions, and vocal and facial expressions." Similarly, with regard to lexicographic purposes and relying on Wittgenstein's notion of the "language game", Plum (1992: 172) argues that "factors that... constitute a stereotypical frame for the use of emotion words" ("Faktoren, die... einen stereotypischen Rahmen für den Gebrauch von Gefühlswörtern konstituieren") should be part of the semantic analysis. These factors are "those behavioural patterns that can be regarded as characteristic manners of expressing a certain emotion" ("jene Verhaltensmuster, die als charakteristisches Ausdrucksbenehmen für eine bestimmte Emotion angesehen werden können") (1992: 176). However, it remains unclear which of those features have to be considered as part of the prototype, and whether there is a clear boundary between prototypical and non-prototypical features. Consider, for example, Russell's "anger" script (1991: 39):

Why Germans don't feel "anger"

1. 2. 3.

4. 5.

121

The person is offended. The offense is intentional and harmful. The person is innocent. An injustice has been done. The person glares and scowls at the offender. The person feels internal tension and agitation, as if heat and pressure were rapidly mounting inside. He feels his heart pounding and his muscles tightening. The person desires retribution. The person loses control and strikes out, harming the offender.

Russell states that "the meaning of each such [emotion] term, the concept it expresses, is a script" (1991: 39; original emphasis). But if this is true, what, then, are the meanings of similar words, such as rage, fury, aggression, and the like? Couldn't the five components posited by Russell be applied to these words, too? Thus, what Russell describes may apply to an artificial supercategory called "anger" alongside four other "basic" emotions ("fear", "sadness", "happiness", "disgust") but not to the meaning of the everyday word anger. This is not to say that the positing of such categories may not be useful for the research on the psychology of emotions. But it should be kept in view that these categories are artificial constructs, and that their labels belong to the technical terminology of a particular discipline - a terminology worthy of linguistic investigation in its own right. While Russell's "anger script" is too extensive from the viewpoint of extensional semantics, it is too specific with regard to the intensional semantics of anger. For example, is an emotion called anger normally caused by an "intentional offense"? Apparently, people often get angry at someone whom they never met, as in the case of political anger [Stearns and Stearns 1986: 25], and a person who is angry at him/herself certainly did not intentionally offend him/herself in a harmful way before. The "desire for retribution" in line 4 seems to belong to words like revenge or vengeance, and what is described in line 5 might be a component of the prototype of the phrase angry outburst rather than of anger. In the NSM approach to semantics, emotion words have been described within a prototypical framework from the very beginning (cf. Wierzbicka 1973). At the present stage, the most general frame for an emotion term includes the following lines (cf. Wierzbicka 1999; for a description of anger and semantically related English emotion words within this framework, see Harkins and Wierzbicka 1997: 323-328.):

122

(7)

Uwe Durst

X feels something because X thinks something sometimes a person thinks

because of this, this person feels something (good/bad) X feels something like this because X thinks something like this Thus, the emotion is specified by what is filled in between the first two lines and the last two or three lines, that is, by the experiencer's (prototypical) thoughts. However, the relation between the thought and the feeling might not be the same in each case and therefore has to be reconsidered in the semantic analysis, so that the frame too might not remain constant for each emotion word.

3. Ärger According to Farrell's (1971: 16) dictionary of German synonyms, the reflexive verb sich ärgern "is the commonest term to indicate in everyday life irritation, annoyance, anger". Zorn and zornig are postulated to "suggest passion and correspond most closely to 'wrath' and 'wrathful'" (15, fn.2), while Wut is not mentioned at all in the paragraph under the lemma "angry, annoyed". Klappenbach and Steinitz (1964 sqq, vol. 1:212) define Ärger as "Unwille, Verdrossenheit" (roughly: 'indignation', 'sulkiness'), as well as Meldau (1978: 20), who adds "zornige Stimmung" ('angry mood'), "Verdruß" ('displeasure') - which, in turn, he defines in terms of "Ärger" and "Unwille" (1978: 20) - and "unangenehmes Erlebnis" ('unpleasant event/experience'). Apart from the fact that all these words are in fact much less common than Ärger in everyday language, we can agree with Farrell insofar as a person reporting his or her own feelings would rather use the verbal form than the noun, particularly when the cause or the causer is also mentioned: (8)

Ich merkte erst später, was mich ärgerte: - Ich war aus dem Bad gekommen, Hanna am Telefon, sie hatte das Hospital angerufen, während ich im Badezimmer war — sie redete mit Elsbeth. Only later did I realise what 'arger.PA.ST me'. When I came out of the bathroom Hanna was on the telephone, she had rung the hospi-

Why Germans don't feel "anger"

123

tal while I was in the bath - she was talking to Elsbeth. [MK1/LFH.OOOOO (Frisch, Homo Faber: 168)] (9)

Sogar Ärzte haben Feierabend, neuerdings sogar die Priester. Darüber ärgere ich mich, sie dürften keinen haben und müßten wenigstens das am Künstler verstehen. Even doctors have time off, and recently even priests. 'About this I ärger.PKES myself, they have no business to and they should be able to understand that about the artist. [MK1/LBC.OOOOO (Boll, Ansichten eines Clowns: 264)]

(10)

Ich ärgere mich über meine Kollegen. I ärger.pRES myself about my colleagues am vexed at my colleagues.'

The preposition über literally means 'about', as in sprechen über ('to talk about'). Therefore, the prepositional complement can be interpreted as a cause of the emotion rather than as its goal, as may be suggested by the English preposition at. Thus, sich ärgern is, first of all, the result of something that has happened before, or is still happening. If in a sentence like (10) the noun Ärger is used, the prepositional complement takes the preposition mit ('with') or in ('in'): (11)

Ich hatte heute Ärger mit den Kollegen. I had today Ärger with the colleagues 'Today I had some trouble with my colleagues.'

(12)

Ich hatte heute Ärger in I had today Ärger in 'Today I had some trouble at work.'

der the

Arbeit. work

Actually, from a purely syntactic point of view, the prepositional phrase is not obligatorily required and therefore could also be treated as an adjunct. However, in a sentence like (13): (13)

Aber sein Ärger ist noch nicht verflogen. But his Ärger is yet not vanished 'But his anger has not vanished yet.' [MK1/ZB 7.03320 (Bildzeitung 26.07.1967: 5)]

it will be clear to the reader why the person in question is still angry, because such a sentence is not conceivable without any preceding context,

124

Uwe Durst

from which the reader will get some evidence for the cause of the person's anger. If the cause is still unknown, the absence of a prepositional phrase would presumably evoke the hearer's reaction in the form of a question for the cause. This also applies to sentences referring to possible future events. In (14) the hearer probably knows to whom the affair in question should not become known, and, if so, why it would have some consequences for the speaker. (14)

Wenn das bekannt wird, bekomme ich If this known becomes get I 'If this becomes known I'll get some trouble.' [MK1/ZB 2.00001 (Bildzeitung 02.02.1967: 6)]

Ärger. Ärger.

Clearly, the verb phrase Ärger bekommen does not mean 'to get angry' but 'to get into trouble'. Thus, it seems as if the word Ärger would be polysemous between Ärger, = 'anger', 'annoyance', andArger2 = 'trouble'. However, the relation between both aspects is obvious. If someone has some trouble (with something or someone else), this person will presumably feel something bad because of this. Therefore, both formulae can start with the line 'someone (X) feels something bad'. Ärger haben mit ('to have Ärger with') can combine with noun phrases referring to persons (individuals as well as a group of persons), concrete objects like the ignition device of one's car, or even abstract entities like norms and rules. Likewise, a person can sich ärgern über ('REFL ärger/i.iNF about') another person (or a group of persons), a person's action, concrete objects, abstract entities, or events (e.g. the weather). It follows that in the component referring to the cause of the feeling (15 c) there is nothing that indicates an action only or an action towards the experiences Nor can the causer be thought of as a person only or a thing only. If someone sich ärgert über ('REFL ärgern.SSG.PRES about') the weather, then all we can say is that 'something is happening', or, as in (10) and (13), that 'something happened'. For the semantic description I would prefer the present tense version because utterances such as (10) or (13) presumably imply the presence of the causing event in the experiencer's thought. In the case of Ärger l sich ärgern there is evidence to suggest that thinking is involved before the feeling occurs. The reflexive form may indicate that the emotion is generated by the experiencer's own thoughts. In the explication this is realised as part of the speaker's assertion (the prototypical frame), rather than part of the experiencer's (prototypical) thought, and has

Why Germans don't feel "anger"

125

to be put straight below the prototypical thought ((e) and (f)). This can also explain the fact that sich ärgern usually takes the preposition über ('about'), which can also be used in combination with the noun Ärger.1 Of course, we have no insight into what a person who feels Ärger really thinks. Therefore it is necessary to present the person's thoughts within a prototypical framework. Thus, the semantic description of X ärgert sich (über Y) can be put as follows: (15)

ÄrgerIX ärgert sich (über Y) (a) X feels something bad (b) sometimes a person thinks (c) something bad is happening (Y) (d) I don't want this to be happening (e) this person thinks about this for some time (f) when this person thinks about this, this person feels something bad because of this (g) X feels something like this (h) because X thinks something like this

In the fifth component (e) 'for some time' is added to indicate that it is not only a flash of a thought. Component (f) indicates that the duration of the feeling of Ärger depends on the duration of the thought. The lines of X's possible thought ((c) and (d)) indicate X's awareness, evaluation, and rejection of Y. But actually, it is not the person, the event, or the entity referred to by Y itself, but it is X's awareness of these things or events, their presence in X's thoughts, which cause the bad feeling, as indicated in (e)-(f). The formula in (15) can also be taken as a basis for the expression Ärger haben (mit) (where Ärger means something like 'trouble'). If someone hat Ärger mit Y ('has Ärger with Y'), then he will inevitably sich über Y ärgern ('REFL about Y ärgern.INF'). However, there seems to be something going on between X and Y because of Y. If you 'have' Ärger mit your ignition device or Ärger mit some rules, then something happens to you because of the ignition, which apparently does not work, or because there are rules due to which something else does not work as you like. Therefore, we have to add a component in the first place (a) which constitutes a causal link between the causing event (Y) and X's feelings. Again, the feeling is based on X's thoughts (f)-(g). Thus, X hat Ärger (mit Y) can be represented as:

126

(16)

Uwe Durst

X hat Ärger (mit Y) (a) something happens to X because of Y (b) because of this X feels something bad (c) sometimes a person thinks (d) something bad is happening (Y) (e) I don't want this to be happening (f) this person thinks about this for some time (g) when this person thinks about this, this person feels something bad because of this (h) X feels something like this (i) because X thinks something like this

From the combinability of Ärger with prepositional phrases, and from the semantics of the subsequent preposition itself, we can draw the conclusion that there is no "goal" component in the meaning of Ärger, that is, we have no evidence for a change in the experiencer's behaviour towards the causer or towards anything else. In particular, there is nothing like "the desire or tendency to aggress against the agent who is blamed for what has happened", which is one of the prototypical components of anger as a "basic emotion" postulated by Ortony and Turner (1990: 322), although Ärger is often used in this "basic" sense in psychological, and even linguistic, literature. (For a general survey of words for basic emotions in psychological literature see Schwarz and Ziegler (1996: 49). In Schreiber, Sommerfeldt, and Starke (1993: 97) Ärger is one of five hyperonyms in the field of emotion words.) To give just a few examples of titles of recent psychological publications: Psychologie des Ärgers ('Psychology of Ärger', U. Mees, J. Chassein eds 1992 Göttingen: Hogrefe) Ärger und Ärgerausdruck ('Ärger and Ärger expression', V. Hodapp 1993. Bern: Huber) Ärger: Psychologie einer alltäglichen Emotion ('Ärger: Psychology of an everyday emotion', H. Weber 1994. Weinheim: Juventa) In this respect, Ärger is treated as a basic or superordinate emotion category, like anger in English book titles, and thus it seems to stand for what in former times presumably would have been called Zorn.

Why Germans don t feel "anger"

127

4. Wut and Zorn versus Ärger Wut (fern.) and Zorn (masc.) differ from Ärger in that they cannot combine with the verb haben ('to have') without an article (Ärger haben, 'to have Ärger'; but *Wut/Zorn haben, 'to have Wut/Zorn'). But when Wut or Zorn are combined with the indefinite article and the verb haben, as e.g. in eine Wut haben ('to have a Wut') or einen Zorn haben ('to have a Zorn'), these phrases are not counterparts of Ärger haben ('to have Ärger') in the sense that the latter normally combines with the prepositions mit or in and is understood as what in English could be called 'trouble'. To describe a person's emotion the nouns Wut and Zorn can be used as well as their corresponding adjectives wütend and zornig. Thus, sentences such as (17) and (18) can both have the same translation in English: (17)

Er hat eine He has a 'He is angry.'

(18)

Er ist wütend/ He is wütend/ 'He is angry.'

Wut l Wut l

einen Zorn. a Zorn

zornig. zornig

What makes it difficult to draw a distinction between the meanings of Wut and Zorn (or wütend and zornig) is the fact that they are interchangeable in a large variety of contexts constituting grammatical sentences. Both nouns can be combined with possessive pronouns or genitive attributes referring to the experiencer, and with a prepositional phrase designating the "object", or the "cause'V'causer" of the emotion. The verb haben can be interpreted as defining a relation that in other terms would be called "alienable possession", that is, the emotions are conceptualised as a sort of "things that people can have", and that are located somewhere between inside and outside the body. Cf. some very common expressions like: (19)

den Zorn l die Wut l den Ärger herauslassen 'to let the Zorn l Wut l Ärger out' den Zorn l die Wut l den Ärger in sich hineinfressen 'to swallow the Zorn l Wut l Ärger'

128

Uwe Durst

While these metaphors are also well applicable to Ärger, the following ones seem to be restricted to Wut and Zorn: (20)

Zornausbruch Wutausbruch Ärgerausbruch

('Zorn -outburst') (' Wut -outburst') ( Ärger -outburst')

(21)

Mich packt ohnmächtiger Zorn. me.ACC seizes helpless.MASC.NOM Zorn.NOM am seized by helpless anger.' [H85/OZ 2.15499 (Die Zeit 23.08.1985: 42)]

(22)

Mich packt die kalte Wut,... me.ACC seizes the.FEM.NOM cold.FEM.NOMWtf.NOM am seized by cold anger,...' [H85/OZ 1.16102 (Die Zeit 15.02.1985: 25)]

In terms of deep case roles, the experiencer in (19) is also the "agent", while the emotion is the "patient" of the action. In (20) it is the emotion that is conceived of as an "agent" of a quite sudden and uncontrollable action, which is indicated by -ausbruch ('outburst'). Similarly, the emotion words in (21) and (22) are the subjects of the sentences, that is, the role of the "agent" would have to be attached to the emotion words, and the role of the "patient" to the sentence-initial personal pronouns. Likewise, Wut and Zorn can jemanden überkommen ('to seize/to overpower s.o.'), and people can get erfaßt ('gripped/seized') or übermannt ('overpowered') by these feelings. As one "falls" or "flies" into a rage in English, one can in Wut/Zorn geraten ('to get into Wut l Zorn') in German,8 which means that there is something that drives a person "into" the state of emotion, not mediated by a thought. Other words within the frame in X geraten are: Furcht ('fear'), Verlegenheit ('embarrassment'), Verzweiflung ('despair'), but also non-emotion words like Zweifel ('doubt'), Panik ('panic'), Streit ('conflict'), or Schwierigkeiten ('difficulties'). What we can conclude from these observations is that in the case of both, Wut and Zorn , the rise of the feeling seems to be out of the experiencer's control. This would explain why the use of the negative imperative of sich ärgern sounds much more natural than the corresponding forms of eine Wut/einen Zorn haben or wütend/zornig sein:9

Why Germans don't feel "anger"

(23)

Mensch ärgere dich nicht. man ärger.iMP yourself not 'Don't get angry' (name of a popular game for four players round a table)

(24)

Ärgere dich heute nicht. ärger.iMP yourself today not 'Don't get angry today.' [MK1/ZB6.01036 (Bildzeitung 10.06.1967: 4)] (horoscope)

(25)

Laß dich nicht (von X) ärgern. let.iMP you.ACC not (by X) ärgern. INF 'Don't let someone or something (X) ärger [trans.] you.'

(26)

?? Sei nicht be.iMP not

(27)

?? Hab keine have.iMP no

129

wütend/zornig. wütend/zornig Wut/keinen Wut/no

Zorn. Zorn

However, (26) sounds more felicitous if the temporal adverb gleich and the modal adverb (or intensifying particle) so, in addition to the modal particle doch, are inserted. But while (24) and (25) would be understood as a good bit of advice, (28) rather indicates that the speaker does not appreciate the hearer's behaviour, because what is rejected in (28) is not the emotion indicated by the adjective, but its suddenness (conveyed by gleich), or its specific manifestation (or intensity) (conveyed by so). In other words, the phrase IMP + doch nicht is not directly related to the emotion words but to their modifiers: (28)

Sei be.iMP

doch MP

nicht not

gleich just

so so

wütend/zornig. wütend/zornig

On the other hand, doch can also be combined with a negative imperative of sich ärgern , but hardly with the bare adjectives wütend and zornig: (29)

Ärgere ärger.iMP

dich yourself

(30)

?? Sei doch be.IMP MP

doch MP nicht not

nicht. not. wütend/zornig. wütend/zornig

130

Uwe Durst

There must be something in sich ärgern that allows utterances such as (29), and/or something in Wut and Zorn (or wütend and zornig) that is responsible for their incompatibility with the construction IMP + doch nicht. As demonstrated by Ickler (1994: 403), the use of doch in imperative sentences presumes that "the hearer's behaviour is clearly shaped and built into a behavioural context that allows the opposition of an alternative behaviour" ("das Verhalten des Hörers eine deutliche Gestalt hat und in einen Verhaltenskontext eingebaut ist, der die Entgegensetzung eines alternativen Verhaltens erlaubt"), where the hearer must be able to recognise what the speaker's call is directed against. That is, the IMP + doch nicht phrase includes a structure like 'you do something', want you to think about it', and want you not to do it'. Hence, the construction IMP + doch nicht is applicable in (29) because there is a semantic component in sich ärgern that complies with this condition. In the formula given in (15), this can only be the process of thinking: an emotion which is generated primarily by a thought can be changed or abandoned via another thought. As mentioned above, the IMP + doch nicht phrase in (28) is not directly related to the emotion words but to their modifiers. Thus, the experiencer in (28) is requested to calm down and to gain control over his or her emotions. Wut and Zorn can be unterdrückt ('suppressed') or gezügelt ('bridled') but not completely abandoned. Therefore, the IMP + doch nicht construction in (28) does not refer to the emotion itself but to the experiencer's ability to deal with the emotional state and thus to control his or her own behaviour. Its semantic structure is incompatible with unmodified Wut or Zorn , presumably because they do not convey the notion of a thought as an antecedent of the feeling, as it is expressed by the components (15e-f) and (15g-h) in the semantic formula of Ärger /sich ärgern. Of course, it is possible that a feeling of Wut or Zorn can be evoked via thinking of something, but this is not necessarily so and therefore does not appear in the semantic explication. Wut and Zorn, as well as their corresponding adjectives, can have a prepositional adjunct with the preposition aw/('on', 'at') or über ('about'). The difference in the use of the prepositions auf and über in this context consists in the difference between a directional orientation and a deliberative process, both of them parts of the meanings of the prepositions, not of the emotion words themselves. People can 'have' eine Wut not only at persons but also eine Wut auf alles ('a Wtf at everything') [H87/FS 6.46036 (Stern, 16.12.1987: 16)]. This may be an extreme situation, but the possibility of

Why Germans don't feel "anger"

131

saying this disproves Schreiber, Sommerfeldt, and Stärke's (1993: 106) claim that Wut auf ('Wut at') only combines with concrete nouns, while Wut über ('Wut about') combines with abstract ones only. So far, we have seen some formal differences between Ärger on the one side and Wut and Zorn on the other side, and I have tried to work out the semantic implications of their linguistic behaviour. In what follows, I will try to distinguish Wut and Zorn, where other contexts must be considered to find useful minimal pairs.

5. Wut versus Zorn As mentioned above, Wut and Zorn are interchangeable in a large variety of contexts. In the following examples (31)-(36) there are no discernible reasons why the author would prefer one of these words over the other. (31)

Er zahlte den ersten Scheck einfach auf das Caritaskonto seiner Pfarre ein, ließ sich an der Sparkasse Zweck und Ort des Verrechnungsschecks erklären, ging dann zu seinem Pfarrer und bat um einen Barscheck über fünfzehn Mark - aber der Pfarrer platzte fast vor Wut. Er erklärte Heinrich, er könne ihm keinen Barscheck geben, weil er die Zweckbestimmung erklären müsse, und so ein Caritaskonto sei eine heikle Sache... The first check he simply paid into the welfare account of his parish, had the bank explain to him the purpose and nature of a crossed check, then he went to his priest and asked him for a cash check for fifteen marks - but the priest nearly exploded with Wut. He told Heinrich he couldn't give him a cash check because he would have to state what it was for, and a welfare account was a ticklish thing... [MK1/LBC.OOOOO (Boll, Ansichten eines Clowns: 193)]

(32)

Ivy starrte mich an, dann schleuderte sie plötzlich ihren Mantel irgendwohin ins Zimmer, stampfend, außer sich vor Zorn... Ivy hatte sich eingerichtet, eine Woche in Manhattan zu verbringen, jetzt gestand sie's, und mein plötzlicher Entschluß, nicht zufliegen wie üblich, sondern morgen schon mit dem Schiff zu reisen, um in einer Woche auch in Paris zu sein, war ein Strich durch ihre Rechnung.

132

Uwe Durst

Ivy stared at me, then she suddenly threw her coat across the room and stamped her foot, beside herself with Zorn... Ivy had arranged to spend a week in Manhattan, she now revealed, and my sudden decision not to fly, but to leave tomorrow by boat so as to reach Paris in a week's time as planned, upset her calculations. [MK1/LFH.OOOOO (Frisch, Homo Faber. 73)] (33)

Leo wurde immer wütend, wenn ich ihn traf und ihn fragte: "Weißt du noch, wie wir das Holz miteinander durchgesägt haben?" Er schreit dann: "Aber wir HABEN das Höh nicht miteinander durchgesägt" Leo always got wütend when I saw him and asked: "Do you remember how we sawed through that wood together?" He exclaims: "But we DIDN'T saw through any wood together." [MK1/LBC.OOOOO (Boll, Ansichten eines Clowns: 219)] (original emphasis in capitals)

(34)

Sabine wurde regelrecht hysterisch und nannte mich "Mörder", Karl schrie mich an: "Du Vagabund - Du Hurenbock", und das machte mich so wild, daß ich ihn "verkrampfter Pauker" nannte, meinen Mantel nahm und in Zorn davonlief. Sabina got positively hysterical and called me a "murderer", Karl shouted at me: "You tramp - you lecher," and that infuriated me so much that I called him a "frustrated pedagogue, " picked up my coat and left in Zorn. [MK1/LBC.OOOOO (Boll, Ansichten eines Clowns: 206)]

(35)

Ich stand nur da, blickte auf die Koblenzer Straße und dachte an Marie. Irgend etwas in meinem Gesicht schien den Mann, neben dem ich stand, wütend zu machen. Er senkte sogar die Zeitung, verzichtete aufsein "Strauß: mit voller Konsequenz!", schob seine Brille vorne auf die Nase, sah mich kopfschüttelnd an und murmelte "Unglaublich" I just stood there, looking out onto Koblenz Street, and thought about Marie. Something in my face seemed to make the man next to me wütend. He even lowered his paper, stopped reading his "Strauss: With unshakeable determination", pushed his glasses down onto his nose, looked at me, shook his head, and murmured "Incredible." [MK1/LBC.OOOOO (Boll, Ansichten eines Clowns: 68)]

Why Germans don t feel "anger"

(36)

133

Der Großvater war dürr und eigensinnig, er geriet leicht in Zorn und trank, je mehr die Mitbürger ihn mieden. The grandfather was withered and stubborn, he'd easily fly into Zorn and he'd drink the more his fellow citizens avoided him. [MK1/LJA.OOOOO (Johnson, Das dritte Buch über Achim: 165)]

Yet there are sentences where it seems to be impossible to replace one of them by the other one: (37)

Mit dieser Entscheidung aber zog er sich den Zorn der guten Bürger zu,... But with this decision he "caught" the Zorn of the good citizens,... [H86/KM 4.12384 (Mannheimer Morgen 21.08.1986: 32)]

(38)

Den Eintritt der Sachsen in die Geschichte bezeichnet das Jahr 286, in dem sie gemeinsam mit den salischen Franken von See her in Nordfrankreich einfielen und die reichen Hafenstädte " mit furchtbarer Wut" plünderten. The Saxons' entrance into history is marked by the year 286, when they invaded from the sea into the north of France together with the Salian Franks and plundered the wealthy seaports "with an awful Wut". [MK1/WPE.OOOOO (Poertner, Die Erben Roms: 339)]

(39)

Er konnte immer noch zuschlagen wie er es gelernt hatte: nicht wütend aber hart und bis der Gegner sah daß er auf den Boden kommen würde unter diesen ausgerechneten zweckdienlichen Stößen. He was still good at dealing a blow the way he had learned it: not wütend, but forcefully until the other guy realised that he'd end up on the floor under these calculated efficient blows. [MK1/LJA.OOOOO (Johnson, Das dritte Buch über Achim: 242)]

The phrase sich jemandes Zorn zuziehen ('to "catch" someone's Zorn'} in (37) is highly idiomatic and does not allow Wut to replace Zorn. In (38) it would be possible to replace Wut by Zorn, but it makes a difference whether people plunder with Wut or with Zorn. In the first case, more uncontrolled and spontaneous action seems to be involved, whereas the second case seems to imply that the plundering is the result of a certain attitude against those seaports and their inhabitants, an attitude by which the plunderers' behaviour is guided and determined. In (39) there is an opposi-

134

Uwe Durst

tion between wütend blows and calculated efficient blows. What is conveyed by wütend is the notion of someone beating his opponent in an uncontrolled manner and breaking into pieces everything that is in his way. This may be the reason why the German psychologist Wilhelm Wundt (1911: 212) put Zorn in the category of "allmählich ansteigende Affecte" ("gradually rising affections"), and Wut in the category of "plötzlich hereinbrechende Affecte" ("suddenly overtaking affections"). But actually, this does not apply to the generation of the emotions as such because a person can be suddenly befallen by either of them (see examples (20)-(22) above); it rather applies to the way the corresponding behaviour is generated. This is why a regent can Zorn und Milde walten lassen ('rule with Zorn and indulgence') [MK1/WPE.OOOOO (Poertner, Die Erben Roms: 259)], but not Wut und Milde, and why it is quite possible to talk about edlen Zorn ('noble Zorn') or heiligen Zorn ('holy Zorn'), while Wut better combines with the adjective rasend ('raging') and can be used to describe not only human behaviour but also the intensity of natural phenomena such as storms or waves. In this respect, Wut seems to be closer to fury, or Italian rabbia (cf. Wierzbicka 1995 a), than to anger. It is interesting in this respect that in contemporary written texts the frequency of wütend is higher than the frequency of zornig, and both are much more common than two other adjectives derived from the same lexemes, wütig and zürnend: Table 1. Frequency of wütend, wütig, zornig, and zürnend (adjectives)

MK

LIM

HBK

Total

165(100%)

10(100%)

260(100%)

wütend

113(68.5%)

8(80.0%)

161(61.9%)

wütig

2(1.2%)

0(0.0%)

4(1.5%)

zürnend

7(4.2%)

1 (10.0%)

2(0.8%)

43(26.1%)

1 (10.0%)

93(35.8%)

zornig

If word classes have their own (language-specific) meanings (see, e.g., Langacker 1987; Wierzbicka 1995b), this distribution can be explained by

Why Germans don't feel "anger"

135

the semantics of the words. The adjectives wütend and zürnend are formally identical with the present participles of the verbs wüten ('to rage', 'to be in a fury') and zürnen (roughly: 'to be angry'). Wütig and zornig are genuine adjective forms with the highly productive denominal derivational suffix -ig, comparable to -y in English. Not surprisingly, earlier German dictionaries offered both of them as seperate lemmata, whereas wütend and zürnend were still considered as verb forms (cf. Adelung 1808, vol. 4: 1733, 1767; Campe 1969-1970[1811], vol.5: 802, 803, 880, 914). According to Grimm and Grimm (1854 sqq, vol. 14/2: 2514), the contemporary use of the present participle wütend in its attributive, predicative, and adverbial function was developed in the second half of the 18th century. This tallies well with the fact that a notion similar to 'anger' was usually expressed in Old and Middle High German by zorn (>Zorn), whereas wuot/wuoti (>Wut) originally was used to designate the symptoms of persons who were seen as demoniacs, and to describe a "state of being bodily and mentally beside oneself, deprived of one's controlling volition" ("zustand des der lenkenden willkür entzogenen leiblich-seelischen auszersichseins"; 2476). This quasi-medicinal aspect is still existent in the words Tollwut ('rabies') and Wutbeere (an older name for Tollkirsche, 'atropa belladonna'), and can be considered as being responsible for the high productivity of -wut as the second element in other compound nouns, while the number of compounds on the head -zorn seems to be comparatively small (118 word forms on -wut with 37 different lexemes, including 46 word forms on -tollwut, versus 8 word forms on -zorn with 4 different lexemes, including 3 word forms of Jähzorn ('irascibility', 'passionateness') in the HBK corpus). To give some examples: (40)

Himmelszorn Bibelzorn Volkszorn

'heaven-zor«'; 'the wrath of the heavens' 'Bible-zo/ ' 'people-zor«'

Arbeitswut

'work-wwf'; 'work mania'

Bauwut Liebeswut Putzwut Sammelwut Zerstörungswut etc.

