Jean Harkins, Anna Wierzbicka - Emotions in Crosslinguistic Perspective

Emotions in Crosslinguistic Perspective
WDE
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. Psycholin-
guistics. 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: Lderitz & Bauer, Berlin
Printed in Germany
Contents
Introduction 1
Anna Wierzbicka and Jean Harkins
Testing emotional universals in Amharic 35

Mengistu Amberber
Emotions and the nature of persons in Mbula 69

Robert D. Bugenhagen
Why Germans don't feel"anger" 115

Uwe Durst
Linguistic evidence for a Lao perspective on facial expression

of emotion 149
N. J. Enfield
Hati: A key word in the Malay vocabulary of emotion 167

Cliff Goddard
Talking about anger in Central Australia 197

Jean Harkins
Meanings of Japanese sound-symbolic emotion words 217

Rie Hasada
Concepts of anger in Chinese 255

Pawei Kornacki
Human emotions viewed through the Russian language 291

Irina B. Levontina, Anna A. Zalizniak
A culturally salient Polish emotion: Przykro (pron. pshickro) . . . . 337

Anna Wierzbicka
vi Contents
An inquiry into "sadness" in Chinese 359

Zhengdao Ye
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 lo-
cations 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 dif-
ferent 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 with-
in 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 ex-
perience of the person who has it, that it has rightly been questioned
whether it is even reasonable to think that there is necessarily, or demon-
strably, 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 con-
text in which one person feels anger may evoke a different kind of emo-
tional response in another person. Even if there is a general social or cul-
tural 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 hap-
pened, 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 physio-
logical 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 lan-
guage, and it is only language that enables people to use the same word,
anger or angry, to designate two different instances of an emotional experi-
ence. 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 dic-
tionaries 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/w-
rieux (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 atten-
tion was devoted to feelings comparable to those described in English as
angry, while more socially cohesive emotions received greater cultural em-
phasis. The passionate feeling of liget described by Rosaldo (1980) in the
Philippines had some similarities to anger, but was seen by Ilongot speak-
ers in a more positive light, as a constructive force motivating much worth-
while human activity.
Links between cultural identity and emotions identified by particular
words, often held to be untranslatable, have long been recognised. They
have often been seen as an important key to understanding cultural patterns

and motivations, as for example in Benedict's (1946) discussion of the Jap-
anese emotion haji in comparison with English shame and guilt. A Japan-
ese psychiatrist has devoted a whole book to analysing "a particularly Jap-
anese emotion" (Doi 1981: 169), amae, which he views as central to per-
sonality structure and social relations in his own culture. This is a good
example of a practically untranslatable emotion concept, discussed by vari-
ous 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 cat-
egory, and there are many groups of people in the world who neither recog-
nise 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 mean-
ings, 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 poststruc-
turalist critiques of essentialist approaches to personality and to culture. If
selves are understood as shifting and plural, it may become difficult to ident-
ify 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 inter-
dependency 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 in-
tended to illustrate that examining these words and the contexts in which
they are spoken can lead to a better understanding of the meanings associ-
ated 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 equival-
ents in other languages or in other cultural contexts. Neither do these
studies assume that any society is so homogeneous as to have a single, uni-
tary 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 ex-
pressions. 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 Eng-
lish word like anger and its equivalents or near-equivalents in other lan-
guages. 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 suffi-
ciently general to serve as a basic equivalent for angry, as each is appropri-
ate in some contexts but not others. The three words focus on different as-
pects of angry feelings and behaviour: pikaringanyi is associated with ac-

tive 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 gener-
alise by giving it a single label the way English speakers do.
Another language, Malay, has a single word, marah, that is usually trans-
lated 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 be-
haviour 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 some-
thing 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. What-
ever does emerge as a possible "emotional universal" in this area will
need to be identified in universal terms, not in terms of the English emo-
tion lexicon.
The papers in this volume focus on words and other linguistic ex-
pressions related to emotions from a diverse selection of languages, and
examine in depth the linguistic evidence concerning their meanings. Atten-
tion 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 de-
ployed 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 el-
ements are shared with similar words and expressions from other lan-
guages.
While this book is written in English, we explicitly recognise the limi-
tations 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 lan-
guage-specific word or concept, such as English anger or Malay marah (in
italics), an English gloss such as 'anger' (in single quotation marks), or el-
ements 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 lan-
guages and cultures.
3. Emotion concepts as cultural artefacts

The semantic differences among words for anger-like feelings in the lan-
guages 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 differ-
ences 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
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 pro-
duce an adequate description of an emotion one must include both situ-
ational or social and cognitive elements in addition to the obvious affective
elements. If this is correct, then we must indeed regard anger, marah, pi-
karinganyi 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 de-
scribed.
There are of course many scholars who claim that nothing is truly self-
explanatory and everything is culture-dependent, and who appear to ac-
cept and even to rejoice in the idea that there is no way out of a "herme-
neutical circle". Charles Taylor (1979 [1971]: 34) applied this idea specifi-
cally 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 can-
not 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 out-
siders are possible at all.
The crucial point is that while most concepts (including toska, zalet',
fago, shame, emotion, implication) are complex (decomposable) and cul-
ture-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', SKA-
ZAT', 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 compari-
sons 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 maxi-
mally 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 "emo-
tion 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 lan-
guage) 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
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 emo-
tions across languages and cultures is still deeply entrenched in emotion re-
search 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 uni-
versal emotion' because the word 'sadness' has a culture- and language-
specific meaning, and can therefore not be used to refer to universal mean-
ings". 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 lim-
ited and universal, but that there is no universality in the corresponding subjective ex-
perience. That is, the universals are of an entirely technical nature, they require par-
ticular technical instruments and theories and do not correspond with intuitive ex-
periential categories accessible to the layperson. [... ] The emotions are experiential
universals, that is, all people are in principle able to entertain a set of similar subjec-
tive 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" experi-
ences: "what a person experiences is not a separable set of such emotion
components, but an inseparable subjective unity. It requires technical ex-
pertise 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 con-
structions) 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 experi-
ence of people in other cultures - or, for that matter, that of most people
in English-speaking cultures. "Technical universals" of emotion (what-
ever they are) cannot throw light on ordinary human experience and con-
ceptualisation 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 de-
pendence on any one natural language as the source of "common sense in-
sights" 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 con-
cepts 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 (Bugen-
hagen 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-
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, lan-
guage-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 culture-
free analytical tools. On the other hand, conceptual primitives such as
'good' and 'bad', or 'want', 'know', 'say' and 'think' are not cultural arte-
facts 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, cul-
ture-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 hy-
pothetical conceptual primitives, lexicalised in all languages of the world,
looks something like this:
Substantives I, YOU, SOMEONE (PERSON), SOMETHING

(THING), PEOPLE, BODY
Determiners THIS, THE SAME, OTHER
Quantifiers ONE, TWO, SOME, MANY/MUCH, ALL
Attributes GOOD, BAD, BIG, SMALL
Mental predicates THINK, KNOW, WANT, FEEL, SEE, HEAR
Speech SAY, WORD, TRUE
Actions, events, movements DO, HAPPEN, MOVE
Existence and possession THERE IS, HAVE
Life and death LIVE, DIE
Logical concepts NOT, MAYBE, CAN, BECAUSE, IF
Time 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 VERY, MORE
Taxonomy, partonomy KIND OF, PART OF
Similarity 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 Eng-
lish, Spanish and Malay words (good and bad, bueno and malo, baik and
buruk), and require that semantic formulae including these words be test-
able 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" sen-
tences 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" con-
cepts, 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 plaus-
ibly 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: some-
thing bad happened" also appear to be universally present and can be as-
signed as plausible semantic components to emotion words such as, for
example, the English sad and the Russian grust', helping to map the simi-
larities 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)
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 ex-
plained 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 dep-
ends 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 as-
sume 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 cer-
tain 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 descrip-
tion we can represent these meanings as follows:
Introduction 15
Pleased (X was pleased)

(a) X felt something because X thought something
(b) sometimes a person thinks:
(c) "something good happened
(d) I wanted this to happen"
(e) when this person thinks this, this person feels something good
(f) X felt something like this
(g) because X thought something like this
Delighted (X was delighted)
(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)
(c) "I don't know what will happen
(d) some good things can happen
(e) I want these things to happen"
(f) when this person thinks this, this person feels something good
(g) X felt something like this
(h) because X thought something like this
Fear (Xfeltfear)
(c) "I don't know what will happen
(d) some bad things can happen
(e) I don't want these things to happen"
(f) when this person thinks this, this person feels something bad
(g) X felt something like this
(h) because X felt something like this
In these semantic formulae, the symbol "X" stands quite simply for "this
person". This symbol is used for convenience, because in reading the for-
mulae quickly, confusion might arise between the "this person" of the ut-
terance 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 represen-
tations 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 intro-
spective vocabulary of the seeker", which in turn depends on his or her lan-
guage and culture (James 1890: 485). James held that feelings represented
the subjective experience of biological (physiological) events, but he re-
cognised 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" ana-
lytical 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 con-
temporary English. "Emotion words" such as anger reflect, and pass on,
certain cultural models; and these models, in turn, reflect and pass on valu-
es, preoccupations and frames of reference of the society (or speech com-
munity) 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 de-
mocratisation of society and the passing of the feudal order (cf. de Tocque-
ville 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" adjec-
tives 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 Hbler 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 en-
coded 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 se-
mantics 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
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',/7os '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, anthropologi-
cal, 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 cul-
tural 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 what-
ever, 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 demon-
strates the vital contributions that can be made by means of detailed analy-
sis 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 cogni-
tive linguists like, for example, Niemeyer (1997), Athanasiadou and Taba-
kowska (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 "rep-
resents 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 (Wierz-
bicka), and the semasiological-structure approach (Geeraerts/Gronde-
laers)." On the other hand, other scholars have strong ideological objec-
tions 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 cog-
nitive linguistics as having two methodological extremes. In the 'good
corner' there are the data-driven, empirically-minded linguists doing psy-
cholinguistics, 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 (Forth-
coming b), that the two approaches should be seen as complementary
rather than inimical:
Just as it is easy to fall into "either-or" thinking in relation to verbal and diagram-
matic representation, so too with the contrast between symbolic (conceptual) mean-
ing and its experiential (pre-conceptual) underpinnings. However, one may very well
grant that embodied, pre-conceptual experiential Schemas (kinaesthetic image-sche-
mas) 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 ap-
proaching 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 lin-
guistics, alongside other models and approaches. The usefulness of this ap-
proach in the area of culture-laden aspects of language, such as the lan-
guage 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 ana-
lyses 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 Nie-
meier 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 dimen-
sions 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 in-
vestigation 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 theor-
etical 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 language-
independent analytical method provided by the Natural Semantic Metalan-
guage, can point the way out of the hermeneutical circle inherent in cul-
ture-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, ex-
plain 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, in-
cluding the existence in all languages of words for cognitively-based feel-
ings, and substantial overlaps between certain elements in the semantic
composition of some of these, overlaps that can be specified in language-
independent 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 the-
matic boundaries.
While the authors share a commitment to seeking language-independent,
non-ethnocentric ways of analysing culturally situated linguistic mean-
ings, grounded in accurate linguistic description and tested against natural
language data, readers will notice also a diversity of perspectives and prac-
tices 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
into the structures of the languages under investigation. These papers dem-
onstrate a greater convergence than previously found between different re-
searchers' 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 posi-
tions 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 pro-
posed, tested against more data, and tested for intelligibility and translat-
ability as shown in several of these studies. It takes many successive revi-
sions 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 accu-
racy than approaches based on partial translation equivalents. The poly-
semy of the verb (t3)s9tnma 'feel; hear' is resolved on formal, language-
internal criteria; and it is found that the Amharic data support the main pro-
posals 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. Ro-
bert 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 occasion-
ally introduce elements from outside the basic NSM set of primes. Al-
though 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 mol-
ecules separately elsewhere.
Turning to a European language, a corpus-based analysis by Uwe Durst
of rger, Wut and Zorn in German illustrates how even a very closely re-
lated language does not have an exact match for the supposedly basic Eng-
lish anger, but that the differences in distribution of these words can be ex-
plained 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 be-
tween emotions and facial gestures, but also a distinctive ethnolinguistic
perspective on what are the significant features or components of facial ex-
pression. At the same time, the Lao evidence provides support for propos-
als concerning the universality of "cry" and "smile" as facial expressions
of feeling (cf. Wierzbicka 1999). While this paper does not propose expli-
cations in the same style as most of the others, it nonetheless demonstrates
the same kind of empirical and non-linguocentric approach to the identifi-
cation 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-
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 em-
ploys a simplified version of the prototype framework explained above. He
also provides a table of the NSM primes in Malay, though he does not ac-
tually give full explications in that language. These latter themes are taken
up in Jean Harkins' exploration of processes and issues involved in devel-
oping 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 embo-
died nature of these concepts raises issues of how to represent auditory el-
ements and physical prototypes in explications (I could hear something
like: "dokiQ"; this person thinks: "if I am in a cold place I will feel some-
thing 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 Le-
vontina and Anna Zalizniak, who cover more than a dozen emotion con-
cepts that have not received in-depth treatment in the already extensive lit-
erature on Russian emotions. By focussing on aspects of the system of cul-
turally significant stereotypes of experience reflected in the language
(which could also be understood in terms of cultural models or image-
schemas), 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 char-

acter".
A single highly salient Polish emotion, przykro, is analysed in depth by
Anna Wierzbicka, who places it in context of the set of interpretive cat-
egories provided by the Polish language. She shows how the different set of
categories provided by English would link similar feelings with several dif-
ferent 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 ethnopsychol-
ogy with a study of three of the most salient words in Chinese literature and
poetry: bei, ai and chou, often glossed interchangeably as 'sadness', 'sor-
row' 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 formu-
lae - 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 - with-
out imposing on those lives a perspective derived from the vocabulary and
other resources of our own native language. Since the descriptions pres-
ented 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 an-
other language, with their own spectacles, tinted by their own native lan-
guage.) 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 in-
cludes Ameka (1990), Bugenhagen (1990), Goddard (1990, 1991, 1995 a, b, 1996,
1997 a, b), Harkins (1990, 1995, 1996), Harkins and Wierzbicka (1997), Hasada
(1997,2000, In press), Kornacki (1995), Mostovaja (1988), Priestley (In press), Tra-
vis (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 se-
mantics 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: Ox-
ford 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: Har-
vard 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 God-
dard & Wierzbicka (eds.), 87-108.
Chappell, Hilary M.
1986 Formal and colloquial adversity passives in standard Chinese. Lin-
guistics 24: 1025-1052.
1991 Strategies for the assertion of obviousness and disagreement in Man-
darin: a semantic study of the particle me. Australian Journal of Lin-
guistics 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.
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. Austra-
lian Journal of Linguistics 10(2): 257-292.
1991 Anger in the Western Desert: a case study in the cross-cultural se-
mantics 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 Psy-
chology 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 Uni-
versity Press.
Forthcoming a Ethnosyntax, ethnopragmatics, indexicality, and culture. In Enfield
(ed.).
Forthcoming b In defence of paraphrase and the place of NSM semantics in cogni-
tive linguistics. In: June Luchjenbroers (ed.) Cognitive Linguistics
Investigations across Languages, Fields, and Philosophical Bound-
aries.
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.
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 Euro-
pean 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 com-
municative behaviour. In Fernando Poyatos (ed.) Nonverbal Com-
munication in Translation, 83-103. Amsterdam: John Benjamins.
1998 Sound symbolic emotion words in Japanese. In Athanasiadou & Taba-
kowska (eds.), 83-98.
2000 An exploratory study of expression of emotions in Japanese: Towards
a semantic interpretation. PhD dissertation, Department of Lin-
guistics, Australian National University.
In press Body part terms and emotion in Japanese. In Enfield and Wierzbicka
(eds.)
Hbler, Axel
(1998) The Expressivity of Grammar: Grammatical Devices Expressing Emo-
tion 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 Jap-
anese Workplace. Chicago: University of Chicago Press.
Komacki, Pawel
1995 Heart and face: semantics of Chinese emotion concepts. PhD disser-
tation, Department of Linguistics, Australian National University.
Kvecses, Zoltn
1995 Metaphor and the folk understanding of anger. In Russell et al. (eds.),
49-71.
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. Philip-
pot, R. S. Feldman and E. J. Coats (eds), The Social Context of Non-
verbal 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: Uni-
versity 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 Com-
parative 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 Fernndez-Dols, The
Psychology of Facial Expression, xii-x. Cambridge: Cambridge Uni-
versity 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 Anthro-
pologist 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: Atha-
nasiadou 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 The-
oretical Foundation. Amsterdam: John Benjamins.
Occhi, Debra J.
1999 Sounds of the heart and mind: Mimetic s of emotional states in Japan-
ese. 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 Univer-
sity 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 grammati-
cal 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 Sub-
strates. 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. Cam-
bridge: Cambridge University Press.
Russell, James A.
1991 Culture and the categorisation of emotions. Psychological Bul-
letin 110:426-450.
Russell, James A., Jose-Miguel Fernndez-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. Chi-
cago: 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. Nor-
ton. [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: Ange-
liki 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 Hon-
our ofKiril Taranovsky, 499-505. The Hague: Mouton.
1986 Human emotions: universal or culture-specific? American Anthropol-
ogist 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 con-
cepts 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. Austra-
lian Journal of Linguistics (Special issue on the semantics of emo-
tions) 10(2): 359-375.
1992 a Defining emotion concepts. Journal of Cognitive Science 16:
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 emo-
tions. In Lawrence A. Hirschfeld and Susan A. Gelman (eds.), Map-
ping 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 As-
sociation.
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.
2000 The semantics of human facial expressions. Pragmatics and Cogni-
tion 8(1): 142-183.
Wilkins, David P.
1986 Particle-clitics for criticism and complaint in Mparntwe Arrernte (Ar-
anda). 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.
In press Different modes of describing emotions in Chinese: Bodily

changes, sensations, and bodily images. In: Enfield and Wierzbicka
(eds.).
Yu, Ning
1998 The Contemporary Theory of Metaphor: A Perspective from Chi-
nese. 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 Am-

haric.1 I assume a distinct cognitive domain which specialises in the emo-
tions, and that the basic architecture of this cognitive domain is innate and
universal. I also assume that the best way of searching for universal emo-
tion 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 frame-
work of Natural Semantic Metalanguage (NSM).
Wierzbicka (1999) proposes the following set of working hypotheses in
her study of emotional universals:
(1) a. All languages have a word for FEEL.

b. In all languages, feelings can be described as "good" or
"bad".
c. All languages have "emotive" interjections (i.e. interjections
expressing cognitively-based feelings).
d. All languages have some "emotion terms" (i.e. terms for cog-
nitively-based feelings).
e. All languages have words overlapping (though not identical)
in meaning with the English words 'angry', 'afraid', and
'ashamed'.
f. All languages have words comparable (though not necess-
arily identical) in meaning to 'cry' and 'smile'.
g. In all languages, people can describe cognitively-based feel-
ings via observable bodily symptoms.
h. In all languages, cognitively-based feelings can be described
via bodily sensations,
i. In all languages, cognitively-based feelings can be described
via figurative "bodily images".
36 Mengistu Amberber
j. In all languages, there are alternative grammatical construc-

tions for describing (and interpreting) cognitively-based feel-
ings.
The main purpose of the present study is to test the above set of hypo-
theses 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 In-
terjections.
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 agree-
ment as shown in (2):
(2) aster wods bet hedged

A. to home go.PF-3F
'Aster went home.'
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- mm
nothing NEG- 3M.IMP- feel- lo- NEG
feel nothing.'
b. min yi- sssmma- h- al
what 3M.IMP- feel- 2M- IMP
'How do you feel?'
The Amharic word is used to express concepts which are undifferenti-
ated 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- ssamma- final
happiness SM.IMP- feel- i- IMP
feel happy.'
b. himam yi- ssdmma- final
pain 3M.IMP- feel- I- IMP
feel pain.'
The Amharic verb for 'feel' cannot be used to translate the English con-
struction such as "I feel the chair" (say as in touching a chair while trying to
find your way in darkness):
38 Mengistu Amber be r
(5) *w9nb3r- u t3- samma- fin

chair- DBF DT- PF.SM- 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 construc-
tions 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 con-
sistent 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. minim a- y- samma- tint- mm

nothing NEG- SM.IMP- feel- 10- NEG
hear nothing.'
b. min yi- ssamma- h- al
what 3M.IMP- feel- 2M- IMP
'What do you hear?'
As the following examples show dQssita 'happiness' and dims' 'noise'
stand in a paradigmatic relationship:
(7) a. d9ssita yi- sssmma- final
happiness 3M.IMP- feel- I- IMP
feel happy.'
b. dims' yi- ssamma- nfi- al
noise SM.IMP- feel- i- IMP
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':
(8) a. ddssita te- samma- w

happiness DT- feel.PF.3M- SMO
'He felt happy.'
b. * 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. dims' te- ssmma- w
noise DT- hear.PF.3M- 3MO
'He heard a noise.'
b. dims' t9- S9mma
noise DT- hear.PF.3M
noise was heard,'
Notice that in (9b), the sentence is still grammatical despite the absence
of the pronominal suffix (-w '3Mo'). However, note that without the agree-
ment 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 val-
ency. The verb corresponding to the meaning 'feel' cannot occur in a ca-
nonical transitive syntactic frame, unlike the verb corresponding to the
meaning 'hear':
(10) a. *temma dassita samma
Lemma happiness feel.PF.3M
b. 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 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- d^ssita as- samma- c i-w
A. Lemma- ACC happiness CAUS- feel.PF- 3r- 3MO
b. aster temma- n zafan as- sdmma- i- w

A. Lemma- ACC song CAUS- hear. PF. SM
(lit. 'Aster let Lemma hear a song.')
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 nomina-
lised forms:
(12) a. simmet 'feeling' (cannot mean'hearing')

b. mosmat 'hearing' (cannot mean 'feeling')
The nominal simmet 'feeling' is productively used in a possessive-at-
tributive phrase to express various emotional concepts.
(13) a. yd- haz9n simmet

of- sadness feeling
'feeling of sadness'
b. 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. aster bird t9- samma- t