'build-wMi' 'love-ww?' 'clean-wwi' ' collect- wuf ' destruction- wut'

136

Uwe Durst

While the compounds on X-zorn can be paraphrased by a genitivus subjectivus /possessivus ('X's Zorn'), or as a metaphor containing the genitive ('something like X's Zorn'), the first element of the compounds on X-wut is usually a verb stem or a nomen actionis, as, for example, Zerstörung, derived from zerstören, 'to destroy'; Liebe is, in this context, not a psychological state but an action (the verb lieben can mean 'to love' as well as 'to make love'). Hence, the compounds on X-wut are better paraphrased as 'Wut to do X', where the behaviour of a person or a group of persons is metaphorically, and hyperbolically, marked as somewhat abnormal. Therefore, they often convey a slightly negative or derogatory connotation, which, however, is not part of their semantics, because the behaviour described by the compound can be a desirable quality of a person: (41)

Zu seinen vorrangigen Eigenschaften müssen Risikobereitschaft, Arbeitswut und Verantwortungsfreude gehören. As for his principal qualities, he is expected to show readiness to incur risks, Arbeitswut, and a willingness to accept responsibility. [H85/CZ 2.30607 (Die Zeit 05.07.1985: 33)]

In other words, the stem Wut-/wut- is related more closely to spontaneous and immediate action than the stem Zorn-/zürn-, which rather focusses on a sort of state including the agent's (= experiencer's) attitude towards something, so that an action which is done out of Zorn seems to be more deliberate rather than uncontrolled. This is why the wrath of God (or God's anger) is in Martin Luther's Bible translation der Zorn Gottes, and not die Wut Gottes, with which he would have conveyed a completely different image of God to his contemporaries. (Another word used by Luther is Grimm 'grimness', 'wrath', which, however, is scarcely used today: 4 items in MK, 5 in HBK, and none in LIM). Likewise, God is zornig or zürnt (zurnen.3sG.pRES), but he is not wütend or wütig, although the Latin translation provides ira Dei as well as furor Dei, the Greek translation orge toü theoü and thymos toü theoü, and the Hebrew text has at least six different nouns.10 It is interesting in this context that two compounds on X-wut according to the word formation pattern 'something like X's Wuf could be found: Furienwut [GOE/AGD.OOOOO (Hamburger Ausgabe, vol. 10: 125)], and the much more common Berserkerwut [Brockhaus-Wahrig 1980 sqq, vol. 1: 620; Duden 1989: 239]. The Furien ('Furies') are the avenging goddesses of the Roman underworld (the Greek Erinyes). Their Latin name Furiae is identical with the noun furia ('fury', 'rage'). The Berserker

Why Germans don't feel "anger"

137

('Berserks') are Odin's warriors in Old Norse mythological literature, while Odin himself is the god of war and ecstasy; his Old High German name Wuotan is obviously derived from wuot/wuoti (> Wut).]' Against this background it becomes clear why the present participle/adjective wütend can be attributed not only to human beings or animals, but also to natural forces, e.g. storms or fires, to Küsse ('kisses'), Leidenschaft ('passion'), or even Pinselstriche ('strokes of the brush') in a description of a painting. It does not seem to be impossible to use zornig in these contexts, but it would convey a meaning including a component like want to do something bad to someone', whereas wütend in this case conveys the notion that the action is rather unconscious, driven by an invisible force that cannot be resisted: have to do something now'. This is verified by the data in table 1: the most common adjective for Wut is the present participle of its corresponding verb, whose verb class meaning can tentatively be described as "momentary action" (that is, 'someone is doing something now', or 'something is happening now').12 Intrinsic adjectives, like zornig or wütig , on the other hand, have the function to attribute a certain property or quality to what is referred to by the head noun. Therefore, ein wütender Mensch ('a wütend.MAiSC.SG.NOM human') is someone who is in a "rage" right in this moment, but ein zorniger Mensch ('a zorwg.MASC.SG.NOM human') can also be someone who is characterised by his "angry" temper, disposition, or attitude.13 This is why John Osborne's angry young men can only be zornig, but not wütend, in German. Likewise, a Christian missionary can be called der zornige, alte Mann ('the z0rmg.MASC.SG.NOM old man') [MK1/WPE.OOOOO (Poertner, Die Erben Roms: 149)] or attributed as der ewig zürnende Bonifatius ('the ever zürnercd.MASC.SG.NOM Bonifatius') [MK1/WPE.OOOOO (Poertner, Die Erben Roms: 247)]. If they were described as wütend, it would be a situational feature of these men rather than a description of their characters, and they would be conceived of as acting in a more spontaneous, and perhaps more aggressive, more "furious" way; and if they were wütig, it would be characteristic for them to act as if they were mad.14 Considering these differences, we can propose the following analyses for the nouns Wut and Zorn:

(42)

Wut (a) X feels something bad (b) sometimes a person thinks:

138

Uwe Durst

(c) something bad is happening (d) I don't want this to be happening (e) because of this I have do something now (f) because of this, this person feels something bad (g) X feels something like this (h) X cannot not feel like this (43)

Zorn (a) X feels something bad (b) sometimes a person thinks: (c) something bad happened because someone did something (d) I don't want things like this to happen (e) because of this I want to do something bad to someone (f) because of this, this person feels something bad (g) X feels something like this (h) X cannot not feel like this

The first two lines within the prototypical thought differ in that the "triggering event" seems to be more present for Wut. This is not to be understood in the sense that it would be happening right now in any case. It is quite possible to say ich bin immer noch wütend ( am still wütend'), but this rather indicates that the cause is still present in the speaker's view. On the other hand, Zorn can be used in the sense of a disposition or attitude, which is reflected in component (e) which, in turn, presupposes a present or past event that is presumably caused by a person's action ('someone did something' in (c)), that is, someone can be blamed for what happened. Of course, you can get wütend 'because someone did something', but this is not necessarily so. For example, if your computer fails to work unexpectedly in the middle of an important operation, your prospective reaction would be described as wütend rather than zornig. There is no reference to the duration of the feeling in the formula. It is clear that both emotions are not just flashes of split-second feelings, but the limitation at the other end of the scale is not clear-cut. A person who is characterised as zornig may have this disposition all his life; and while the author of the following dialogue could write wütend in all three contexts, the English translation turns from furious to angry where the duration of the emotion would be too much extended:

Why Germans don t feel "anger"

(44)

139

"Ich hob immer gemeint, du bist wütend auf mich", sage ich, "wegen damals'.' Hanna nur verwundert. "Wieso wütend? Weil wir nicht geheiratet haben?" sagt sie. "Das wäre ein Unglück gewesen -" Sie lachte mich geradezu aus. "Im Ernst", sagt sie, "das hast du wirklich gemeint, daß ich wütendbin, Walter, einundzwanzig Jahre lang?" [MK1/LFH.OOOOO (Frisch, Homo Faber. 165)] "I always thought you were furious with me," I said, "because of what happened." Hanna was merely surprised. "Why should I be furious? Because we didn't marry?" she said, "that would have been a disaster." She positively laughed at me. "Seriously," she said, "did you think I was angry with you, Walter, for twenty-one years?" [Frisch 1974: 142]

Line (e) in Wut refers to the somewhat uncontrolled ( have to do something') and spontaneous ('now') character of the subsequent action, while in Zorn it indicates a deliberate decision ( want to do something'), and the consciousness that the intended action will do harm to someone, even if the agent/experiencer is not conceived of as doing harm to people in principle (cf. God's Zorn, and expressions like gerechter Zorn ('justified Zorn"), heiliger Zorn ('holy Zorn') etc.). This person, however, remains unspecified in the semantic formula. In this respect Zorn differs from words like Haß ('hatred') or Rache ('revenge'), which seem to include the notion of an action that is directed towards specific persons ( want to do something bad to this person'). The differences in (42e) vs. (43 e) also account for the fact that people can talk about the Wut of the waves, or of wütende kisses, whereas the notion of a Zorn of the waves would be much more figurative and imply a demonistic conception of the world, while zornige kisses either does not make sense or implies violence in the sense that want to do something bad to someone' (43e). The last components (42h) and (43 h) refer to the fact that Wut and Zorn can take the role of an agent in (semi-) idiomatic phrases, and that their corresponding adjectives cannot be combined with the negative imperative.

140

Uwe Durst

6. Semantics, metaphors, and etymology The prototype approach to emotions is related to the metaphor approach in that prototypes and metaphors are both defined via the notion of comparison. In a cross-linguistic study on anger metaphors Kövecses (1995: 186) finds that the metaphorical understandings of anger "are in part based on shared ideas about the human body and certain physiological processes that are associated with anger." He concludes that the terminology of anger is not only characterised by various container metaphors, but also takes the form of "conceptual metonymies" (191) describing physiological responses like body heat, internal pressure, and redness in the face and the neck area. "It would be odd", he argues, "to conceptualise anger as, say, softly falling snow, an image completely incompatible with what our body is like and what our physiology does in anger." (192) Not surprisingly, these three types of physiological responses can also be found in combination with our three German 'anger' words, and, in particular, with Wut. But Wut can also be grünlich ('greenish') [MK1/LGB.OOOOO (Grass, Die Blechtrommel: 223)] and kalt ('cold') [H85/OZ 1.16102 (Die Zeit 15.02.1985: 25)], and one can be rot ('red') as well as bleich vor Wut ('pale with Wut' [Klappenbach and Steinitz 1961 sqq, vol. 6: 4409]). Of course, these data could be explained by assuming a further metaphorical process, but in any case they illustrate that we cannot rely upon metaphors and metonymies alone in order to find out the meaning of the respective words. According to Geeraerts and Grondelaers (1995: 155) "the possible role of cultural traditions as a source of emotion concepts" should also be taken into consideration. The tradition they find is the ancient theory of the four humours, that is, the four basic fluids, where the choleric temperament is determined by a preponderance of yellow bile, whose characteristics are warmth and dryness, whose taste is bitter, whose element is the fire, and which is located in the liver or the stomach (158). The influence of this cultural tradition in the European history of civilisation can hardly be denied, but again, it only elucidates the formation and the use of a couple of favoured metaphors. Following Geeraerts' and Grondelaers' claim of "the methodological importance of a diachronic perspective for linguistic studies with a cultural orientation" (1995: 155), I will conclude this paper with a brief look at the histories of the three German words.

Why Germans don't feel "anger"

141

As mentioned before, the attributive, predicative and adverbial use of the present participle wütend expanded first in the 18th century. Similarly, there are few examples of the noun Wut in Old and Middle High German literature; the lexeme was most commonly realised as an intransitive verb and thus conceptualised as describing an action. Its original meaning, where it describes the mad behaviour of a demoniac, is also proved for the etymologically linked Gothic present participle wods, in a translation of the Greek word daimonistheis ('he who had been demon-possessed') in Mark 5:18 by saei was wods (cf. also Grimm and Grimm 1854 sqq, vol. 14/2: 2475). Subsequently, it could be used to describe a certain kind of behaviour, where spontaneous and uncontrolled action was involved, so that finally it was suitable to refer to the assumed feeling of a person who behaved, or could behave, in this way. In the course of this "psychological shift" this characteristic remained and is still alive in the prototypical component "I have to do something now". As for Zorn, it is controversial whether it descends from the Indo-European root *der , and thus is related to Modern English to tear, or whether it belongs to a group of words whose common core meaning could be paraphrased as ('push', 'shove', 'blow') (see Grimm and Grimm 1854 sqq, vol. 16: 90-91). In any case, it originally describes an action that is either destructive or harmful, and that entails a certain effect on a person or an object. In Middle High German it is still used in the sense of 'fight', 'encounter', 'affray' (Grimm and Grimm 1854 sqq, vol. 16: 90-91), besides its early use as a pure emotion word, where it was applied to translate, in particular, the Latin ira. Although the process of a "psychological shift" is harder to trace than in the case of Wut, it is still evident. The historical constant here is want to do something bad to someone'. Finally, the noun Ärger came into use not before the 18th century and is mentioned in Grimm and Grimm (1854 sqq, vol. 1: 548) only in a very brief note as quite a "strange" ("befremdliche") word formation. The transitive verb ärgern is derived from ärger, which is the comparative form of the adjective arg ('bad/evil/harmful'), and thus originally meant 'to make worse', 'to aggravate'. The reflexive verb, then, could be interpreted as 'to make oneself worse', that is, to aggravate one's own situation by doing something. As a result of the "psychological shift", the effect was attributed to the experiencer's bad feeling, while the action was reduced to the process of thinking about a certain ("annoying") event. Thus, the his-

142

Uwe Durst

torical constant is the notion of a process of "getting worse", which can be mediated by the experiencer's own behaviour.

7. Conclusion There are three conclusions that can be drawn from the observations reported above. First, none of the three most frequent 'anger' words in German matches the meaning of the English word anger. Second, each of them has its own semantic structure, determining their ranges of use and their restrictions in certain contexts. And finally, the meanings of these words as analysed here are supported by the evidence from etymology if one assumes a "psychological shift" in their histories, which is responsible for their semantic change and explains the relation between constancy and variation in the diachronic process. What has been preserved throughout the centuries are those components that indicate how a person to whom the respective emotion is attributed is likely to behave, or how the person would be affected by his/her own behaviour. It would be interesting, in this context, to compare the proposed meanings with those of the etymologically related Dutch words zieh ergeren, woede, and toorn, and to extend the analysis to further German 'anger' words in order to verify the correctness of the semantic formulae proposed in this paper.

Notes 1. This paper was written while I was a Visiting Fellow at the Australian National University in Canberra. I thank the members of the Linguistics Department for their hospitality. I also thank Jean Harkins and Anna Wierzbicka for helpful comments on an earlier version of this paper, presented at the University of New England in Armidale in March 1998. 2. Ut scias autem non esse sanos quos ira possedit, ipsum illorum habitum intuere; nam ut furentium certa indicinia sunt audax et minax vultus, tristis frons, torva fades, citatus gradus, inquietae manus, color versus, crebra et vehementius acta suspiria, ita irascentium eadem signa sunt: flagrant ac micant oculi, multus ore toto rubor exaestuante ab imis praecordiis sanguine, labra quatiuntur, denies comprimuntur, horrent ac surriguntur capilli, spiritus coactus ac stridens, articulorum se ipsos torquentium sonus, gemitus mugitusque et parum explanatis vocibus sermo praeruptus et conplosae saepius manus etpulsata humus pedibus et totum concitum corpus "magnasque irae minas agens,..." (De Ira I.I: 3—4)

Why Germans don t feel "anger"

143

3. The prepositions von (Of, 'from') and über ('about', Over', 'above') seem to be mutually interchangeable within this syntactic frame. Their use in this context can be compared to that of the English words on and about, although, on a closer look, some differences will certainly be found. 4. ... etiam si amarum nos acerbumque dicimus, nee minus stomachosum, rabiosum, clamosum, difficilem, asperum, quae omnia irarum differentiae sunt; (De Ira I.IV: 2) 5. The analysis here is based on an online research of selected text corpora held by the Institut für Deutsche Sprache (IDS) in Mannheim/Germany, by means of the Corpus Storage, Maintenance, and Access System (COSMAS), developed at IDS; along with some random observations and intuitions by the author as a German native speaker. The corpora are abbreviated as follows: GOE: Goethe-Korpus, 1 709663 words from J. W. von Goethe's novels and theoretical/scientific writings (assembled at the University of Kyushu/Japan); GRI: Grimm-Korpus, 518827 words from Grimm and Grimm's collection of sagas and fairy tales (assembled at the Ryukoku University, Kyoto/Japan); HBK (H85-H88): Handbuch-Korpus, 13911305 words from newspapers, in particular from Mannheimer Morgen, Die Zeit, and Stern, 1985-1988 (assembled by IDS); LIM: Limas-Korpus, 1348813 words, passages from a large variety of text genres, 1970-1971 (assembled by the Forschungsgruppe LIMAS, Bonn und Regensburg/ Germany); MK (MK1-MK2): Mannheimer Korpus, 3 291 136 words from a large variety of text genres, including novels, scientific and popular books, manuals, hand-bills, instructions for product use, etc., 1949-1973 (assembled by IDS). Interlinear glosses are given where the argumentation is based on a particular structure. Abbreviations used in glosses are: ACC = accusative, FEM = feminine, IMP = imperative, INF = infinitive, MASC = masculine, MP = modal particle, NOM = nominative, PAST = past tense (Präteritum), PRES = present tense, REFL = reflexive pronoun, SG = singular. If not indicated otherwise, emphases in examples are added by the present author. 6. This also applies to studies in other languages. For survey and discussion see Schmidt-Atzert (1980); Debus (1988), who argues for an orientation along Biihlers trichotomic model of communication; and Schwarz and Ziegler (1996), who further try to establish a link between neuropsychology/neurolinguistics and lexicology. 7. Über seems to be used where a deliberative process is involved in the action referred to by the verb: sprechen über ('to talk about'), nachdenken über ('to think about'). It must be mentioned, however, that there are reflexive emotion words in German which cannot combine with this preposition, e.g. sich fürchten ('to be afraid'). Thus, the semantics of the preposition must not be confused with the semantics of the reflexive verb. The semantic interrelation between (reflexive) verbs and their prepositional complements requires further investigation. 8. Actually, two examples for the phrase in Ärger geraten ('to get into Ärger') were found, one of them in a text from the early 19th century [GOE/AGD.OOOOO (Dichtung und Wahrheit, Hamburger Ausgabe, vol. 9: 73)], another one in a contemporary

144

9.

10. 11.

12.

13.

14.

Uwe Durst

novel [MK1/LJA.OOOOO (Johnson, Das dritte Buch über Achim: 211)]. However, this use seems to be out of date. Two imperatives of the verb zürnen (corresponding to Zorn) were found in the corpus, one of them in Grimm and Grimm's collection of legends [GRI/SAG.00542 (Lohengrin zu Brabant; Sagen: 542)], the other one in Goethe's Werther [GOE/ AGW.OOOOO (Die Leiden des jungen Werther, Hamburger Ausgabe, vol. 6: 91)]. To use these words in contemporary German would sound quite obsolete, if not ridiculous. As Anna Wierzbicka pointed out to me, it is also quite common to use another 'anger' word, böse (see example (4) above), with the negative imperative: Sei mir nicht böse (lit.: 'be.IMP me.DAT not bad'; hence: 'don't be angry with me'). The semantics of the phrase jemandem böse sein (lit.: 'someone.DAT bad be.INF'; 'to be angry with someone') were not investigated, but I would expect a similar result as for Ärger, as far as the notion of an intermediate thought, and thus the possibility of controlling the rise of the feeling, is concerned. 'op, za'am, hemäh, hdron, qce$cep, and 'cebrah. For a comparative survey of their meanings, see Bergman and Johnson (1973), and Myhill (1997). Considering the religious usage of Zorn, Philipp Lersch (1956: 235) stated in his phenomenological analysis that "When we are zornig, it is always in the name of what is to happen according to generally binding norms... But we get into Wut only in the name of our individual self-being." ("Wir sind zornig immer im Namen dessen, was nach allgemein gültigen Normen geschehen soll... In Wut aber geraten wir immer nur im Namen der Belange unseres individuellen Selbstseins.") However, this seems to be too schematic and does not account for examples such as (34) (for Zorn), where the experiencer apparently feels that he was personally offended, and for (35) (for Wut), where the man's reaction described by the speaker seems to rely on a certain idea of how a decent citizen should look. This is not a complete analysis of the German present participle (for example, the aspect of duration is not considered here), but sufficient to draw a distinction between the word class meanings of present participles and adjectives (see also note 14). To avoid misunderstandings: I am not saying that Zorn is a "disposition" in the sense in which Wittgenstein (1967: 87; number 491) called Liebe ('love') and Haß ('hate') dispositions. In German, neither love nor hate can "seize" someone, nor can they suddenly "break out", whereas Zorn, as well as Wut, can (see examples (20)-(22) above). The semantic difference between adjectives on -ig and adjectives on -end is still transparent in some cases. For example, ein tollwütiger Hund ('a tollwutig.MA.SC. SG.NOM dog') is a dog that suffers from rabies, which is thought of as an incurable disease rather than a momentary state. Again, a person can have zittrige Hände ('shaky hands', adj.), which implies, to a certain extent, general weakness, or old age etc., whereas zitternde Hände ('trembling hands', pres. part.) describes a temporary state; consequently, zitternd can combine in such contexts with causative prepositional phrases indicating a temporary deviation from the normal, e.g. zitternd vor Angst/vor Kälte ('trembling with fear'/'shivering with cold').

Why Germans don if eel "anger"

145

References Adelung, Johann Christoph 1808 Grammatisch-kritisches Wörterbuch der Hochdeutschen Mundart, mit beständiger Vergleichung der übrigen Mundarten, besonders aber der Oberdeutschen. Mit D. W. Soltau's Beyträgen, revidirt und berichtiget von Franz Xaver Schönberger. Wien. Bergman, J. and E. Johnson 1973 "änap. In: G. Johannes Botterweck and Helmer Ringgren (eds.), Theologisches Wörterbuch zum Alten Testament, Band I, 376-389. Stuttgart u.a.: W. Kohlhammer. Biblia Hebraica Stuttgartensia 1967/1977 K. Elliger and W. Rudolph (eds.). Stuttgart: Deutsche Bibelstiftung. Biblia Sacra luxta Latinam Vulgatam Versionem 1926-1995 Romae Typis Polyglottis Vaticanis; Romae Libreria Editrice Vaticana. Böll, Heinrich 1988 The Clown. (Translated by Leila Vennewitz). London/New York: Marion Boyars Brockhaus-Wahrig 1980-1984 Deutsches Wörterbuch. Edited by Gerhard Wahrig, Hildegard Krämer, Harald Zimmermann. Wiesbaden: F. A. Brockhaus; Stuttgart: Deutsche Verlags-Anstalt. Campe, Joachim Heinrich 1969-1970 Wörterbuch der Deutschen Sprache. Mit einer Einführung und Bibliographie von Helmut Henne (Documenta Linguistica; Reihe II) [Reprografischer Nachdruck der Ausgabe Braunschweig 1807-1811]. Hildesheim; New York: Georg Olms Verlag. Debus, Günter 1988 Psychologie der Gefühl s Wörter: Empirisch-experimentelle Untersuchungsansätze und -ergebnisse. In: Jäger, Ludwig (ed.), 95-138. Die gotische Bibel 1960. Edited by Wilhelm Streitberg. Heidelberg: Carl Winter, Universitätsverlag. 3./4. Aufl. Duden 1989 Duden Deutsches Universalwörterbuch. Mannheim, Wien, Zürich: Dudenverlag. 2. Aufl. Farrell, R. B. 1971 Dictionary of German Synonyms. Cambridge: University Press. 2nd edition. Frisch, Max 1974 Homo Faber. A Report. (Translated by Michael Bullock). Harmondsworth: Penguin.

146

Uwe Durst

Geeraerts, Dirk and Stefan Grondelaers 1995 Looking back at anger: Cultural traditions and metaphorical patterns. In: Taylor and MacLaury, (eds.), 153-179. Grimm, Jacob and Wilhelm Grimm 1854-1954 Deutsches Wörterbuch. Leipzig: S. Hirzel Harkins, Jean and Anna Wierzbicka 1997 Language: A key issue in emotion research. Innovation 10/4: 319-331. Ickler, Theodor 1994 Zur Bedeutung der sogenannten 'Modalpartikeln'. Sprachwissenschaft 19: 374-404. Jäger, Ludwig (ed.) 1988 Zur historischen Semantik des deutschen Gefühlswortschatzes: Aspekte, Probleme und Beispiele seiner lexikographischen Erfassung. Aachen: Alano. Jäger, Ludwig and Sabine Plum 1989 Probleme der Beschreibung von Gefühlswörtern im allgemeinen einsprachigen Wörterbuch. In: Franz Josef Hausmann et al., (eds.), Wörterbücher. Dictionaries. Dictionnaires: Ein internationales Handbuch zur Lexikographie, vol. l, 849-855. (Handbücher zur Sprach- und Kommunikationswissenschaft; 5.1) Berlin, New York: de Gruyter. Klappenbach, Ruth and Wolfgang Steinitz, (eds.) 1961-1977 Wörterbuch der deutschen Gegenwartssprache. Berlin: AkademieVerlag. Kövecses, Zoltän 1995 Anger: Its language, conceptualization, and physiology in the light of cross-cultural evidence. In: Taylor and MacLaury (eds.), 181-196. Langacker, Ronald W. 1987 Nouns and verbs. Language 63/1: 53-94. Lersch, Philipp 1956 Aufbau der Person. München: Barth. 7. Aufl. Luther, Martin 1968/72 D. Martin Luthers Werke: Kritische Gesamtausgabe, vol. 6-12: Die Deutsche Bibel 1522-1546. Unveränderter Abdruck der bei Hermann Böhlaus Nachfolger, Weimar (1929-1961), erschienenen Ausgabe. Graz: Akademische Druck- u. Verlagsanstalt. Marx, Wolf gang 1982 Das Wortfeld der Gefühlsbegriffe. Zeitschrift für experimentelle und angewandte Psychologie XXIX/1: 137-146. Meldau, Rudolf 1978 Sinnverwandte Wörter und Wortfelder der deutschen Sprache: Ein Handbuch für den Deutschunterricht. Paderborn: Schöningh.

Why Germans don't feel "anger" Myhill, John 1997

147

What is universal and what is language-specific in emotion words? Evidence from Biblical Hebrew. Pragmatics and Cognition 5/1: 79-130. Ortony, Andrew and Terence J. Turner 1990 What's basic about basic emotions? Psychological Review 97/3: 315-331. Plum, Sabine 1992 Gefühlswörter im Wörterbuch: Überlegungen zur lexikographischen Bedeutungserläuterung des emotionalen Wortschatzes. In: Gregor Meder und Andreas Dörner (eds.), Worte, Wörter, Wörterbücher. Lexikographische Beiträge zum Essener Linguistischen Kolloquium, 169-182. (Lexicographica: Series Maior; 42). Tübingen: Niemeyer. Russell, James A. 1991 In defense of a prototype approach to emotion concepts. Journal of Personality and Social Psychology 60/1: 37-47. Schmidt-Atzert, Lothar 1980 Die verbale Kommunikation von Emotionen: Eine Bedingungsanalyse unter besonderer Berücksichtigung physiologischer Prozesse. Ph. D. Dissertation, Universität Gießen. 1987 Zur umgangssprachlichen Ähnlichkeit von Emotions Wörtern. Psychologische Beiträge 29: 140-163. Schmidt-Atzert, Lothar and Walter Ströhm 1983 Ein Beitrag zur Taxonomie der Emotionswörter. Psychologische Beiträge 25: 126-141. Schreiber, Herbert, Karl-Ernst Sommerfeldt and Günter Starke 1993 Deutsche Substantive: Wortfelder für den Sprachunterricht. Leipzig: Langenscheidt. Schwarz, Manuela and Wolfram Ziegler 1996 Emotionen in Neurolinguistik und Lexikologie: Ein Forschungsüberblick. Lexicology 2/1: 34-62. Seneca, Lucius Annaeus [1970] Moral Essays. Vol. I. With an English translation by John W. Basore (Loeb classical library; 214). London: William Heinemann; Cambridge: Harvard University Press. [1989] Philosophische Schriften. Bd. 1. Übersetzt, eingeleitet und mit Anmerkungen versehen von Manfred Rosenbach. Darmstadt: Wissenschaftliche Buchgesellschaft. 4. Aufl. Septuaginta [1962] Edidit Alfred Rahlfs. Stuttgart: Württembergische Bibelanstalt. Editio Septima. Stearns, Carol Zisowitz and Peter N. Stearns 1986 Anger: The Struggle for Emotional Control in America's History. Chicago: University of Chicago Press.

148

Uwe Durst

Taylor, John R. and Robert E. MacLaury, (eds.) 1995 Language and the Cognitive Construal of the World. (Trends in Linguistics. Studies and Monographs; 82). Berlin/New York: Mouton de Gruyter The Holy Bible 1995 New International Version. London/Sydney/Auckland: Hodder and Stoughton. Wierzbicka, Anna 1973 The semantic structure of words for emotions. In: R. Jakobson, C. H. van Schooneveld and D. S. Worth (eds.) Slavic Poetics: Essays in Honor ofKiril Taranovsky, 499-505. The Hague: Mouton. 1995 a Everyday conceptions of emotion: a semantic perspective. In: James Russell, Jose-Miguel Fernändez-Dols, Antony S. R. Manstead and J. C. Wellenkamp (eds.) Everyday Conceptions of Emotion: an Introduction to the Psychology, Anthropology and Linguistics of Emotion, 17-47. Dordrecht/Boston: Kluwer Academic Publishers. 1995 b Adjectives vs. verbs: The iconicity of part-of-speech membership. In: Marge E. Landsberg (ed.) Syntactic Iconicity and Linguistic Freezes, 223-245. (Studies in Anthropological Linguistics; 9). Berlin, New York: Mouton de Gruyter. 1996 Semantics: Primes and Universals. Oxford: Oxford University Press. 1999 Emotions Across Languages and Cultures: Diversity and Universals. Cambridge: Cambridge University Press. Wittgenstein, Ludwig 1967 Zettel. Oxford: Basil Blackwell. Wundt, Wilhelm 1896 Grundriß der Psychologie. Leipzig: Wilhelm Engelmann.

Linguistic evidence for a Lao perspective on facial expression of emotion1 N. J. Enfield

In the ongoing debate about emotions and their relationship with facial expression, James Russell has recently campaigned for "the gathering of new evidence" (1995: 382) concerning the nature and meaning of facial expressions across cultures and across languages. In this paper, I present data on some ways in which Lao people describe facial expression, as well as some of the ways they attribute inner states to people making particular facial displays. I first discuss briefly a number of words and expressions available in Lao for the description of emotions and other inner states. In the second section, I describe some of the ways in which Lao speakers isolate, recognise and describe particular facial expressions, as well as making some comments on their attribution of inner states (of thought, feeling, and/or emotion) to people making the expressions in question. One point I want to make is that "folk" analyses of semiotic phenomena, revealed in linguistic semantics, are fundamental to a well-informed comparative science of emotion and nonverbal communication.

1. "Emotions" in Lao There are no exact equivalents in Lao of the English words sad, angry, disgusted, happy, surprised, and afraid. In some cases, there are obvious rough translations of these so-called "basic emotions" (Ekman 1992), while in others the equivalents are not so clear. Let me introduce some of the problems.

1.1. 'Sad' There are several possible translations into Lao of the English word sad. The most likely to correspond to the canonical sense of English sad is sia-

150

N.J.Enfield

caj, literally 'lost-heart'. Sia-caj shares with sad the idea of bad feelings as a response to some unfortunate event. Thus, one can be sia-caj 'about' something, or sia-caj 'that' something has happened, apparently restricted to cases in which something happens to the experiencer. For example, a speaker says he was sia-caj when, having hurried home from the market to get money to buy a bargain-priced turkey, returns to find that the creature has already been sold. Another common term often translatable as sad is sook-saw, which is more likely to carry the connotation of a longer term state of unhappiness, due to some sorry state the subject finds themselves in, rather than some specific event. A third term which overlaps to some extent with sad is ngaw, which most typically expresses a bad feeling of loneliness or aloneness, especially when a person is too often by themselves in overly quiet surroundings. The feeling of ngaw is no doubt related to the premium in Lao culture on muan 'fun', and being with muu 'friends, the group'. Lao people are most comfortable spending time in larger groups of people with plenty of activity going on. Feeling ngaw is a response to aloneness in this cultural context.

1.2. 'Angry' There are two expressions in Lao which can be routinely translated, in various contexts, by the English word angry. The first is caj-haaj Offensiveheart' (in the sense of 'heart-on-the-offensive'), which expresses the emotional state of a person who is responding to some incident which directly and negatively affects them. Someone who is caj-haaj is preferably avoided, since they would be likely to say (or perhaps do) something bad to anyone who they may consider responsible for whatever it is they are reacting to. In this sense, caj-haaj is quite like English angry, but the two differ in that caj-haaj cannot usually be used to express an idea of being angry about some situation which does not directly affect the person involved. It is more immediate. (Note that the component haaj may be used as a speech-act verb - to 'tell somebody off' - and Lao speakers normally find it upsetting, as recipient or bystander, when this behaviour occurs.) The second common 'anger' term is khiat, where the negative feeling is typically not overtly expressed, similar in some contexts to Australian English pissed off. In contrast with caj-haaj, it is «on-expression which char-

A Lao perspective on facial expression of emotion

151

acterises a typical state of being khiat. If someone is khiat with/at someone else, they are likely not to speak to that person, nor to do anything with or for that person. In the kinds of close-knit and cooperative village settings typical in Lao society, to ignore someone is to send a strong signal. The idea of khiat as 'non-confrontational anger' relates to a strong value in Lao culture of avoiding overt confrontation (especially the possibility of people "saying things") where possible. A key expression encapsulating this is the ubiquitous bo-pen-nang [not be anything] 'no problem, it's nothing', which epitomises a core ideal that one should tackle problems with a degree of acceptance and tolerance, and without koo-luang [build-issue] 'creating an issue'. If Lao people are 'angry' at each other, they are not as likely to make their bad feelings explicitly known in confrontational ways, but will rather "express" them through, literally, non-expression. This remains an effective way of getting the message across.

1.3. 'Disgusted' There is a fairly clear translation into Lao of English disgusted, namely khidiat. The prefix khi- (related to khii 'shit') is used widely in the language for a range of idiomatic expressions to do mainly with different kinds of effluent material and waste products (khi-muuk [muuk = nose] 'mucus'; khithaw [thaw = 'ash'] 'ashes'; khi-luaj [luaj = 'mill'] 'sawdust'), as well as negative personal characterisations (khi-khuj [khuj = 'chat'] 'snobby'; khilak [lak= 'steal'] 'thief; khi-qaaj [qaaj= 'shy'] Overly shy'). (The morpheme -dial has no independent meaning outside of the expression khi-diat 'disgusted'.) The usage of khi-diat is more restricted than that of the English term disgusted, since the Lao term is only used to refer to responses evoked by things which are literally physically disgusting, such as the idea of eating certain things, or of dealing with physically revolting things such as rotten food, etc. In Lao, for example, one cannot be disgusted 'at' someone, or 'that' something has happened.2

1.4. 'Happy' The English expression happy may be translated into Lao as dii-caj 'goodheart', sabaaj(-caj) 'content(-heart)' or mii khuam-suk 'have contented-

152

N.J.Enfield

ness'. Dii-caj has a close affinity with the English expression glad, being a transient positive emotional response to some specific fortunate incident. Thus, one may be dii-caj 'that' something has happened, but unlike English happy, dii-caj cannot describe a feeling of general contentedness (as in John is happy with his new job or Mary seems happy these days). This latter idea is partly covered by the term sabaaj 'content/comfortable', a cultural key word for Lao speakers (as for speakers of neighbouring Thai). Sabaaj itself refers to a feeling rather than an emotion, since it does not entail a cognitive component - it often means 'comfortable', for example with regard to seating, clothing, or ambience. However, in a compound with -caj 'heart', the meaning does include the cognitive evaluation leading to the feeling. This prior cognitive evaluation is the hallmark of emotion (according to many; but cf. e.g. Zajonc 1980, 1984). Another expression corresponding in certain cases to happy is mii khuam-suk, literally 'to have happiness'. (Khuam 'sense' acts here as a nominalising prefix to the verb-like bound element suk 'content/happy'.) This expression is used in many cases which would also be compatible with sabaaj, but it does not permit reference to exclusively physical comfort which sabaaj 'content' allows. Rather, it is a profound state of wellbeing which normally refers to one's general condition of mind and soul. While one may be dii-caj 'glad' to see an unexpected visitor, it would be odd to mii khuam-suk 'have happiness' in the same way. On the other hand, an aim in life could be to mii khuam-suk 'have happiness', but it would make little sense for one to generally pursue being dii-caj 'glad'.

1.5. 'Surprised' Two Lao translations are possible for English surprised. These are tok-caj 'fall-heart' and tuun 'awaken'. Both involve a feeling of reaction to the unexpected, but the second entails an instinctive physical reaction (a 'start'), and there is no entailment of any thought or evaluation preceding and/or causing the feeling of 'surprise'.3 (Cf. Goddard 1997 on a similar distinction in Malay.) Tok-caj, however, can include the idea of a cognitive evaluation leading to the feeling of 'surprise'. Thus, one can be tok-caj, but not tuun, 'that' something has happened. This element of evaluation leading to feeling would seem to point to tok-caj as an emotion, and tuun as a mere feeling or instinctive physical reaction. Even so, the term tok-caj is not pre-

A Lao perspective on facial expression of emotion

153

cisely equivalent to surprise. For example, tok-caj cannot express the kind of 'surprise' expressed in English I'm surprised you're still here. With respect to tok-caj, some specific event must accompany the evaluation leading to the feeling of 'surprise'.

1.6. 'Afraid' The most obvious translation of afraid or scared into Lao is jaan. It is unclear to me what difference exists between Lao jaan and English afraid or scared. It would seem that jaan is closer to afraid in its being less suggestive of immediate and present danger. Jaan is also used in Lao as a complement-taking predicate expressing what the speaker merely suspects rather than literally fears (cf. English I'm afraid she's not coming), and unlike English it need not have a negative connotation (cf. Lao jaan phoo [afraid enough] '(I) suspect (there'll be) enough [e.g., rice to eat]' vs. English ?I'm afraid there'll be enough).