A. cold DT- feel.PF.3M- 3ro
'Aster felt cold.'
b. aster rshab t9- samma- t
A. hunger DT- feel.PF.3M- 3FO
'Aster felt hunger.'
c. aster him9m t9- samma- t
A. pain DT- feel.PF.3M- 3ro
'Aster felt pain.'
It is important to note that physiological experiences such as 'be hun-

gry', 'be cold', 'be in pain/be sick' can also be expressed without the use of
the verb 'feel'.
(15) a. aster (in) bdrr^d- at

A.- (ACC) be.cold.PF.3M- 3ro
'Aster is cold.'
b. aster (in) rabb- at
A.- (ACC) be.hungry.PF.3M- 3ro
'Aster is hungry.'
c. aster (in) amm^m- at
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. t'iru simmet te- samma- t

good feeling DT- feel.PF-3M- 3FO
'She feels good.'
(lit. 'She feels good feeling.')
b. mot'fo simmet te- ssmma- t
bad feeling DT- feel.pF-3M- 3ro
'She feels bad.'
(lit. 'She feels bad feeling.')
The above expressions refer to a bipolar opposition between positive
emotions and negative emotions.
The predicate which is normally used to express the emotion 'happiness'

can be used in a generic sense to express 'good' feeling, whereas its ne-
gation is used to express roughly 'bad' or 'unpleasant' feeling:
(17) a. dass ya- mmil simmet
happy RC- AUX. feeling
'good feeling'
b. d9ss y9- mmayil simmet
happy RC- AUX.NEG feeling
'bad feeling' (or 'unpleasant feeling')
The form glossed as ' 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 mor-

phologically isolated words which do not belong to any word class. There
is a small class of interjections which can refer to cognitively-based feel-
ings. The following list is gleaned from Leslau (1995: 900ff):
(18) a. ahe 'exclamation of surprise, sorrow'
b. ahehe 'exclamation of sorrow, grief, or worry'
c. ih 'sorrow, distress, or pain'
d. Hill 'joy, ululation' (trilling sound often made by
women)
e. 93 'expresses surprise at something unexpected'
f. issay 'exclamation of joy, pleasure, happiness'
g. awwoy 'exclamation of sorrow' ('woe is me!')
h. ec' ' 'expression of disgust, of impatience'
i. iffoy 'sigh of relief
j. wa 'exclamation of sorrow or sadness, expression of
warning of threat ('watch out!')'
k. wiy 'exclamation of (negative) surprise, pain, or sor-

row' ('my goodness!, alas!')
1. gud '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 Am-
haric point of view. Looking at the use of these interjections from an Am-
haric 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 Eng-
lish emotion term. In some cases, however, it does seem justfiable on lan-
guage-internal grounds to posit polysemy. I would suggest that this applies,
in particular, to ahe, whose two meanings (ahe} and ahe^ could be repre-
sented 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
something bad is happening now
because of this, this person feels something bad
I feel something like this
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 im-
peding 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 cat-
egory: roughly 'be happy' and 'be sad' respectively.
For ease of exposition the emotion predicates are presented in the (un-
marked) intransitive syntactic frame. Other possible grammatical strat-
egies 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 ta- dassdta-
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 cor-
responds to 'be happy', 'be joyful' can be presented as follows:
(22)
X feels something
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 judge-
ment 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) intran-
sitive stem:
(24) aster azzans- cc

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
can only present a brief general commentary. I will examine the relation-
ship between the Amharic expressions which roughly correspond to 'sad-
ness', '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) 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.'
(26) 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 par-
ticular 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 'sym-
pathy' is expressed by grammatical means. Thus, when the verb azzand oc-
curs with an argument marked by the preposition h 'for' (with or without a
corresponding agreement marker on the verb) it has the approximate mean-
ing 'sympathise (with)' or more literally 'be sad for someone':
(27) temma h- hit- u azzana- (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 re-
siding in the shape of the prepositional suffix, as shown in (28)-(29):
(28) azzdnd- 11- at
be.sad.PF.3M- FOR- 3ro
'He feels sorry for/sympathises with her.'
(lit. 'He is sad for her.')
(29) azz9n^- bb- at
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 follow-

ing explications (-PER is an abbreviation used for the relevant person-
number-gender marking; it refers to the component 'that person' in the
semantic explication):
(30)
X feels something
because X thinks something
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
X feels something like this
because X thinks something like this
(31) azzdti9-bb-PER
X feels something

something bad happened (to that person)
it happened because that person did something
I don't want this person to do things like this
Note that in (30) - the meaning that roughly corresponds to English
'sympathise' - the explication has the component "something bad hap-
pened to that person". In (31) on the other hand - the meaning that corre-
sponds 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 dis-
appointed with' them. For example, if had an expectation that X will
execute (or not execute) some task and X failed to do so, can say (31) -
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 con-

stitute 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 language-
specific 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:
(32) a. aster bat'am ferra- cc

A. very be.frightened.PF- SF
'Aster is very frightened.'
b. aster wissa- w- in farra -cc/- w
A. dog- DBF- 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 as- fyrra- t
dog- DBF A.- ACC caus- fear- 3M-3FO
The dog frightened Aster.'
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 ti- fara- Ite- cc
A. dog IMP- fear- IMP- 3r
'Aster fears dogs.'
When somebody is overcome by fear, a special construction which in-
volves the use of the nominal firhat 'fear' and the verb yazd 'hold, catch' is
used:
(35) aster- in firhat yaz- at
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 'cow-
ardly 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
danger. A typical example of a powerful (but non-adversative) authority

can be represented by one's father or God:
(37) aster abbat-wa- n ti- fara- lid-

A. father-DBF- ACC IMP- fear- IMP- 3F
'Aster fears her father.'
(38) aster igziyabher- in ti- fsra- Ih- cc
A. God- ACC IMP- fear- IMP- 3r
'Aster fears God.'
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, in-
cluding 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 tenaddsda- d

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):
(40) 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-judge-
ment component. This can be summarised as follows:
(41) a. with no negative-judgement component

b. 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 re-
lated (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 arr9r9- cc
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 ex-
press 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 ver-
sion 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 ihitwa- n te- k'ot't'a- cd- 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
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
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 says something like this to this
other person
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 some-
thing 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 reflex-
ively, 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 -

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 aff9ra- cc

A. be.embarrassed.PF- 3r
'Aster is embarrassed/ashamed.'
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
The verb affara can also have a polysemous meaning comparable to the
English 'be shy' (see Harkins 1990 for the expression of 'shame' and 'shy-
ness' in Australian Aboriginal English):
(52) aster afford- cc

A. be.shy.pf- 3
'Aster became shy.'
When people feel 'shy' it is not necessarily because they think that some-
one thinks bad about them. The feeling may be due to general 'self-con-
sciousness' as pointed out by Darwin (1872/1965).6 Thus, the semantic ex-
plication 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
Incidentally, in Amharic being shy is typically regarded as a positive cul-
tural 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 re-
ferring 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'
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 transi-
tive verb can be translated into English as 'make X be ashamed/embar-
rassed/shy' and also 'humiliate'.
(56) y9- iht- wa sira aster- in as-
POSS- sister-poss.3p deed A.- ACC CAUS-
affar- at
be.embarrassed.PF- 3r
'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. aster fygsg ah- co
A. smile.vn aux.pf.- 3r
'Aster smiled.'
b. aster alsk'k'dsa- co
A. cry.pf.- 3r
'Aster cried.'
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 as-
sociated 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 el-
ement 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 ob-
servable bodily images or movements. For example, one can say some-
thing like 'his face became ashen' to express sudden shock or fright:
(60) 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 symp-
toms 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 ex-
pression. 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
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.'
Notice that the emotion occurs as an oblique manner adverbial bo-dinig-

gat'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 follow-
ing interrogative sentence refers to either an auditory perception (HEAR) or
an emotional state (FEEL):
(64) min yi- ssdmma- h- al
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 sen-
sation 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 'taste bitter'

b. FEEL t3m9rr3r3 'be irritated', 'be embittered'
(66) a. Smell kanakkans 'tickled (as in pepper in the nostrils)'
b. 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. Touch t9n9kka 'be touched'
b. FEEL tanakka 'be (emotionally) moved'
(68) a. See ayya 'see'
b. FEEL ayy9 'have an emotional experience'
The last two occur in constructions such as (69)-(70):
(69) aster ba- hazan tanakka- do
A. with- sorrow (touch)feel.PF.- 3r
'Aster is touched with sorrow.'
(70) aster mdkarawa- n ayya-c

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 ono-
matopoeic 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 feel-
ing 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 tambbo'abbs'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
sometimes a person thinks:

something bad happened to this other person
I don't want things like this to happen
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. aster t9- dassata- co

A. DT- be.happy.PF- 3r
'Aster is happy/joyful.'
b. temma aster- in as- dassat- at
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 'hap-
piness/joy') is realised in the S function, whereas in the second example
the experiencer is realised in the function. For ease of exposition we can
refer to these patterns as SUBJECT EXPERIENCER (SubjExp) and OBJECT EX-
PERIENCER (ObjExp) respectively (cf. Pesetsky 1995).
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- cci- w
A. L- ACC love.PF.- 3rs- 3MO
'Aster loved Lemma.'
b. Pattern B: lamma ta-
L. DT-worry.PF.3MS
'Lemma is worried.'
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 obliga-
tory 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 ObjExp
W9dd9d9 as-w9dd9d9
'love y' 'make love y' 'worry' 'make worry'
f9rra as-f9ira t9-d9SS9t9
'fear y' 'make fear y' 'be pleased' 'make pleased'
t'slla as-t'alla t9-k'ot't'a as-k'ot't'a
'hate y' 'make hate y' 'be angry' 'make angry'
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.'
b. aster temma-n as- dassata- oc- iw

A. L.-ACC CAUS- be.please.pf.- SFS- SMO
'Aster pleased Lemma.'
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. *ddss3t9 > te-dassata- 'be happy'
b. *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. aster- (in) c'snnsk'- at
A.- (ACC) worry.pf.3ms- 3ro
'Aster is worried.'
b. aster to- o'anndk'd- oc
A. Dt- worry.pf.- SFS
'Aster is worried.'
(81) a. aster- (in) ddnndk'- at
A.- (ACC) astonish.pf.3MS- 3ro
'Aster is astonished.'
b. 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 inter-
esting 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.'
The presence (albeit optionally) of the accusative case marking on the

subject noun phrase in (82) is unusual in Amharic, as only (definite) ob-
jects 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 aster- in matt- at