1.7. A note on caj, the Lao 'heart/mind' The term caj, which elsewhere refers to the physical 'heart' or 'centre', occurs in most Lao expressions related to emotion, and denotes the primary 'seat' of cognitive and emotional activity for the Lao, the so-called 'heart' or 'mind'. This word is cognate with Thai cay (Diller and Juntanamalaga 1990), and shares many of the same idiomatic and grammatical properties. Both the Lao and Thai terms also show extensive semantic and grammatical similarity with Malay hati and comparable terms in other Southeast Asian languages (cf. Matisoff 1986, Oey 1990, Jaisser 1990 and Goddard this volume - Enfield 2001 provides a methodological and theoretical discussion). There are dozens of expressions in Lao involving caj, of roughly two main types. A set of terms describing lasting character traits treat caj as a regular nominal head, with some quasi-adjectival modifier following (e.g. caj-dam [heart-black] 'ruthless', caj-kuang [heart-broad] 'generous', cajkaa [heart-dare] 'daring, courageous'). Most emotion terms, referring to more transient characterisations, place caj second in the combination (e.g. nooj-caj [small heart] Offended', naak-caj [difficult-heart] 'undeciding,

154

N.J.Enfield

tortured', tok-caj [fall-heart] 'surprised'). There are, however, exceptions to these generalisations (cf. caj-haaj 'angry' § 1.2., above). The issue ofcaj compounds and their grammatical and semantic properties deserves an extensive study.

1.8. Comment We began with the question of correlations across languages between emotions and facial expressions. We are yet to consider facial expressions, but it is already clear that particular facial expressions cannot be said to universally express particular emotions. This is because when eliciting and/or describing Lao speakers' attribution of inner states to people making various facial displays, the researcher cannot investigate ideas such as angry, happy, or sad, since Lao people do not use these English words when speaking Lao, and there is no guarantee that they would understand them if they were used. Instead, the terms used are Lao. And as discussed above, the Lao terms do not have equivalents in English. So, even if we accept that the Lao and English categories are indeed "emotions", they are not the same emotions.

2. Conventional description of facial expression in Lao I now describe some ways of speaking about facial expression in Lao, concentrating on lexical items and idioms which can be used to describe particular facial expressions (or features thereof). I will note the kinds of attributions of inner/emotional states which Lao people make, in their own terms, to people with these facial expressions. I am not concerned here with such questions as the factual accuracy of attribution of a particular emotional state to some facial expression, nor am I interested here in correlation, causal or otherwise, of inner states and facial displays. The aim is to present linguistic evidence for the Lao speaker's conventional analysis of the face.

A Lao perspective on facial expression of emotion

155

2.1. Grammatical properties of facial expression descriptions in Lao While English has a range of simple words for facial expressions (e.g. scowl, grin, smile, grimace), Lao notably has few. The only one attested below is mm 'smile'. Many of the rest are more explicitly descriptive polymorphemic expressions, such as 'rotten face', 'smelly face', 'stuck eyes', and 'clenched mouth'. In between these extremes are a number of 'cranberry' expressions (i.e. in which one component does not independently occur), such as naa-beq 'ready-to-cry face' and taa-loo 'exposed-whitesabove-iris eyes', in which the modifying elements (beq and loo, respectively) have no independent meaning. Syntactically, these facial expression descriptions involving head nominals such aspaak 'mouth', taa 'eye(s)', or naa 'face' may pattern either as (a) noun + predicate, in a kind of relative clause or noun + attribute construction (§ 2.4.4., below): taa-khaang eye-stuck 'stuck eyes'; 'eyes which are stuck'; or (b) a kind of incorporating verb + object construction (§ 2.3.4., below): nom-paak clench-mouth 'clench-mouthed'.

2.2. The face The face is referred to in Lao as naa, or baj-naa (where baj elsewhere means 'leaf and serves as a classifier for flat handle-able things such as sheets of paper). The word naa performs a range of other semantic functions, including a role as a locative marker meaning 'front'/'in front of, and a temporal marker meaning 'next' (as mpii naa 'next year', literally 'year face/front'). There is a range of various 'face' expressions, in the noun + predicate pattern, in which naa 'face' is noun phrase head, with a following verbal modifier.

156

N.J.Enfield

2.2.1. naa-buut 'rotten-face' Elsewhere, buut means 'rotten', with reference especially to food which has gone bad, is repulsive, and no longer edible. The mood often attributed to someone with a 'rotten face' is khiat, which (as described above) involves brooding, directed non-involvement with someone who is the target of some ill-feeling. It would seem that the 'rotten-face' provides the khiat person with one way to explicitly show their ill-feeling (i.e. in the presence of the person towards whom they are khiat). There is some latitude with regard to defining what exactly constitutes naa-buut 'rotten face'. An English speaker might call it a scowl or a sour face, but it seems to me best described as an anti-smile. By this I mean that a naa-buut 'rotten face' expression aims to look least like a smile as possible. A number of informants have defined naa-buut in just this way - as 'never smiling' - and accordingly, all speakers I consulted used the term naa-buut (or just buut alone) to describe this schematic illustration (the abstract "opposite" of the smile "

Note also that being naa-buut 'rotten-faced' does not entail an emotional state (such as khiat 'angry'), but may simply describe perceived ugliness or generally poor temperament (i.e. as a description of the kind of person who never smiles).

2.2.2. naa-men 'smelly-face' The face which Lao speakers call 'smelly face' is essentially the classical "disgust face", with wrinkled nose, and raised upper lip. The element men 'smelly', however, refers not to a property of the face itself (in contrast to naa-buut 'rotten face', § 2.2.1., above), but to something smelly (i.e. rotten) which would cause someone to pull such a face. While naa-buut 'rotten face' is typically used to describe the face of someone who is of poor temperament, or who is khiat, or even caj-haaj (§ 1.2., above), the ex-

A Lao perspective on facial expression of emotion

157

pression naa-men 'smelly-face' would not be used to describe someone in such a mood. Rather, naa-men 'smelly-face' describes the face of a physically disgusted person only.

2.2.3. naa-khüm

'furious-face'

The term khum refers to a kind of furious anger or rage, although certainly less strong than these English terms. To have a khum face is to look like someone in such a state, typically manifest in classical signs of 'rage' in the face, such as exposed incisors, square mouth, frowning, and perhaps also exposed-whites-above-iris eyes (cf. §2.4.1., below).

2.2.4. naa-beq/naa-nuq 'ready-to-cry face' The face of someone who is about to cry may be described as naa-beq or naa-nuq, the latter term being less common, perhaps of regional origin. Neither beq nor nuq have related meanings alone or in compound expressions other than these. The typical naa-beq includes a combination of slightly protruding, and perhaps trembling, lower lip, along with knitted and lowered eyebrows. The defining factor of naa-beq is evidently this lower lip pose (cf. sop-beq; § 2.3.2., below).

2.3. The mouth There are a number of terms in Lao for facial expressions involving the mouth and parts of the mouth, mostly including the terms paak 'mouth', sop 'lips', and kheew 'teeth'. It is worth noting, incidentally, that the term paak is only approximately translated as 'mouth' - it can be used to refer to the lips and outer part of the mouth as well, as in the expression paak teek [mouth break] 'broken mouth' (referring to split or chapped lip(s)).

158

N.J.Enfield

2.3.1. mm 'smile' The term mm 'smile' is apparently equivalent in meaning to English smile. The crucial idea is that the corners of the mouth are raised (and naturally this includes associated visible muscle movements around the eyes, cheeks, and jaw). In attributing an emotional state to a smiling person, Lao speakers generally describe them as dii-caj 'glad', and occasionally sabaaj(-caj) 'content' (if the smile in question is not judged to be a transient response to some specific happy stimulus).

2.3.2. sop-beq 'about-to-cry lip' As described above (§ 2.2.4.), someone who is about to cry may have a 'beq face' (naa-beq). This facial expression is localised in the lower lip, since it may also be described by the term sop-beq, i.e. having a 'beq lip'. The lower lip protrudes, typically trembling. In attributing an inner state to someone with this expression, most informants use the terms sia-caj 'sad', sook-saw 'unhappy', and occasionally bo-sabaaj 'uncomfortable, ill'. Most often the subject is said to be 'about to cry' (kamlang-ca-haj [progressive-irrealis-cry] or ca-haj-leew [irrealis-cry-perfective]).

23.3. qaa-paak Open-mouthed' The expression qaa-paak, referring to wide-openness of the mouth, is a verb + object construction somewhat like its English noun-incorporating counterpart open-mouthed. A doctor speaking Lao asks patients to qaapaak (where in English one would say open wide or say 'ah'). The qaapaak Open-mouthed' facial expression is not necessarily associated with emotion, since of course one may hold one's mouth wide open when at the doctor, when putting food in there, or in many other situations. When asked to attribute an emotional state to someone qaa-paak Openmouthed', Lao speakers usually use the term tok-caj 'surprised'. Note that when one is qaa-paak Open-mouthed', one cannot speak, a fact which would seem to be consistent with the idea of being 'surprised'. It would be unusual, if possible at all, to say laaw qaa-paak vaw [s/he speak openmouth] '?S/he spoke open-mouthed'.

A Lao perspective on facial expression of emotion

159

2.3.4. nom-paak 'clench-mouthed' The expression nom-paak 'clench-mouthed' refers to tight compression of the lips, as in an expression of great determination. This may also be referred to as kat-kheew 'biting-the-teeth'. When shown a picture of somebody with tightly compressed lips, informants used both terms nompaak 'clench-mouthed' and kat-kheew 'biting-the-teeth', and invariably attributed great determination to the subject, for example by saying that the person jaak het qan.daj.qan.nung [want do something] 'wants to do something', or tang-caj [set.in.place-heart] 'intends (to do something), is determined'. Indeed, it would seem that the idea of 'clenching' is suggestive of readiness for springing into action. (One piece of related linguistic evidence is the four-syllable expression kat-kheew-nom-kon, literally 'bite-teeth-clench-bottom', which refers to a state of great determination and/or readiness for action.)

2.4. Eyes and eyebrows There are a number of descriptions of facial expression in Lao involving taa 'eye(s)' and khiw 'eyebrow(s)'.

2.4.1. taa-loo 'exposed-whites-above-iris eyes' The expression taa-loo (where loo does not occur independently) refers to wide-open eyes, where the whites of the eyes above the iris are visible. Informants most often describe someone with this expression as jaan 'afraid', especially when other aspects of the facial expression support (or at least do not contradict) this, for example if the subject is also qaa-paak Open-mouthed' (§2.3.3., above). If a subject combines taa-loo 'exposedwhites-above-iris eyes' with some incongruous pose, such as a smiling mouth, Lao informants attribute to the person a state of insanity (baa 'insane'), or else describe them as a malevolent spirit (phii 'spirit, ghost', phii-baa [spirit-insane] 'lunatic').

160

N.J.Enfield

2.42. het taa-naj 'make big-eyes' To 'make big-eyes' - het taa-naj - is to open the eyes wide, with associated raising of the eyebrows. One who 'makes big eyes' does not necessarily expose the whites of their eyes above the iris. Most informants describe a face with taa-naj 'big eyes' as that of someone who is tok-caj 'surprised', although it may also mean that they are merely son-caj 'interested' in something.

2.4.3. taa-tii 'eyes held open!apart' The term taa-tii has a similar meaning to taa-naj, but with a connotation of less control over the facial expression. The verb tii means 'hold an opening apart so as to provide access through', for example when holding a plastic bag open for someone to put something in. If you are taa-tii '(with) eyes held open/apart', the connotation is that what you are seeing is so surprising or engaging that it 'holds your eyes open'.

2.4.4. taa-khaang 'stuck-eyes' The term khaang conveys the idea of being 'stuck' or 'remaining, left over', as for example with reference to water which has not drained from a flat rooftop. To say that someone's eyes are khaang 'stuck' is to say that they are wide open and stuck there, i.e. that they do not close at all, not even momentarily for blinking. The open eyes of a corpse can be described as khaang, as well as the open eyes of someone sleeping or in a coma. When taa-khaang 'stuck eyes' is taken to be an index of some inner state, the idea is that the experiencer is so amazed or interested in what is happening in front of them that they cannot look away or even blink. This points to what may be a general parameter in the description of facial expressions, namely the question of whether the pose is extended over time or not. The expression taa-khaang 'stuck-eyes' entails the pose being held over a stretch of time, while other terms describing facial expressions may make no such specification. Grin, for example, entails neither a momentaneous nor extended time profile. Thus, taa-khaang 'stuck-eyes' is not associated simply with being tok-caj 'surprised', but with this plus

A Lao perspective on facial expression of emotion

161

something more, which accounts for the persistence of the pose over time. For example, the person who is taa-khaang 'stuck-eyes' may be son-caj 'interested', or the like.

2.4.5. nik-khiw 'raise-eyebrows' While Lao has a basic transitive verb nok 'to raise something', this verb cannot be used for raising one's eyebrows, unless the meaning intended is that the person physically pulls up their own eyebrows with the hands. To raise one's eyebrows as a facial gesture is referred to as nik-khiw (in the verb + object pattern, perhaps best literally glossed as 'raise-eyebrowed'). (The verb nik does not occur elsewhere.) Informants have responded in a number of ways in attributing inner states to subjects with nik-khiw 'raised eyebrows', and it seems that accompanying features of facial expression are important factors in choosing between alternatives. Someone who is merely nik-khiw 'with raised eyebrows', and without specific expression elsewhere in the face, may be described as son-caj'interested\jaak-thaam 'wanting to question', or boo-sua 'non-believing, doubtful'. The common theme here seems to be that the subject wants to know more about what is going on. When nik-khiw 'raised eyebrows' is accompanied by taa-naj 'big eyes', taa-khaang 'stuck eyes', or especially taa-loo 'exposed-whitesabove-iris eyes', this is associated with being jaan 'afraid', as well as tokcaj 'surprised'.

2.4.6. hetkhiwson 'making-eyebrows-collide' Lao does not have a term equivalent to English/rovwi, and the idea of bringing one's eyebrows together is expressed as het khiw son 'make the eyebrows collide'. (The verb son is used elsewhere most notably for the butting of buffaloes' heads when they clash.) Inner states of son-caj 'interested', or kamlang-khit [progressive-think] 'thinking' are attributed to someone displaying this facial expression.

162

N.J.Enfield

2.4.7. het khiw cööt-kan 'making-eyebrows-be-parked-together' The term het khiw coot kan means literally 'to make the eyebrows be parked together'. This is an alternative description of the 'knitted brow' just described as 'making the eyebrows collide' - the same comments apply.

2.5. Comment The evidence presented in § 2 has shown that Lao speakers' habitual ways of talking about facial expression differ from those of English speakers. It appears that for certain English expressions which seem to describe the face as a whole - such as frown or grimace - the Lao equivalents explicitly refer to relevant parts of the face, such as the eyebrows or the lips. Similarly, certain specific aspects of facial expression associated with certain specific emotions are singled out in Lao, such as for example the eyebrows, and the exposed-whites-above-iris eyes. Thus, it is not simply expressions on whole 'faces' that reveal and/or communicate feelings and thoughts, but rather expressions involving features or components of the face, such as the eyes, eyebrows, lips, mouth, or some combination of these. Certain descriptions of facial expressions make specific reference to components of the face, and I suggest that in further research, it would pay to consider facial expression as having a greater componentiality than is sometimes assumed. (Cf. Wierzbicka 1999, Chapter 4.)

Conclusion While we may be led to attribute thoughts and/or feelings to a person on the basis of the expression on their face, it is not the case that a particular set of facial expressions conveys, either universally or specifically, a particular set of inner states or emotions. Firstly, there is no universal set of "emotions". What are denoted by words for 'emotions' are not actual phenomena, but conventional and emic descriptions of actual phenomena. Emotion terms denote concepts, by particular names in particular languages, such as angry, pissed off, or caj-haaj. When we work on the description of emotion, we cannot escape our reliance on language. Scientific inquiry is

A Lao perspective on facial expression of emotion

163

chronically linguocentric, and very often hopelessly monolinguocentric as well. (For these terms, see Enfield 2000.) It is unacceptable to adopt Ekman's (1973) perspective that '[r]egardless of the language, of whether the culture is Western or Eastern, industrialized or preliterate,... facial expressions are labeled with the same emotion terms: happiness, sadness, anger, fear, disgust, and surprise' (Ekman 1973: 220). It is simply not true that in all societies these labels are applied, because these labels are English, and not all societies know and/or use English. It is interesting to consider the parallel between research on facial expression of emotion, and research on the semantics of colour terms. (See Foley 1997: Ch. 7 for a useful recent review of colour research.) It has been argued (e.g. by Lucy 1996 and Wierzbicka 1996) that colour terms are not labels for mathematically defined frequency bandwidths, but are labels for concepts, which of course may in turn relate to, or be derived from, those natural phenomena. Research on colour semantics has been traditionally fraught with problems of translation, and the unquestioned assumption that the biological basis of the 'meaning' intended by a colour term has provided researchers with a way of anchoring language in uncontroversial human universals. But the problem lies in the very assumption that these universal biological phenomena - such as visual perceptual apparatus (in the case of colour) or empirically definable visceral responses (in the case of "emotion") - are what linguistic expressions actually label. There is no evidence that they are.4 The data presented here supports a view that different ethnolinguistic systems analyse the expressive face using somewhat different conventions. Universals of thought and feeling undoubtedly exist, along with ethologically grounded explanations for visceral phenomena and universal aspects of facial behaviour. But these have no direct link to what we refer to, by lexical items in natural languages, as emotions. We cannot take "basic emotions" for granted, and certainly not as something that pre-defined facial expressions (if these could be identified) universally signify. And it is essential to take an empirical and maximally assumption-free approach to understanding what "facial expressions" really consist of, cross-culturally and cross-linguistically. Hence the value of "folk" analyses of the data, such as I have described here. I therefore reiterate Russell's recent call for more empirical data on the possible interpretations of facial expression which natural languages reveal.

164

N.J.Enfield

Notes 1. Transcription of Lao is a simplified system, with tone-marking omitted. Note that the consonant symbol 'j' in this system (as in the IPA system) is pronounced like English 'y' in you and boy; 'c' is pronounced in a similar manner to English 'j' in Jill; 'q' represents a glottal stop. Symbols ' ', 'o', 'e' represent high-back-unrounded, low-back, and low-front vowels, respectively. All data are from fieldnotes, and from field interviews conducted in Sisavat and Saphang Mo villages in Vientiane, Lao PDR, January-May, and July-August 1998. Syban Khoukham's assistance is gratefully acknowledged. Pitsana Vayaphanh also provided generous consultation, and Sotaro Kita gave useful comments on a late version of the text. 2. Disgust is a particularly physical emotion, with a genuine functional/biological aspect to it, and one could imagine that certain things - such as, say, rotten food - would evoke disgust in anyone with a human body, in any cultural sphere. But it is interesting to note that culture can differently define what disgusts people in this physical way. Thus, one can easily find everyday practices which are embraced by one group, yet physically revolting for another. Rural Lao people enthusiastically snack on insects of different kinds (crickets, grasshoppers, beetles), a practice which the average urban Australian, say, would find physically disgusting. On the other hand, the popular practice of oral sex in Anglo culture is regarded similarly as physically disgusting in many other settings (e.g. rural Burma, Spiro 1977: 232). These fascinating culture-specific visceral responses, showing the influence that culture may have on our biological responses, are worth closer and broader research. 3. Τύύη 'awaken' has an interesting argument structure, being S=A ambitransitive, where the subject is semantically a patient; thus, cf. khooj tuun hok moong [I awaken six o'clock] Ί woke up at six o'clock', khooj tuun caw [I awaken you] Ί (got a) start (from) you'/'You gave me a start'. In the second example, the syntactic object caw 'you' is the stimulus for the event of surprise predicated by tuun 'awaken' of its subject khooj T. 4. Even so, colour research has resulted in interesting findings, such as the various implicational statements which Berlin and Kay (1969) originally put forward. While there are serious methodological problems with their research (Lucy 1996), it is nevertheless interesting to consider the kinds of questions being asked. If a language has only three "basic colour terms" (as Berlin and Kay define these), can we predict what will they be? Berlin and Kay argued in the affirmative for LIGHT ('white'), WARM ('red'), and DARK/COOL ('black/blue/green'). If a language had five terms, they argued, then 'yellow' would appear, along with a split of DARK/COOL into 'black', and 'green/blue'. Could we follow such a line with respect to facial expressions? If a language had only one "basic facial expression term", would it be 'smile'? (Cf. Wierzbicka 1999: 275, 282.) But this may be the wrong kind of question to ask (cf. Lucy 1996 on colour).

A Lao perspective on facial expression of emotion

165

References Berlin, Brent, and Paul Kay 1969 Basic Color Terms. Berkeley: University of California Press. Diller, Anthony V. N., and Preecha Juntanamalaga 1990 'Full hearts' and empty pronominals in Thai. Australian Journal of Linguistics 10 (2): 231-256. Ekman, Paul 1973 Darwin and Facial Expression: A Century of Research in Review. New York: Academic Press. 1992 An argument for basic emotions. Cognition and Emotion. 6 (3/4): 169-200. Enfield, N. J. 2000 On linguocentrism. In: Martin Pütz and Marjolijn H. Verspoor (eds.), Explorations in Linguistic Relativity, 125-157. Amsterdam: John Benjamins. 2001 Semantic analysis of body parts in emotion terminology: avoiding the exoticisms of "obstinate monosemy" and "online extension". Pragmatics and Cognition 9.2. Foley, William A. 1997 Anthropological Linguistics. London: Blackwell. Goddard, Cliff 1997 Contrastive semantics and cultural psychology: 'Surprise' in Malay and English. Culture and Psychology 3 (2): 153-181. Jaisser, Annie 1990 DeLIVERing an introduction to psycho-collocations with SIAB in White Hmong. Linguistics of the Tibeto-Burman area 13(1): 159-177. Lucy, John 1996 The linguistics of 'color'. In: C. L. Hardin and Luisa Maffi (eds.), Color Categories in Thought and Language, 320-346. Cambridge: Cambridge University Press. Matisoff, James A. 1986 Hearts and minds in Southeast Asian languages and English: an essay in the comparative lexical semantics of psycho-collocations. Cahiers de Linguistique Asie Orientale 15(1): 5-57. Oey, Eric. M. 1990 'Psycho-collocations' in Malay: a Southeast Asian areal feature. Linguistics of the Tibeto-Burman Area 13(1): 141-158. Russell, James A. 1995 Facial expressions of emotion: What lies beyond minimal universality? Psychological Bulletin 118 (3): 379-391.

166

N.J.Enfield

Spiro, Melford E. 1977 Kinship and Marriage in Burma. Berkeley: University of California Press. Wierzbicka, Anna 1996 Semantics: Primes and Universals. Oxford: Oxford University Press. 1999 Emotions across Languages and Cultures: Diversity and Universals. Cambridge: Cambridge University Press. Zajonc, R. B. 1980 Feeling and thinking. Preferences need no inferences. American Psychologist 35 (2): 151-175. 1984 On the primacy of affect. American Psychologist 39(2): 117-123.

Hau: A key word in the Malay vocabulary of emotion Cliff Goddard

Introduction1 The word hau is one of the "key words" of Malay culture. By this I mean, following Wierzbicka (1997), that hati functions as a conceptual focal point for an entire complex of characteristically Malay values, attitudes, and expectations; and that by studying the meaning and uses of this one word we can learn a surprising amount about Malay culture - in particular, about the conceptualisation of emotion in Malay culture. Though the nearest English gloss for hati is 'heart' (in its emotionalmoral sense) the two words are not semantically identical, if only because the Malay hati is significantly more active, and more cognitive, than the English heart. A more revealing, but still inadequate, gloss for hati would be 'the sensitive part of a person'. Hati has a high frequency in any discourse about human interaction - partly on account of its participation in dozens (if not scores) of fixed expressions concerning attitudes, moods, and personal traits, e.g. susah hati 'troubled, worried', hati keras 'determined', rendah hati 'humble, modest', partly because the hati is the locus for feelings (especially feelings about other people), and partly because emotional reactions are often presented in terms of the hati "speaking". As one might expect, the word occurs frequently in traditional sayings and poems, and in popular song titles. It is no exaggeration to say that one cannot approach an understanding of Malay attitudes about human nature and about social life without understanding this quintessentially Malay concept. It would be impossible in one short paper to detail the full range of usage of hati. I attempt here to do the following: first, to outline the range of use and collocational possibilities of hati, informally comparing and contrasting it with English heart, second, to advance and argue for an explicit semantic explication of hati in its core or central meaning (as in an expression like hati orang 'a person's hati'); third, to explicate the semantics of five

168

Cliff Goddard

common fixed expressions involving hati, all of which designate what we might term feeling-states or emotional reactions: susah hati 'troubled, worried', senang hati 'relaxed, easy at heart'; sakit hati 'annoyed, offended', puas hati 'satisfied (with someone)', kecil hati 'feel hurt'. The example sentences given throughout come from a corpus of contemporary informal Malay writing.

1. The uses and meaning of hati To begin with, it is useful to highlight the differences between Malay hati and the emotional-moral sense of English heart, as in expressions such as broken-hearted, matters of the heart, a good heart, and so. (For a comparable study, see Diller and Juntanamalaga's (1990) study of the syntax and semantics of Thai cay 'heart, mind, disposition'. Limitations of space prohibit consideration of various intriguing parallels between the situation in Malay and in Thai, which are geographically contiguous but genetically unrelated languages. Actually, the parallels extend to many languages of Southeast Asia, cf. Matisoff (1986), Oey (1990), which forms a "linguistic area" in many respects.) One difference between heart and hati, perhaps not terribly significant for our purposes, is although both words have polysemic senses referring to parts of the body, anatomically hati refers not to the heart (which isjantung), but to the liver.2 A second, more significant, difference has been mentioned above: that Malay hati is more cognitive and more active than English heart. To illustrate the first of these properties, consider example (1) below. It would hardly be possible, in English, to say that one's heart was filled with thoughts, or with memories (one's mind, perhaps, but not one's heart). On the same point, it is noteworthy that certain fixed expressions involving hati are more felicitously translated into English using mind than heart. For example, senang hati (senang 'relaxed, easy') as 'peace of mind'. (1)

... seribu satu masalah memenuhi hatinya. Ingatannya pada ibunya di kampung, ingat adik-adiknya yang masih kecil-kecil, ingat pula pesanan ibunya. '...a thousand and one troubles filled her hati. Memories of her mother in the village, of her brothers and sisters who were still little, memories also of her mother's exhortations.'

H at i: A key word in the Malay vocabulary of emotion

169

The more active nature of the Malay hati is shown by the existence of expressions such as those in (2) and (3), in which the hati is depicted as tertanya 'spontaneously asking' an emotionally loaded question, or as bisik 'whispering' an emotional reaction. In general, the hati is capable of "expressing" itself in ways which can be reported with a wide range of speechact verbs, includingpujuk 'console, persuade', rungut 'grumble', and plain kata 'say', aside from those verbs illustrated below. (Note that in examples throughout the paper, one sees hati suffixed with pronominal clitics: -ku Ί, my', -mu 'you, your', -nya 'he/she, his/her'.) (2)

Ay u ... Hatiku tertanya-tanya mengapakah kau menghubungiku lagi? 'Ayu... My hati keeps asking why do you keep (trying to) contact me?'

(3)

Ini bukan wayang kulit, bisik hatinya kecewa. 'This isn't shadow theatre, he thought to himself (lit. whispered his hati) with disappointment.'

Even more striking, it is quite common to see emotional reactions attributed directly to a person's hati, as though the hati itself were the "experiencer". That is, rather than saying the equivalent of Ί am sad', one says 'my hati is sad' or 'this hati is sad'; rather than saying 'it hurt me', one says 'it hurt my hati' or 'it hurt this hati'. Examples are given in (4)-(6) below. (4)

Roslina tidak dapat berkata apa-apa lagi, hatinya terasa diirisiris, pedih, pahit. 'Roslina couldn't say anything more, her hati felt cut up, smarting, bitter.'

(5)

Pada masa yang sama teruskan usaha membahagiakan dan menggembirakan hati ibu dengan bantuan yang terdaya. 'At the same time carry on your effort to make your mother's hati happy and pleased with the best help you can give her.'

(6)

Tapi Chan tak tau pedihnya hati ni bila mendapati Chan tak ada di tempat biasa... Tak boleh takjumpa. Rosak hati ni. 'But you (Chan) don't know how sad this hati was on realising that you (Chan) weren't at the usual place... We have to meet. (If not) it hurts this hati.'

170

Cliff Goddard

The hau may also be the locus of desire or intention, especially if the desire or intention concerns another person. Not only may one speak of niat hati 'desire in one's hati', hasrat hau 'hati's desire', kehendak hau 'hati's wish, desire', but also of the hati actively wanting or intending to do something. For example: (7)

Keesokan harinya Rizal sudah awal menanti kedatangan Rina. Hatinya sudah tak tahan lagi untuk menatap wajah ay u isterinya. 'The following day Rizal was awaiting his wife's return all day. He (lit. his hati) couldn't wait (lit. withstand) any longer to set eyes on his wife's lovely face.'

(8)

Sebenarnya sejak kita berjumpa dulu lagi hati say a terbetik hendak membuat cadangan. 'As a matter of fact, ever since we've been reunited my hati has been itching to make a suggestion.'

Along with these differences between hati and heart, there are also important affinities between the two concepts. Like the heart, the hati is the realm of feelings, especially sustained feelings about other people, such as cinta 'romantic love', rindu 'longing', cemburu 'jealousy, envy', and sedih 'sorrow'. The generic expressionperasaan hati (whereperasaan 'feelings' is a nominalisation of rasa 'feel'), corrresponds closely to feelings of the heart. That is, it designates feelings which are relatively longstanding, involve evaluation, and are directed toward another person (as opposed to transient or impersonal feelings such as terkejut 'being startled', takut 'fear'), cf. Heider (1991), Goddard (1995). Also like English heart, the hati is by nature a private, inner realm. Others cannot know the isi hati 'contents of the hati' unless the subject chooses to luahkan 'let (it) out' or to mencurahkan 'pour (it) out'. Consequently, one often finds hati used in contexts like those in (9)-(l 1) below, which concern private, hidden emotions. I believe there is a subtle difference between the expressions di hati (lit. 'at hati') and (di) dalam hati (lit. 'inside hati'}, the latter expression conveying a greater sense of concealment, or at least, greater contrast with external appearances; but we can let this pass here. Notice also the existence of expressions such as di sudut hati 'in the depths of the hati', terpendam di hati 'buried in the hati', and hati kecil 'innermost hati' (lit. 'small hati'} - all of which highlight the hidden, inaccessible depths of the hati.

Hau: A key word in the Malay vocabulary of emotion

171

(9)

Jauh di sudut hatinya, kasih mula dipintal, sayang mula dianyam dan cinta mula dipilin. 'Deep in the recesses of her heart, love began to wring out, care began to be woven, and passion began to be plaited.'

(10)

Namun di dalam hatinya dia merasa bangga kerana dipuji, kerana ramai yang terpesona pada kejelitaannya. 'But in her heart she felt proud of the compliment, because there were many who were enchanted by her beauty.'

(11)

Perasaan marah kerana ayah berkasar dengan ibu itu memang biasa dalam hati seorang anak. 'The feeling of annoyance on account of a father's harsh treatment of a mother is simply normal in the hati of a child.'

Consistent with the private, inacessible nature of the hati is the fact that a person can say things (di) dalam hati 'inside the hati', meaning to say things to oneself, as illustrated in (12) and (13) below; cf. also the expression membaca di dalam hati 'read to oneself, as opposed to reading aloud. (12)

Tidak ada yang menarik, katanya dalam hati. 'Nothing interesting here, he said to himself (lit. inside hati).'

(13)

Ah, gila aku macam nil Bolehjadi gila! Jerit Mira, tetapi di dalam hatinya saja. Oh, I'm going crazy. I really will go crazy! Screamed Mira, but only to herself (lit. inside her hati).'

A reference to the hati belonging to someone, being given to someone, or being attracted to or ensnared by someone, indicates romantic love or attraction. For example, Hatinya terpikat kepada Jasni 'her heart was snared by Jasni', i.e. she was secretly in love with him. Sundry other hati expressions also concern love; for example, buah hau (buah 'fruit') means 'sweetheart'; jatuh hati (jatuh 'fall') is 'to fall in love with'. It may be worth noting in passing that romance has always been a deeply ingrained tradition in Malay rural society (cf. Karim 1990), and in modern times Malay pop music maintains this tradition, with an even greater concentration on sweetness, unfulfilled yearning, sacrificing for love, etc. than Western pop music.