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 predi-
cates (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). Con-
sider 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 gram-
matical 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 predi-
cate such as 'worry' is an affected argument with less volition, thus alig-
ning itself with prototypical objects (cf. Hopper and Thompson 1980)
rather than prototypical subjects. In other words, the "experiencer" argu-
ment of the emotion predicates has much in common with typical "patient"
arguments.
The primary cognitive component connected to the "patient" role is ex-
pressed 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
"patient" argument of typical action predicates (e.g. 'hit') because the ex-
periencer 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 assum-
ing 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 ex-
press the 'target' of emotion. Thus, for example with the intransitive verb
k'9nna which roughly corresponds to 'be jealous', or 'be envious', the tar-
get of the emotion occurs in a prepositional phrase:
(85) aster ba- (i)htwa k'anna-
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'9nna- cc- 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 ar-
gument 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 proto-
typical "patient" arguments rather than with prototypical "agent" argu-
ments. However, the alignment is partial, manifesting itself mainly in mor-
phology: the occurrence of an accusative case morpheme which is other-
wise restricted to definite object NPs of agentive verbs. As far as the overall
clausal structure is concerned there is no evidence which suggests that ex-
periencers are grammatical objects.
4. Conclusion
This study has presented a general overview of the semantics of some emo-
tional (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 mis-
match between concepts expressed by the individual languages, the meta-
language 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 con-
cepts 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 'sym-
pathy' is expressed by grammatical means, i.e. attaching a prepositional
suffix on the verb which corresponds to 'sad/sorry'. It has been demon-
strated 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 ten-
tative and only a detailed study of a wider range of data will provide a more
in-depth insight into the cognitive domain of emotions.
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 re-
sult 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 "disap-
pointment" 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 "mod-
esty" 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... grad-
uates 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.
Grimes, Barbara
1996 Ethnologue. 13th edn. Dallas: Summer Institute of Linguistics.
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
Hirschfeld, Lawrence A. and Susan A. Gelman (eds.)
1994 Mapping the Mind: Domain Specificity in Cognition and Culture.
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.
Zaenen, Annie, Joan Maling and Hskuldur Thrinsson.
1985 Case and grammatical functions: The Icelandic passive. Natural Lan-
guage 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 nur-
ture and mind, with different socio-cultural groups dividing up the emo-
tional 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), melan-
cholic (black bile), choleric (yellow bile), and phlegmatic (phlegm) (Zim-
bardo and Ruch 1976: 225). Darwin (1872) held that emotional patterns are
largely inherited, innate responses that have had biological utility in evol-
ution. William James (1884) held that emotions are perceptions of physio-
logical 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 gas-
tric acid, inter alia. Ekman, Sorenson, and Friesen (1969) report that emo-
tions like happiness, sadness, fear, anger, and disgust are commonly dis-
played 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 bio-
logically 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 re-
lief 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 some-
thing 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 re-
cognise as emotions." Similarly, a number of the articles in Harre (ed.)
(1986) speak about stereotypical situations or scenarios which are ex-
pected to call forth emotions.
Recent anthropological research like that of Lutz (1982, 1987) and Ro-
saldo (1980) has demonstrated the importance of an individual's socio-cul-
tural 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 speak-
ers' self-reports of their emotional experiences, one fruitful investigative
strategy has been a lexicographic one: "In the 1970s and early 1980s cul-
tural anthropologists and like-minded philosophers elaborated the position
that since we can't get inside other people's heads and hearts, we must in-
stead 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." (Lea-
vitt 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:
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 com-
pare meanings expressed in different languages and different cultures, one needs a
semantic metalanguage independent, in essence, of any particular language or cul-
ture - 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 per-
spective; 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 ex-
pressing experiential notions in the Mbula language - and an eighth one, to
be discussed below. Bugenhagen (1990) describes the syntax and seman-
tics of most of these in some detail.
1. Uninflected experiential verbs with coreferential Experiencer Sub-
jects 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 en-
coding emotions: menmeen 'be happy' and kaipa 'selfishly rejoice'. These
two semantic groups are distinguished syntactically by their ability to co-
occur 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
am very happy about that talk of yours (so).'
(2) Ni petel=i.
SS.NOM be.hungry=3s.ACC
'He is hungry.'
2. Inflected experiential verbs with Experiencer Subjects. These fall into

the following groups: 1) cognition: -ute 'know', -kilaala 'recognise, per-
ceive', 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
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 yo.
medicine this 3s-burn IS.ACC
'This medicine burns me.'
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 'em-
barrassment', sir) 'blood, bleeding', motoijana 'fear' - a nominalised form
from the verb -moto 'fear', and dadaru 'doubt', and the Experiencer is en-
coded as the direct Object.
(6) Mete i-kam yo.

disease 3s-get IS.ACC
am sick.'
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 un-
pleasant 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 thinks about , feels something like this
b. -ur koi pa ':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
Regarding such explications, it will be observed that they express a

prototypical situation resulting in some sort of feeling, this being em-
bedded under the component "sometimes people think something like
this". A person is then said to "feel something like this". The approxi-
mative "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 emo-
tion 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 means to feel as a person does who has certain (specifi-
able) 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
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-
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 under-
stand this enigma and says it was "burdensome/laborious/oppressive to his
eyes." In New Testament Greek emotions like compassion, longing, and re-
freshment 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 He-
brew and Greek, since his/her own language also abounds in such ex-
pressions.
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 emo-
tions 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, en-
code emotional or physical experiences.
In the preceding section, most of the listings of lexical items and con-
structions 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 'disap-
pointment' . And yet, living among Mbula speakers over the past seventeen
years, I can vouch for the fact that they certainly seem to experience feel-
ings which strike me, albeit I am a cultural outsider, as bearing a consider-
able 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 fre-
quently occur are: mata- 'eye', lele- 'inside(s), feelings', kete- 'liver', ni-
'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 construc-
tions, 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 syn-
tactic 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 ad-
verbs 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 as-
sociated 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
person', zog mataana men 'the sun is very bright'.) The least extended ex-
periential 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 speaker'/'look
down' /' look up' /' look back'
(14) mata-ikam Z
(lit. eye do/get/receive)
'involuntarily happen to see something'
something happens in someone Y's eyes
because of this, sees something
not because wants to see it
(15) mata- mbeij
(lit. eye night)
'to not see something which one should be able to see' (used to
scold someone)
someone does not see something.
I think: if wanted to see this, could see this
because of this, I think does something bad
I say is like someone who cannot see things because it is night3
(16) mata-pisana
(lit. eye-closed.one)
'blind'
The next class of experiential examples with mata- have to do with life
and consciousness, the capacity to think and know.
(17) mata- iyaryaara

(lit. eye is shining)
'be alive/be awake with one's eyes open'
someone/something is alive
(now 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. 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"
is like this
(22) mata- kankaana pa Z
(lit. eye confused Oblique Z)
is confused and as a result accidentally makes a mistake having
to do with Z'
(23) mata- mburri
(lit. 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
when these people think like this they feel something (bad)
feels something like this
(There is another sense for this expression, which appears to be essen-
tially 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) mata- molo
(lit. eye long)
when things happen in places far away from here,
some people know these things happen
is like this
('omniscient' when used of God)
(26) a. Y mata- iygal
(lit. eye pierces Z)
before now, did not think about something
now thinks about this thing
b. mata- iygal Z:personal noun (pa W)
(lit. eye pierces Oblique W)
before now did not think about person
now thinks about this person
wants to do good things for this person
because of this, does something good for this person
(27) mata- iggal itu-
(lit. eye pierces self)
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
does something like this

I could do something bad
if I do some other thing, I will not do this bad thing
because of this, these people do this other thing
does something like this
(The expression mata- ise kor 'eye ascends above' has a similar meaning
of being careful or wary.)
(28) - ZseY mata-

(lit. ties ascend eye)
'memorises'
thinks:
after now, I want to be able to think about something
because of this, does something
because of this, afterwards, can think about
(29) mata- imiili pa
(lit. eye returns Oblique Z)
something happened before now
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) mata- imbeleele

(lit. eye shrivels Z)
forgets Z'
(32) mata- imunlmata munrjaana
(lit. eye be.closed/eyes closed.ones)
'ignorant'
sometimes a person does not know something
other people know this thing
I think it is bad if a person does not know this thing

person is like this
(33) mata- leyaleija
(lit. 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 is like this
(34) mata- ise ki Z
(lit. eye ascend Locative Z)
'look up to, respect, admire'
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 some-
thing, I want to do this thing
when these people think like this, they feel something good
when thinks like this about someone Z
(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) mata- pasom

(lit. eye cause.not Z)
despises Z as being worthless/insignificant'
The third class of experiential expressions involving mata- are more
clearly emotional in nature.
82 Robert D. Eugenhagen
(36) mata- putput

(lit. Y eye bulging)
'furious and about to act'
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 feels something like this now
(37) mata- pambaara
(lit. eye crossed)
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 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) mata- berber pa

(lit. 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 sees something , feels something like this
(39) mata- imapalpaala pa
(lit. eye be.breaking Oblique Z)
'intense coveting to the point of being unable to concentrate, ob-
sessed'
sometimes if people see something, they think like this:
this thing is very good
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
(40) mata- kinigana
(lit. 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
is like this
Other items occurring in this mata- -...-gana construction are: moni
'money', pat 'stone/money', and moon 'woman'.
(41) mata- mburmbur (pa Z)

(lit. eye jealous Oblique Z)
1. 'jealous of another's good fortune'
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
bad
2. 'jealous of one's relationship with someone else',
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
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 vil-
lage. 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) mata- iyoyou pa
(lit. eye hurts Oblique Z)
'not want to see something anymore'
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
(43) mata- pot pa Z:event NP
(lit. eye float Z)
'do something reluctantly and poorly'
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 thinks about , thinks like this
(44) mata- sirjsirj pa Z

(lit. eye red Oblique Z)
'eager, enthusiastic'
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 does something like this
(45) iur mata- (pa Z)

(lit. 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
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 experi-
ential 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 mata-
pa, someone desires to see something. In other instances, like mata- put-
put and mata- pambaara, the eyes have a signalling function. One recog-
nises the person's emotional state from the appearance of their eyes/face.
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
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 lele-rj ambai paso,

IS.NOM insides-ls.GEN be.good because
le-T) pata-rja-na sa som.
RECX-lS.GEN heavy-NMZ-3s.GEN NON.REF NEG
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.'
Unlike mata- , lele- is rarely used to express any sort of physical sen-
sation 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 ex-
erting oneself. Having a temperature/fever is expressed differently as kuli-
ibayou '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. Y lele- iurpa Z.NP

(lit. insides puts Oblique Z)
b. lele- iur be Z:S
(lit. insides puts NAF Z)
1. 'someone decides to do Z'
before now, someone did not want to do something Z
now wants to do this thing
2. chooses Z and is determined to do good things for Z'
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
thinks like this about someone Z
This "choosing" is more along the lines of the biblical concept of elec-
tion; i.e. chooses to have a special relationship with Z in which is com-
mitted to Z and is determined to do good things for Z.
(54) lele- iwe ru pa Z

(lit. insides become two Oblique Z)
'doubt'
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 thinks about something , thinks like this
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
thinks like this about
(55) lele- iur pa Z sorok
(lit. insides put Oblique Z without.basis)
'to suspect that someone has done something wrong'
I think someone did something bad
I don't know if this person did this thing.
thinks like this about something
(56) lele- iurur
(lit. 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. insides Oblique Z)
sometimes people feel something inside
because of this, they want something
because of this, wants something
b. 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
(58) lele- be Z:S

(lit. insides NAF Z
when thinks about , feels something inside
because of this, wants to happen
(59) a. Y lele- ambai (pa Z:S)
(lit. insides be.good Oblique Z)
'happy, contented, at peace, approving, grateful, free from
worry, care, anger, and sorrow'
does not feel anything bad inside
feels something good inside (because of Z)
b. lele ambai pa Z:NP
(lit. insides be.good Oblique Z)
'be pleased with someone, approve'
does not feel anything bad towards someone Z
when thinks about , feels something good inside
(60) lele- ipuru
(lit. 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 vi-
olate their good principles.
(62) a. Y Me- isaana pa Z:personal noun

(lit. insides deteriorate Oblique Z:N)
'feel sorry for someone and want to help them, compassion'
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
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. lele- isaana (pa .-non-personal noun)
'feel bad about, be upset by'
someone did something bad to something
this person did something bad to me
when people think like this, they feel something bad inside
c. lele- isaana (pa :non-personal event noun)
someone did something bad
d. lele- isaana paso Z:S
(lit. insides deteriorate because Z:S)
'feel bad about, be upset'
someone did something

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 'rebel-
lion/stubbornness'. An example with an event noun (62 c) would be: Ni le-
leene isaana pa sua kizin 'He is upset because of their talk'.
(63) lele- ipata pa

(lit. insides be.heavy Oblique Z)
'very sad or troubled and helpless to do anything about the situ-
ation'
something bad happened
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 thinks about someone/something Z,
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.
(64) lele- imbukmbuk pa Z

(lit. 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
(65) lele- sa pa
(lit. insides non.referential Oblique Z)
= ni- sa pa Z
(lit. being non.referential Oblique Z)
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 thinks about someone , 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) does not look Z directly
in the eye, 2) says as little to Z as possible, and 3) does not smile. Lele-
sapa 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) lele- ipasiksik Z

(lit. insides compel Z)
'unable to refrain from doing something'
(67) lele- ikam uraata
(lit. insides do work)
'be emotionally moved'
someone feels something
This expression is the most generic way of all of expressing that someone
is feeling emotion. It indicates nothing whatsoever about the kind of emo-
tion being experienced.
(68) lele imbai pa

(lit. insides be.poor Oblique Z)
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
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 vil-
lage are upset with someone and therefore refuse to cooperate with him in
traditional ceremonies or large work projects.
(69) lele- iluumu (pa Z)

(lit. insides be.cool Oblique Z)
before now, someone felt something bad inside (because of Z)
now does not feel like this any more
Lele- iluumu is a generic expression indicating the cessation of unpleas-
ant emotional states. When one is no longer angry or sad, their insides are
said to be 'cool'.
(70) lele- imiili pa Z:personal noun

(lit. insides return Oblique Z)
'reconciled and no longer angry'
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
I wanted to do bad things to this person

now I don't feel like this any more
thinks like this about Z
(71) YiraoZlele-
(lit. be.adequate Z's insides)
'something satisfies someone Z'
(72) a. Y lele- imet (kat) pa Z:event NP
(lit. insides snap.in.two (really) Oblique Z)
b. lele- imet (kat) be Z:S
(lit. insides snap.in.two (really) NAF Z)
'someone becomes determined to do something Z, regard-
less of any possible negative consequences doing it might
have for Y'
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
inside because of this
when Y thinks about , 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 ex-
perience implied by the use of the noun lele-. Explications for the other
body image expressions will also frequently make use of such a compo-
nent, with "inside" replaced by the corresponding body part term. Such ex-
plications 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, ap-
proving, grateful, content, at peace, free from worry, care, sorrow, and
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 de-
lineating 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 co-
operation. Being able to precisely talk about bad feelings can be viewed as
having considerable adaptive value, since it allows one to diagnose effi-
ciently the causes and current or envisioned consequences of an individ-
ual'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 accept-
able, 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 ex-
tremely 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 vil-
lage, lele keteenerjana is a flat plain.) Unlike mata- and lele-, body image
expressions containing kete- never express pure cognition. The term is
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 stom-
ach', kete- isalaklaaga 'be nauseated', kete- kutkut 'chest/liver be.pulsat-
ing - 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) kete- guruklpitpit (pa Z)

(lit. 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 feels something like this
(74) kete- pitpit pa Z

(lit. chest/liver be.jumping Oblique Z)
'respond to something too quickly without thinking of the poten-
tial 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
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 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) kete- ise (pa Z)

(lit. 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 thinks about something , feels something like this
(77) kete- malmal (pa Z:personal noun) (pa W:event noun)
(lit. chest/liver fights Oblique Z Oblique W)
is angry atZ'
when these people think something like this, they feel something
bad in the part of them where their liver is
these people want to do something bad to this person

when thinks about someone Z (and something W), Y feels
something like this
(78) kete- ibayou (pa Z:personal noun) (pa W:non-personal noun)
(lit. chestAiver be.hot Oblique Z Oblique W)
= kete- kembeita lai
(lit. chest/liver like wild yam)
= kete- ibeleu (pa Z) (pa W)
(lit. chest/liver swirls Oblique Oblique W)
= kete- ipakpak (pa Z) (pa W)
(lit. chest/liver be sour Oblique Z Oblique W)
' is furious at someone Z (because of what Z did to something W)'
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 thinks about someone Z (and something W), 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 kete-
(lit. removes Z's chest/liver)
'Z enrages Y'
someone did something/ something 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
The expression -pas kete- is used to talk about someone causing some-
one else to become angry and focuses upon the transition from a state of
not being angry to one of being angry.
(80) kete- imap

(lit. chest/liver ends)
'be astonished/dumbfounded because of something unusual hap-
pening 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 being is.smooth 'feeling fit or fresh'
ni- itekteege being is.holding 'shiver with cold'
ni- ambai being is.good 'healthy'
ni- iyoyou being hurts 'physically suffer, ache all over'
ni- isaana being deteriorates 'be exhausted, worn out'
ni- ipeele being opens 'feel fully awake'
(A near equivalent of ni- isaana would be mbura- imap 'strength be fin-
ished'.)
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-
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 fol-
lowing explications the component "feels something in a part of F" is ab-
sent.
(82) F ni- pa Z som
(lit. being Oblique Z not)
does not want Z'
(83) F ni- gesges pa
(lit. being tedious Oblique Z)
finds Z tedious, irksome'
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
(84) -sapaZ
(lit. being non-referential Oblique Z)
See explication for lele- sa pa.
(85) F ni- sanaana pa Z
(lit. being sin Oblique Z)
= F ni- puliizi pa Z
(lit. being [?] Oblique Z)
'upset with'
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
The kinds of things done are: 1) does not look Z directly in the eye,
2) says as little to Z as possible, 3) does not smile, and 4) when is in
Z's presence, does not come close to Z.
(86) ni- ise (pa Z)
(lit. being ascends Oblique Z)
'be visibly exultant over one's good fortune'
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
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 some-
thing 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 ex-
press various physical experiences in expressions like: kuli- isu 'skin de-
scends 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
shivers - have shivers up one's spine (in response to being cold or en-
countering 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 physi-
cal 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 kek.

skin-ls.GEN 3s-sense/3s-get go thinking SP.LOC PERF
somehow sensed/felt what they were thinking.'
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
The term miarj subsumes a whole range of English terms: shame, embar-
rassment, shyness. The experience is an uncomfortable one of being the
focus of someone's attention when one does not want to be under the "spot-
light", 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.
b. kuli- ka- miar) pa Z:non-personal noun

'shame'
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
(89) kuli- puliizi
(lit. skin embarrassed)
'be embarrassed because of being talked about'
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
(90) kuli- imoto (pa Z)
(lit. skin be.afraid Oblique Z)
something bad will happen to me because of something
I don't want this
skin
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,
kuli- is being used in a wider sense than just 'skin', but 'skin' is included in
the prototype proposed here.
(91) kuli- mburaana pa Z som
(lit. skin its-strength Oblique Z NEG)
feels something in 's skin
because of this, does not want Z
(92) a. Y kuli- imet (kat) pa Z:NP
(lit. skin snaps.in.two (really) Oblique Z)
b. kuli imet (kat) be Z:S
(lit. skin snaps.in.two (really) NAF Z)
1. 'someone 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 uncom-
fortable in their presence'
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 Y thinks about someone , 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 ex-
pressions 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
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) kopo- rru pa

(lit. stomach be.seeking Oblique Z)
is anxious about Z'
after now something bad could happen to/because of some-
thing
I don't want this
stomach
(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 rest-
lessly 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
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'
(96) 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) kwo- iyoyou pa Z

(lit. mouth hurts Oblique Z)
'tired of talking about something'
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
now these people do not want to say anything more about this thing
(98) 7 kwo- itaanda pa
(lit. mouth be.wide.open Oblique Z)
'be astonished by something unusual with the result that one opens
one's mouth widely'
(99) taiga- iyoyou pa Z
(lit. ear hurts Oblique Z)
'tired of listening to something'
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
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 ring-
ing - ringing in ears that hinders one's hearing'.
(100) taiga- iluumu
(lit. ear be.cool)
'not have to hear something anymore'
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
(The source of the annoyance could be an unpleasant, irritating noise, or
it could be things people were saying which were unpleasant.)
(101) taiga- ituntun pa
(lit. ear itches Oblique Z)
'always wanting to hear about something'
if people say things about something , wants to hear these
things
(102) ikam ggure-
(lit. does/gets throat)
= izuuru ggure-
(lit. pushes 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
feels something like this when thinks about
Qgure- refers to a cylindrically shaped part connecting other parts. In ad-

dition 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 some-
thing.'
(103) rjgure- isaana pa

(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) zuru- imbukmbuk pa
(lit. lip be.bulging Oblique Z)
'pouting and sticking out one's lip as an indication that one is upset
with someone/about something'
(105) ikarrut zuru-
(lit. 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 ex-
pressions 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 corre-
sponding English emotional terms. In other cases like ni- ise, as would be
expected, it is difficult to find precise English equivalents, because they in-
dicate a combination of semantic components unlike that of any English
emotional expression. Perusing the different explications, we find the fol-