172

CliffGoddard

Fixed expressions involving hati are extremely numerous, and can be classified into several groups on both formal and semantic grounds (cf. Karim 1990). Among the more revealing verbal expressions involving hati are those listed in (14a)-(14f). Abdullah's (1990) Karnus Simpulan Bahasa [Dictionary of Fixed Expressions] lists another two dozen or so. (14)

a. b. c. d. e. f.

jaga hati 'to be protective of (someone's) feelings' (lit. 'watch over hati') ambil hati 'to charm (someone), win over' (lit. 'get hati') suka hati 'do as you please, do whatever you feel like' (lit. 'please hati') tidak sampai hati 'can't stand to (do something)' (lit. 'not reach hati') cuba hati 'test (someone's) honesty or commitment' (lit. 'try hati') ubah hati 'change attitude, switch allegiances' (lit. 'change hati')

Taken together, these expressions illustrate the rather rich nature of the hati, combining aspects of feeling, wanting, and thinking. To begin with, there is the expression jaga hati (orang) 'be protective of (people's) hati', which can be regarded as designating one of the central themes of Malay culture. Cultural commentators invariably emphasise the great value placed on "showing consideration and concern, anticipating the other... and, above all, being sensitive to the other person" (Wilson 1967: 131-2), lest the other person's feelings be hurt (tersinggung). In the expression jaga hati, hati seems nearly equivalent to the word perasaan 'feelings'. Indeed, jaga hati is often paraphrased as memelihara perasaan 'tend to feelings' or bertimbang rasa 'take account (lit. weigh) feelings'; and the combination hati dan perasaan 'hati and feelings' is not uncommon. In the expressions in (14b)-(14d), however, we find hati being employed in a way which seems to combine reference both to feelings and to intentions. To ambil hati is to gain an influence over someone by being sweet and charming; suka hati means to do as one's pleases (generally regarded, in Malay culture, as a risky and immature way to act), and the expression tidak sampai hati indicates something that one cannot stand to do. Finally, in the expressions in (14e)-(14f) hati seems to designate an even broader, and more enduring, aspect of a person: something akin to integrity or one's degree of commitment to a cause or to another person.

Hau: A key word in the Malay vocabulary of emotion

173

The same broad range of semantic effects is found among fixed nominal expressions involving hati. Most of these have a two-part structure, consisting of hati and an adjective. They can be grouped formally according to whether the adjective follows or precedes the noun hati. Attributive adjectives normally follow their head nouns in Malay, but hati + adjective is the less frequent of the two patterns. Abdullah's (1990) Kamus Simpulan Bahasa [Dictionary of Fixed Expressions], for example, lists around 15 hati + adjective expressions, but over 50 adjective + hati expressions. Both these figures are likely to be far short of the full number, as several of the combinations listed below are not on Abdullah's lists, but the relative preponderance of the adjective + hati pattern is clear. While the semantic difference between the meanings expressed via these two patterns is not perfectly clear to me, it seems that while the hati + adjective pattern encodes an attributive relationship (as one would expect), the adjective + hati pattern is more like a "locus" relationship, at least when it designates a feeling, as in the examples listed in (15 a). That is, a preposed adjective does not "describe" the hati, but rather it indicates a feeling or attitude which is being located, so to speak, in the hati (somewhat similar in construction to English phrases such as sick at heart). The locus construction with hati also seems somehow to "personalise" the state being described. A second group of adjective + hati expressions seem to designate moral aspects of a person's character. Some examples are listed in (15 b). These might seem to be straightforwardly attributive, but my impression is that the meaning structure is more subtle than this: that these expressions are not describing the person's hati as such, but are describing the person him or herself- by means of saying something about the hati (perhaps a bit like English expressions such as pure of heart). More research is obviously needed. Examples of several semantic subcategories of the two patterns follow. Notice that the adjectives involved also fall into several types. There are explicitly moral terms such as murni 'noble' and murah 'generous', there are more general evaluative terms such as baik 'good', tulus 'pure', kotor 'dirty,' and busuk 'putrid', there are terms for sensation-states such as panas 'hot', sejuk 'cold', sakit 'hurting', and puas 'sated', and there are terms for physical dimensions and states such as besar 'big', kecil 'small', lembut 'tender', keras 'hard', berat 'heavy', andputus 'broken'.

174

CliffGoddard

(15)

Adjective + hati

a. Emotions and moods: besar hati 'proud, thrilled with' (besar 'big') kecil hati' feeling hurt' (kecil 'small') geli hati 'amused, tickled' (geli 'ticklish') iri hati 'envious, spiteful' panas hati 'angry, worked up' (panas 'hot') sejuk hati 'calmed, reassured' (sejuk 'cool, cold') susah hati 'troubled' (susah 'difficult') senang hati 'contented, relaxed' (senang 'easy') hancur hati 'devastated' (hancur 'crushed') puas hati 'satisfied' (puas 'sated') sakit hati 'annoyed, resentful' (sakit 'sick, pain') putus hati 'despairing' (putus 'broken asunder') lapang hati 'content' (lapang 'vacant, open')

(16)

b. Aspects of moral character: baik hati 'kind, good-hearted' (baik 'good') busuk hati 'vicious, mean' (busuk 'putrid') tulus hati 'sincere, honest' (tulus 'pure, unadulterated') murah hati 'generous hearted', (murah 'generous') rendah hati 'humble, modest' (rendah 'low') tinggi hati 'proud' (tinggi 'high') putih hati 'sincere, honest' (putih 'white') suci hati 'pure of motive' (suci 'morally pure') tabah hati 'determined' (tabah 'resolute, persevere') lembut hati 'tender-hearted' (lembut 'tender') tawar hati 'lost interest' (tawar 'bland') keras hati 'stubborn' (keras 'hard')

hati + Adjective

a. Moral states (often in religious contexts): hati bersih (lit. 'clean') hati kotor (lit. 'dirty') hati nurani 'enlightened, virtuous' (nurani 'bright') hati murni 'noble heart' (murni 'noble') hati tawajuh 'devout' (tawajuh 'devoted to God') c. States of suffering: hati berat (berat 'heavy') hati luka (luka 'wounded') hati terkilan 'upset, aggrieved' hati mutu/walang 'melancholy, depressed'

b. States of mind: hati keras 'determined' (keras 'hard') hati terbuka '(with) an open mind' (terbuka Open') hati waja 'resolute, determined' (waja 'steel')

Hau: A key word in the Malay vocabulary of emotion

175

To round out this quick picture of the uses ofhati, it is instructive to look at someperibahasa 'traditional sayings' andpepatah 'maxims'. Traditionally, there were literally hundreds of peribahasa, which to a large extent encapsulated the distinctive Malay adat 'customs, rules for living'. These days many of the old peribahasa have fallen out of use, but even so scores of them survive in everyday use. In (17) the reference to hau carries moral overtones. The hati is the source of morally correct action, so better one's eyes be blind than one's hati. (18) reflects the view that people's hati can be reached by tutur yang manis 'sweet words', which can soften even the hardest hati. In example (19), however, the reference to hati is less favourable. The sentiment is cautionary: we will suffer if we ikut hati 'follow hati' without careful reflection, another persistent theme of Malay culture. Perhaps the allusion here is primarily directed towards the role of the hati in romantic matters. In (20), the expression dekat di hati 'close at hati' refers to fondness: someone may be physically far away and yet be held in fond remembrance. The expressions in (21)-(22) employ hati from yet another angle, referring here to a general alignment of views and attitudes. People who are sehati One hati' are all of the one mind; those who are hati Iainlain 'different hati' cannot get on harmoniously. (17)

Biar buta matajangan buta hati. 'Let the eyes be blind, (but) not the hati'

(18)

Batu pun empuk, jangankan hati manusia. '(If) even a rock can be softened, why not the hearts of people.'

(19)

Ikut rasa binasa, ikut hati mati. 'Follow feelings suffer, follow hati die.'

(20)

Jauh di mata, dekat di hati. 'Far from sight, (but) close at heart.'

(21)

Rambut sama hitarn, hati Iain-lain. 'The same black hair, (but) different at heart.'

(22)

a. b.

Sehati sejiwa. One hati, one soul.' Sama sehati. 'at one' (lit. 'same one hati')

176

CliffGoddard

2. Meaning of hati Three things should be evident at this point: (1) that hati is one of the key ethnopsychological terms of the Malay language and culture; (2) that hati has certain similarities with the English term heart', (3) that hati also differs in important respects from English heart. The question now is: Is it possible, in a principled and rigorous fashion, to state the meanings of words like hati and heart, so that the singularities and differences can be made be explicit? The most important scholarly precedent bearing on this question is Wierzbicka's (1992: 31-63) study of the Russian concept of dusa, roughly, 'soul', and comparable concepts in several other European languages, such as Russian serdce, English heart, mind, and soul, and German Seele. For purposes of comparison, it is useful to consider Wierzbicka's (1992: 48) explication (below) of English heart. It is framed within the "natural semantic metalanguage" (NSM) approach to semantic description, according to which meanings are stated in the form of explanatory paraphrases composed in a small, standardised and translatable metalanguage based on natural language. For the purposes of this paper, it is not necessary to describe this methodology or the theory behind it in any detail (cf. Wierzbicka 1996, Goddard and Wierzbicka eds, In press). A list of NSM semantic primes, in English and Malay, is given in the Appendix. All the explications should be readily translatable into Malay, cf. Goddard (In press). (23)

heart (a) a part of a person (b) one cannot see it (c) one can imagine that it is a part of the person's body (d) (in the middle of the upper part of the body (e) one can hear its movements) (f) because of this part, a person can feel good things and bad things (g) because of this part, a person can feel good things towards other people

It will be be helpful if we work through this explication line by line. Lines (a)-(c) state that a heart is a 'part of a person', which we cannot see but which we can 'imagine' to be part of the person's body. The wording

Hali: A key word in the Malay vocabulary of emotion

177

here is intended to capture the fact that although we recognise that heart (in its emotional-moral sense) is not part of a person's body it can nonetheless be thought of as such. Lines (d)-(e) indicate which particular physical part is implied. The key components of the explication are, however, the final two lines. Line (f) states that it is the heart which enables a person to 'feel good things and bad things'. This reflects the fact that the heart is linked only with "valenced" feelings such as joy, bitterness and sadness, and not with neutral feelings such as surprise and interest, which make no reference to the concepts of 'good' and 'bad' (one can 'feel' surprise and interest, but not in one's heart). Line (g) goes on to state more specifically that the heart enables one to 'feel good things towards other people'. This reflects the fact that, so far as feelings directed towards other people are concerned, the heart is specifically linked with the potential for positive feelings (love, pity, admiration, etc.). Since the above explication was originally proposed, there have been some significant refinements to the natural semantic metalanguage theory, so not surprisingly there are several matters of phrasing which now appear less than optimal. In particular: (a) the term imagine is no longer regarded as a viable semantic prime, (b) the expression feelings towards is now recognised not to be a universal syntactic possibility for the primitive feel. In addition, (c) it is preferable to avoid using the term and (as in good things and bad things), since and is not a linguistic universal either. Rephrasing to dispense with these features would make the heart explication somewhat longer, but content-wise it is, in my opinion, substantially correct, and forms a useful point of comparison for our account of Malay hau. In view of the discussion of hau in the preceding section, I would propose the following explication. (24)

hau orang (a person's hau) (a) a part of a person (b) people think it is like a part of a person's body a big part inside the middle of the body (c) when a person thinks about someone else for some time, something can happen in this part (d) when this happens, this person feels something because of it sometimes something good, sometimes something bad (e) because of this, this person wants to do something sometimes something good, sometimes something bad

178

CliffGoddard

(f)

when something happens in this part of a person, this person can think: this is like something inside me is saying something to me (g) other people can't know what happens in this part of a person, if this person doesn't want them to know The first two lines are similar in intention to their counterparts in Wierzbicka's explication for heart. Component (a) just states that the had is a part of a person. Component (b) conveys the notion that the hau is thought of as 'like' a part of a person's body, and therefore, by implication, not literally a body-part. The second line of component (b), which characterises the physical part of the body involved, i.e. the liver, is provisional at this stage. It describes the liver simply as a large internal organ. It would be good to find out more about traditional Malay folk beliefs about the bodily hati, as reflected, for example, in traditional healing practices. Intriguingly, Peletz (1996: 206) remarks in passing that: "Some Malays refer to the liver as the 'ruler' (raja) of the human body and note that it 'governs' or 'regulates' (merintah) the rest of the body, much like a ruler or commander governs his army." Karim (1990: 26) describes the liver as a "mysterious organ, which is believed to control the moods and emotions of humans and to command more permanently their psyche and personality in both a psychological (zahir) and spiritual (batin) sense." The bulk of the explication attempts to articulate the rich cultural construal of this "mysterious organ". Component (c) states that sustained thought about another person can be associated with an event ('something happening') in the hati. This already gives the hati a more dynamic quality than the English heart. The internal events of the hati, according to component (d), give rise to feelings which may be either good or bad. These feelings, according to component (e), have motivational consequences, i.e. they can lead to the person wanting to do something, which, again, may be something good or something bad. This set of components, i.e. components (c)-(e), spells out the role of the hati as the wellspring of interpersonal feelings and intentions. The specification that the intended actions are 'sometimes good, sometimes bad' embues the hati with a certain moral ambivalence. If we were to follow the urgings of the hati alone, therefore, the results would not always be good (hence the need for careful monitoring and regulation of the hati, as mentioned in the earlier discussion).

Hati: A key word in the Malay vocabulary of emotion

179

Component (f) characterises an important aspect of the phenomenology of the hau, the fact that it can seem to the experiencer that the hati is 'saying something' to him or her. This component reflects the relatively active and cognitive role of the hati (as compared with English heart}. The final component (g) specifies that the workings of the hati are inaccessible to other people, unless the experiencer wishes other people to know of them.

3. Feeling states of the hati Given that the hati is the source of emotional responses to other people, it is not surprising that there are so many emotional expressions based on hati. In the remainder of the paper we will look at five such expressions, seen in (15 a) above, which are common in everyday discourse. (25)

susah hati senang hati sakit hati puas hati kecil hati

'troubled, worried, preoccupied' (susah 'difficult') 'relaxed, comfortable, happy' (senang 'easy') 'annoyed, offended, resentful' (sakit 'sick, pain') 'satisfied, pleased' (puas 'satisfied, sated') 'feeling hurt, disappointed' (kecil 'small')

Each of these expressions is commonly verbalised by means of the productive causative derivation, which is achieved by simultaneous affixation with the transitive prefix meN- and the valency-increasing suffix -kan. The prefix meN- requires some brief description. Basically the final segment is a nasal whose realisation depends on the nature of the initial segment of the root. If the initial segment of the root is a voiced obstruent, N appears as a homorganic nasal (e.g. besar 'big', membesarkan 'make big'), but if it is a voiceless obstruent, N appears as a homorganic nasal which replaces the voiceless obstruent (except if the obstruent is palatal stop c, in which case meN- appears as men-}. For example, corresponding to puas we have memuaskan, and corrresponding to senang we have menyenangkan. Thus: (26)

menyusahkan hati menyenangkan hati menyakitkan hati memuaskan hati mengecilkan hati

'make troubled, preoccupied' 'make comfortable, please' 'annoy, offend, embitter' 'satisfy, please' 'hurt, disappoint'

180

CliffGoddard

In the following, I will assume that the verbal expressions incorporate the same meaning as the base expressions.3 Susah hati Susah hati designates a troubled state of mind. This may be focused either on oneself or on someone else one cares about. A person who is susah hati is preoccupied with the idea that something bad and undesirable may happen to someone. The experiencer feels under pressure to do something to prevent it, but, unfortunately, he or she cannot think what to do. Hence overall one is left in an unpleasant state of troubled preoccupation. (27)

Person-X susah hati (a) for some time, X thinks something like this: (b) something bad can happen to someone (c) I don't want this (d) I have to do something (e) I can't think what I can do (f) because of this, X feels something bad

In support of this explication, I can offer the following observations. First, susah hati may occur with a complement introduced by memikirkan (tentang) 'thinking about'. This is consistent with the active cognitive scenario depicted in the explication. Someone who is susah hati is indeed thinking about something, as stated in component (a). (28)

Dia susah hati memikirkan anaknya yang pergi ke bandar mencari kerja. She was worried thinking about her son who'd gone to the city looking for work.

Second, the range of worries which can induce susah hati is quite varied. It could be a potential problem of just about any kind, and, as shown in examples (29)-(31), this potential problem may concern either oneself or someone else one cares about, e.g. a family member. For this reason, component (b) is worded in deliberately general and open-ended fashion. The explication depicts someone who is susah hati as dwelling on the thought that 'something bad can happen to someone' (where 'someone' could include oneself). This is a prospect which, as set out in component (c), the experiencer wishes not to happen. (Note that in the context of (29), the old

Hati: A key word in the Malay vocabulary of emotion

181

man was definitely ruminating on what was going to happen to him. His illfated life was evidence for him that there was worse to come.) (29)

Lelaki tua semakin susah hau. Dia sentiasa mengeluh mengenangkan nasibnya yang malang. 'The old man got more and more troubled. He was forever sighing as he recalled his misfortune.'

(30)

Perkara ini akan say a selidiki nanti. Bersabarlah, dan jangan susah hati. 'I'll sort this matter out directly. Be patient, and don't worry.'

(31)

Bila dengar Sarha cakap macam tu, say a susah hati takut Sarha akan bunuh din. 'When I heard Sarha talking like that, I was susah hati fearing that Sarha would kill herself.'

Given the experiencer's awareness of the possibility of a bad and unwanted event befalling someone, component (d) spells out the consequent sense of responsibility for finding a solution or resolution (Ί have to do something'). Component (e) states that, for the time being, at least, the experiencer is at a loss about what to do (Ί can't think what to do'). These components are not as readily or directly illustrated by means of textual examples, though they emerge clearly from consultations with native speakers. The following two examples do bear indirectly on components (d) and (e), however. In (32) the protagonist is a village girl who has, for want of money, fallen behind in her studies at a college in the city. The sight of her books reminds her both of difficulties and of her obligations. (32)

Buku-buku berselerakan di meja belajarnya rasanya sangat menyusahkan hati. 'The books scattered over her desk made her feel heavy-hearted (susah hati).'

Example (33) is advice being given to a girl who is having problems coping with the task of looking after her two mischievous younger siblings. The author (an advice columnist, Kak Nor) assures the girl that the major responsibility for disciplining the kids lies with the parents. It is enough for her, as older sister, to give her younger siblings advice to the best of her ability. The essence of Kak Nor's counsel is to release the troubled young

182

CliffGoddard

writer from her state of susah hati by releasing her from the pressure of sole responsibility. (33)

Oleh yang demikian adik jangan susah hati. Memadai dengan memberi nasihat ala kadar. 'Given that this is the case, you (lit. younger sibling) needn't be too troubled (susah hati). It's enough to give (them) advice to the best of (your) ability.'

Senang hati The expression senang hati, roughly 'contented, having peace of mind', has been identified by some commentators as designating a traditional Malay social ideal. Djamour (1965: 145-6) says it would be difficult to overemphasise its importance: "it permeates all fields of human behaviour. When a Malay contemplates a change of residence, marriage, divorce, or the exercise of a profession, the primary consideration is... serenity of mind". Djamour and others (e.g. Maeda 1975) note that Malays would often rather disregard grievances or dissatisfactions than disturb their "peace of mind". I would advance the following explication, much of which is, in effect, the converse of susah hati. Whereas someone who is susah hati is concerned about possible misfortune (thinking 'something bad can happen to someone; I don't want this') and feels under pressure to act (thinking Ί have to do something'), the person who is senang hati is free from such concerns and pressure. There is also a positive dimension, stated in component (e). The person who is senang hati has a sense of being able to do what he or she wants. This cognitive state brings about a 'good feeling'.4 (34)

Person-X senang hati (a) for some time, X wasn't thinking something like this: (b) something bad can happen to someone (c) I don't want this (d) X thinks something like this: (e) I don't have to do anything now (f) if I want to do something I can do it (g) because of this, X feels something good

Hau: A key word in the Malay vocabulary of emotion

183

Some features of this explication can be justified by reference to the following naturally occurring examples. Firstly, one's senang hati may be threatened by a concern of any kind, whether focused on the experiencer himself or herself, or another person, typically a family member (or someone else one feels loyal toward or protective of). (35)

Anak rusa berasa senang hati kerana tuannya sentiasa menjaganya daripada dibunuh anjing-anjing. The young deer felt very secure (senang hati) because his master was always watching over him to protect him from being killed by dogs (from a children's story).'

(36)

Dia senang hati bila mendapat tahu anaknya sudah tiba di rumah sepupunya dengan selamat. 'He was relieved (senang hati) when he found out his child had arrived at his cousin's house safely.'

As suggested by these examples, and also by the following one, the expression senang hati is typically found in situations which focus on being free of particular concerns or troubles. (37)

"Kenapa kau kata begitu Nadia?" Sambil mengemas-ngemas rambut isterinya, Rizal cuba untuk menyenangkan hati Nadia. '"Why do you say such things Nadia?" As he stroked his wife's hair, Rizal tried to calm her (lit. calm her hati)'

When one speaks of wishing to bring about a state of senang hati in one's parents (a wish commonly expressed by young adults), what is being referred to is not only the wish that the parents be free of financial and other worries, but also that they should feel free of the various pressures and responsibilities which they have been subject to all this time as parents - they should now be able to do as they please. (Cf. also the expression senang hati, goyang kaki 'carefree at heart, swinging one's feet', which depicts the carefree attitude of someone who is free of the pressure to work.) (38)

Aku berkerja rajin-rajin untuk menjimatkan duit. Aku hendak membeli rumah untuk menyenangkan hati ibu baba. 'I'm working really hard to save up money. I want to buy a house to make my parents' lives easier.'

184

CliffGoddard

Notice that the expression tidakltak ('not') senang hau is often used as a euphemism to refer to emotions and reactions which are presumably more serious in reality than simply an absence of senang hau. There is nothing particularly surprising about this (much the same happens with the English word happy, as the translations below illustrate). Example (39) is addressed by a girl to a persistent, unwelcome suitor. In example (40) a divorcee is relating how people reacted when, after some years of financial hardship, she was finally able to afford to smarten up her image. (39)

Sungguh aku tak senang hati setiap detik melihat tingkah laku kau yang menjemukan itu. 'Really I'm not happy every time I see the tiresome way you carry on.'

(40)

Tetapi ada orang tidak senang hati, kalau orang perempuan, mereka mula memikirkan mungkin saya melaram untuk memerangkap suami mereka. Bagi orang lelakipula mereka berpendapat mungkin saya memang hendak menarik perhatian mereka. 'But there some people who aren't happy (with me), if they're women, they've begun to think that maybe I am showing off with a view to snaring their husbands. For the men, they think maybe I am trying to attract their attentions.'

Sakit hati The primary meaning of sakit is 'in pain, sick, ill', as in expressions like sakit kepala 'head ache' (kepala 'head') and sakit kuning 'diabetes' (kuning 'urine'), so on the basis of their English counterparts one might imagine that sakit hati is something like English sick at heart. But this would be mistaken: sakit hati may perhaps imply something like pain, but nothing like incapacity. If anything, it has a rather active flavour, suggesting the possibility of some kind of reprisal. An interesting interpretation of sakit hati comes from an old medical work (Ellis 1893) on the amuk syndrome, i.e., the insane 'killing frenzy' to which Malays occasionally fall victim. Ellis (cited in Winzeler 1990: 106) makes the observation that the mental state of a person preceding an attack was invariably described, by Malays, as sakit hati, a state which he describes as involving 'depression and brooding over wrongs or supposed wrongs'.

Hati: A key word in the Malay vocabulary of emotion

185

Various bilingual dictionaries give 'resentful, angry' (Coope 1991: 307), 'annoyed' (Haji Abdul Rahman 1994: 231), 'annoyed, irritated, resentful' (Awang and Yusoff 1990: 942). Definitions in monolingual Malay dictionaries tend to combine a reference to marah 'angry, offended' with tersinggung 'have one's feelings hurt'. For example: berasa marah atau tersinggung kerana dihina atau ditipu 'feel offended or hurt because being maligned or cheated' (Abdullah 1996: 265); berasa tersinggung atau marah kerana sesuatu sebab atau perbuatan orang lain 'feel hurt or angry because some reason or action of another person' (Zakaria 1988: 118). Notice that these two sources differ on whether the cause of sakit hau is typically or predominantly verbal (as implied by Abdullah's reference to being maligned or cheated), or whether the cause might be an action of another kind. According to the explication below, sakit hau designates a reaction to someone else's behaviour (typically but not necessarily verbal), behaviour which is seen as indicative of this person's lack of esteem for the experiencer. It is coming face to face with this unpleasant fact which causes an unpleasant feeling in the experiencer (and explains the affinity with being offended) and which provokes an urge to do something (typically but not necessarily retaliation). (41)

Person-X sakit hati (a) for some time X thinks something like this about someone else: (b) this person did (said) something (c) if this person thought good things about me, he/she wouldn't have done this (d) I want to do something to someone because of this (e) because of this, X feels something bad

Various features of this explication can be supported by the examples below. Examples (42) and (43) show typical contexts for sakit hati (and its verbal derivative menyakitkan hati). They describe a person's reaction to being maligned, or to being the butt of sarcasm or insinuation. In both cases the stimulus is a verbal event. Note, though, that the speech-act may or may not be addressed to the experiencer him or herself.

186

CliffGoddard

(42)

Paling menyakitkan hati dia tuduh kita curi duit dia, macamlah kita ini pencuri besar. 'What really makes us mad (causes sakit hati) was him accusing us of stealing his money, as though we were a terrible thief.'

(43)

Panas telinga Mira mendengar sindiran Razi. Tetapi, bersabarlah wahai hati, biarlah dia hendak menyindir atau menyakitkan hati aku, biarlah, asal dia bahagia. 'Mira's ears burnt listening to Razi's needling. But, be patient hati, let him if he wants to needle me or pain me (causes sakit hati), let him, as long as he's happy.'

Still with verbal stimuli, (44) is an interesting example. In this case, what provokes the protagonist's sakit hati is the way his rival is publicly praised by the manager. Although the speech-act in this case is not actually about the experiencer, it still reflects on his performance and on his standing in the eyes of his superior, the manager (if the manager thought him the equal of his rival, why not praise him at the same time?). (44)

Menurut pengarah, Hasnul seorang pekerja yang baik dan amanah. Lebih menyakitkan hatinya pula, apabila pengarah tanpa segan-silu memuji kebaikan Hasnul di hadapannya. 'According to the manager, Hasnul was a good and honest worker. That only further inflamed his hatred (of Hasnul), especially when the manager praised Hasnul in front of him.'

The following examples show that sakit hati can be induced by non-verbal events - by a social event such as divorce and its attendant complications, as in (45), or a physical event such as carelessly bumping into someone, as in (46). (45)

Lagipun saya rasa lebih baik bercerai waktu muda, hati tak sakit sangat, kalau dah tua nanti lebih menyakitkan hati. 'Besides I reckon it's better to divorce when you're still young, it doesn't make (one) so bitter, but if you're already old it makes one more bitter.'

(46)

Mohamad sakit hati melihat Mahmud yang melanggarnya tanpa sebab. (Zakaria Salleh 1998: 18) 'Mohamad was annoyed seeing that Mahmud had bumped into him without any reason.'

Hati: A key word in the Malay vocabulary of emotion

187

Puas hati The expression puas hati designates a good feeling resulting from someone's behaviour, when this behaviour is something that the experiencer has wanted for some time. (47)

Person-X puas hati (a) X thinks something like this about someone: (b) this person did something (c) for some time, I wanted this person to do this (d) because of this, X feels something good

Typical examples involve a family member, employee, or boss fulfilling expectations, as in examples (48)-(50) below. However, one can also be puas hati in relation to one's own behaviour; for example, if one has managed to achieve something one has long strived for, as in (51). (48)

... apa yang say a buat tak kena pada padangannya. Hinggakan cara saya gosok bajupun dia tak puas hati dan tak kena pada matany a. '...nothing I did was any good in his eyes. Even to the point where the way I ironed a shirt didn't please him and wasn't good enough.'

(49)

Dia berasa puas hati dengan kerajinan Jali. Orangnya agak lurus dan raj in. 'She felt pleased with Jali's diligent (work). He seemed an honest and hard-working sort of person.'

(50)

"Cakaplah! Cakap apa sahaja yang kamu semua tidak puas hati!" sambung Tuan Setiausaha lagi. '"Speak up! Say whatever, whoever isn't contented!", added the Chief Secretary.'

(51)

Saya tidak puas hati hingga saya menyelesaikan kerja saya. Ί won't be satisfied until I complete my work.'

Generally speaking, the action or behaviour which gives rise to the puas hati reaction can itself be described as good - but not always. In example (52), the speaker is expressing her pleasure at the suffering of her rival. The explication is worded so as to accommodate such uses; it does not describe the person's behaviour as, so to speak, good in itself, but merely as something which the experiencer wants.

188

(52)

Cliff Goddard

Biar dia cemburu, biar dia marah, biar dia tak tidur malam... biarkan. Barupuas hati aku! 'Let her be jealous, let her be angry, let her not sleep at night... so be it. Finally I'm content.'

Kecil hati When someone unexpectedly does something (typically, but not necessarily, says something) and this makes us realise that this person doesn't actually think much of us, the unpleasant feeling this induces is the feeling of kecil (lit. 'small') hati. (53)

Person-X kecil hati (a) X thinks something like this about someone: (b) this person did something (c) I didn't think this person would do something like this (d) I know now that this person doesn't think good things about me (e) X feels something bad because of this

Some of the key points about the above explication can be illustrated with the following examples. In (54) and (55) we see typical examples in which the triggering event is someone saying something which betrays their negative attitude towards us. Notice, though, that the speaker's dictum does not have to be explicitly about the experiencer (though it very often is). For example, in (55) Razi's words are certainly unkind and reflect a dismissive attitude towards Mira, but they are not directly about Mira. This example can also be used to illustrate another point, which is that the speaker need not be explicitly addressing the experiencer; if Mira had overheard Razi talking about her like that, she would feel kecil hati. These considerations explain why component (b) is phrased simply as 'this person did something' (and not, for instance, 'someone said something about me' or 'someone said something to me'). (54)

Saya rasa begitu kecil hati dan tak sanggup melihat mereka kerana setiap kali berjumpa mereka akan mengeluarkan perkataan yang bukan-bukan tentu juga kita rasa kecil hati. Ί feel so hurt and I don't want to see them because every time we meet they come up with false allegations (against me). Of course one (lit. we) is hurt.'

Hati: A key word in the Malay vocabulary of emotion

(55)

189

"Itulah sebabnya aku tanya kau! Adakah dia bercuti saja atau dah berhenti kerja. Sekurang-kurangnya kalau dia berhenti, hilanglah satu tugas aku untuk bertegang urat dengan dia!". Menggelegak dada Ezza mendengar jawapan Razi. Kalaulah Mira dengar katakata Razi itu, tentu kecil hati Mira. Kesian Mira. '"That's why I asked you! Has she just gone on a holiday or has she quit. At least if she's quit, it'll be the end of my job of tussling with her!" Ezza's blood boiled when she heard Razi's reply. If Mira had heard Razi say that, she'd have been so hurt. Poor Mira.'

Staying with the wording of component (b), it may be asked why it does not contain the specification that what was done or said was 'something bad'. The reason for preferring the more general phrasing is that kecil hati can result from actions which are not necessarily being seen as bad in themselves. For example, in (56) what Pak Ngah Russin did to cause the speaker's mother to feel kecil hati was to turn her out of his house, telling her that he was not willing to help her (despite being her neighbour and much better off than she). The same point is apparent from example (57). The key thing about the triggering stimulus is not that it is bad in itself but simply that it makes us realise that the speaker does not hold us in high regard after all; hence, component (d) Ί know now that this person doesn't think good things about me'. (56)

Nol, Pak Ngah Hussin tidak ben kita pijak tangga rumahnya lagi. Lagipun emak melarang kita ke sana. Kau tahu Nol, emak benarbenar kecil hati. 'Nol, Pak Ngah Hussin won't let us step foot in his house again. Plus, mother has forbidden us to go there. You know Nol, mother was really hurt.'

(57)

Ini mesti ada cerita best kat panggung ni, sikit pun tak nak sound aku... ini yang kecik hati aku ni. 'This (film) must be the best story in the cinema at the moment, but they didn't even say a word to me (about their plan to go)... that's what hurt me so.'

Example (57) raises another consideration which may seem problematical for the explication, because it shows that sometimes one can feel kecil hati on account of something which has not been said - in this case, the

190

Cliff Goddard

lack of an invitation to go along with friends to the movies. However, I don't think this really poses a significant problem. It is a very general fact about semantic explications that an omission to do something can itself count as doing something. That is, if an explication identifies 'someone did something' as the triggering event, then in practice this condition can be also satisfied by someone not doing something, as the result of a more general principle of construal. This concludes our semantic study of five emotive expressions formed with hati. It perhaps bears repeating that there are dozens of other, similar expressions.