lowing 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:
does things
because does these things, people know that 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)
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
mata- 'eye', kete- 'liver/chest', and kopo- 'stomach'. With regard to the de-
bate 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 de-
fined functions as the western traditions seem to do. For example, if mata-
can refer to one's physical eyes which see or experience pain, or experi-
ences 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 de-
scribed 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/de-
sires 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
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 liv-
ing 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 Malayo-
Polynesian, Eastern Malayo-Polynesian, Oceanic, Western-Oceanic, North New
Guinea Cluster, Ngero-Vitiaz Family, Vitiaz linkage. The language exhibits Sub-
ject-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) Demon-
strative. Pronouns and verbal affixation distinguish: 1) first, second, and third per-
son, 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 mor-
phology doesn't. There is no grammatical category of tense. Use is made of tem-
poral adverbials to fix the temporal deictic centre whenever it changes, but most sen-
tences have no indication whatsoever of the time of occurrence of the situation en-
coded. 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. Non-
transparent abbreviations used in this paper are: Accusative, CSB cosubordinating
conjunction, DEMonstrative, EXC speaker excluded, GENitive, INC speaker in-
cluded, LOCative, NMZ nominalisation, NOMinative, OBLique, PASSive, PERFect,
RECX Recipient-Experiencer, REDuplication. I would like to very gratefully ac-
knowledge 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 con-
taining 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.
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
1994 The exponents of semantic primitives in Mangap-Mbula. In: Cliff
Goddard and Anna Wierzbicka (eds.), Semantic and Lexical Univers-
als, 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. Am-
sterdam: 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.
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 Eth-
nologist 23(3): 514-539.
Leibniz, Gottfried Wilhelm
1903 Opuscules et fragments inedits de Leibniz. Paris: Louis Couturat. [Re-
printed 1961, Hildesheim: Georg Olms.]
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. Cam-
bridge: 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. Austra-
lian 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 Com-
pany.
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 phe-
nomenon of anger as well as other emotions, pointing to typical character-
istics of behaviour, causal events, and physical incidents in a more or less
scientific way. It could easily be shown that most of the symptoms de-
scribed by Seneca (and other philosophers, doctors, and psychologists)
would be attributed by people around the world to an angry or choleric per-
son 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 ap-
plications, 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 ber 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 emo-
tions. Although Seneca notes that Latin does not have the multiplicity of
distinctive words for different kinds of anger that there is in Greek, he ob-
viously uses the word ira as a hyperonym for various types of angry behav-
iour (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 estab-
lished in philosophical and theological writings throughout the centuries
(cf. also der Zorn Gottes, 'the wrath of God'), recent books on psychologi-
cal topics, on academic as well as popular levels, have included the word
rger (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, psycho-
logical state, the same "referent" in the non-linguistic reality, so that the
words Zorn, Wut, and rger could be seen as interchangeable synonyms?
Or do we have to consider one of them as a hyperonym, labelling a cat-
egory 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
zurck im Zorn, whereas the title of Theodore Isaac Rubin's The Angry
Book presumably could not, so that one German translation takes the noun
rger, another one the verb sich rgern ('to be annoyed/angry'), instead of
the adjective corresponding to Zorn, namely zornig? (These two trans-
lations of The Angry Book are: Mach deinem rger Luft, Mnchen: Mvg-
Verlag 1989, and rgere 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 Spa,
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 prop-
erly. [Frisch 1974: 90]
(2) Irgendeiner hatte ihn sogar zum Minister machen wollen, aber er
war wtend geworden und hatte gesagt: "Ich bin Lehrer, und ich
mchte wieder Lehrer sein". [MK1/LBC.OOOOO (Bll, 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" ein-
studiert, lange daran gearbeitet, und als ich sie auffhrte, wurde es
das, was man in unseren Kreisen einen Erfolg nennt: d.h. die rich-
tigen Leute lachten, und die richtigen rgerten sich. [MK1/
LBC.OOOOO (Bll, 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 bse.
[MK1/LBC.OOOOO (Bll, 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 Ger-
man expressions: the adjectives zornig (derived from Zorn), wtend (re-
lated to Wut), and bse (lit.: 'bad'), and the reflexive verb form sich rgern
(which corresponds to rger). In fact, however, angry is by no means the
only word that can be used to translate one of these four German ex-
pressions into English. Nor is anger the only noun that fits, for example, for
the German word Zorn:
118 Uwe Durst
(5) Ivy starrte mich an, dann schleuderte sie pltzlich ihren Mantel
irgendwohin ins Zimmer, stampfend, auer 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 con-
text where the German author uses the noun Wut:
(6) Ich heulte vor Wut, zertrampelte die Schiescheiben und schrie
den Jungen drauen, die mich bewachten, immer wieder zu: "Ihr
Nazischweine". [MK1/LBC.OOOOO (Bll, 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 mean-
ings 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 rger, Wut, and Zorn, which constitute
the most frequent and most common 'anger' words in German, will be sub-
jected 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
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. "un-
angenehm" ("pleasant" vs. "unpleasant") and "krperliche 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 lin-
guistic 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" ("exci-
tation"), close to Zorn, both differing in this regard from the less "excited"
states of rger and Aggressionslust. In another study presented by Wolf-
gang Marx (1982: 143), who investigated 20 German emotion words, the
highest degree of "Aktivation" ("activation") is attached to rger, followed
by Aggression (3rd position) and Wut (5th position), while Zorn is not
found at all. Nor is it found in Schmidt-Atzert and Strhm (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', 'resent-
ment'), whereas in another study (Schmidt-Atzert 1987: 148, 150, 153) it
was either Groll or rger, depending on whether Groll was part of the
given set of emotion words or not. In the latter study (1987: 140) the par-
ticipants 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 in-
vestigations 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 hnlichkeit

bzw. Unhnlichkeit von Wortbedeutungen'.'} But all that can be yielded by
this procedure are certain degrees of similarity between two or more emo-
tion words, and thus a taxonomy that consists in a hierarchical order of hy-
ponymy relations. In particular, we are not told what the words really
mean, why they differ in their ranges of use, and which semantic compo-
nents 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 ac-
cording to the linguistic conventions." ("Die Verwendung eines Gefhls-
wortes erfolgt nicht unter dem Gesichtspunkt, ob eine klar abgrenzbare Er-
lebnisqualitt vorhanden ist, sondern ob die Verwendung des Wortes ent-
sprechend den Sprachkonventionen in der jeweiligen Situation pat.")
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 ac-
cepted 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 ac-
tions, and vocal and facial expressions." Similarly, with regard to lexico-
graphic purposes and relying on Wittgenstein's notion of the "language
game", Plum (1992: 172) argues that "factors that... constitute a stereotypi-
cal frame for the use of emotion words" ("Faktoren, die... einen stereoty-
pischen Rahmen fr den Gebrauch von Gefhlswrtern konstituieren")
should be part of the semantic analysis. These factors are "those behaviou-
ral patterns that can be regarded as characteristic manners of expressing a
certain emotion" ("jene Verhaltensmuster, die als charakteristisches Aus-
drucksbenehmen fr eine bestimmte Emotion angesehen werden knnen")
(1992: 176).
However, it remains unclear which of those features have to be consider-
ed as part of the prototype, and whether there is a clear boundary between
prototypical and non-prototypical features. Consider, for example, Rus-
sell's "anger" script (1991: 39):
1. The person is offended. The offense is intentional and harmful.

The person is innocent. An injustice has been done.
2. The person glares and scowls at the offender.
3. The person feels internal tension and agitation, as if heat and press-
ure were rapidly mounting inside. He feels his heart pounding and
his muscles tightening.
4. The person desires retribution.
5. The person loses control and strikes out, harming the offender.
Russell states that "the meaning of each such [emotion] term, the con-
cept 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, ag-
gression, 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" emo-
tions ("fear", "sadness", "happiness", "disgust") but not to the meaning of
the everyday word anger. This is not to say that the positing of such cat-
egories 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 dis-
cipline - a terminology worthy of linguistic investigation in its own right.
While Russell's "anger script" is too extensive from the viewpoint of ex-
tensional semantics, it is too specific with regard to the intensional seman-
tics 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 inten-
tionally offend him/herself in a harmful way before. The "desire for retri-
bution" 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 in-
cludes 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 Uwe Durst
(7) X feels something because X thinks something

sometimes a person thinks
because of this, this person feels something (good/bad)

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 (prototypi-
cal) 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. rger
According to Farrell's (1971: 16) dictionary of German synonyms, the re-

flexive verb sich rgern "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 rger as
"Unwille, Verdrossenheit" (roughly: 'indignation', 'sulkiness'), as well as
Meldau (1978: 20), who adds "zornige Stimmung" ('angry mood'), "Ver-
dru" ('displeasure') - which, in turn, he defines in terms of "rger" 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 rger in everyday language, we can agree with Farrell insofar as a per-
son 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 spter, was mich rgerte: - Ich war aus dem Bad
gekommen, Hanna am Telefon, sie hatte das Hospital angerufen,
whrend 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-
tal while I was in the bath - she was talking to Elsbeth.

[MK1/LFH.OOOOO (Frisch, Homo Faber: 168)]
(9) Sogar rzte haben Feierabend, neuerdings sogar die Priester. Dar-
ber rgere ich mich, sie drften keinen haben und mten we-
nigstens das am Knstler verstehen.
Even doctors have time off, and recently even priests. 'About this I
rger.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 rgere mich ber meine Kollegen.
I rger.pRES myself about my colleagues
am vexed at my colleagues.'
The preposition ber literally means 'about', as in sprechen ber ('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 rgern is, first of all, the result of some-
thing that has happened before, or is still happening.
If in a sentence like (10) the noun rger is used, the prepositional com-
plement takes the preposition mit ('with') or in ('in'):
(11) Ich hatte heute rger mit den Kollegen.

I had today rger with the colleagues
'Today I had some trouble with my colleagues.'
(12) Ich hatte heute rger in der Arbeit.
I had today rger in the work
'Today I had some trouble at work.'
Actually, from a purely syntactic point of view, the prepositional phrase
is not obligatorily required and therefore could also be treated as an ad-
junct. However, in a sentence like (13):
(13) Aber sein rger ist noch nicht verflogen.
But his rger 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, be-
cause 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 rger.

If this known becomes get I rger.
'If this becomes known I'll get some trouble.'
[MK1/ZB 2.00001 (Bildzeitung 02.02.1967: 6)]
Clearly, the verb phrase rger bekommen does not mean 'to get angry' but
'to get into trouble'. Thus, it seems as if the word rger would be polyse-
mous between rger, = 'anger', 'annoyance', andArger2 = 'trouble'. How-
ever, 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'.
rger haben mit ('to have rger 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 rgern ber ('REFL r-
ger/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 no-
thing 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 rgert ber ('REFL rgern.SSG.PRES about') the weather, then all we
can say is that 'something is happening', or, as in (10) and (13), that 'some-
thing 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 rger l sich rgern there is evidence to suggest that think-
ing is involved before the feeling occurs. The reflexive form may indicate
that the emotion is generated by the experiencer's own thoughts. In the ex-
plication this is realised as part of the speaker's assertion (the prototypical
frame), rather than part of the experiencer's (prototypical) thought, and has
to be put straight below the prototypical thought ((e) and (f)). This can also
explain the fact that sich rgern usually takes the preposition ber
('about'), which can also be used in combination with the noun rger.1
Of course, we have no insight into what a person who feels rger really
thinks. Therefore it is necessary to present the person's thoughts within a
prototypical framework. Thus, the semantic description of X rgert sich
(ber Y) can be put as follows:
(15) rgerIX rgert sich (ber 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 rger 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 re-
ferred 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 rger
haben (mit) (where rger means something like 'trouble'). If someone hat
rger mit Y ('has rger with Y'), then he will inevitably sich ber Y rgern
('REFL about Y rgern.INF'). However, there seems to be something going
on between X and Y because of Y. If you 'have' rger mit your ignition de-
vice or rger 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 rger (mit Y) can be represented as:
126 Uwe Durst
(16) X hat rger (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 rger with prepositional phrases, and from
the semantics of the subsequent preposition itself, we can draw the con-
clusion that there is no "goal" component in the meaning of rger, 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
rger is often used in this "basic" sense in psychological, and even lin-
guistic, literature. (For a general survey of words for basic emotions in psy-
chological literature see Schwarz and Ziegler (1996: 49). In Schreiber,
Sommerfeldt, and Starke (1993: 97) rger is one of five hyperonyms in the
field of emotion words.) To give just a few examples of titles of recent psy-
chological publications:
Psychologie des rgers ('Psychology of rger', U. Mees, J. Chas-
sein eds 1992 Gttingen: Hogrefe)
rger und rgerausdruck ('rger and rger expression', V. Hodapp
1993. Bern: Huber)
rger: Psychologie einer alltglichen Emotion ('rger: Psychology
of an everyday emotion', H. Weber 1994. Weinheim: Juven-
ta)
In this respect, rger is treated as a basic or superordinate emotion cat-
egory, 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 rger
Wut (fern.) and Zorn (masc.) differ from rger in that they cannot combine
with the verb haben ('to have') without an article (rger haben, 'to have
rger'; 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 rger haben ('to have rger') 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 per-
son's emotion the nouns Wut and Zorn can be used as well as their corre-
sponding adjectives wtend and zornig. Thus, sentences such as (17) and
(18) can both have the same translation in English:
(17) Er hat eine Wut l einen Zorn.
He has a Wut l a Zorn
'He is angry.'
(18) Er ist wtend/ zornig.
He is wtend/ zornig
'He is angry.'
What makes it difficult to draw a distinction between the meanings of
Wut and Zorn (or wtend and zornig) is the fact that they are interchange-
able in a large variety of contexts constituting grammatical sentences.
Both nouns can be combined with possessive pronouns or genitive at-
tributes referring to the experiencer, and with a prepositional phrase desig-
nating 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 be-
tween inside and outside the body. Cf. some very common expressions
like:
(19) den Zorn l die Wut l den rger herauslassen
'to let the Zorn l Wut l rger out'
den Zorn l die Wut l den rger in sich hineinfressen
'to swallow the Zorn l Wut l rger'
128 Uwe Durst
While these metaphors are also well applicable to rger, the following
ones seem to be restricted to Wut and Zorn:
(20) Zornausbruch ('Zorn -outburst')

Wutausbruch (' Wut -outburst')
rgerausbruch ( rger -outburst')
(21) Mich packt ohnmchtiger 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 berkommen ('to seize/to overpower s.o.'), and people
can get erfat ('gripped/seized') or bermannt ('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 me-
diated 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 ('con-
flict'), 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
rgern sounds much more natural than the corresponding forms of eine
Wut/einen Zorn haben or wtend/zornig sein:9
(23) Mensch rgere dich nicht.

man rger.iMP yourself not
'Don't get angry'
(name of a popular game for four players round a table)
(24) rgere dich heute nicht.
rger.iMP yourself today not
'Don't get angry today.'
[MK1/ZB6.01036 (Bildzeitung 10.06.1967: 4)] (horoscope)
(25) La dich nicht (von X) rgern.
let.iMP you.ACC not (by X) rgern. INF
'Don't let someone or something (X) rger [trans.] you.'
(26) ?? Sei nicht wtend/zornig.
be.iMP not wtend/zornig
(27) ?? Hab keine Wut/keinen Zorn.
have.iMP no Wut/no Zorn
However, (26) sounds more felicitous if the temporal adverb gleich and
the modal adverb (or intensifying particle) so, in addition to the modal par-
ticle 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 appreci-
ate the hearer's behaviour, because what is rejected in (28) is not the emo-
tion 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 doch nicht gleich so wtend/zornig.
be.iMP MP not just so wtend/zornig
On the other hand, doch can also be combined with a negative imperative
of sich rgern , but hardly with the bare adjectives wtend and zornig:
(29) rgere dich doch nicht.
rger.iMP yourself MP not.
(30) ?? Sei doch nicht wtend/zornig.
be.IMP MP not wtend/zornig
130 Uwe Durst
There must be something in sich rgern that allows utterances such as

(29), and/or something in Wut and Zorn (or wtend and zornig) that is re-
sponsible for their incompatibility with the construction IMP + doch nicht.
As demonstrated by Ickler (1994: 403), the use of doch in imperative sen-
tences presumes that "the hearer's behaviour is clearly shaped and built
into a behavioural context that allows the opposition of an alternative be-
haviour" ("das Verhalten des Hrers eine deutliche Gestalt hat und in einen
Verhaltenskontext eingebaut ist, der die Entgegensetzung eines alter-
nativen 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 r-
gern that complies with this condition. In the formula given in (15), this
can only be the process of thinking: an emotion which is generated pri-
marily 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 unterdrckt ('suppressed') or gezgelt ('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 , pre-
sumably because they do not convey the notion of a thought as an anteced-
ent of the feeling, as it is expressed by the components (15e-f) and
(15g-h) in the semantic formula of rger /sich rgern. Of course, it is
possible that a feeling of Wut or Zorn can be evoked via thinking of some-
thing, 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 prep-
ositional adjunct with the preposition aw/('on', 'at') or ber ('about'). The
difference in the use of the prepositions auf and ber in this context con-
sists 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
saying this disproves Schreiber, Sommerfeldt, and Strke's (1993: 106)

claim that Wut auf ('Wut at') only combines with concrete nouns, while
Wut ber ('Wut about') combines with abstract ones only.
So far, we have seen some formal differences between rger 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 rea-
sons 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, lie sich an der Sparkasse Zweck und Ort des Verrech-
nungsschecks erklren, ging dann zu seinem Pfarrer und bat um
einen Barscheck ber fnfzehn Mark - aber der Pfarrer platzte
fast vor Wut. Er erklrte Heinrich, er knne ihm keinen Barscheck
geben, weil er die Zweckbestimmung erklren msse, und so ein
Caritaskonto sei eine heikle Sache...
The first check he simply paid into the welfare account of his par-
ish, had the bank explain to him the purpose and nature of a cross-
ed 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 pltzlich ihren Mantel
irgendwohin ins Zimmer, stampfend, auer sich vor Zorn... Ivy
hatte sich eingerichtet, eine Woche in Manhattan zu verbringen,
jetzt gestand sie's, und mein pltzlicher Entschlu, nicht zufliegen
wie blich, sondern morgen schon mit dem Schiff zu reisen, um in
einer Woche auch in Paris zu sein, war ein Strich durch ihre Rech-
nung.
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.
(33) Leo wurde immer wtend, wenn ich ihn traf und ihn fragte: "Weit
du noch, wie wir das Holz miteinander durchgesgt haben?" Er
schreit dann: "Aber wir HABEN das Hh nicht miteinander durch-
gesgt"
Leo always got wtend when I saw him and asked: "Do you re-
member how we sawed through that wood together?" He ex-
claims: "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 "Mrder",
Karl schrie mich an: "Du Vagabund - Du Hurenbock", und das
machte mich so wild, da 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 Strae und dachte an
Marie. Irgend etwas in meinem Gesicht schien den Mann, neben
dem ich stand, wtend zu machen. Er senkte sogar die Zeitung,
verzichtete aufsein "Strau: mit voller Konsequenz!", schob seine
Brille vorne auf die Nase, sah mich kopfschttelnd an und mur-
melte "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 wtend. 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)]
(36) Der Grovater war drr und eigensinnig, er geriet leicht in Zorn
und trank, je mehr die Mitbrger 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 ber 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
Brger 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 Hafenstdte
" mit furchtbarer Wut" plnderten.
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
wtend aber hart und bis der Gegner sah da er auf den Boden
kommen wrde unter diesen ausgerechneten zweckdienlichen
Sten.
He was still good at dealing a blow the way he had learned it: not
wtend, 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 ber 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 un-
controlled and spontaneous action seems to be involved, whereas the sec-
ond case seems to imply that the plundering is the result of a certain atti-
tude 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 wtend blows and calculated efficient blows. What is con-
veyed by wtend is the notion of someone beating his opponent in an un-
controlled 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 "allmhlich ansteigende Affecte"
("gradually rising affections"), and Wut in the category of "pltzlich here-
inbrechende 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 in-
dulgence') [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 be-
haviour 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 wtend is higher than the fre-
quency of zornig, and both are much more common than two other adjec-
tives derived from the same lexemes, wtig and zrnend:
Table 1. Frequency of wtend, wtig, zornig, and zrnend (adjectives)
MK LIM HBK
Total 165(100%) 10(100%) 260(100%)
wtend 113(68.5%) 8(80.0%) 161(61.9%)
wtig 2(1.2%) 0(0.0%) 4(1.5%)
zrnend 7(4.2%) 1 (10.0%) 2(0.8%)
zornig 43(26.1%) 1 (10.0%) 93(35.8%)
If word classes have their own (language-specific) meanings (see, e.g.,

Langacker 1987; Wierzbicka 1995b), this distribution can be explained by
the semantics of the words. The adjectives wtend and zrnend are for-
mally identical with the present participles of the verbs wten ('to rage', 'to
be in a fury') and zrnen (roughly: 'to be angry'). Wtig 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 wtend and
zrnend 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 con-
temporary use of the present participle wtend in its attributive, predi-
cative, 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 willkr entzogenen leiblich-seelischen aus-
zersichseins"; 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 com-
paratively small (118 word forms on -wut with 37 different lexemes, in-
cluding 46 word forms on -tollwut, versus 8 word forms on -zorn with
4 different lexemes, including 3 word forms of Jhzorn ('irascibility',
'passionateness') in the HBK corpus). To give some examples:
(40) Himmelszorn 'heaven-zor'; 'the wrath of the heavens'

Bibelzorn 'Bible-zo/ '
Volkszorn 'people-zor'
Arbeitswut 'work-wwf'; 'work mania'

Bauwut 'build-wMi'
Liebeswut 'love-ww?'
Putzwut 'clean-wwi'
Sammelwut ' collect- wuf
Zerstrungswut ' destruction- wut'
etc.
136 Uwe Durst
While the compounds on X-zorn can be paraphrased by a genitivus sub-

jectivus /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, Zerstrung, de-
rived from zerstren, 'to destroy'; Liebe is, in this context, not a psycho-
logical 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. There-
fore, they often convey a slightly negative or derogatory connotation,
which, however, is not part of their semantics, because the behaviour de-
scribed by the compound can be a desirable quality of a person:
(41) Zu seinen vorrangigen Eigenschaften mssen Risikobereitschaft,
Arbeitswut und Verantwortungsfreude gehren.
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-/zrn-, 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
zrnt (zurnen.3sG.pRES), but he is not wtend or wtig, although the Latin
translation provides ira Dei as well as furor Dei, the Greek translation orge
to theo and thymos to theo, and the Hebrew text has at least six differ-
ent 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 god-
desses of the Roman underworld (the Greek Erinyes). Their Latin name
Furiae is identical with the noun furia ('fury', 'rage'). The Berserker
('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/ad-
jective wtend can be attributed not only to human beings or animals, but
also to natural forces, e.g. storms or fires, to Ksse ('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 wtend in this case conveys the no-
tion 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 mean-
ing can tentatively be described as "momentary action" (that is, 'someone
is doing something now', or 'something is happening now').12 Intrinsic ad-
jectives, like zornig or wtig , on the other hand, have the function to at-
tribute a certain property or quality to what is referred to by the head noun.
Therefore, ein wtender Mensch ('a wtend.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 Os-
borne's angry young men can only be zornig, but not wtend, 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 zrnende Bonifatius ('the ever zr-
nercd.MASC.SG.NOM Bonifatius') [MK1/WPE.OOOOO (Poertner, Die Erben
Roms: 247)]. If they were described as wtend, 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 wtig, it would be char-
acteristic 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
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
(h) X cannot not feel like this
(43) Zorn
(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
(h) X cannot not feel like this
The first two lines within the prototypical thought differ in that the "trig-
gering event" seems to be more present for Wut. This is not to be under-
stood in the sense that it would be happening right now in any case. It is
quite possible to say ich bin immer noch wtend ( am still wtend'), 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 wtend 'because someone did something', but this is
not necessarily so. For example, if your computer fails to work unexpect-
edly in the middle of an important operation, your prospective reaction
would be described as wtend 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 wtend in all three contexts,
the English translation turns from furious to angry where the duration of
the emotion would be too much extended:
(44) "Ich hob immer gemeint, du bist wtend auf mich", sage ich,
"wegen damals'.'
Hanna nur verwundert.
"Wieso wtend? Weil wir nicht geheiratet haben?" sagt sie. "Das
wre ein Unglck gewesen -"
Sie lachte mich geradezu aus.
"Im Ernst", sagt sie, "das hast du wirklich gemeint, da ich w-
tendbin, Walter, einundzwanzig Jahre lang?"
"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 some-
thing') 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"), hei-
liger Zorn ('holy Zorn') etc.). This person, however, remains unspecified
in the semantic formula. In this respect Zorn differs from words like Ha
('hatred') or Rache ('revenge'), which seem to include the notion of an ac-
tion 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 wtende 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 compari-
son. In a cross-linguistic study on anger metaphors Kvecses (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 re-
sponses 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 grnlich ('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 mean-
ing 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 civili-
sation 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.
As mentioned before, the attributive, predicative and adverbial use of the