4. Concluding discussion What lessons can be drawn from this brief inquiry into the Malay hati? One lesson is that apparent translation equivalents must be handled with great care. In a sense, this is a familiar adage, at least to those working in crosslinguistic semantics. Numerous studies have shown that seemingly "basic" English words such happiness, anger, fear, love, and pride do not have precise equivalents in other languages, and, conversely, that the apparently "basic" emotion vocabulary of other languages often resists straightforward translation into English. By and large, however, the studies which have demonstrated this fact have focused on primary (i.e. monolexemic) emotion terms. There is still a tendency, at least in some quarters, to assume that secondary emotion terms, especially those involving descriptive words such as 'sick', 'small', or 'soft', can be understood in a transparent fashion. The present study has shown that any such view would be mistaken. For example, the meaning of the Malay expression sakit hati (lit. sick/pained heart) cannot be deduced, on the basis of an English speaker's semantic intuitions, from the combination of sick and heart, e.g. sick at heart or heartsick. Similarly, it would be incorrect to assume that kecil hati designates either fear or stinginess, although these are the interpretations which the "translation" small heart tends to suggest to English speakers. I do not want to suggest that the meanings of "figurative" expressions like these vary without limit between languages. No doubt there are some cross-linguistic tendencies to be discovered; I doubt, for example, that in any language an expression corresponding to 'small heart' would designate anything like bravery, generosity, or joy. But to establish these tendencies,

Hati: A key word in the Malay vocabulary of emotion

191

what is needed is a series of detailed semantic studies of secondary lexemes for emotions (and related concepts) in a range of languages. A second lesson is that to arrive at an authentic picture of the Malay vocabulary of emotion, it is not enough to stop at the one or two dozen monolexemic terms, such as kasih 'care, love', marah 'angry, offended', malu 'shame, embarrassment', etc. (cf. Goddard 1996, 1997b). We have to come to grips also with the myriad of emotion expressions based on hati, which equip the Malay language with a versatile resource for describing subtle nuances of feeling. No doubt many languages have a similarly elaborate vocabulary of emotion; for example, Thai, cf. Diller and Juntanamalaga (1990). If we identify these languages and study these elaborated systems, we may expect new insights into the human "semantic potential", so far as emotion is concerned - just as by studying elaborated systems of positional and directional terms, for example, we can gain insights into human spatial abilities, and by studying elaborate classifier systems we can gain insights into categorisation. Finally, if our objective is to understand culture-specific perspectives on emotion (including gaining a better perspective on the idiosyncracies of the English emotion system), it will prove extremely valuable to move beyond focusing mainly or wholly on emotion expressions per se. Ethnopsychology is much broader than this. Wierzbicka (1992) has shown that studying the Russian key concept of duSa 'soul' opens up a uniquely Russian perspective on human nature. I have no doubt that studying the Malay key concept of hati can open up a uniquely Malay perspective on human nature, and, in particular, on the role of interpersonal feelings in the human condition.

APPENDIX: Semantic Primes - English and Malay Exponents Substantives

I-AKU, Ύου-ΚΑυ SOMEOWE-SESEORANG SOMETHING-SESi/ArC/ PEOPLE-ORANG

Time

WHEN/TIME-BILA/MASA(WAKTU) NOW-SEKARANG BEFORE-SEBELUM, AFTER-SELEPAS A LONG TIME-LAMA, A SHORT TIMESEKEJAP FOR SOME ΎΙΜΕ-BEBERAPA LAMA

192

Cliff Goddard

Mental predicates

Space

WANT-MA//t/, THINK-/?/™

WHERE /PLACE-D7MANA/TEMPA T,

KNOW-rA/f[7, FEEL-RASA

HERE-SINI

SEE-NAMPAK,

ABOVE-(£>/) ΑΓΑ5, BELOW-(/?/)

HEAR-DENGAR

BAWAH

Speech

INSIDE-O/) DALAM

SAY-ΑΓΑ ΓΑ

ON (ONE) SIDE -(£>/) SEBELAH

V/OKD-PERKA T A AN

NEAR-DEKAT,

FAR-JAUH

TRUE-BENAR

Logical concepts Actions, Events, Movement

BECAUSE-SEA4.B, NOT-TIDAK

OO-BUAT,

IF-KALAU,

HAPPEN-TERJADI

ΜΑΎΕΕ-MUNGKIN

MOVE-BERGERAK

CAN-BOLEH

Existence and Life

Evaluators and Descriptors

THERE IS-ADA

GOOD-BAIK, BAD-BURUK

LIVE-HIDUP, DIE-MA 77

BIG-BESAR,

Determiners and Quantifiers

Intensifier and augmentor

SMALL-KECIL

THIS-/7V7 THE SAME-( YANG) 5AMA,

MORE-LAG/

OTHER-LA/TV ONE-SATi/, TWO-Dt/A,

Taxonomy, Partonomy, Similarity

MUCH/MANY-BANyA/C

PART-BA

SOME-BEBERAPA,

LIKE-MACAM

A.LL-SEMUA

4G7A7V, KIND-/BN/5

Notes 1 . For helpful comments and advice I would like to thank Nick Enfield, Hazidi Abdul Hamid, participants in the Symposium on The Semantics of Emotions held at Australian National University in 1999, and the editors and reviewers of this volume. 2. Even though its anatomical sense is 'liver', hati is not altogether unconnected with the heart. For example, the verb debar 'beating of the heart' can be combined with hati, in the expression berdebar hati, roughly, 'to thrill with feeling, to have one's heart throb with feeling'. 3. This is not as straightforward an assumption as it may seem. The complicating factor is that one may form any number of causative expressions which have hati as ob-

Hati: A key word in the Malay vocabulary of emotion

193

ject, without the need for any comparable fixed expression. For example, one may form the expressions menggembirakan hati (from gembira 'happy, thrilled') and membahagiakan hati (from bahagia 'happy, content in life'). These look exactly parallel with menyenangkan hati but there are no fixed expressions *gembira hati and *bahagia hati, corresponding to senang hati. Nor (consequently) could you describe a person as *bergembira hati or *berbahagia hati, though one can describe a person as bersenang hati. In view of these facts, it is not necessarily the case that expressions like menyenangkan hati and memuaskan hati (for example) are based directly on the fixed expressions senang hati andpuas hati. Even so, semantically the alignment does seem to hold; that is, menyenangkan hati means 'to make someone senang hati', memuaskan hati means 'to make someone puas hati'. 4. The meaning is different if senang 'at ease, comfortable' is applied attributively; for example, Hati saya tak berapa senang 'My hati isn't very at-ease'. This usage suggests the transient state of the experiencer's hati, and would be appropriate in reporting a reaction to an event. For example, in (a) below the protagonist is being faced with unwelcome questioning from an older person (questions which, given Malay cultural values, could not be ignored). In example (b), the protagonist is a child. The disturbance to his senang hati is induced by his joining Koran-reading classes for the first time, and encountering his new strict teacher Tuan Syed. (58) a. "Siapa? Saya?" tanya Hilmy, memperhatikan muka orang lua itu tepattepat. Hatinya tidak begitu senang ditanya-tanya begitu waktu sedang sibuk-sibuk bekerja. ' "Who? Me?" retorted Hilmy looking fixedly at the face of the old man. He didn't feel too comfortable being questioned while he was busy working.' b. Hati teman tak berapa senang juga melihat Tuan Syed buat kali pertamapetang tu. Ί was not very happy to see Tuan Syed for the first time that afternoon.'

References Abdullah Hussain 1990 Kamus Simpulan Bahasa. Kuala Lumpur: Dewan Bahasa dan Pustaka. Awang Sudjai Hairul and Yusoff Khan 1990 Kamus Lengkap. Petaling Jaya: Pustaka Zaman. Coope, A. E. 1991 Malay-English, English-Malay Dictionary. (Revised edition). Kuala Lumpur: Macmillan. Diller, Anthony V. N. and Preecha Juntanmalaga 1990 "Full hearts" and empty pronominals in Thai. Australian Journal of Linguistics 19(2): 231-256.

194

Cliff Goddard

Djamour, Judith 1965 [1959] Malay Kinship and Marriage in Singapore. London: The Athlone Press. Goddard, Cliff 1995 'Cognitive mapping' or 'verbal explication'?: Understanding love on the Malay archipelago. Semiotica 106(3/4): 323-354. 1996 The "social emotions" of Malay (Bahasa Melayu). Ethos 24(3): 426-464. 1997 a Cultural values and "cultural scripts" of Malay (Bahasa Melayu). Journal of Pragmatics 27(2): 183-201. 1997b Contrastive semantics and cultural psychology: 'Surprise' in Malay and English. Culture and Pyschology 3(2): 153-181. In press The syntax of universal semantic primes in Malay (Bahasa Melayu). In: Cliff Goddard and Anna Wierzbicka (eds.), Meaning and Universal Grammar: Theory and Empirical Findings. Amsterdam: John Benjamins. Goddard, Cliff and Wierzbicka, Anna (eds.) In press Meaning and Universal Grammar: Theory and Empirical Findings. Amsterdam: John Benjamins. Haji Abdul Rahman bin Yusof 1994 Bahasa Malaysia-English, English-Bahasa Malaysia Dictionary. (Revised edition.) Glasgow: Harper Collins. Heider, Karl G. 1991 Landscapes of Emotion. Mapping Three Cultures of Emotion in Indonesia. Cambridge: Cambridge University Press. Karim, Wazir Jahan (ed.) 1990 Emotions of Culture. A Malay Perspective. Singapore: Oxford University Press. Maeda, Narifumi 1975 Family Circle, Community, and Nation in Malaysia. Current Anthropology 16(1): 163-166. Matisoff, James A. 1986 Hearts and minds in Southeast Asian languages and English: An essay in the comparative lexical semantics of psycho-collocations. Cahiers de Linguistique Asie Orientale 15(1): 5—57. Oey, Eric M. 1990 'Psycho-collocations' in Malay: A Southeast Asian areal feature. Linguistics of the Tibeto-Burman Area 13(1): 141-158. Peletz, Michael G. 1996 Reason and Passion: Representations of Gender in a Malay Society. Berkeley: University of California Press.

Hati: A key word in the Malay vocabulary of emotion

195

Wierzbicka, Anna 1992 Semantics, Culture, and Cognition: Universal human concepts in culture-specific configurations. New York: Oxford University Press. 1996 Semantics: Primes and Universals. Oxford: Oxford University Press. 1997 Understanding Cultures through their Key Words: English, Russian, Polish, German, Japanese. New York: Oxford University Press. 1999 Emotions Across Languages and Cultures: Diversity and universals. Cambridge: Cambridge University Press. Wilson, Peter J. 1967 A Malay Village and Malaysia. New Haven: Hraf Press. Winzeler, Robert. 1990 Amuk: Historical, Psychological, and Cultural Perspectives. In Wazir Jahan Karim (ed.), 96-122. Zakaria Salleh 1988 Simpulan Bahasa Praktik. Kuala Lumpur: Berita Publishing.

Talking about anger in Central Australia Jean Harkins

A search for better understandings of human emotions, whether in general terms or in specific social and cultural contexts, is inevitably faced with issues of ethnocentrism and monolinguocentrism (the problem of privileging one language, such as English, over others; Enfield 2000: 150). Even the notion of "emotions" or human feelings as a topic of investigation, as a category of things that people would want to know more about, and to read about in this and other volumes, arises from a particular cultural perspective or perspectives. While methods of study and investigation vary across academic disciplines, the activities of collecting data, observing, experimenting, discussing, reading, writing and so on are, themselves, cultural practices. This may seem so obvious as to be unworthy of much discussion. But if we are making serious attempts to describe emotions in terms of a language-independent and culture-neutral metalanguage, as most of the studies in this volume do, we need to consider that this is a culturally situated activity. Even if our metalanguage can be culture-free, when we engage in constructing definitions or explications, trying to arrive at precise statements of meaning that represent the necessary and sufficient conditions for the use of a linguistic expression, formulating and discussing semantic components, using various oral and written means - aren't all of these practices specific to certain cultures, and indeed, subcultures? Aren't the resulting definitions or explications inevitably cultural artefacts? And could the aim of culture-independent semantic description thus be somewhat misguided, if not wholly chimerical? Such big questions certainly couldn't be answered in a single, small study. Only as a variety of methods of investigation and description are carried out in a wide range of linguistic and cultural contexts, by people of diverse linguistic and cultural backgrounds, can we get a better idea of which ways of examining and describing emotions really work across cultures. Moreover, diversity within cultural groups should never be underestimated. Even if broad social or cultural trends promote, or discourage, thought and discussion about emotions (or language, or any other topic),

198

Jean Harkins

there will always be individuals and groups who have lesser or greater interest in such matters than their fellows. This exploration of a small group of emotion expressions in the Arrernte language of Central Australia1 takes the view that it is useful and illuminating, when investigating emotional or other meanings, to cultivate awareness of the cultural and intercultural dimensions of the enterprise, and to see both the processes and outcomes of the investigation in this light. Aiming for complete scientific objectivity in a culturally situated study would be not only foolish, but less informative than taking into account the cultural and linguistic realities that present themselves in the course of such investigation. However, aiming for interculturally comprehensible ways of expressing what people mean is both practicable and worthwhile.

1. Talking about people's feelings It cannot be assumed that discussion and investigation of emotions as a category, or of people's feelings in general, will be considered interesting or appropriate in every cultural or social setting. The idea of looking closely at emotion talk in Arrernte had some of its origins in earlier work on concepts of "shame" in educational contexts (Harkins 1990, 1996), an issue of considerable interest to Arrernte teachers and parents. Further examination of the language of feelings was proposed partly as a further contribution to understanding emotional interaction in similar contexts, and partly as a semantic field in which it would be possible to experiment with formulating definitions or statements of meaning in the Arrernte language itself, without using English or outside technical terminology to define Arrernte words. The Eastern and Central Arrernte to English Dictionary (Henderson and Dobson 1994) is rightly hailed as one of the best works of its kind ever compiled, but some Arrernte speakers would hope also for definitions of Arrernte words framed in the Arrernte language itself. A central contention within the NSM theory of semantic primes is that every human language contains a subset of basic, indefinable terms that can serve as the language's own definitional metalanguage, and hence, that words in any language can be defined in that same natural language, without resort to either a foreign or an artificially constructed metalanguage (Wierzbicka 1996b). Crosslinguistic research has provided evidence in support of this idea (Goddard and Wierzbicka 1994 and Forthcoming;

Talking about anger in Central Australia

199

Goddard 1997), and explications have now been formulated in a good many languages other than English, particularly in Goddard and Wierzbicka's (Forthcoming) book on meaning and universal grammar. Working with Chinese, Zhengdao Ye (in this volume) has found the process of translating definitions into the language under investigation and back to English, often more than once, very helpful in clarifying and refining the semantic formulae. However, this Arrernte study may be one of the first in which a group of speakers has formulated statements of meaning in their own language, without proceeding via English. The domain of emotion talk, or words and other linguistic expressions used in talking about people's feelings, is a fruitful area for exploration of this kind, provided such exploration can be approached in a culturally appropriate way. The semantic properties of emotion expressions are well enough explored in a wide range of languages for us to know that they can usually be defined in explications of manageable length, and their semantic content often includes many extremely interesting elements, such as prototype structures; hypothetical, conditional and causal reasoning; affective, bodily and behavioural components; and semantic links to related cultural concepts, values and Schemas. The Arrernte language has a large vocabulary of words and phrases related to emotional feelings, and rich discourses about people's feelings and their effects on behaviour and interaction. This proved to be a fascinating and highly appropriate area for discussion and exploration among a group of mature professional women. The research participants were three firstlanguage Arrernte speakers living in Alice Springs, all with some experience in bilingual education and Arrernte language work: Margaret Heifernan, Rosie Ferber and Anne McMillan; and myself, a learner of Arrernte who was designated scribe for the group because of an ability to write rapidly in the language. These speakers are literate in the language, and Margaret Heffernan as the senior participant checked all the data that we recorded. Most of our research and discussions were carried out over four weeks in 1997, during which time the participants consulted other Arrernte speakers as well as drawing upon their own knowledge. While there is no word for "emotion" as a category in Arrernte, there is, as Wierzbicka (1999: 276) predicts, a word for "feel", the intransitive verb awelhe-(me) 'feel-(NPp)'2 (see also Harkins and Wilkins 1994: 290). The etymology of this verb, from awe- 'hear' and reflexive -Ihe, although fairly transparent, is not particularly salient for contemporary speakers (just as

200

Jean Harkins

the etymology of emotion, from e- Outward' and motion, isn't salient for most English speakers). Awelheme 'feeling' is intuitively related more to sensations in the middle part of the body than the ears, despite the attractiveness of the image of listening to one's own inward state (cf. also Goddard 1989: 54). Participants referred to the topic of our discussions as angketye tyerrtye awelhe-me akerte 'word people feel-NPP PROP', which could be glossed as 'words for people's feelings', or more formally, 'linguistic expressions of emotion'. Awelhe- 'feel' refers to physical sensations such as heat, cold, pain, as well as to emotional or cognitively-based feelings, but not to intentionally feeling an object (cf. the transitive verb anpe-me 'touch-Npp'). "Feeling something good/bad" can be physical or emotional: (1)

a.

b.

Ay enge akurne awelhe-me. IsS bad feel-NPP Ί feel (something) bad.' Ayenge mwarre awelhe-me. IsS good feel-NPP Ί feel (something) good.'

Taken out of context, these expressions can be interpreted as purely physical or purely emotional, or a more general bad or good inner feeling, while in contexts like "feeling bad because I'm ill" versus "feeling bad because you won't talk to me" the physical or emotional interpretation is unambiguous. Thus awelhe- is clearly equivalent to the NSM prime FEEL. Over several days, the group collected over eighty Arrente words and phrases for various emotional feelings. It quickly became apparent that this collection was just the tip of a vast iceberg of expressions and discourse related to emotions, all of which would be worthy of extensive study (cf. Wilkins' outstanding 1986 analysis of some emotion-related Arrernte particles). With limited time for our work, we collected three or more examples from people's everyday conversation for each of 55 words and phrases, and made rough notes on their semantic content, before narrowing our focus for a detailed analysis of meanings in just eight words. We considered whether the expressions collected could be arranged into subgroups of related feelings, and found that it was easy for speakers to arrange them intuitively into seven groups. These are shown below, with some typical members (from various word-classes) of each subgroup. The

Talking about anger in Central Australia

201

rough glosses give some indication of the meanings, but for many of these words there is no adequate English equivalent. (2)

a.

b.

c.

d.

e.

f.

g.

Feeling good (mwarre awelheme) akangkeme 'happy, having a good time' ilkelheme 'enjoying' atnewe-atnewe 'relaxed, calm' atneratye 'trust, confidence in someone' Wanting something (ahentye-aneme) ahentye 'wanting, liking' alhirreme 'admiring, fancying' kwakwe 'keen' ingketye 'desirable' Feeling bad (akurne awelheme) because something happened alhwarrpe 'sadness, sorrow, grief irrare 'loneliness, yearning' irrareltheme 'depressing' Feeling bad towards someone else ahe 'anger' ahakngerre 'aggressive' akiwarre 'infuriated' arnkelye 'resentful' ingkerte-ingkerte 'jealous' Feeling bad about something akurne awelheme 'regretting' apurrke-apurrke 'bored, exasperated' ilkelhetyakenhe 'uncomfortable, fed up' Feeling bad because something bad could happen akngerre anthurre itirreme 'worrying' atere 'fear' apure 'shame' Feeling because you didn't know something apateme 'astonished, confused' atnulkeme 'shocked, surprised'

Each of these categories has many more associated words and phrases, and each could be the focus of an extended study in cultural semantics. Our

202

Jean Harkins

preliminary analysis considered just a few of the good feelings (2 a), and anger-like feelings (2 d) which are the focus of the present discussion. Wilkins (To appear) has examined atere and apure (2f) in depth, and Harkins (1995; Harkins and Wilkins 1994: 291) discusses ahentye 'want'.

2. Angry words in Arrernte Anger-like feelings and their interpersonal manifestations are a frequent and lively part of social and emotional life for a great many Arrernte speakers. These feelings arise from transgressions of social and cultural norms and relationship boundaries, and from severe pressures imposed by invasion of and damage to Arrernte country by outsiders. English lacks precise equivalents or ways of expressing most of them, with the dual outcomes of frequent intercultural misunderstanding in difficult situations, and significant semantic differences between some Aboriginal English words and their non-Aboriginal English counterparts. Although this analysis deals mainly with Arrernte, Aboriginal English and other local languages are mentioned where relevant to the general topic of talking about anger in this Central Australian context. Goddard's (1991) analysis of anger-like words in the immediately adjacent Western Desert language provides a particularly interesting comparison, because these words and concepts are quite language-specific and do not match up exactly even between these two closely linked languages and cultural groups, despite frequent interaction and kinship bonds between them. As a first example, consider the following examples from Arrernte narrator Willie Rice (1986: 28-31), who provided the English glosses in consultation with his daughter Rosie Ferber. He refers here to two groups of Arrernte people who encountered the first non-Aboriginal visitor to the Alice Springs area. (3)

Artwe itne ne-ke Iwepetheke-arenye mape, man 3pS be-pc Jay.Creek-ASSOC PL(g) artwe mwarre mape, ahe-le-irre-tyekenhe. man good PL(g) anger-ADV-iNCH-VbNEG 'These men belonged to Jay Creek and they were good men, not angry.'

Talking about anger in Central Australia

(4)

203

...itnenhe are-tye.lhe-ke re. 3pO see-oo&DO-pc 3sS line artwe mperlkere ikwere ahe-le-irre-tyekenhe. 3pS man white 3SDAT anger-ADV-iNCH-VbNEG 'He met up with them but they weren't cheeky to him.'

Here we see the same word ahele-irre- (with negative verbal suffix) translated using the English words angry and cheeky respectively. This reflects the Aboriginal English usage, where hostile could be a closer non-Aboriginal English equivalent. In Aboriginal English, cheeky often refers not to casual insolence but to a serious potential for violence or harm (Harkins 1994:155) - a semantic difference that can produce significant intercultural misunderstanding in situations of actual or potential interpersonal conflict. This is probably the most frequently occurring anger-like word in Arrernte, with a fairly broad semantic scope. While the base, ahe- 'anger', is almost certainly etymologically related to 'throat' (as is ahentye 'want', Harkins 1995: 46-54), synchronically its semantic content doesn't include any body part reference (nor does the probable derivation of English anger from Old Norse 'narrow' form part of its current semantic composition). The multifunctional inflectional suffix -le has ergative, locative, instrumental and other functions in addition to its function here as a manner adverb formative (Wilkins 1989: 176). This is followed by the inchoative derivational suffix -irre, whose occurrence with many emotion predicates (including ahentye 'want', apure 'shame', at ere 'fear' and others) may signal a conceptualisation of them as processes; this matter requires some further study (cf. Harkins 1995: 51, Wilkins To appear). Along with this word, four other anger-like words were selected for semantic comparison and analysis. These are listed below, with English glosses from Henderson and Dobson (1994): (5)

a. b.

c.

ahele-irreme 'get angry, be angry at someone' arnkelye 'not talking to people because you're unhappy about something; sulking, in a sulk' akiwarre 'showing anger; in a tantrum, sulking, having lost your temper, pissed off, flying off the handle'

204

Jean Harkins

d.

e.

ahakngerre 'aggressive, angry, violent, tough, mean, savage, vicious, dangerous, touchy. In local Aboriginal English, "cheeky" has this meaning, and not its standard English meaning.' atnerte ampeme ' 1. have a burning feeling in the stomach, usually from hunger. 2. be really angry.'

The English glosses give an indication of the various ways these words tend to be translated in different contexts; all of them are sometimes translated as angry, but obviously there are important differences in meaning among them. The Aboriginal English cheeky is associated here with ahakngerre, while in (4) above it was used to translate ahele-irreme. In (5 b) and (5 c), sulk is also used in its Aboriginal English sense (cf. Harkins 1994: 157). Non-Aboriginal English speakers can fail to appreciate the seriousness of this emotional state, because in non-Aboriginal usage sulking tends to be viewed as rather petty and self-indulgent, something that a person can and should fairly easily "get over", while arnkelye is more like a prolonged state of brooding resentment, fuming over some significant offence, and carries a real danger of erupting into violence. The usual linguistic method of collecting examples of each word in a variety of natural utterance contexts was followed, aiming for at least twenty examples of each in order to observe its range of use as broadly as possible, and to consider what elements of meaning were common to all uses of the word. Some examples, like (3) and (4) above, are seen in written Arrernte texts, but most of the available writings were factual or eventoriented narratives, containing less emotion vocabulary than is found in everyday verbal interaction. Ordinary family and community life, on the other hand, provided a wealth of examples every day. Often, more than one of the words under investigation were found in the same utterance, indicating that each contributes some meaning more specific than a generalised angry feeling, for example: (6)

Nhakwe ame akiwarre arnkelye-irre-me-le that Q infuriated resentful-iNCH-NPP-ss ahe-ke-athathe ane-rlane-me, anger-DAT-before be-coNT-npp

Talking about anger in Central Australia

205

atwe-tyenhe-le-athathe re ane-pe-ane-me? hit-Npc-ss-before 3sS be-FREQ-RDP-NPP 'Is that person building up anger inside, heading for a fight?' The picture here is of a person who is really "pissed off' (akiwarre) at someone for something they've done, fuming (arnkelye) continuously over it and working herself up into a real temper (ahe), likely to be vented in a fight (i.e. hitting somebody).

3. Scenarios, scripts and Schemas As the examples were assembled, it soon became apparent that each one stemmed from some real life incident, a social context or scenario in which the emotion was triggered by some event or action. Moreover, the emotional response - whether a person felt akiwarre or arnkelye or ahele-irreme or atnerte ampeme or more than one of these feelings, or whether on the other hand they didn't get angry or upset over it - depended largely on how the person "took" the situation, that is, how they interpreted it, or thought about it. Trying to tease out the fairly subtle differences in meaning and use between arnkelye, akiwarre and ahele-irreme, in particular, thus provided a startling degree of support for the view that these emotions are cognitively based, from the viewpoint of Arrernte social perception as well as that of NSM semantic theory. Some examples highlighting the nature of each of these three words may be helpful here (recalling also those given above). The translations are those provided by the contributors of the examples, containing some of the Aboriginal English usages mentioned earlier. (7)

Ahe-ke-athathe ame unte arnkelye-irre-me? anger-DAT-before Q 2sS resentful-iNCH-Npp 'Are you getting sulky so you can get angry at me?'

(8)

Re atherre untye-me-ke-untye-me ane-rre-me 3S DU face.away-NPP-DAT-RDP sit-RECip-NPP angke-rre-tye + arlke + akenhe, talk-RECiP-Vb+too+NEC arnkelye akwele atwe-rre-rne-arle + iperre. resentful QUOT hit-RECiP-P.imm-after

206

Jean Harkins

'The two of them are sitting facing away from each other and not talking to each other. They're sulking because they had a fight.' (9)

Re akiwarre-irre-me arelhe arrpenhe-le 3sS infuriated-iNCH-NPP woman other-ERG renhe terangke ake-ke-le-iperre. 3sO drunk call-pc-ss-after 'He lost his temper because that woman said he was a drunk.'

(10)

Ayenge ngkwenge akiwarre ake-lhe-me-le ilelhe-me! IsS 2SDAT infuriated call-REFL-Npp-ss admit-NPP 'I'm telling you I'm really pissed off with you!'

(11)

Ahe-le-irre-tyele-aye, urreke imp-0-aye! anger-ADV-iNCH-NegiMP-EMPH later leave-iMP-EMPH 'Don't get angry, leave it for later!'

(12)

Artwe ampwe nhakwe ahe-le-irre-me man old that anger-ADV-iNCH-Npp renhe arle ampe urreye mape-le 3sO THAT child male PL(g)-ERG ingwe-le apwerte-le we-ke-nge. night-LOC stone-iNST throw-pc-ABL 'That old man there is really angry because some young boys threw stones at him last night.'

As each example and the real life scenario it came from - the ayeye 'story' behind it - were discussed, the participants were able to make some generalisations about what all these situations had in common. For each word, these took the form of ayeye akweke angketye nhenhe-ke 'story small word mis-DAT', that is, a little story, script or mini-narrative for each word. These ayeye akweke, arrived at by discussion among the speakers, show a surprising degree of convergence with the kinds of semantic explications produced through more formal NSM analysis. Perhaps even more surprisingly, participants wanted to use false initials to represent the protagonists in each scenario and ayeye akweke, for reasons of anonymity. In a close-knit community, it could be undesirably obvious to any reader of our examples which real individuals, family and friends, had been involved in each of the disputes that produced the examples, and we could be accused of gossiping about these people instead of getting on with our linguistic work. Hence the familiar X and Υ of other

Talking about anger in Central Australia

207

NSM explications came into play, though for reasons that had more to do with social realities than with issues of indexicality in semantic representation (cf. Introduction to this volume, p. 20). Thus, a tentative ayeye akweke for arnkelye-irreme took the following form: (13)

X arnkelye-irre-me 'X is getting arnkelye' a. X re awelhe-me-le itirre-ke Y-ke alakenhe X 3sS feel-NPP-ss think-pc Y-DAT thus 'X is feeling something because X thought like this about Y' b. Y-le akenge mpware-ke-nge Y-ERG rotten make/do-pc-os Ύ did something very bad' c. the akurne ile-tyenhenge Υ renhe ikwere-nge IsA bad do-SBSQ Υ 3sO SSDAT-ABL Ί will do something bad to Υ because of this' d. X alakenhe itirre-me-le, X akurne awelhe-me Y-ke X thus think-NPP-ss X bad feel-ΝΡΡ Y-DAT ' when X is thinking like this, X feels something bad towards Y'

This is a very preliminary attempt at identifying the relevant semantic components; it may be far from a precise statement of the necessary and sufficient conditions for the use of the word arnkelye. Nevertheless, it is an important first step toward Arrernte language definitions. It indicates that an Arrernte understanding of the relationship between thoughts and feelings is far from incompatible with the cognitive perspective of NSM explications for other languages: component (a) expresses a concept of the experiencer's feeling something as the result of her own thoughts. This component also indicates that arnkelye is directed specifically at some other person (Y), who is the target of the experiencer's resentful feeling. It is an essentially interpersonal emotion, as confirmed in the last component (d). Components (b) and (c) identify the most typical thoughts of someone experiencing arnkelye: a strongly negative evaluation of the offender's action, and a burning desire for retaliation. Notice that the offending action doesn't have to be something that personally affects the experiencer of the arnkelye; of course this is often the case, but one can also feel arnkelye about other kinds of very bad deeds.

208

Jean Harkins

An equally tentative ayeye akweke for akiwarre has a similar format and is also directed at a specific offender (Y), but has a slightly different set of typical thoughts, indicating that the experiencer is personally affected by the offending action, is indeed too angry for words, and intends to retaliate, though not as immediately as in the case of arnkelye: (14)

X akiwarre 'X is akiwarre' a. X re awelhe-me-le itirre-ke Y-ke alakenhe X 3sS feel-NPP-ss think-pc Y-DAT thus 'X is feeling something because X thought like this about Y' b. Y-le akenge mpware-ke-nge atyinhe-ke Y-ERG rotten make/do-pc-os ISPOSS-DAT Ύ did something very bad to something of mine' c. ayenge ahentye kwenye angke-tyeke nhenhe ikwere-nge IsS want NomNEG say-PURp this SSDAT-ABL Ί don't want to say anything because of this' d. the urreke akurne ile-tyenhe Y renhe ikwere-nge IsA later bad do-Npc Y 3sO SSDAT-ABL Ί will do something bad to Y later on because of this' e. X alakenhe itirre-me-le, X akurne awelhe-me Y-ke X thus think-Npp-ss X bad feel-ΝΡΡ Y-DAT 'when X is thinking like this, X feels something bad towards Y'

The 'something of mine' in (b) can be one's personal property (e.g. mutekaye atyinhe 'my car') one's family members (e.g. akngeye atyinhe 'my father'), one's reputation or feelings (examples 9-10 above) - anything that the experiencer values enough to feel infuriated or severely aggrieved by some offence against it. This feeling is often "bottled up" (component c, which should perhaps also include anteme 'now'), and vengeful impulses (d) can persist inwardly or find expression in a show of temper. Interestingly, although the evidence from (5 b) and (8) above suggests that not talking is also associated with arnkelye, these speakers felt that it was an essential element of akiwarre in a way that it wasn't for arnkelye. This is an example of the kind of issue that may need to be explored further before optimal definitions are arrived at. The connotation of uncommunicativeness probably also accounts for the use of the English word sulk in relation to these much less petty emotional states; an explication for the Aboriginal

Talking about anger in Central Australia

209

English sulky might be closer to these Arrernte meanings than to the nonAboriginal English definition. Another example scenario or ay eye 'story' also illustrates the kind of contrast reflected in the ayeye akweke atherre 'two little stories' above: (15)

Re atherre ilte-rre-ke TV-ke arrkngele. 3S DU argue-RECiP-PC television-DAT reason Marie re arnkelye arrate-le alhe-ke, female 3sS resentful go.out-ss go-PC kele re kethe-le anteme ane-tyeme. then 3sS outside-LOC now sit-pp Kele anewe ikwere akiwarre-le then spouse SSPOSS infuriated-ss apmere-nge alhe-ke akartne-nge. home-ABL go-PC area-ABL 'The couple were arguing over the TV. The wife went out in a sulk and was sitting outside. Then her husband stalked off in a really bad temper.'

The third ayeye akweke, shown below, indicates that ahele-irreme is also directed at a target person (a) who has done something that negatively affects something important to the experiencer (b). Alhough the offending action is not as strongly condemned (akurne 'bad' cf. akenge 'very bad, rotten'), it is still unwanted (c). There is a desire for retaliation (d), but not as concrete an intention as suggested in (13 c) and (14d) above. (16)

X ahele-irreme 'X is getting angry' a. X re awelhe-me-le itirre-ke Y-ke alakenhe X 3sS feel-NPP-ss think-pc Y-DAT thus 'X is feeling something because X thought like this about Y' b. Y-le atyinhe akurne ile-ke Y-ERG ISPOSS bad do-pc Ύ did something bad to (something of) mine' c. Y-ke ay enge ahentye kwenye alakenhe Y-DAT IsS want NomNEG thus mpware-tyeke atyinhe-ke make/do-PURp ISPOSS-DAT Ί don't want Υ to do (like) this to something of mine'

210

Jean Harkins

d.

ayenge ahentye ane-me the akurne ile-tyeke IsS want be-npp IsA bad do-purp Ύ-0 renhe ikwere-nge Y-ABS

e.

3SO

3SDAT-ABL

Ί want to do something bad to Υ because of this' X alakenhe itirre-me-le, X akurne awelhe-me Y-ke X thus think-NPP-ss X bad feel-NPP Y-DAT 'when X is thinking like this, X feels something bad towards Y'

Interestingly, it appears that all three of these anger-like feelings are inherently interpersonal, directed at some specific offender. Participants insisted that Y-ke Ύ-DAT' was a necessary element in the first and final components of each of the three ayeye akweke. A non-personal grievance might be more likely to be spoken of in terms of akurne awelheme (2 e above), something like 'regret'. This is also in accord with speakers' intuitive categorisation of bad feelings towards someone (2 d) in a separate group from feeling bad about something (2e) or as the result of some event (2c). The Arrernte components proposed here differ in some important ways from the anger-like emotions analysed by Goddard (1991, 1998), particularly in focussing negative judgement on the offenders' actions rather than their character, although there are equally important similarities, such as the interpersonal emphasis (Y-ku Ύ-DAT'), and the desire for retaliation in mirpanarinyi but not kuyaringanyi. Space does not permit a full analysis here of ahakngerre or atnerte ampeme, but we can observe that ahakngerre, somewhat like the Western Desert pikaringanyi (Goddard 1991,1998), seems not to be an emotional feeling but a general propensity to do something bad to someone. The fact that ahakngerre, like Aboriginal English cheeky, can be predicated of non-humans and inanimates, such as snakes or poisonous plants, raises an interesting problem of whether or not this should be represented as a metaphorlike "wanting to do something bad to someone", or in terms of the possibility that the person or thing can or will do something bad to someone. And the emotional sense of atnerte ampeme, lit. 'stomach burning' (5e above) raises issues of semantic representation of emotions described via bodily sensations (Wierzbicka 1999: 295-297). Some of these issues are discussed further in other papers in this volume. The interpersonal feelings and transactions reflected in these examples and partial explications occur within a broader sociocultural context,

Talking about anger in Central Australia

211

against a rich background of culturally situated experiences, beliefs, values, attitudes and practices. For example, the basis for judging someone's action as akurne 'bad' or akenge 'very bad, rotten' are only partly a matter of personal opinion; they are strongly linked with wider cultural values and attitudes about what behaviour is and isn't acceptable from others, what is or isn't a potential trigger for each of these anger-like feelings. The 'script' for getting angry or akiwarre etc. is made explicit in the statement of the word's meaning; in this way the NSM approach (Wierzbicka 1972, 1999 etc.) is compatible with but more definitionally rigorous than, for example, Lakoff and Kövecses' (1987: 213) anger scenario or Frijda's (1986) eliciting conditions. But each emotion 'script' is also connected to other cultural concepts (for example, what counts as atyinhe 'mine' to an Arrernte person), and also to related discourses and recurrent patterms of cultural practice, or Schemas, such as how anger is expressed and how disputes are aired and resolved, or escalated (cf. examples 6-12 and 15 above). Some of these cultural concepts are lexicalised in the form of cultural key words (Wierzbicka 1997), while others can be identified in the form of 'cultural scripts' (Wierzbicka 1996 a, 1998, Forthcoming), and in terms of broader societal and cognitive Schemas as suggested in recent work on Aboriginal narrative discourse by Malcolm (1999) and Sharifian (2000) (cf. also Fillmore and Atkins' 1992 'frames'). These connections between and among cultural meanings can be investigated systematically in terms of similarities and overlaps between individual semantic components of the words, scripts, and perhaps also the schema representations: this dimension of cultural semantics has only just begun to be explored.

4. Practical semantics This small study has found confirmation for several of the hypotheses about emotional meanings put forward by Wierzbicka and other practitioners of the NSM approach to semantic analysis, and has raised some questions about other parts of this theory. But perhaps more importantly, it has demonstrated something of the practicality of the NSM approach in facilitating intercultural discussion and understanding of people's emotional life and behaviour in cultural context, and, furthermore, as a tool for stating meanings in the language of inquiry. This is a practical tool that field-

212

Jean Harkins

workers and research participants can use in their own language and cultural setting - a far cry from some popular notions of semantics as an abstract, ivory-tower activity. There was surprising convergence between these Arrernte perceptions and the NSM picture of emotions as cognitively based feelings, and the script-like nature of their meanings, with a series of typical thoughts giving rise to the affective state. These anger-like feelings all contain the impulse to act, proposed as a universal of anger-like feelings by Wierzbicka (1999: 288). The Arrernte ayeye akweke did not have the full prototype structure for cognitively based feelings ("sometimes a person thinks...", etc.) proposed by Wierzbicka (1999, see also our Introduction to this volume). A first attempt to apply such a frame rendered the ayeye akweke unintelligible in Arrernte, and it is not entirely clear how this problem could be resolved, or whether it would be appropriate to do so. The papers by Wierzbicka and by Goddard in this volume employ a similarly simplified frame for practical definitional purposes, and it may be that the full prototype structure is a little too abstract for workable natural language definitions. The outcomes of this modest study were of interest to the participants, but it was clear that such a process of group discussion might be an unmanageably time-consuming way to produce Arrernte language definitions for more than a handful of Arrernte words. Nonetheless, the intercultural insights gained through the use of this method of analysing meanings, particularly for important social and cultural concepts, were valued by all concerned. These outcomes are at least as important, indeed more so for most practical puposes, as the artefacts of semantic formulae that were produced. The fact that a group of interested speakers could gain some satisfying insights through practical experimentation with this methodology is promising, and suggests that the approach has much to offer in linguistic fieldwork and other intercultural communiation applications. For semantics to be of use outside of academic circles, it needs to offer something that fairly ordinary people with an interest in meaning and communication can do, and gain some benefit thereby. And for the academic study of meaning and communication to make progress, it needs to move beyond the hermeneutical circle of linguocentric discussion and representation of meanings. While some aspects of the NSM approach remain controversial, many issues in semantic analysis will only be resolved through the experiences of people analysing meanings in many different human languages and cultural environments, and thus discovering what is and

Talking about anger in Central Australia

213

isn't valid, what does and doesn't work in describing linguistic and cultural meanings. This study has shown just a little of how such investigation can proceed in a practicable way, and it is hoped that others may also find this kind of approach a useful and practical tool for learning more about emotional and other meanings within and across languages and cultures.

Notes 1. This study was conducted in Alice Springs (Mparntwe), among speakers of Eastern and Central (Mparntwe) Arremte (described by Green 1994, Henderson & Dobson 1994, Wilkins 1989), and was funded by the Australian Research Council through the University of New England. The participants wish to thank the Institute for Aboriginal Development, the Arandic Dictionaries Project, and Pamela Ditton for practical assistance that made this work possible. 2. Abbreviations used in morphemic glosses (mostly following Wilkins' 1989 analysis) are: 1, 2, 3[person]s[ingular]/p[lural]; ABLative; ABSolutive; ADVerb; Associative; Afgent]; CONTinuous; DATive; DS different subject; DUal; EMPHatic; ERGative; FREQuent; IMPerative; INCHoative; INSTrumental; LOCative; Neg[ative]IMPerative; NMZ nominaliser; Nom[inal]NEGator; NPC non-past completive; NPP non-past progressive; O[bject]; PC past completive; PLural(g[roup]); POSSessive; PP past progressive; PURPosive; P[ast].imm[ediate]; QUOTative; Q[uestion]; RDP reduplication; RECIProcal; REFLexive; SBSQ subsequent; SS same subject; S[ubject]; THAT complement marker; V[er]bNEGator.

References Enfield, Nicholas J. 2000 On linguocentrism. In: Martin Pütz and M. H. Verspoor (eds.), Explorations in Linguistic Relativity, 125-157. Amsterdam: John Benjamins. Fillmore, Charles H. and Beryl T. Atkins 1992 Toward a frame-based lexicon: The semantics of RISK and its neighbours. In: Adrienne Lehrer and E. F. Kittay (eds.), Frames, Fields and Contrasts. Hillsdale, NJ: Erlbaum. Frijda, Nico H. 1986 The Emotions. Cambridge: Cambridge University Press. Goddard, Cliff 1989 Issues in Natural Semantic Metalanguage. Quaderni di Semantica 10(1): 51-64. 1991 Anger in the Western Desert. Man 26: 602-619.

214

Jean Harkins

Goddard, Cliff (ed.) 1997 Studies in the Syntax of Universal Semantic Primitives. Special Issue of Language Sciences 19(3): 197-288. Goddard, Cliff 1998 Semantic Analysis: A Practical Introduction. Oxford: Oxford University Press. Goddard, Cliff and Anna Wierzbicka (eds.) 1994 Semantic and Lexical Universals: Theory and Empirical Findings. Amsterdam: John Benjamins. Forthcoming Meaning and Universal Grammar: Theory and Empirical Findings. Green, Jenny 1994 A Learner's Guide to Eastern and Central Arrernte. Alice Springs: IAD Press. Harkins, Jean 1990 Shame and shyness in the Aboriginal classroom. Australian Journal of Linguistics 10: 293-306. 1994 Bridging Two Worlds: Aboriginal English and Crosscultural Understanding. St Lucia: University of Queensland Press. 1995 Desire in language and thought. PhD thesis, Australian National University. 1996 Linguistic and cultural differences in concepts of shame. In: David Parker, R. Dalziell and I. Wright (eds.), Shame and the Modern Self, 84-96. Melbourne: Australian Scholarly Publishing. Harkins, Jean and David P. Wilkins 1994 Mparntwe Arrernte and the search for lexical universals. In: Goddard and Wierzbicka 1994: 285-310. Henderson, John and Veronica Dobson 1994 Eastern and Central Arrernte to English Dictionary. Alice Springs, NT: IAD Press. Lakoff, George and Zoltan K vecses 1987 The cultural model of anger inherent in American English. In: Dorothy Holland and Naomi Quinn (eds.), Cultural Models in Language and Thought, 195-221. Cambridge: Cambridge University Press. Malcolm, Ian G. and Judith Rochecouste 1999 Event and story Schemas in Australian Aboriginal English. English World-Wide 22: 95-123. Rice, Willie 1986 Ay eye artwe mperlkere arrwekelenye akerte: The story of the first white man. In: John Henderson (εά.), Arrernte Ay eye: Arrernte Stories, 26-35. Alice Springs: Yipirinya School Council and IAD. Sharifian, Farzad 2000 Aspects of schematic processing in Indigenous speakers of Aboriginal English: An initial exploration. Paper presented at 5th annual Round

Talking about anger in Central Australia

215

Table of the Centre for Applied Language and Literacy Research, Edith Cowan University, Claremont, Western Australia, 1 July. Wierzbicka, Anna 1972 Semantic Primitives. Frankfurt: Athenäum. 1996a Japanese cultural scripts: Cultural psychology and "cultural grammar". Ethos 24(3): 527-555. 1996 b Semantics, Primes and Universals. Oxford: Oxford University Press. 1997 Understanding Cultures through their Key Words: English, Russian, Polish, German, Japanese. New York: Oxford University Press. 1998 German cultural scripts: Public signs as a key to social attitudes and cultural values. Discourse and Society 9(2): 265-306. 1999 Emotions across Languages and Cultures: Diversity and Universals. Cambridge: Cambridge University Press. Forthcoming Cultural Scripts: Theory and Case Studies. Wilkins, David P. 1986 Particle/clitics for criticism and complaint in Mparntwe Arrernte (Aranda). Journal of Pragmatics 10: 575-596. 1989 Mparntwe Arrernte: Studies in the Structure and Semantics of Grammar. PhD thesis, Australian National University. To appear Fear and shame as a covert category in Mparntwe Arrernte. MS.

Meanings of Japanese sound-symbolic emotion words Rie Hasada

1. Introduction1 It has often been noted that the Japanese language is rich in sound-symbolic words, which form a conspicuous group in the Japanese lexicon (cf. Shibatani 1990: 153). Japanese onomatopoeic words are generally referred to as giongo-gitaigo, and divided into three classes. The first class is phonomimes (giseigolgiongo), which imitate sounds. The second class is phenomimes (gitaigo) which describe appearances, states, conditions of the external world. The third class is psychomimes (gijoogo) which express one's inner feelings or mental conditions. Phenomimes and psychomimes are often called mimetic words, as opposed to phonomimes which are called (sound-)onomatopoeia (cf. Shibatani 1990: 154; Martin 1964). Japanese everyday conversation is full of these sound-symbolic words. In contrast to the Indo-European languages whose sound-symbolic words are mostly phonomimes (sound-onomatopoeia), Japanese has more mimetic words than sound-onomatopoeic words (cf. Tamamura 1982).2 It has also been noted that Japanese is rich in psychomimes which describe various emotion/sensation states. While non-native speakers of the Japanese language will certainly face difficulties in acquiring psychomimetic words, they need to master the correct use of them if they wish to communicate effectively with Japanese people. Psychomimetic words play a vital role in Japanese everyday language life, and thus they are an indispensable key for outsiders to understand the Japanese people and culture.

218

Rie Hasada

1.1. Examining the meanings of psychomimes Japanese sound-symbolic words are very difficult to translate directly into another language (e.g. English), and likewise are difficult for non-native speakers to learn. Indeed, many advanced students of the language do not fully understand where and how to use them correctly. Among these words, psychomimes are the most difficult to grasp, as they refer to somewhat intangible inner feelings. The real meaning and the range of usage for each psychomime can never be captured from bilingual dictionaries. This is because the dictionaries offer only loosely equivalent English emotion words or phrases, instead of explicating the meaning. Sometimes they give the identical English gloss for more than one psychomime. For example, the largest current JapaneseEnglish dictionary Kenkyuushas New Japanese-English Dictionary [KNJED] provides "startled" for both gyoQ and dokiQ; "be afraid" and "be in fear" for both hiya-hiya and biku-biku. However, the meanings involved in these different psychomimes are not identical, and accordingly their ranges of use are different. As shown in the following sentences, different psychomimes cannot always occur in the same context, even though they seem to have almost identical meanings.3 (1)

a) b)

(2)

a)

b)

HaQI*gyoQ to shita shunkan meian ga ukanda. Ά good idea occurred to me when I felt haQ/*gyoQ.' *HaQ/gyoQ to shite sono ba ni tachisukunda. 'Feeling *haQ/gyoQ, I stood there paralysed.' Uso ο tsuita no ga baresoo na n de. hiya-hiyal?biku-biku na n da. Ί feel hiya-hiy al?biku-biku since it seems my lie will be found out.' Uso ο tsuita no ga bareta node. * hiya-hiy a/biku-biku na n da. Ί feel * hiya-hiy a/biku-biku since my lie was found out.'

The sketchy definitions given in the above dictionary cannot explain these differences between analogous psychomimes. The purpose of this paper is, therefore, to explicate the meaning of selected psychomimetic words, so that learners can actually use them correctly. Space permits the examination of only seven psychomimes here.

Meanings of Japanese sound-symbolic emotion words

219

These seven were chosen, because: (1) they are very commonly used, and thus deserve to be studied; (2) they are, in various ways, related to each other semantically and it is interesting and revealing to compare them; and (3) they reflect important characteristics of the Japanese way of conceptualising emotions. The semantic metalanguage developed by Wierzbicka and colleagues is employed as a tool to define these words. This metalanguage enables us to explicate complex concepts such as psychomimes by means of simple words and grammar which are understandable to anyone, and thus clarifies the similarities and dissimilarities between the components involved in different psychomimes. In order to comment on the variety of sound-symbolic vocabulary, data have been drawn from a varied range of sources. Most of the Japanese examples are from actual materials: newspapers, journals (mainly interviews), novels (most of which are written in the first person spoken style), magazines, advertisements, TV commercials, popular songs, scholarly papers, etc. I have also taken examples from English texts or books, in order to show how Japanese sound-symbolic words are translated into English, as well as what kind of English expressions have been translated into Japanese sound-symbolic words. Examples are also taken from several dictionaries in order to supplement my data, particularly from the Japanese—English Onomatopoeia and Mimesis Translation Dictionary [JEOMTD], whose examples were all taken from novels or film scripts. The source is indicated in abbreviation in square brackets at the end of each example. Details of abbreviations are given in the list of sources at the end of this paper. Unreferenced Japanese examples are from the author who is a native speaker. The subtle differences of meaning between psychomimes are often said to be impossible to translate appropriately into a foreign language. Furthermore, the differences between emotions, for example, between hiyahiya and biku-biku, which are usually identically explained as "be afraid" or "be in fear" in Japanese-English dictionaries, are hard to explain even for the Japanese. Nevertheless, these words are not synonymous and Japanese people use them distinctively in different situations. As illustrated above, adequate explanations for the use and nuances of these words are rarely found in dictionaries. Although it is the purpose of dictionaries to give a definition which captures the invariable semantic aspects of a word, they do not usually succeed in doing this. Many JapaneseEnglish dictionaries often do not even try to give a definition, but instead only give a single English word or a list of possibly analogous words. For

220

Rie Hasada

example, Nichiei-Taishoo: Giseigo Jiten [Japanese-English onomatopoeia dictionary] (Mito and Kakehi 1981) gives "feel nervous, be afraid" for the meaning of hara-hara. However, 'feel nervous' or 'be afraid' cannot indicate the proper meaning and usage of hara-hara since the range of meaning which hara-hara covers is not only something like 'feel nervous', not only something like 'be afraid', but both of them plus the feeling of something like 'be frightened' and more than that. Thus, the list of supposed synonyms given by the dictionary as a definition is in fact far from a meaningful explication of the meaning of hara-hara. Some Japanese-English sound-symbolic dictionaries offer paraphrasal explanations for the meaning of words. For example, in Practical Guide to Japanese-English Onomatopoeia and Mimesis [PG] (Ono 1989), a phrase such as "to be frightened, nervous, or anxious about something" is given for the meaning of hara-hara. This kind of explanation is problematic in several respects. First of all, this explanation is not capturing the semantic invariant of the meaning of these words. The given definition with the disjunction Or' does nothing but enumerate the different possibilities covered by the concept of hara-hara, instead of capturing its invariant meaning. In order to capture the semantic invariant, the definition should be minimal (not too broad, not redundant, not using elegant variation), but sufficient. Moreover, the biggest problem is offering English emotion words for the meaning of Japanese psychomimes. Although this may seem to be the main purpose of a bilingual dictionary, it is in fact misleading to the users because, for example, saying "nervous" for hara-hara means interpreting this Japanese concept through the prism of the English language. Some people like Kimizuka (1967: 16) suggest that giving many sample sentences with the appropriate sample context is important for teaching sound-symbolic words so that the student can learn the appropriate semantic situation in which it is used. In fact, most Japanese-English dictionaries or sound-symbolic dictionaries list many sample sentences in order to illustrate the actual use of those words. However, as Wierzbicka points out, quoting Leibniz, "an accumulation of examples cannot replace a general or formal signification for them" (Leibniz 1949: 365-366 in Wierzbicka 1986: 522). Wierzbicka (1986: 522) says that "if a series of examples provides a shortcut to a purely intuitive knowledge acquired by lengthy 'immersion in a language', a clear and general explanatory formula may provide a shortcut to a semi-intuitive knowledge acquired through lengthy study of examples, supported by time-consuming exercises".

Meanings of Japanese sound-symbolic emotion words

221

Neither vague analogous words from another language, nor lengthy explanations, nor accumulations of sample examples can show us clearly and precisely the unique nature or concept of Japanese psychomimes. If we want to define Japanese psychomimes in a way which is truly free of ethnocentrism and explanatory, we can do this using explicit semantic formulas, couched in terms of universal semantic primes (cf. Wierzbicka 1986: 530; 1992a). In this approach, a natural semantic metalanguage (NSM) is used as a tool for explicating the meanings of words in any language. These words are "maximally clear, maximally simple, and maximally universal" in the sense of having absolute equivalents in all the languages of the world, and thus are intuitively understandable (non-technical). By using these words which themselves are not names of emotions in specific languages, we can also avoid ethnocentric 'definitions' of words. In this paper, I attempt to explicate the meaning of Japanese psychomimes by using this metalanguage. In my explication of Japanese psychomimes, however, in order to avoid long and unwieldy definitions, I have used an expanded list, including words which, although not universal, recur widely in the languages of the world as separate lexical items. As a framework for the explication of psychomimes, I will employ the prototype scenario proposed by Wierzbicka (1992c: 540). This explicates the meaning of emotions by means of the framework showing "to feel a certain emotion means to feel like a person does who has certain (specifiable) thoughts, characteristic of that particular situation (and to undergo some internal process because of this)" (Wierzbicka 1994: 437). Since the whole range of emotions which psychomimes can represent is too large to be analysed in this paper, the discussion is limited to seven representative psychomimes which are frequently used. They are ordered and grouped in such a way as to show the differences and similarities of meaning between loosely related concepts. Three of them, representing momentary mental activity, will be discussed in section 2. They are gyoQ, haQ, dokiQ for something like 'surprise' feelings. Although there is a standard verb odoroku or a quasi-soundsymbolic verb bikkuri-suru for the expressions of surprise in Japanese, the Japanese language possesses those variant psychomimes such as gyoQ, dokiQ, haQ, which describe the exact moment of being surprised. One of the characteristics of Japanese psychomimes is that many of them represent restless emotions. In section 3,1 will examine four psychomimes which represent continuous restless emotional activity. They are hara-

222

Rie Hasada

hara, hiya-hiya, biku-biku and odo-odo, for bad feelings about an anticipated bad happening. Hara-hara, hiya-hiya, biku-biku, and odo-odo, are often translated or glossed as "nervous" in Japanese-English dictionaries. According to Morita, a particular type of neurosis called shinkeishitsu or shinkeishitsu-shoo, which is translated as "nervosity" or "nervosis", is relatively common among Japanese (Lebra 1976: 216). Fujita (in Lebra 1976: 217-8) states that the shinkeishitsu patient is torn apart by strong anxiety, tenseness, and fear on the one hand, and by the desire to overcome these and to participate in human interaction, on the other. Psychomimes such as hara-hara, hiyahiya, biku-biku, odo-odo all connote some bad feelings, which are related to this shinkeishitsu-shoo. The meaning of these psychomimes cannot be explained by English emotion terms, since they are indeed Japanese language-specific. Also, while psychomimes of the same group seem to have almost identical meanings, in fact they do mean different things and function differently in Japanese. When they are translated into the universal semantic metalanguage, the similarities and the differences between those concepts become apparent, and the meaning of each psychomime becomes comprehensible to outsiders.

2. Momentary emotions HaQ, gyoQ, dokiQ are psychomimes denoting something like the feeling of 'surprise' about an unexpected happening. The quality of immediate reaction involved in all of these psychomimes is expressed by the component "this person feels something for a very short time". While, for example, Japanese-English dictionaries give the English word "startled" for the meaning of Japanese psychomime haQ, "startled" cannot always occur in the context where haQ occurs. For instance, haQ cannot be replaced by "startled" in the following sentence: Natsukashii kyoku ο kiite haQ to shitari... Ί feel haQ, hearing the music of old times'. Moreover, dictionaries often offer one English emotion word for more than one Japanese psychomime. For example, the major Japanese-English dictionary KNJED offers "be startled" for all three psychomimes haQ, gyoQ, and dokiQ; "feel a shock" for both gyoQ and dokiQ; "get a start" for both haQ and dokiQ. Looking at those lists of English emotion words for each

Meanings of Japanese sound-symbolic emotion words

223

psychomime, the reader just gets confused, wondering how each word is different from the others. Izard (1977: 277) says that the feeling of 'surprise' is difficult to describe because the feeling does not last long. However, the common features as well as differences of haQ, gyoQ, and dokiQ can be clearly explicated if we examine the underlying concepts of these psychomimes in terms of intuitively understandable words. For the discussion of each psychomime, examples in context are given first, then my proposed definition, followed by amplification or justification of particular elements in the definition.

2.1. HÜQ Examples: (3)

HaQ to shita shunkan no kimi wa tenshi sa... [From a popular song by T. Tahara] RH4: 'You look like an angel at the moment I feel haQ.'5

(4)

Kanojo no utsukushisa ni mitorete, sono ba ni kugizuke ni natte ita watashi wa, musume no yobu koe ni haQ to ware ni kaetta. [PG] B: 'Transfixed, lost in admiration for her beauty, I suddenly \haQ to] came to myself at the sound of my daughter calling me.'

(5)

Futon o mochiageru toki, yoisho to kakegoe shite, haQ to omotta. Watashi wa, ima madej'ibun ga yoisho nante gebita kotoba o iidasu onna da to wa omotte inakatta. [HM: 71] J-»E: 'As I lift it (my futon) to put it away, I hear myself go "Oof!" and I think, what? \haQ to}. I've never thought of myself as the type of girl who'd come out with something like "oof'.' [RM: 44]

(6)

Tsumi to batsu, Dosutoefusukii, chira to sore ga zunoo no katasumi o kasumete toon, haQ to omoimashita. Moshimo ano Dosutoshi ga tsumi to batsu o shinonimu to kangaezu, antonimu to shite narabeta mono to shitara? [NS: 104] J -» E: 'Crime and punishment. Dostoyevsky. These words grazed over a corner of my mind, startling me [making me feel haQ]. Just supposing Dostoyevsky ranged 'crime' and 'punishment' side by side not as synonyms but as antonyms?' [NLH: 147]

224

Rie Hasada

(7)

[Kita Morio-shi no sakebigoe ni] Kupukupu wa haQ to omoidashita. Taihen-datta. Ima wa konna tokoro de guzu-guzu shite wa irarenai toki na no da. [FKB: 145] J-»E: ['To hear the cry of Morio Kita.JSpeaking of which, Kupukupu suddenly remembered [remembered haQ-to]. This was no time for idle chatter.' [AKS: 133]

(8)

HaQ to shita mimi ni, ame ga shito-shito to noki o nurasu oto ga kikoe... [T: 103] RH: 'My ears which felt haQ heard the sound of rain falling shitoshito on the eaves...'

(9)

Natsukashii kyoku o mimi ni shite, haQ to shitari... [A phrase in the advertisement of Gunze (underwear company) 4/7/1993] RH: Ί feel haQ. hearing the music of old times.'

(10)

Densha ο orita totan, kasa ya kaimonobukuro o shanai ni okiwasureta koto ni haQ to kizuita. [PG] B: 'Just as she got off the train it suddenly \haO to\ occurred to her that she had left her umbrella and shopping bag on the train.'

(11)

Nemuri ni ochiru toki no kimochi tte, hen na mono da. Funa ka, unagi ka, gui-gui tsuriito ο hipparu no ni, nandaka omoi, namari mitai na chikara ga, ito de motte, watashi no atama o, guQ to h te, watashi-ga toro-toro to nemukeru to, mata, chotto ito o yurumeru. Suruto, watashi-wa haQ to ki o torinaosu. [HM: 112] J-»E: 'Falling asleep is a strange sensation. It's like something very heavy, like lead, pulling on my head with a string, like a carp or eel jerking on a fishing line, pulling me down. You start to nod off, and then the line goes slack and I snap back awake. ΓΙ pull myself together haQ-to\:\RM: 101]

(12)

Uoorugai de wa subete no hito ga Uoorujaanaru o maiasa yomu. Toki ni wa "haQ" to suru yoo na kiji ga dete iru-shi, heion-buji na hi mo am. [TB: 230] RH: On Wall St. everybody reads the Wall Street Journal every morning. Sometimes there is an article which makes me feel haQ, and sometimes there are only tranquil articles.'

Meanings of Japanese sound-symbolic emotion words

225

(13)

Tomodachi ga hootai de gurugurumaki ni shite, doa no tokoro ni arawareta no de, watashi wa haQ to shita. [PG] Β: Ί was startled for a moment [felt haQ] when my friend appeared at my door with his head all bandaged up.'

(14)

Aete (ie ni iru toki wa) mecha-kucha na kakkoo o shite, joyuu ni nattara, HaQ to shokku o ataeru. Sono rakusa o tanoshinde iru n dem. [FK January 1993: 345] RH: Ί tend to dress awfully on purpose when I am at home, then when I change into an actress, I give other people a shock which makes them feel haQ. I enjoy this difference.'

Meaning of haQ (a) (b) (c) (d) (e) (f) (g) (h) (i)

X feels something for a very short time because X thinks something sometimes a person thinks something like this: "something happened now I didn't think before that this would happen I know something now because of this" because of this, this person feels something for a very short time X feels something like this because X thinks something like this

According to Morita (1988: 83), the Japanese word odoroku, a nonsound-symbolic emotion word for a 'surprise' feeling, originally had milder connotations than the present odoroku. For example, in the Heian period (794-1192 AD), even the feeling experienced when one is woken from sleep was described as odoroku. Nowadays, only one psychomime haQ can be used for such an occasion of awakening from sleep, as seen in sentence (11). Other psychomimes or emotion words for 'surprise'-like feelings including odoroku are used for referring to more intense feelings. • something happened now • I didn't think before that this would happen This feeling occurs suddenly when one reacts to some happening immediately. There is an unexpected feeling in haQ, although its degree is not as strong as that of gyoQ (Ί didn't know before that something like this

226

Rie Hasada

could happen') and dokiQ (Ί didn't think before that this could happen'). The trigger of this haQ feeling can be anything. It does not have to be remarkable, and often is a very trivial thing, such as someone's calling one's name as seen in sentence (4). Moreover, the unexpectedness included in the haQ feeling is not necessarily caused by an external thing. This could be one's own voice as seen in sentence (5); one's mental activity as seen in sentence (6); or one's own action of getting off the train as seen in sentence (10). • I know something now because of this Izard (1977: 277-8) describes the thinking activity at the moment of 'surprise' as follows: "our mind seems to be blank. ... It is as though ordinary thought processes are momentarily stopped. Thus there is very little thought content associated with surprise, and virtually none with startle". However, the intensity of unexpectedness of the haQ feeling is not strong enough to stop one's thinking. For example, one could say that one gets an idea at the moment of this haQ feeling as follows: (15)

HaQ to shita shunkan subarashii kangae ga omoiukanda. Ά bright idea occurred to me at the moment when I felt haQ'

Therefore, the time when one feels haQ is rather the crucial moment to start the active movement of one's mind. At this moment, we recognise or know something: for example, we come to reality from an enchanted state as in sentence (4); we recognise something we didn't recognise before, such as our own habit as seen in sentence (5) or get a flash of an idea as seen in sentence (6); we remember something we have forgotten as in sentence (7) and (10); we recognise something anew, such as someone's beauty as seen in sentence (3) and (14). Therefore, haQ is often used together with the verbs kizuku 'recognise' in (10), or omou 'think' in (5), (6), (7), or with phrases such as ware ni kaeru 'come to oneself in (4), or ki ο torinaosu 'pull oneself together' in (11). There are a lot of sound-symbolic words which refer to the state of absent-mindedness, day-dreaming, or unconsciousness in Japanese: uttori 'be transported', ukkari 'inadvertently', uka-uka 'to dream away one's time'; booQ, pokan, bokeQ, bosa-bosa, bonyari 'to look vacantly', and so on. HaQ is the feeling experienced when one comes to oneself from these absent-minded, delusive, or unrecognised feelings.

Meanings of Japanese sound-symbolic emotion words

227

• because of this, this person feels something for a very short time

This psychomime haQ captures the feeling experienced in a short time. The resulting feeling does not include a bad or a good component. One may feel something good if what one discovers at the moment of haQ is good, such as when one becomes conscious of another person's beauty in sentence (3). One may feel something bad if what is discovered is something bad, such as one's own undesirable habit in sentence (5) or a friend's injury in sentence (13). However, these are not components of the meaning of haQ.

2.2. GyoQ Examples: (16)

Watashi wa Tsugumi no ude o tsukami gyoQ to shita. Sono hitomi wa reisei ni watashi ο mitsumekaketa ga, ude wa monosugoi atsusa datta. [T: 159] RH: Ί felt gyoQ when I held Tsugumi's arm. She stared at me calmly, but her arm was awfully hot.'

(17)

Tsugumi mo sukoshi waratte miseta. Soshite, itta. "Ano na, omae ni dake iu kedo na, atashi, dame kamo shirenai. Kitto shinu." Watashi wa gyoQ to shita. ΓΤ: 203] RH: Tsugumi smiled a bit. Then she said, "Well, I will tell only you. I might be hopeless. I am sure I will die." I felt gyoQ? Sukoshi zutsu, jigazoo no seisaku ni torikakatte mimashita. Jibun demo gyoQ to shita hodo. insan na e ga dekiagarimashita. Shikashi, kore koso munasoko ni hitakakushi nojibun no shootai na no da. [NS: 34] J-»E:'... I began to draw a few self-portraits... The pictures I drew were so heart-rending as to stupefy even myself [make even myself feel gyoQ]. Here was the true self I had so desperately hidden.' [NLH: 55]

(18)

(19)

(Ani wa) kaoiro mo waruku, gyoQ to suru hodo yasete, kewashii yooboo ni narimashita. [HM: 214] J->E: '[My oldest brother's] face had a stern look, and I was shocked [I felt gyoQ] to see how thin he was.' [DO: 166]

228

Rie Hasada

(20)

Nedan ga takai no ni gyoQ to shita. [KNJED] Β: Ί was staggered Tfelt gvoO^ at the price.'

(21)

Kare wa gyoQ to shite, kos hi o nukashita. [PG] RH: 'He felt gyoQ and was unable to stand up.'

(22)

(a) He, ano hito, okama? gyoQ. [PG] B: 'What! That guy's a fag? Gee! \gyoQ\V (b) GyoQ! Koko gajisatsu no meisho no dangai? [GGT: 98] RH: 'GyoQ\ Is this the famous cliff where many people commit suicide?' (c) Hee, ano hito sonna toshi nano? gyoQ. [GGJ: 84] RH: 'Is that person really that age? gyoQ.'

(23)

GyoQjojsono ba ni kugizuke ni natta. [PG] B: 'Shocked [feeling gyoQi she was riveted to the spot.'

(24)

Amari shizuka ni tatte iru mono desu kara,futari ga ikite im koto mo sukkari wasurete shimaimashita. Sorede, meimei no eri no ushirogawa ni, 'Tsuiidoru to iuji ga kaite aru ka doo ka tashikameyoo to, gurutto ushiro no hoo ο miyoo to shita toki, 'DAMU' no shimshi no aru hoo ga koe o dashita node, [Arisu-wa] gyoQ to shimashita. [KKA: 45-6] J-«-E: 'They [DUM and DEE] stood so still that she [Alice] quite forgot they were alive, and she was just going round to see if the word "TWEEDLE" was written at the back of each collar, when she was startled [felt gyoQ] by a voice coming from the one marked "DUM"'. [TLG: 165]

Meaning of gyoQ (a) (b) (c) (d) (e) (f) (g) (h) (i)

X feels something for a very short time because X thinks something sometimes a person thinks something like this: "something happened now I know: this is bad I did not know before that something like this could happen I can't think now I can't do anything now because of this" because of this, this person feels something bad for a very short time

Meanings of Japanese sound-symbolic emotion words

(j) (k)

229

X feels something like this because X thinks something like this

• something happened now. • I know: this is bad The feeling of gyoQ occurs when we encounter some unexpected thing which we think is "bad". For example, we feel gyoQ when our friend suddenly says she will die soon as in sentence (17); when we see something terribly bad as in sentence (18), when we find out the incredibly high price of some merchandise as in sentence (20). One cannot feel gyoQ when the discovery is something good or neutral (neither good nor bad). For example: (18')

Sukoshizutsu, jigazoo no seisaku ni torikakatte mimsashita. Jibun demo gyoQ to shita hodo, insan na[* subarashii/*heibon na] e ga dekiagari-mashita... Ί began to draw a few portraits... The pictures I drew were so dreary T*great/* ordinary 1 as to make even myself feel gyoQ.'

The picture which is subarashii 'great' or heibon na Ordinary' cannot be the trigger of gyoQ, since this feeling requires the connotation of discovering something "bad". The psychomime gyoQ came to be used much more widely and commonly when a disk jockey regularly used this word in his radio program in 1949 (Inagaki 1982). Inagaki mentions that this word became popular at this time because many Japanese people were actually feeling gyoQ about the occurrence of unexpected and unbelievable bad events, such as the Korean war, or other subsequent political and social affairs. People at present have much more choice for expressing the feeling of something like 'surprise' than people in ancient times who did not have words such as gyoQ or bikkuri, but had only milder emotion words such as haQ or odoroku. This might be because there are more events/things in the modern world that are likely to cause various "surprise-like" emotions. • I didn't know before that something like this could happen The degree of the experiencer's unexpected feeling towards the occurrence of something is very strong.

230

Rie Hasada

• I can't think now • I cannot do anything now because of this The feeling of gyoQ is so strong and intense as to make the experiencer sometimes utter this word spontaneously and unconsciously like an interjection as seen in (22). Such a spontaneous expression of the HaQ feeling never occurs, since it is a much milder feeling. In contrast to the feeling of haQ, which activates one's mental activity, the feeling of gyoQ leaves the experiencer feeling at a loss about how to deal with this unexpected happening, and stops his/her mental or physical activity for a moment. Therefore, the person feeling this gyoQ is often described as having stopped his/her movement with phrases such as sono ba ni kugi-zuke ni narultachi-sukumu 'stand paralysed' or as having become blank in the mind with phrases referring to blank facial expression such as [kuchi] anguri 'be open-mouthed being unaware it is open'. • because of this, this person feels something bad for a very short time The resultant feeling after the discovery of an incredibly, unexpectedly bad happening is 'bad'.

2.3. DokiQ Examples: (25)

... to igai na koto ο osshatta. Watashi wa dokiQ to shite... [Dazai "Shayoo (setting sun)" in JEOMTD] J-»E: Ί was taken aback Tfelt dokiQ^ by this unexpected remark.'

(26)

Kurejitto kaado no seikyuu ni dokiQ to shitari... [A phrase in the advertisement of Gunze, 4/7/1993] RH: Ί feel dokiQ to see the demand for payment of my credit card.'

(27)

Tetsuo ga itta. "Chigau, nigeru na'.' dokiri6 to shite kare o mitsumeruto... [KY: 158] RH: Tetsuo said, "No, don't run away (from your trouble)." I felt dokiri. then I gazed at him.'

Meanings of Japanese sound-symbolic emotion words

231

(28)

'Tokoroga, shoojiki na hanashi, ome ni kakatte boku wa dokiQ to shita no da.' [Osaragi "Kikyoo (homecoming)" in JEOMTD] J-»E: 'Honestly, though, something struck me [I felt dokiQ] when I saw you.'

(29)

Tsuma gajisatsu shita to iu shirase o kiita toki, kare wa karada ga tobiagaru-hodo mune ga dokiQ to shite, omowazu iki o nonda. [PG] B: 'Upon hearing the news that his wife had committed suicide, he got such a shock [felt ftofa'QI that he leapt up taking a startled breath.'

(30)

DokiQ to shita no wa, tsuini sore rashii shinshi no sugata o mitsuketa kara de atta. [Genji "Hoopu-san" in JEOMTD] J-*E: 'He gave a start [he felt dokiQ]. as he spotted someone who looked as if he might be his man.'

(31)

"DokiQ to sase-yagaru ze. Koko de nani shiteru n da" to yaban wa tomodachi ni itta. [PG] B:' "You scared the hell out of me [made me feel dokiQ] \ What are you doing here?" the night watchman said to his friend.'

Meaning of dokiQ (a) (b) (c) (d) (e) (f) (g) (h) (i) (j)

X feels something for a very short time because X thinks something sometimes a person thinks something like: "something happened to me now I didn't think before now that this could happen something happened to my heart now because of this I could hear something like this: dokiQ" because of this, this person feels something for a very short time X feels something like this because X thinks something like this

• something happened to me now • I didn't think before that this could happen

232

Rie Hasada

Like haQ and gyoQ, the discovery of the occurrence of something is sudden [now]. The negative or positive judgement does not need to be attached to the trigger of this emotion. Sometimes the trigger could be something bad, such as the large amount for payment of one's credit card as seen in sentence (26) or bad news of somebody's death as in sentence (29). It could also be something good such as the feeling experienced when one is given a word from an admired person suddenly as seen in the following sentence: (32)

Akogare no kanojo ga, Ohayoo' to surechigai ni itta node, dokiQ to shita. Ί felt dokiQ. since the girl I admire said "good morning" to me when she passed by.'

One also could feel dokiQ when the trigger of that feeling cannot be judged good nor bad, such as the sudden unexpected discovery of someone's presence in a place as seen in sentence (31). Unlike gyoQ, whose cause could be anything as long as the experiencer thinks it bad and incredibly unexpected, the cause of dokiQ is restricted to something that will affect the experiencer somehow. People feel dokiQ when they think about something that concerns them [i.e. "to me"]. For example, we feel dokiQ upon discovering an extremely high balance due on our own credit card bill as in sentence (26). We would not feel dokiQ, while we might feel gyoQ, if that is the bill of our friend's credit card. (26')

? Tomodachi no Kurejitto kaado no seikyuu ni dokiQ to shitari... ? I feel dokiQ to see the demand for payment of my friend's credit card.

This is because we know that the discovery of an unexpected happening does not concern us, but our friend. We might feel dokiQ, if we know that the bill might affect us somehow: e.g. we have the responsibility for paying the bill in case of the bankruptcy of our friend; or the friend will find out that we have used that friend's credit card secretly before. • something happened to my heart now because of this It is our mune 'heart' where the physical reaction of this dokiQ feeling occurs, as seen in mune ga dokiQ to suru 'my heart feels dokiQ' in sen-

Meanings of Japanese sound-symbolic emotion words

233

tence (29). Other parts of the body would not feel dokiQ: *atama/hara/karada ga dokiQ to sum '*My mind/belly/body feels dokiQ'. The degree of unexpectedness of dokiQ is strong enough to make the experiencer think he/she had a strong heart beat. The intensity of this unexpectedness is stronger than that ofhaQ, where no particular physical manifestation is seen except for one's mental activation, but this is weaker than that of gyoQ, where the unexpected shock is so strong as to stop one's thinking as well as physical activity. • I could hear something like: dokiQ The experiencer of a dokiQ feeling who physically felt a strong beat of his heart thinks as if he/she actually heard the sound of its beating. The sound of one's repeated heart beat is described as the phonomime dokidoki in Japanese, as in the following sentence: (33)

Chooshinki o mune ni ateru to doki-doki to iu kodoo ga kikoeru. 'When we apply a stethoscope to the heart, we can hear the sound doki-doki of the heart beating.'

DokiQ is the emphatic form of a single occurrence of this heartbeat dokidoki. • because of this, this person feels something for a very short time The resulting feeling can be either good or bad. The experiencer might feel something good when what he/she finds at the moment of dokiQ is something good, such as a word from a girl he admires as seen in sentence (32). One might feel something bad if his/her discovery is something bad, such as news of someone's death as in sentence (29). However, this good and bad feeling is not a component of the meaning of dokiQ, but is derived from the context of the sentence.

3. Continuous restless emotions In this section, semantic aspects of psychomimes for continuous restless emotions will be examined. The nature of continuity involved in all these psychomimes is identified by the common component "this person feels this FOR SOME TIME". The selected continuous emotional psychomimes are

234

Rie Hasada

those expressing negative feelings about expected bad happenings: harahara, hiya-hiya, biku-biku, odo-odo. As usual, existing bilingual dictionaries merely list possibly related English emotion words, each of which is often used as the gloss of more than one psychomime. For example, in KNJED the English gloss 'be nervous' is given for hara-hara, hiya-hiya, biku-biku, and odo-odo. The gloss 'be afraid' is given for the meaning of hara-hara, hiya-hiya, and biku-biku; 'be in fear' and 'be timid' are offered for hara-hara, hiya-hiya, biku-biku, and odo-odo. Readers thus will not understand how these psychomimes are similar and different to each other. Similarities and dissimilarities among all these psychomimes can be revealed more precisely by explications phrased in the Natural Semantic Metalanguage. I have classified hara-hara, hiya-hiya, biku-biku, and odo-odo, in one group, since they seem to have a common semantic core of the experiencer's expectation of some bad happening and "bad" feeling as a result.

3.1. Hara-hara Examples: (34)

Sore kara futari wa yashizake o nomihajimeta. Kupukupu wa hara-hara shita. Nazette.futari tomo, taihen na ikioi de gabugabu nomidashita kara datta. [FKB: 87] J-»E: 'The two began drinking the wine. It made Kupukupu nervous [feel hara-hara] to see how fast they were gulping it down.' [AKS: 80]

(35)

Yakuruto, kotoshi mo hara-hara yakyuu. [AS: 5/4/1993: 21] RH: 'Yakuruto [baseball team] plays the game making the audience feel hara-hara this year, too.'

(36)

Mozuku daihoosaku: Sanchi hara-hara [AS: 11/6/1993: 17] RH: 'The people in the production area ofmozuku feel hara-hara because too much is being harvested [since the price of mozuku might greatly fall].'

(37)

Hajimete Shaarokku Hoomuzu o yonde, hara-hara shite kara to iu mono wa, yamitsuki ni natte shimatta no de aru. [CS: 71]

Meanings of Japanese sound-symbolic emotion words

235

RH: 'He had been keen on them ever since he read Sherlock Holmes for the first time and felt hara-hara.' (38)

[Sakki kara uma kara ochite bakari iru] Kare gafutatabi uma ni noru no ο hara-hara shinagara. mimamotte imashita. [KKA: 215] J*-E: ['The knight was falling down from the horse many times.] [Alice] watched him with some anxiety [feeling hara-hara]^ he mounted again.' [TLG: 116]

(39)

Watashi wa mae kara, anta-tachi no mamagoto asobi mitai na kurashikata ο hara-hara shinagara mite ita n desu. [Dazai "Shayoo" in JEOMTD] (J-»E: Tve been watching with my heart in my mouth [feeling hara-hara] the way you two have been living.')

(40)

Watashi wa takai tokoro ni nobotte iru hito o miru to, ochinai ka to omotte, hara-hara shite shimau. [JF: 60] RH: 'When I look at a person who is climbing a high mountain, I feel hara-hara, wondering if he/she might fall down.'

(41)

Kabu ο katta hito wa, ichinichi ichinichi no kabu no neugoki ni hara-hara shiteiru soo da. [JF: 60] RH: Ί hear that the people who bought the stocks are feeling harahara about the movement of the price of stock day by day.'

Meaning of hara-hara (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k)

X feels something because X thinks something sometimes a person thinks something like this: "I see: something is happening to someone/something (Y) now something bad could happen to Υ in a very short time because of this I don't want this to happen because of this, I want to do something I don't know what I can do" because of this, this person feels something bad this person thinks: "I will see Υ (something/someone) is falling down [hara-hara] into something bad"

236

(1) (m)

Rie Hasada

this person feels this for some time X feels like this

• I see: something is happening to someone/something (Y) now • something bad could happen to Υ in a very short time because of this The feeling of hara-hara occurs during the on-going process of some happening or situation, and when the experiencer is the objective observer of it (Y cannot be the experiencer of the hara-hara feeling herself/himself). It has to happen while some process is on-going; one cannot feel hara-hara seeing somebody before he/she does something or after he/she did something, as seen in the following sentences. (42)

*Takai tokoro ni noboru to ketsui shita hito ο mite hara-hara sum. *'I feel hara-hara seeing a person who decided to climb the high mountain.'

(43)

*Takai tokoro ni nobotta hito ο mite hara-hara sum. *'I feel hara-hara seeing a person who climbed the high mountain.

Therefore, for example, in sentence (35), the audience feels hara-hara, watching their team (Yakuruto) play, assuming they will lose if they keep playing like that. In sentence (40), the person feels hara-hara at seeing somebody in the process of climbing a high mountain, anticipating the possibility of them falling down at any time. In sentence (36), the farmers feel hara-hara at seeing the on-going too abundant harvest of mozuku (nemacystis decipiens), knowing this could cause a great fall in price at any time. In sentence (39), the person felt hara-hara seeing the young couple's way of living, because she thought their life was like playing at housekeeping, thus it could collapse at any moment. In sentence (41), the person feels hara-hara because stock prices are changing everyday, which could make the price of their own stocks fall. In all these examples, the experiencers of hara-hara feeling presuppose the possibility of something bad happening to a particular third person/object (Y) because of something in the on-going process. They feel this possible danger in the immediate context ("now"). • I don't want this to happen

Meanings of Japanese sound-symbolic emotion words

237

A person who feels hara-hara for Υ does not want something bad to happen to Y. A person who does not care if something bad happens to Υ does not feel hara-hara. For example, in sentence (35), a fan of Yakuruto's opposition team will not feel hara-hara about Yakuruto's unstable playing during the baseball game, since they want something bad to happen to Yakuruto so that their favourite team will win. • because of this, I want to do something • I don't know what I can do The experiencer feels like doing something about the possibility of something bad happening to Y. However, since the experiencer of the harahara feeling is a passive observer of the on-going situation, often he/she can do nothing but watch helplessly, just hoping that something bad will not happen to Y. For example, in sentence (36), the farmers feel hara-hara, but cannot do anything about the increase of mozuku, which is a natural phenomenon. In sentence (37), the reader of the Sherlock Holmes story can do nothing about the development of the plot in the story. Since the experiencer of hara-hara feeling cannot do anything about the on-going situation, the psychomime hara-hara often collocates with phrases like ki ο momu 'to fidget', or mimamoru 'to watch over'. • because of this, this person feels something bad: The experiencers feel "something bad", since they cannot do anything, while they are faced with possible danger toward Y, which they do not want to happen. • this person thinks: I will see Y (something/someone) is falling down into something bad7 Hara-hara originally describes the manner of something small and light falling down, which causes the observer to feel sad or melancholy, as seen in the following sentences. (44)

Hara-hara to sakura no hana ga ochiru no ο mite iru to, konoyo no mujoo o kanjiru ne. [FSB: 92] B: 'Watching the cherry blossoms flutter down \hara-hara to\ makes you realise how fleeting life is.'

238

(45)

Rie Hasada

... to ine, Etsuko wa hara-hara to namida o furikoboshita. [Enchi "onnazaka": 199/197 in JEOMTD] J-»E: 'The anxious tears spilled from \hara-hara te>] Etsuko's eyes as she spoke.'

This image of seeing something falling down is closely linked with the emotional feeling of hara-hara which refers to an anxious feeling towards Υ (someone/ something) falling into a dangerous or frightening situation ("I will see Υ is falling down into something bad").

3.2. Hiya-hiya Examples: (46)

Oya-kyoodai-tachi wa, byoonin ο kinodokugam yon mo, sono toki no surudoi sakebigoe o kikasareru no ga yarikirenai no de, hiyahiya shita ga. ... [Tanizaki "Sasameyuki" : 291 in JEOMTD] RH: 'The family seemed less concerned for the patient than for themselves, and were feeling hiya-hiya since they would have to listen to his shriek.'

(47)

Yappari uso nanka tsuku mon ja nai naa. Itsu barer u ka to hiyahiya shita yo. [FSB: 58] B: 'Lying just isn't worth it, after all. I was scared stiff [felt hivahiyq] that I'd get caught any minute.'

(48)

Hahaoya no saifu kam ichiman'ensatsu ο nuita no ga baresoo de, hiya-hiya nan da. [BP: 83] RH: I'm scared stiff [am feeling hiya-hiya] 'cause it looks like my mum found out I swiped 10,000 yen outa her purse.'

(49)

Renzoku hooka ni hiya-hiya I Suginami no juutakugai. [AS: 12/5/1972 in GG] RH: 'The people in the Suginami residential area feel hiya-hiya. because of the constant occurrence of arson.'

(50)

Watashi wa hajime wa hiya-hiya shinagara damatte Karube no dabette iru koto o kiite ita no da... [Yokomitu "Kikai": 139 in JEOMTD] J-*E: 'At first I listened nervously [feeling hiva-hiva] to the

Meanings of Japanese sound-symbolic emotion words

239

chatter of Karube.'/'At first I could feel myself going cold all over [feel hiya-hiya\ as I stood there listening to Karube chatting away.' (51)

Kare ga sono koto o ukkari shaberi wa shinai ka to hiya-hiya UKNJED] Β: Ί was in great fear [felt hiya-hiya} lest he should blurt it out.'

(52)

Hiya-hiya shinagara gakeppuchi no michi o susunda. [PG] B: 'We timidly [feeling hiya-hiya\ made our way along the path on the edge of the cliff.'

(53)

Yaku sennin no yajiuma wa, te ni se o nigitte hiya-hiya. [Yomiuri (evening) 14/3/1972 in GG] RH: 'About one thousand curious bystanders feel hiya-hiya in breathless suspense.'

(54)

Itai n ja nai ka to, naishin hiya-hiya datta. [AS: 15/11/1972 in GG] RH: Ί was feeling hiya-hiya inwardly, thinking it might be painful for you.'

Meaning of hiya-hiya (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k) (1) (m)

X feels something because X thinks something sometimes a person thinks something like: "something bad can happen I don't want this because of this, I want to do something I don't know what I can do" because of this, this person feels something bad this person thinks: "if I were in a cold place, I would feel like this [hiya-hiya]" this person feels this for some time X feels something like this because X thinks something like this

• something bad can happen Unlike hara-hara, where the possible danger is always directed to a third person, hiya-hiya has no such restriction. It could be directed to a third per-

240

Rie Hasada

son. For example, in sentence (53) curious bystanders feel hiya-hiya in observing some event. In sentence (54) the person felt hiya-hiya anticipating the possibility of something bad happening (pain) to another person. On the other hand, the patient of the bad happening could also be the experiencer. For example, in sentence (46), the family members feeling hiya-hiya are not concerned about the condition of the third person (the patient) but about the undesirable possibility of they themselves hearing the patient's shriek. In sentences (47) and (48) the experiencer feels hiya-hiya about him/herself for fear of being accused of the deed he/she did. The experiencer feels hiya-hiya when he/she thinks the anticipated bad happening is still at the level of possibility ('something bad CAN happen') and when the possibility of danger is not so imminent or urgent. Therefore, if the people are sure about the occurrence of a bad happening, hiya-hiya cannot be used. This point can be illustrated by comparing the following sentence with the sentence (48) above. (48')

*Hahaoya no saifu kara ichimariensatsu o nuita no ga bareta kara hiya-hiya nan da. *'I am feeling hiya-hiya 'cause my mum found out I swiped 10,000 yen outa her purse.'

In the above sentence (48'), one's sense of danger about the bad happening to oneself (e.g. one's mother's scolding or accusation) after the revealing of his bad deed is more imminent than that conveyed by hiya-hiya. • I don't want this The experiencer does not want the possibility of something bad happening to them to be realised. • because of this, I want to do something • I don't know what I can do The experiencer wishes to do something about the possibility of danger. However, the experiencer of the hiya-hiya feeling does not really feel able to control the situation. • because of this, this person feels something bad

Meanings of Japanese sound-symbolic emotion words

241

Because of the possibility of a bad happening as well as recognising the lack of power of controlling the situation, the experiencer feels "something bad". • this person thinks this if I were in a cold place, I could feel like this [hiya-hiya] Hiya of hiya-hiya is associated with the transitive verb hiyasu 'to cool'. The biological sense of hiya-hiya refers to a 'chilly' feeling as seen in the psychomime in the following sentence: (55)

Reika to nisshoo-busoku no sei daroo. Amido dake de wa hiyahiya sum. (B: 'It is probably due to the cold summer and the lack of sunny days that I feel chilly \hiya-hi\a\ with just the fly screen shut.') [PG]

This concrete image of a physical chilly feeling is closely linked with the emotional hiya-hiya feeling. When one is feeling hiya-hiya emotionally one often thinks as if one's body is cooled. While this physical experience does not necessarily actually occur when one feels emotional hiya-hiya, its chilly image is reflected in the meaning of the psychomime hiya-hiya.

3.3. Biku-biku Examples: (56)

Kowai mono ni hikizurareru yoo ni, suru-suru to soba e itte suwatta. Naishin biku-biku shinagara. "Okaasan wa dooshitemo dame desu ka?" to itta. [HM: 224] J -» E: 'As though drawn by some fearful force, I went towards him and sat down. I felt an inward tremor [felt biku-biku inwardly! as I asked, "Will she (mother) die?'" [DO: 173]

(57)

Hidoku shikarareru to omotte, biku-biku mono de kita n desu kedo, kite yokatta wa. [Inoue "obasute" p. 23/89 in JEOMTD] J -> Ε: "Ί wondered if I was going to be scolded severely and I was nervous ΓΙ was in the state of feeling biku-biku] about coming - but I'm glad I came," she said.'

242

Rie Hasada

(58)

G an ja nai ka to biku-biku shite inai de, hayaku byooin e ikinasai yo. [GGT] RH: 'Don't keep feeling biku-biku, anticipating the possibility of cancer, but just go to the hospital to get it checked.'

(59)

Tada no uwasa ni biku-biku nasaru koto wa gozaimasen. [Mishima "Ham no yuki" : 289/269 in JEOMTD] J-»E: 'There is absolutely no need to get upset [feel biku-biku] over what's only a rumor.'

(60)

Ore ga sake o nonde, mata daishippai demo shitara, sore koso ichidaiji da to bikubiku shite im n daroo. [Miura "Shiogaritooge": 281/222 in JEOMTD] J-*E: 'You're scared [feel biku-biku] that I'll get drunk and make a big mistake and there'll be a row.'

(61)

Kono toki, oosama wa kenka o yamesaseyoo to, aida ni hairimashita ga, biku-biku shite, koe wa gata-gata furuete imashita. [KKA: 113] J*-E: 'Here the King interrupted, to prevent the quarrel going on: he was very nervous [feeling biku-biku]. and his voice quite quivered.' [TLG: 113]

(62)

Gomufuusen ga itsu yabureru ka biku-biku shinagara, yukkuri iki ofukikonde iru. [PG] B: 'Anxious [feeling biku-biku] about when the rubber balloon might break, the girls are slowly blowing it up.'

(63)

/ no warui kekka ο yosoo-shite, byooin e iku no wa, mattaku bikubiku-mono datta. [GGJ] RH: Ί was really in a state of feeling biku-biku when I went to the hospital, anticipating a bad result of the examination of my stomach.'

(64)

Kare wa kubi ni nan wa shinai ka to biku-biku shite iru. B: 'He is afraid Tfeeling biku-biku\ of being sacked.' [KNJED]

(65)

Majime ni renshuu ni konai aitsu wa, koochi ni yaki o irerareru koto ο yosoo shite, biku-biku shi-dooshi da. [PG] B: 'That guy who does not come to practice regularly is a bundle of nerves [keep feeling biku-biku]. realising that the coach will probably come down on him like a ton of bricks.'

Meanings of Japanese sound-symbolic emotion words

(66)

(67)

243

Umarete hajimete no supiichi da kara, jumban ga kuru made wa, biku-biku no shidooshi datta. [GGJ] RH: 'Since it was my first speech, I kept feeling biku-biku until my turn came.' Hato ga hey a made hairikomu. Akachan no iru katei wa biku-biku desu. [AS: 13/6/1972 in GG] RH: 'The doves even come into the room. The family with a baby feels biku-biku.'

Meaning of biku-biku (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k) (1) (m)

X feels something because X thinks something sometimes a person thinks something like this: "something bad will happen to me I don't want this because of this, I want to do something I cannot do anything" because of this, this person feels something very bad this person thinks: "I am trembling/shaking [biku-biku]" this person feels this for some time X feels something like this because X thinks something like this

• something bad will happen to me Unlike hara-hara or hiya-hiya, the occurrence of danger of biku-biku is always specifically directed to the experiencer him/herself, as seen in sentence (56) to (66) ("to me"). It could be directed to someone/something which one regards as a part of oneself. For example, in sentence (67) the family feel biku-biku about the possible attack on their baby by the dove. The experiencer's expectation that something bad will happen to him/her is much higher for the person feeling biku-biku than it is for the person feeling hiya-hiya or hara-hara ("something bad WILL happen" rather than "something bad COULD/CAN happen"). One of the big differences between biku-biku and hiya-hiya is this degree of sense of danger felt by the experiencer. When one feels the danger very close and immediate, a person would say biku-biku, but not hiya-hiya. For example:

244

Rie Hasada

(68 a)

Uso ο tsuita no ga baresoo na n de. ?biku-bikulhiya-hiya na n da. Ί am feeling ?biku-bikulhiya-hiya, since it seems that my lie will be found out.'

(68 b)

Uso o tsuita no ga bareta no de. biku-bikul*hiya-hiya na n da. I am feeling biku-bikul*hiya-hiya since my lie was found out.

When one feels the danger of something bad happening very soon, as when one is sure of being accused by somebody about one's lie as in the sentence (68 b), hiya-hiya is inappropriate. Where the occurrence of danger is regarded as just a possibility as in sentence (68 a), however, hiyahiya is more appropriate than biku-biku. Therefore, biku-biku occurs often when one is near or getting closer to someone/something which one thinks will bring some dangerous situation upon one (as seen in sentences (56), (57), and (61)). For example, in sentence (60), the speaker's conversation partner feels biku-biku while being close to the speaker who usually does something bad after drinking. In sentence (61) the king feels biku-biku as he gets closer to the person involved in a quarrel, who he thinks will embroil him in the quarrel. Often, the belief or thought that we are going to be in a threatening situation is a sufficient trigger of biku-biku as seen in sentence (58), where the person feels biku-biku only by his obsessive thinking about having cancer. • I don't want this to happen The experiencer does not want this to happen. • because of this, I want to do something • I cannot do anything While the experiencer aspires to do something about the imminent danger, at the same time he/she feels the occurrence of the dangerous situation to be inevitable. For example, in sentence (61) the king would like to avoid being embroiled in a quarrel, but as a responsible person, he has to enter into a quarrel in order to stop it. Similarly, in sentence (65) the man who has neglected to practice wants to avoid seeing the coach who would scold him about his lack of practice, but he feels biku-biku since he knows that it will not be possible as long as he belongs to the club. • because of this, this person feels something very bad

Meanings of Japanese sound-symbolic emotion words

245

The intense bad feeling is stronger than that of hara-hara or hiya-hiya. • this person thinks: I am trembling/shaking The non-emotional sense of biku-biku refers to a shaking or twitching movement as seen in the following sentence. (69)

Oyoide iru toki ashi ga tsutte, biku-biku keiren shita koto ga aru. [PG] B: Once, while swimming I got a twitch [biku-biku] cramp in my leg.'

The emotional biku-biku feeling is associated with this image of a shaking movement. People do not necessarily actually tremble when they experience the emotional biku-biku feeling, like biku-biku in the phrase naishin biku-biku shinagara 'feeling an inward tremor' [the inward biku-biku} of sentence (56): this trembling is metaphorical, and not a physical movement. However, the experiencers of the biku-biku emotion feel as if they are trembling, and they sometimes actually physically tremble. Thus, this image of a 'trembling' movement is reflected in the emotional biku-biku feeling. When the person actually trembles because of this biku-biku feeling, that movement is easily observable outwardly. Therefore, the prohibition form biku-biku suruna! 'Don't feel/be biku-biku!' is possible for this feeling. By contrast, Ί hiya-hiya suruna! 'Don't feel/be hiya-hiya!' sounds odd, since the feeling is manifest only to the experiencer him/herself, and a third person cannot tell whether another person is really feeling hiya-hiya or not from external appearance. Because of the possible objective observation of biku-biku, it is often used for describing a third person's feeling manifested as outward behaviour.

3.4. Odo-odo Examples: (70)

Watakushi, nandaka, shingakki o anata to issho ni hajimeru shinnyuusei no yoo na kimochi ga shimasu wa. To itte mo, odo-odo

246

Rie Hasada

shita kimochi nankaja nakutte, chotto keii o arawasu kimochi desu no. [CS: 28] J*-E: Ί feel rather like a new boy beginning his first term with you. Not scared \odo-odo feelingl. mind you - but just, for once, in a thoroughly respectful mood...' [GMC: 36] (71)

Jidoosha no nai tokoro e itte mitai. Watashi-tachi hokoosha wa itsumo inochigake de, odo-odo shite aruite iru. [PG] Β: Ί want to go to a place where there are no cars. We pedestrians are always nervous [feeling ofto-oifo]. risking our lives when walking.'

(72)

Hitomae ni deru to odo-odo suru. [cf. KNJED] Β: Ί am shy and timid [feel odo-odo] in another's presence.'

(73)

Fuyaku no ryuukeishuu-tachi wa, odo-odo shita manazashi de hey a e haitte iki. ... [Kaikoo "ryuubooki": 230/96 in JEOMTD] J -» E: 'We exiles under sentence of hard labour came into the room with terror \odo-odo expression! in our eyes „.'

(74)

Kanojo wa nando mo kyooin-shiken ni ochi, shibaraku wa odoodo shite ita. [PG] B: 'She failed the teaching examination four times, and lacked confidence [felt odo-odo] for some time after.'

(75)

(Tomioka wa) karadatsuki to wa hantai ni, me no iro wa itsumo odo-odo shite ite. aite no kao o tadashiku seishi dekinai ki no yowasa ga am. J-*E: 'But his eyes, in spite of this well-built body had an expression of timidity [odo-odo feeling], unable to look straight at people.' [FC: 75]

(76)

Sono shoonen wa kateinai de booryoku ofuruu no de, hahaoya wa itsumo odo-odo shite iru. [PG] B: 'Since the boy behaves violently at home, this mother is always on edge, [feeling odo-odo].'

(77)

Kootoo-shimon de, kare wa odo-odo shite, toi ni kotaeru dokoro de wa nakatta. [PG] B: 'During the oral examination, he was all uptight [feeling odoodo}. and wasn't in a state to answer any questions.'

Meanings of Japanese sound-symbolic emotion words

(78)

247

Yamada wa shikenchuu kanningu no gemba o osaerare, odo-odo suru bakari de atta. [PG] B: 'Yamada was caught red-handed cheating in the exam, and was at a loss for what to do [was feeling odo-odo].'

Meaning of odo-odo (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k) (1)

X feels something because X thinks something sometimes a person thinks something like this: "something bad can happen to me at any time because I can't do much I don't want this because of this, I want to do something I don't know what I can do" because of this, this person feels something bad this person feels this for some time X feels something like this because X thinks something like this

• something bad can happen to me at any time • because I can't do much We feel odo-odo when we are lacking confidence in ourselves and presuppose something bad might happen to us because of it. For example, as seen in sentence (70), new students typically could feel odo-odo, since they feel small in an unfamiliar place and amongst unfamiliar people, anticipating something bad could happen in that circumstance. The pedestrian in sentence (71) feels odo-odo, thinking that one is powerless in comparison with the cars and that one would be easily killed if run over by a car. In sentence (74), the girl is feeling odo-odo after failing the examinations, since she feels herself incapable and without confidence, and imagines somebody might do something bad (sneering, backbiting, etc.) to her. In sentence (77), the boy feels odo-odo because he feels incompetent before the examiners, and supposes that it is possible the examiners will fail him. • I don't want this • because of this, I want to do something • I don't know what I can do.

248

Rie Hasada

While the experiencer of the odo-odo feeling definitely does not want something bad happen to them, and thus wants to take some action about it, they cannot think of a way out of the "upset" situation. • because of this, this person feels something bad The resulting feeling is "something bad".

4. Conclusion The major goal of this paper was to explicate the meaning of selected psychomimes to an extent that can give non-native speakers a greater understanding of the meaning and correct use of these words. The problems with presently available definitions of psychomimes given in dictionaries were discussed above. It was shown that neither vague analogous words from other languages, nor lengthy explanations, nor accumulations of sample examinations can show the unique nature or concept of Japanese psychomimes clearly and precisely. I have proposed the use of the NSM approach to solve this problem. In sections 2 and 3 I hope to have demonstrated that the complex Japanese-specific meanings involved in selected psychomimes can be clearly shown and made comprehensible to outsiders, when they are translated into the Natural Semantic Metalanguage based on universal human concepts and represented in the framework of a "prototype scenario". I believe that I have succeeded in showing that the complex and unique semantic concepts of Japanese psychomimes, which are usually described as 'untranslatable', are in fact translatable on the level of semantic explication with a language-independent semantic metalanguage. The extent of similarities and dissimilarities in labelling and the conceptualisation encoded in different psychomimes becomes apparent with the use of the Natural Semantic Metalanguage. It can also be seen from the above examples that many psychomimes are associated with concrete images such as those of: (1)

sound impressions The surprise-like dokiQ feeling and excited-like doki-doki feeling are linked to the sound image of one's heart beating.

Meanings of Japanese sound-symbolic emotion words

249

(2)

visual impressions The perturbed feeling of hara-hara was associated with the visual image of the manner of small light thin objects falling down.

(3)

the meaning of non-onomatopoeic words

The psychomime hiya-hiya was associated with the concrete referential image evoked by the meaning of the transitive verb hiyasu 'to cool'. These associated concrete images also contribute to the meaning of the psychomimes. They are important components involved in the meaning of psychomimes. And this concreteness helps in the understanding of the relatively abstract concepts of emotional onomatopoeic words. The onomatopoeic words constitute a fully established independent lexical group in Japanese, and are continually acquired and used throughout life by Japanese speakers. Understanding the correct use of Japanese soundsymbolic words is necessary for non-native speakers if they are to grasp some essential qualities of Japanese language, people and culture. I hope that the explication of psychomimes presented in this paper has made their meaning intelligible, and helps non-native speakers to get a good command of them. I also hope that the results of this paper will increase non-native speakers' interest in and understanding of Japanese psychomimes, which will help them to have successful heart-to-heart communication with Japanese people.

Notes 1. I would like to express my gratitude to Prof. Anna Wierzbicka and Dr. Anthony E. Backhouse for their comments on an earlier draft of this paper. It was partially prepared with the financial assistance of the Toyota Foundation. 2. The direction of transfer is always from phonomimes to mimetic words (phenomimes/psychomimes), and not vice versa (Kakehi 1992: 18). 3. The Romanisation used here follows the Hepburn system. The Hepburn system indicates a phonetically close approximation of Japanese pronunciation (and thus spells shi, chi, tsu, etc.) (for further details see cf. e.g. Backhouse 1993: 60-63). However, I use /aa, ii, uu, ee, oo/ for long vowels instead of the macron. Also, for the final mora obstruent in sound symbolic words, /Q/ is used instead of/t/, for example: jit-to -*jiQ-to; mukat-to -*· mukaQ-to 4. All translations are my own (marked RH) except when otherwise indicated. When translations of original texts were available I have used those translations. In these cases I have shown which part of the English sentence corresponds to a Japanese

250

Rie Hasada

onomatopoeic expression by inserting that expression in square brackets [ ]. I have marked the gloss for each referenced example as follows. J -> E: This indicates a translation into English of a Japanese speaker or text. J«-E: This indicates a translation from English to Japanese. B: English translation of a Japanese sentence is given in the same book/text. 5. 'There are moments when I suddenly see you are an angel.' 6. The /-ri/ suffix connotes the softer quality of the root /doki/. 7. The following words are used for the complements of associated images. They are "down", "fall" for hara-hara; "cold" for hiya-hiya; and "tremble" for biku-biku. These words can be further defined by means of NSM primes, but this is not necessary for the present purpose.

References Backhouse, Anthony E. 1993 The Japanese Language: An Introduction. Oxford: Oxford University Press. Goddard, Cliff and Anna Wierzbicka 1994 Introducing Lexical Primitives. In: Cliff Goddard and Anna Wierzbicka (eds.), Semantic Primitives and Lexical Universals, 31-54. Amsterdam: John Benjamins. Izard, Carroll E. 1977 Human Emotions. New York: Plenum Press. Inagaki, Yoshihiko 1982 Gendai no Shingo, Ryuukoogo [New words and words in vogue]. In Yoshihiko Inagaki (ed.), Gendai no Goi. Tokyo: Meiji Shoin. Kakehi, Hisao 1986 The function and expressiveness of Japanese onomatopoes. Kobe Daigaku bungakubu kiyoo, 1-12. Kimizuka, Sumako 1967 A Contrastive study of the Japanese Structural Device Onomatopoeia + verb' and Certain English Verbs. Journal-Newsletter of the Association of Teachers of Japanese. Vol. III. No. 3. September: 12-16. Lebra, Takie Sugiyama 1976 Japanese Patterns of Behaviour. Honolulu: University of Hawaii Press. Leibniz, Gottfried Wilhelm 1949 New Essays Concerning Human Understanding. La Salle: Open Court Publishing Company. [1704]

Meanings of Japanese sound-symbolic emotion words

251

Martin, Samuel E. 1964 Speech labels in Japan and Korea. In: Dell Hymes (ed.), Language in Culture and Society: A reader in linguistics and anthropology. New York: Harper and Row. Mito, Y. and H. Kakehi 1981 Nichi-Ei Taishoo Onomatope Jiten [Japanese-English Comparative Onomatopoeic Dictionary]. Tokyo: Gaku Shoboo. Morita, Kiichi 1988 Gitaigo no tokusei ni tsuite [On the characteristics of mimetic words]. Iwate-ikadai-kyooyoo-nempoo 23: 107-118. Ono, Shuuichi (ed.) 1989 A Practical Guide to Japanese-English Onomatopoeia and Mimesis. Tokyo: Hokuseidoo. Ootsubo, Heiji 1989 Giseigo no Kenkyuu [A Study of Onomatopoeia]. Tokyo: Meiji Shoin. Shibatani, Masayoshi 1990 The Languages of Japan. Cambridge: Cambridge University Press. Tamamura, Fumio. 1982 Goi oyobi go no koozoo [Vocabulary and word structure]. In: Kooza Nihongogaku, vol. 12. Gaikokugo to no Taishoo, 1-21. Tokyo: Meiji Shoin. Wierzbicka, Anna 1986 Introduction. Journal of Pragmatics 10 (5): 519-534. 1991 Japanese Key Words and Core Cultural Values. Language in Society 20: 333-385. 1992a Semantics, Culture, and Cognition: Universal Human Concepts in Culture-Specific Configurations. New York: Oxford University Press. 1992b Talking about Emotions: semantics, culture and cognition". Cognition and Emotions 6(3/4): 285-319. 1992 Defining Emotion Concepts". Cognitive Science 16(4): 539-582. 1994 Cognitive domains and the structure of the lexicon. In: Lawrence Hirschfeld and Susan Gelman, eds. Mapping the Mind: Domain Specificity in Cognition and Culture, 431-452. Cambridge: Cambridge University Press.

SOURCES CITED - JAPANESE AS BP

Asahi Shimbun: Shukusatsuban [Asahi Newspaper: compact edition]. Tokyo: Asahi Shimbunsha. Beyond Polite Japanese: A Dictionary of Japanese Slang and Colloquialisms. 1993. Yoneyama, Akiko. Tokyo: Koodansha.

252

Rie Hasada

CS

FK FKB FSB GG GGJ GGT

HM JEOMTD

JF KKA KNJED KY NS PG

T TB

Chippsu Sensei Sayoonara [Good-bye, Mr. Chips]. 1956. Hilton, James. Translated from English by Jyuuzaburoo Kikuchi. Revised ed.; Tokyo: Shinchoosha, 1966. Fujin Kooron [Women's Public Opinions]. Tokyo: Chuuoo Kooronsha. Funanori Kupu-Kupu no booken [The Adventures of Kupukupu the Sailor]. 1975. Kita, Mono. Tokyo: Shinchoosha Flip, Slither, and Bang. 1993. Fukuoka, Hiroko. Tokyo: Koodansha Giongo-Gitaigo Jiten [Onomatopoeia and Mimesis Dictionary]. 1974. Yasushi Amanuma (ed.). Tokyo: Tokyoo-doo. Giongo-Gitaigo Jiten [Onomatopoeia and Mimesis Dictionary]. 1978. Asano, Tsuruko and Haruhiko Kindaichi (eds.). Tokyo: Koodansha. Giseigo-Gitaigo Tsukai-kata Jiten [Usage Guide to Japanese Onomatopoeias]. 1993. Atooda, Toshiko and Kazuko Hoshino (eds.). Tokyo: Sootakusha. Hashire, Merosu [Run, Melos]. 1967. Dazai, Osamu. Tokyo: Shinchoosha. "Kokyoo" [Homecoming]. 207-224. Wa-Ei Giongo-Gitaigo Honyaku Jiten [Japanese-English Onomatopoeia and Mimesis Translation Dictionary]. 1984. Fjita, Takashi and Shinichi Akiho, (eds.). Tokyo: Kinseidoo. Japanese for Foreigners: adverbs. 1987. Kayaho, N. et al Tokyo: Aratake Shuppan. Kagami no Kuni no Arisu. 1959. Carroll, Lewis. (Translated from English by Tadataka Okada.) Tokyo: Kadokawa Shoten. Kenkyuusha's New Japanese-English Dictionary (Fourth edition/ 27th Impression) 1993. Masuda, Koh (ed.). Tokyo: Kenkyuusha. Kanashii Yokan.[Melancholic Premonition]. 1988. Yoshimoto, Banana. Tokyo: Kadokawa. Ningenshikkaku [No Longer Human]. 1952. Dazai, Osamu. Tokyo: Shinchoosha. Nichi-Ei Giongo-Gitaigo Katsuyoo Jiten [A Practical Guide to Japanese-English Onomatopoeia and Mimesis]. 1989. Ono, Shuuichi (ed.). Tokyo: Hokuseidoo. Tsugumi. 1989. Yoshimoto, Banana. Tokyo: Chuuoo Kooronsha. THINK BIG! 1988. Terasawa, Yoshio. Tokyo: Shufunotomosha.

SOURCES CITED - ENGLISH AKS

The Adventures of Kupukupu the Sailor. 1985. Kita, Morio. (Translated from Japanese edition by Ralph F. McCarthy.) Tokyo: Koodansha International.

Meanings of Japanese sound-symbolic emotion words DO

FC GBC NLH RM

TLG

253

Dazai Osamu: Selected stories and sketches. 1983. Dazai, Osamu. (Translated from Japanese edition by James O'Brien.) New York: China-Japan Program, Cornell Univ. "Homecoming." 161-173. Floating Clouds. 1965. Hayashi, Fumiko. (Translated from the Japanese edition by Y. Koitabashi. and M. C. Collcutt.) Tokyo: Harashoboo. Good-Bye, Mr. Chips. 1937. Hilton, James. 8th ed. London: St. Paul's House. No Longer Human. 1958. Dazai, Osamu (Translated from Japanese by Donald Keene.) New York: New Directions Publishing Corporation. Run, Melos!: and other stories. 1988. Dazai, Osamu. (Translated from Japanese by Ralph F. McCarthy.) Tokyo: Koodansha International, Ltd. "School Girl". 43-101. "Through the Looking-Glass." In Lewis Carroll, The Complete Works of Lewis Carroll. 1939. London: Nonesuch Library. 126-250.

Concepts of anger in Chinese Pawel Kornacki

When his mother saw me, her eyes immediately became two fireballs. She jumped at me and jabbed at my forehead with her finger. "So you are Li Lu! You are Li Lu!" She grabbed me and hit me again. How I hated her! My face was burning with rage. At any other time I would have bitten the fat hand that was grabbing my collar. But I remembered my teacher's story and suppressed my anger. (Li Lu Moving the mountain, 1990: 42)

1. Introduction This paper focusses on five Mandarin Chinese words - nu, sheng/qi, nao(huo),fen, taoyan - as well as their figurative associations and elaborations, which are pertinent to the conceptualisation of the "emotions" (qingxulqinggan) often rendered with, or comparable to the English words 'angry' or 'anger'. Ever since Darwin's (1872/1979) classic treatment of emotions, "anger", "something like anger", or "a family of anger concepts/ expressions" have been recurrently proposed by a number of Western psychologists as one of the "fundamental", "universal", "primitive" or "basic" human "emotions" (cf., e.g., Ekman 1992, 1993, Oatley 1992 for the pro arguments, and Russell 1991, 1994, van Brakel 1994, Wierzbicka 1992 a and b, Wierzbicka 1999 for a recent contra). My aim here will be to examine some of the insights which the examination of the Chinese lexical data might contribute to this debate.

2. Nu Tian zi zhi Heaven son of bai wan, hundred 10 thousand

nu, nu liu flow

fu s hi lie corpse xue qian li blood thousand mile

256

Pawet Kornacki

Yong fu zhi nu, mian guan common man of nu remove hat tu xian, yi tou qiang di bare foot with head scratch ground 'When the Son of Heaven becomes nu, corpses fall in millions, and the blood flows for thousands of miles. When a common man becomes nu, he throws off his hat, bares his feet, and scratches the ground with his head.' (from Tang Sui's "A Discourse on nu", 3rd cent. B.C., quoted in Zhang 1947/1985: 154) Nu is a powerful, overwhelming and possibly violent feeling. Han-Ying Cidian (The Chinese-English Dictionary 1988: 502) glosses it as 'anger, rage, fury', and Bucher (1986: 422), Matthews (1960: 661) and Morohashi (1960), suggest 'anger, passion, rage' as the most suitable equivalents. Although this paper deals with the contemporary concept, it is perhaps worth mentioning that nu could be consistently found on the different lists of the (universal) "human feelings" (ren qing} which appeared in various Classical Chinese writings and - even though nu is certainly not the most common of the Chinese 'anger'-concepts, it is nevertheless a fairly basic word, and it can be found in such dictionaries as Xinbian Xiao Xuesheng Zidian (The New Character Dictionary for Primary School Students, 1986: 262), and Hanyu Changyongzi Zidian (The Dictionary of Frequently Used Characters, 1977: 333). Interestingly, nu is also the word with the biggest (of all five) number of figurative associations which testify to its salience, provide valuable clues to its meaning, and constitute a fairly coherent cultural model of the manifestations, symptoms and effects of nu. Before proceeding with the explication of this concept, I am, therefore, going to analyse a number of such expressions, subsuming them under three "narrative slots" (cf. Shweder 1991: 425-426), which reflect, roughly, the communicative, controlling, and phenomenological aspects of nu. The following "fixed sayings" (chengyu) describing nu were gathered from Hanyu Chengyu Fenlei Da Cidian (The Big Chinese Dictionary of the Classified Chengyu 1987: 204-209).

Concepts of anger in Chinese

257

2.1. Communicating nu As it appears from a number of expressions referring to nu, this emotion can rather easily be noticed in another person. Generally speaking: (1)

nu xing yu se nu form/appear in face 'Nu appears in one's face/one becomes nu'

Nu, in fact, may 'hold' or 'fill' one's face completely: (2)

nu rong man nu hold full/whole 'The face is full of nu:

mian face

In such a case, one can also observe redness in the face and bulging of the neck: (3)

nu de lian hong bozi cu nu ASSOC face red neck thick 'So nu that one's face becomes red and the neck bulges.'

The face is an especially prominent part of the body where nu appears, and the following expressions refer to the specific parts of the face - eyes, eyebrows, and teeth - which are associated with nu, and (presumably) by virtue of which nu can be discerned: (4)

nu de heng mei shu yan nu ASSOC horizontal eyebrow vertical eye 'So nu that one frowns and scowls.'

(5)

heng mei nu mu horizontal eyebrow nu eye/look 'frowning brows and nu eyes/look'

Gnashing, as well as showing one's teeth, are signs of nu: (6)

nu de yao ya nu ASSOC grit tooth 'gnash one's teeth in nu'

qie gnash

chi tooth

258

(7)

Pawel Kornacki

nu ya tu nu tooth protrude 'showing teeth in nu'

zui mouth

Yet another aspect of one's countenance involved in showing nu is hair, as in (8) and (9): (8)

nu de ling ren fa zhi nu ASSOC make person hair stand 'so nu that it makes one's hair stand on end/bristle'

(9)

nu fa chong guan nu hair raise cap 'so nu that hair lifts one's cap'

Nu can also be communicated with one's hands, firstly, as reflected in the expression nubi (lit. «w-raise hands) -by raising one's hands 'in nu', or, by clenching one's fists and also baring the teeth, as in (10): (10)

nu de wo quan tou zhao nu ASSOC clench fist show fang 'so nu that one clenches the fists and bares the teeth'

Giving expression to nu may involve an amount of 'verbal violence' - cursing and shouting, as in (11), (12), and (13):

(11)

nu ma bu xiu nu curse not rest 'cursing nu-\y without stopping, revile'

(12)

nu de pao xiao ru nu ASSOC roar roar like 'to roar like thunder with nu'

(13)

nu de da fa lei ting nu ASSOC big send thunder- clap 'to thunder with nu'

lei thunder

However, nu can also be communicated with the eyes only, and the silence which ensues then, heightens the tension of the whole situation:

Concepts of anger in Chinese

(14)

259

nu s hi bu yu nu look not speak 'to give nu looks and say nothing'

2.2. Management and control of nu Nu is hardly an easy emotion to subdue. This idea is expressed in the following two phrases: (15)

nu bu zi sheng nu not self conquer 'nu beyond self-control'

(16)

nu bu ke e nu not can stop 'so nu that it can't be stopped'

The idea that nu implies lack of control can further be corroborated by the fact that nu can be ascribed to a number of the natural phenomena which are characteristically violent and uncontrollable. In such contexts the difference between nu and 'anger' is especially prominent, since in translating these expressions one is usually forced to speak of 'fury' or 'rage', rather than simply 'anger'. Thus, for instance, water, wind and fire can be said to exhibit nu, as in the following stock descriptions: (17)

kuang feng da zuo, hai shui nu hou mad wind big rise sea water nu roar Ά violent wind is rising, the sea roars with nu.' (?angrily/with rage)

(18)

nu tao xiong yong nu wave rush tempestuous "The nu (?angry/raging) waves rush forcibly.'

(19)

kuang feng nu hao mad wind nu howl Ά violent wind is howling with nu.' (?anger/rage)

(20)

ji feng nu wild wind nu

huo fire

260

Pawel Kornacki

'a surge of wild wind and nu (?angry/raging) fire' (which may be used both about a natural calamity and a "raging" person A quite interesting piece of lexical evidence which also favours the hypothesis that the meaning of nu contains a component referring to the impossibility of self-control comes from the compound numa (lit. «w-horse), which describes a horse that is hard to mount or ride, kicks, and is otherwise "hard to control", or "spirited". Continuing with the idea of the essential impossibility to control oneself while being nu, it might seem quite illuminating to observe that, as mentioned before, when nu, people can either stare speechlessly at each other (as if doing something bad to another person with one's eyes), as in (21): (21)

nu mu xiang shi nu eye mutually look 'to eye each other with nu\

or else, curse, shout, and otherwise "thunder", i.e., do something bad (possibly to another person) with one's words. One is, however, hardly able to speak intelligibly "with nu" (cf. one of the previous examples), and, even more tellingly, one can be so overwhelmed by nu, that one cannot give a coherent reply to the opponent at all, as the following chengyu suggests: (22)

nu bu neng hui nu not can answer 'to be so nu that one can't reply

Congruent with the above suggestions is also the fact that nu is also connected with notions such as the inability to control one's speech, thinking or actions through the associations with insanity. One can thus speak of kuangnu (lit. mad-Λ«), in which case, according to Zhang (1947/1985: 162), "a person cannot withhold the destructive impulse, wounds or kills people, sets fire, commits suicide, or scatters and destroys things".

Concepts of anger in Chinese

261

2.3. The feelings of nu Nu is like fire - both in the experience of the person who feels it, and in its effects on the people or things that are affected by it. There are quite a few expressions which metaphorically link nu with fire and smoke, the most common of them being probably this: (23)

nu huo zhong shao nu fire middle burn 'to be burning with nu'

Although this metaphor appears to be parallel with the English 'to be burning with anger', there seems to be a subtle, but nevertheless interesting difference between the two. The Chinese expression suggests that one is as if in the middle of the burning fire (i.e., the fire is Outside' of the person), into which one is perhaps momentarily thrown - it is also important to note here, that the Chinese expression fa nu (lit. send nu, i.e., get nu - with the inceptive meaning) differs from its usual English translation 'get angry' in that it cannot be combined with the durative aspect suffix -zhe - so that the translation of an English sentence like He is getting angry about it has to be done either with the perfective aspect suffix -le, or the inceptive affix -qilai (in both cases the Chinese sentence means then something like 'He has become m/')> whereas the following sentence, which would be a literal translation of the English one, appears to be unacceptable: (24)

*Ta yin na-jian shi er he because that-CL thing CONJ '*He is getting nu about it.'

fa send

-zhe nu -ASP nu

Thus, nu is also a very sudden feeling - a conjecture which may be supported by an expression such as baonu (lit. violent/explode nu), where again nu is conceptualised as a momentary and violent occurrence. Yet another construal, with the nu person as the source of fire and smoke (and, by extension, a harmful and destructive agent), is also possible, as the following expressions indicate: (25)

nu de huo mao san zhang nu ASSOC fire send three zhang 'so nu that one sends fire for three zhang (about ten meters)'

262

(26)

Pawel Kornacki

nu de bitou chu huo nu ASSOC nose come out fire 'so nu that fire comes out of the nose'

Again, eyes can be symbolically linked with nu, as in (27): (27)

nu de Hang yan mao nu ASSOC two eye send 'so nu that one's eyes send fire'

huo fire

So that ultimately: (28)

nu de qi qiao sheng yan nu ASSOC seven aperture produce smoke 'so nu that smoke comes out from the "seven apertures'" (eyes, ears, nostrils and the mouth)

In the system of correspondences underlying traditional Chinese medicine (cf. Porkert 1976, Ots 1990: 40) nu was associated with the liver- as evidenced in sayings like gan zhu nu (lit. liver hosts nu), or nu shang gan (lit. nu hurts the liver). Yet, as pointed out by a number of scholars (cf., e.g., Lin 1981: 101), nu (on a par with other feelings) was not considered necessarily harmful, except when excessive and upsetting the balance of yin - yang. It may be, therefore, interesting to observe that in the modern colloquial perception nu seems to be portrayed as having a rather generally bad - physical and social - influence. Thus, for example, Wang (1993: 75) in a popular 'self-help' article on "Psychological methods of controlling nu" suggests such remedies as repeating to oneself injunctions like 'Don't become nu, cool down' (Bu yao fa nu, lengjing xialai), 'Getting nu can make things worse' (Fa nu hui ba shiqing ban huai), or, putting up comments such as 'Calm your nu' (Xinu), and 'nu hurts your body/health' (Nu shang sheri) above one's office-desk.

2.4. nu vs. angry A Chinese psychologist Zhang Yaoxiang (1947/1985: 155-156) offers the following revealing comment on the reasons and causes for becoming nu:

Concepts of anger in Chinese

263

The causes of nu are most complex. Being insulted or cheated, disappointments (such as in love), being hampered in one's initiatives, experiencing opposition or resistance to one's words, unfairness, envy, the situation where our good intentions are misunderstood by others, having one's weaknesses pointed out by other people, being informed against after committing an unlawful deed, having one's secrets revealed by others, one's rights infringed, observing certain striking physical characteristics of someone whom we habitually cannot stand (taoyan) (for example, especially tall or short, fat or thin, big-eyed, with a pockmarked face), or some of his peculiar movements (such as, for instance, limping, unfocused eyes, manner of swallowing, weird pronunciation, etc.), bizarre clothes, smoking...

To this list Zhang (1947/1985: 155-156) adds also fatigue resulting from menial jobs, quoting the following saying:

(29)

ren kun ze duo person fatigue/stressed therefore much 'Much nu results from fatigue.'

nu nu

Gathering the above clues together, and in order to proceed with further discussion, I would propose the following explication:

nu (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k)

X feels something X thinks something like this: something happened this is bad I don't want this because of this, I want to do something very bad to someone now I can't not do this X feels something very bad because of this because of this, X can't think now something is happening inside X because of this something bad can happen to X because of this

Arguably, the prototypical case of 'anger' evokes the idea of "bad action" (i.e., "someone did something bad" - see, for example, Wierzbicka 1992b: 569, Goddard 1991: 268). This, however, is not necessarily the case

264

Pawel Kornacki

with nu (although, as Zhang's list makes clear, it is by no means incompatible with it). Consider the following example: (30)

Yi jian qing laozi hing one see young mule illness cheng na-ge yangzi, zhe-wei lao become that-CL shape this-CL old siyangyuan like fa-qi nu -lai animal-keeper at once send-raise nu come 'Upon seeing that the young mule's illness became so grave, the old animal-keeper immediately became nu.' (You et al., 1986: 362)

As this example shows, neither the idea that "someone did something bad" nor that "something bad happened to me" are necessary for nu - the mule's illness wasn't, so to speak, somebody's action ("against" the keeper), neither was it something that actually happened TO HIM (although he was, of course, affected by it). To see that it is also not necessarily the case that "something BAD happened" when one becomes nu, let us reconsider some of the situations mentioned by Zhang. Thus, for instance, when a person becomes nu upon seeing someone 'hateful/disgusting' (taoyan) to them, or when he or she catches sight of a weirdly dressed individual, the focus seems to be on the feelings which arise in that person because of these events, and which are caused by that person's "negative assessment" ("this is bad") and "disagreement" ("I don't want this") in the situation, rather than an inherent feature of it - in which case nu seems to resemble something closer to 'shocked' rather than to 'angry' - hence it seems more accurate to "split" an otherwise plausible component "something bad happened". It is, moreover, not really clear whether "something" should "happen" necessarily "NOW" - since the onset of nu could be either an unexpected, singular, and salient event (such as someone's insult, a quarrel, finding oneself abandoned, noticing something hateful), or a protracted period of toil. Nu can be a powerful and uncontrollable emotion. The metaphorical references to the natural phenomena beyond human influence (cf. the nu of the waves, wind, and fire versus the unacceptable *'anger' of the waves, wind, or fire) on the one hand, and certain fairly unequivocal non-verbal manifestations (bared teeth, clenched fists) of nu, embedded in a number of lexicalised expressions, strongly suggest the will of immediate aggression

Concepts of anger in Chinese

265

directed at a person - rather than at an object - as it is spelled out in component (f), which is certainly more suggestive of something like 'rage' rather than just 'anger'. The sense of being as if possessed or overwhelmed by nu (i.e., by something like a natural element over which people have no real control) is expressed in the rest of the explication. Thus, component (g) represents the irrepressibility of the urge to DO something destructive right away - NOW - (as suggested in Section 2.2.). Components (h) - (k), which hint at the more "phenomenological" aspect of nu, portray it first of all as something that happens TO a person (as well as, perhaps more specifically, INSIDE a person), and convey its sudden, harmful ("something bad can happen to X"), and rather debilitating character (inhibiting the ability to think/speak coherently) (see also Roth 1996: 138-139).

3. Shenglqi The concept of qi constituted one of the basic notions of traditional Chinese medicine and philosophy (cf. Porkert 1976: 167-8). As far as its contemporary colloquial usage is concerned, Han-Ying Cidian (The ChineseEnglish Dictionary, 1988: 535-6) lists eleven related meanings of this word: (1) gas, (2) air, (3) breath, (4) smell/odour, (5) weather, (6) airs/ manners, (7) spirit/morale, (8) make angry/enrage/annoy, (9) be angry/enraged, (10) suffer wrong [as in shou qi, lit. 'receive qi'], and (11) energy of life. According to Ots (1990: 40): Qi resembles the Greek concept ofpneuma and the Indian prana. The air we inhale is at the same time our life giving force. Qi is a monistic concept that transcends the modern notions of organic and inorganic, material and abstract, psyche and soma. SHENGqi literally means "to give birth (rise) to qi'; its semantic meaning is 'to get annoyed', or 'to get angry'.

Shenglqi is, arguably, the most common, everyday Chinese word roughly comparable to the English 'angry'. Shaver et al. (1992: 193), who in their sorting study of the Chinese emotion concepts gloss it simply as 'anger', found it to be one of the emotion words "understood by a majority of children in the 30- to 35-month group". Xiandai Hanyu Cidian (The Dictionary of Modern Chinese, 1988: 1026) defines sheng/qi as follows:

266

(31)

Pawel Kornacki

Yin bu he xinyi er bu yukuai because not suit intention CONJ not happy 'unhappy as the result of things not following one's intentions/ wishes'

According to several informants, shenglqi refers first of all to the bad feelings caused by the unpleasant shenghuo de xiaoshi ('small things of everyday life'), things which in the first instance make one bu gaoxing ('not happy'), i.e., things which generally qiren (lit. ^/-person, 'make one qi'), such as: unruly children, long queues, disagreements with parents (or children), having lazy pupils, bumping into someone while riding a bike, having something of one's personal possessions (e.g., a pen, a newspaper, or a book) borrowed by someone around without asking for permission, being teased, or insulted. Shenglqi seems to be associated especially often with children - their (typical) misbehaviour is certainly more likely to cause their parents to feel this emotion, rather than, say, nu. Thus, for instance, a father who has just noticed that his son broke their neighbours' window with a stone, exclaims: (32)

Erzi! son xue blood

Ta he yi one

qi qi re, hot

de hun shen de ASSOC whole body ASSOC erzi gan shenme, ni! son do what you

'Son! He was so qi that the blood in his whole body became hot. Son! What are you doing!' (Xiaoshuo Lin 1987/4: 48) Similarly, one is quite likely to say that: (33)

Zhe haizi you qi-le ta mama le this child again qi-ASP s/he mother PRT 'This child made his/her mother qi again.'

whereas it would be very strange indeed to say: (34)

UZhe haizi you shi-le ta mama fa nu this child again make-AP s/he mother sendnw '??This child made his/her mother nu again.'

Concepts of anger in Chinese

267

Children are therefore often reminded that:

(35)

qi laoshi I fumu bu dui qi teacher parents not correct 'It is wrong to qi one's teachers/parents.'

Conversely, shenglqi is also associated with a typically "childish" way of reacting to unwanted events - as one informant reminisced, once, upon not being allowed to go with his sister to visit the relatives: (36)

Wo zai wu-li sheng-zhe qi ne I there.be room-in rise-Asp qi ASP Ί was in the room being shenglqi' (sheng/qi-ing)

The aspectual marking of shenglqi (especially when contrasted with nu) provides some interesting clues to its meaning. Firstly, it can be combined with both durative aspect markers zai and -zhe (cf. the above example), in which case the emotion is conceptualised as extending over a period of time, so that one can say that, for example: (37)

Ta jintian zai shenglqi s/he today ASP riseAjri 'S/he is being shenglqi today.' (roughly similar to the English 'sulking')

Furthermore, shenglqi can be the object of conscious restraint and control. According to Ots (1990:40), 'When a Chinese says wo sheng qi le (literally: Ί got angry'), this does not mean that he was able to vent his anger, but that the anger was kept inside. An interesting corroboration of Ots's statement can be found in the expression: (38)

bie-zhe yi duzi qi hold.back-ASp(dur) one stomach qi 'to be keeping one's qi held back in one's stomach'

The phrase points to two important aspects of shenglqi: first, that it can be naturally seen as extending over a period of time - which makes it different from nu, typically construed as a momentary, or a short-time event, and second, that it can be plausibly viewed as something that happens IN-

268

Pawel Kornacki

SIDE a person. The relevant word here is bie, translatable as 'suppress, hold back, suffocate'. According to Bucher (1986: 338), it has strongly "physiological" associations, as it is used first of all about the suppressed natural needs, as in (39): (39)

Fujin mei you cesuo, ta xiang around not have toilet he want xiaobian wu nai zhi de bie-zhe urinate no way only must hold-DUR 'There was no toilet around, and there was no way but to hold it.'

Interestingly, Lin (1972: 374) glosses bie-zhe qi as 'to smoulder with resentment', which clearly focuses on the possible long-term duration of this feeling. There is also a contrast between this expression and the apparently parallel xinu (lit. cease-ww) in that the former evokes the image of storing an expanding amount of 'air/energy' (qi) inside one's stomach, whereas the latter refers simply to the cessation of««, rather than, so to speak, its 'storage', which is confirmed by the fact that one cannot combine this phrase with the durative aspect marker, cf. the unacceptable *xi-zhe nu. Several lexicalised expressions pointing to shenglqi also appear to suggest that whereas nu was conceptualised first of all as oriented Outwards' in its destructive/aggressive aspects, qi is rather self-oriented, and harmful mainly to the experiences Thus, to express that one was qi to an extreme degree one speaks of 'dying with qi" (qisi, lit. '