present participle wtend 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 com-
ponent "I have to do something now".
As for Zorn, it is controversial whether it descends from the Indo-Euro-
pean 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 para-
phrased 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 ob-
ject. In Middle High German it is still used in the sense of 'fight', 'en-
counter', '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 con-
stant here is want to do something bad to someone'.
Finally, the noun rger 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 rgern is derived from rger, which is the comparative form of the ad-
jective 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 at-
tributed 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 re-
ported above. First, none of the three most frequent 'anger' words in Ger-
man matches the meaning of the English word anger. Second, each of them
has its own semantic structure, determining their ranges of use and their re-
strictions in certain contexts. And finally, the meanings of these words as
analysed here are supported by the evidence from etymology if one as-
sumes 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 etymologi-
cally 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 Uni-
versity 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 Ar-
midale 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 compri-
muntur, 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: 34)
3. The prepositions von (Of, 'from') and ber ('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 fr 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 theor-
etical/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 par-
ticular 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, instruc-
tions for product use, etc., 1949-1973 (assembled by IDS).
Interlinear glosses are given where the argumentation is based on a particular struc-
ture. Abbreviations used in glosses are: ACC = accusative, FEM = feminine,
IMP = imperative, INF = infinitive, MASC = masculine, MP = modal particle,
NOM = nominative, PAST = past tense (Prteritum), 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. ber seems to be used where a deliberative process is involved in the action referred
to by the verb: sprechen ber ('to talk about'), nachdenken ber ('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 frchten ('to be afraid'). Thus,
the semantics of the preposition must not be confused with the semantics of the re-
flexive verb. The semantic interrelation between (reflexive) verbs and their preposi-
tional complements requires further investigation.
8. Actually, two examples for the phrase in rger geraten ('to get into rger') were
found, one of them in a text from the early 19th century [GOE/AGD.OOOOO (Dich-
tung und Wahrheit, Hamburger Ausgabe, vol. 9: 73)], another one in a contemporary
144 Uwe Durst
novel [MK1/LJA.OOOOO (Johnson, Das dritte Buch ber Achim: 211)]. However,
this use seems to be out of date.
9. Two imperatives of the verb zrnen (corresponding to Zorn) were found in the cor-
pus, 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 rid-
iculous. As Anna Wierzbicka pointed out to me, it is also quite common to use an-
other 'anger' word, bse (see example (4) above), with the negative imperative: Sei
mir nicht bse (lit.: 'be.IMP me.DAT not bad'; hence: 'don't be angry with me').
The semantics of the phrase jemandem bse sein (lit.: 'someone.DAT bad be.INF';
'to be angry with someone') were not investigated, but I would expect a similar re-
sult as for rger, as far as the notion of an intermediate thought, and thus the possi-
bility of controlling the rise of the feeling, is concerned.
10. 'op, za'am, hemh, hdron, qce$cep, and 'cebrah. For a comparative survey of their
meanings, see Bergman and Johnson (1973), and Myhill (1997).
11. 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 gltigen Normen geschehen soll... In Wut aber geraten
wir immer nur im Namen der Belange unseres individuellen Selbstseins.") How-
ever, 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.
12. 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 be-
tween the word class meanings of present participles and adjectives (see also
note 14).
13. 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 Ha
('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).
14. The semantic difference between adjectives on -ig and adjectives on -end is still
transparent in some cases. For example, ein tollwtiger 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 Hnde
('shaky hands', adj.), which implies, to a certain extent, general weakness, or old
age etc., whereas zitternde Hnde ('trembling hands', pres. part.) describes a tem-
porary state; consequently, zitternd can combine in such contexts with causative
prepositional phrases indicating a temporary deviation from the normal, e.g. zit-
ternd vor Angst/vor Klte ('trembling with fear'/'shivering with cold').
Why Germans don if eel "anger" 145
References
Adelung, Johann Christoph

1808 Grammatisch-kritisches Wrterbuch der Hochdeutschen Mundart, mit
bestndiger Vergleichung der brigen Mundarten, besonders aber der
Oberdeutschen. Mit D. W. Soltau's Beytrgen, revidirt und berichtiget
von Franz Xaver Schnberger. Wien.
Bergman, J. and E. Johnson
1973 "nap. In: G. Johannes Botterweck and Helmer Ringgren (eds.),
Theologisches Wrterbuch 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.
Bll, Heinrich
1988 The Clown. (Translated by Leila Vennewitz). London/New York: Ma-
rion Boyars
Brockhaus-Wahrig
1980-1984 Deutsches Wrterbuch. Edited by Gerhard Wahrig, Hildegard Krmer,
Harald Zimmermann. Wiesbaden: F. A. Brockhaus; Stuttgart: Deut-
sche Verlags-Anstalt.
Campe, Joachim Heinrich
1969-1970 Wrterbuch der Deutschen Sprache. Mit einer Einfhrung und Biblio-
graphie von Helmut Henne (Documenta Linguistica; Reihe II) [Repro-
grafischer Nachdruck der Ausgabe Braunschweig 1807-1811]. Hil-
desheim; New York: Georg Olms Verlag.
Debus, Gnter
1988 Psychologie der Gefhl s Wrter: Empirisch-experimentelle Untersu-
chungsanstze und -ergebnisse. In: Jger, Ludwig (ed.), 95-138.
Die gotische Bibel
1960. Edited by Wilhelm Streitberg. Heidelberg: Carl Winter, Universitts-
verlag. 3./4. Aufl.
Duden
1989 Duden Deutsches Universalwrterbuch. Mannheim, Wien, Zrich:
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). Harmonds-
worth: 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 Wrterbuch. Leipzig: S. Hirzel
1997 Language: A key issue in emotion research. Innovation 10/4:
319-331.
Ickler, Theodor
1994 Zur Bedeutung der sogenannten 'Modalpartikeln'. Sprachwissen-
schaft 19: 374-404.
Jger, Ludwig (ed.)
1988 Zur historischen Semantik des deutschen Gefhlswortschatzes: Aspek-
te, Probleme und Beispiele seiner lexikographischen Erfassung. Aa-
chen: Alano.
Jger, Ludwig and Sabine Plum
1989 Probleme der Beschreibung von Gefhlswrtern im allgemeinen ein-
sprachigen Wrterbuch. In: Franz Josef Hausmann et al., (eds.), Wr-
terbcher. Dictionaries. Dictionnaires: Ein internationales Handbuch
zur Lexikographie, vol. l, 849-855. (Handbcher zur Sprach- und
Kommunikationswissenschaft; 5.1) Berlin, New York: de Gruyter.
Klappenbach, Ruth and Wolfgang Steinitz, (eds.)
1961-1977 Wrterbuch der deutschen Gegenwartssprache. Berlin: Akademie-
Verlag.
Kvecses, Zoltn
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. Mnchen: Barth. 7. Aufl.
Luther, Martin
1968/72 D. Martin Luthers Werke: Kritische Gesamtausgabe, vol. 6-12: Die
Deutsche Bibel 1522-1546. Unvernderter Abdruck der bei Hermann
Bhlaus Nachfolger, Weimar (1929-1961), erschienenen Ausgabe.
Graz: Akademische Druck- u. Verlagsanstalt.
Marx, Wolf gang
1982 Das Wortfeld der Gefhlsbegriffe. Zeitschrift fr experimentelle und
angewandte Psychologie XXIX/1: 137-146.
Meldau, Rudolf
1978 Sinnverwandte Wrter und Wortfelder der deutschen Sprache: Ein
Handbuch fr den Deutschunterricht. Paderborn: Schningh.
Myhill, John
1997 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 Gefhlswrter im Wrterbuch: berlegungen zur lexikographischen
Bedeutungserluterung des emotionalen Wortschatzes. In: Gregor
Meder und Andreas Drner (eds.), Worte, Wrter, Wrterbcher. Lexi-
kographische Beitrge zum Essener Linguistischen Kolloquium,
169-182. (Lexicographica: Series Maior; 42). Tbingen: 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 Bercksichtigung physiologischer Prozesse. Ph. D.
Dissertation, Universitt Gieen.
1987 Zur umgangssprachlichen hnlichkeit von Emotions Wrtern. Psycho-
logische Beitrge 29: 140-163.
Schmidt-Atzert, Lothar and Walter Strhm
1983 Ein Beitrag zur Taxonomie der Emotionswrter. Psychologische Bei-
trge 25: 126-141.
Schreiber, Herbert, Karl-Ernst Sommerfeldt and Gnter Starke
1993 Deutsche Substantive: Wortfelder fr den Sprachunterricht. Leipzig:
Langenscheidt.
Schwarz, Manuela and Wolfram Ziegler
1996 Emotionen in Neurolinguistik und Lexikologie: Ein Forschungsber-
blick. 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; Cam-
bridge: Harvard University Press.
[1989] Philosophische Schriften. Bd. 1. bersetzt, eingeleitet und mit Anmer-
kungen versehen von Manfred Rosenbach. Darmstadt: Wissenschaft-
liche Buchgesellschaft. 4. Aufl.
Septuaginta
[1962] Edidit Alfred Rahlfs. Stuttgart: Wrttembergische Bibelanstalt. Editio
Septima.
Stearns, Carol Zisowitz and Peter N. Stearns
1986 Anger: The Struggle for Emotional Control in America's History. Chi-
cago: 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 Lin-
guistics. 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 Fernndez-Dols, Antony S. R. Manstead and
J. C. Wellenkamp (eds.) Everyday Conceptions of Emotion: an Intro-
duction 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.
1999 Emotions Across Languages and Cultures: Diversity and Universals.
Wittgenstein, Ludwig
1967 Zettel. Oxford: Basil Blackwell.
Wundt, Wilhelm
1896 Grundri 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 ex-
pression, James Russell has recently campaigned for "the gathering of new
evidence" (1995: 382) concerning the nature and meaning of facial ex-
pressions 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 fa-
cial 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 iso-
late, 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, dis-
gusted, 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 alone-
ness, 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 vari-
ous contexts, by the English word angry. The first is caj-haaj Offensive-
heart' (in the sense of 'heart-on-the-offensive'), which expresses the emo-
tional 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 react-
ing 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 Eng-
lish 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 no-
thing', 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 re-
mains an effective way of getting the message across.
1.3. 'Disgusted'
There is a fairly clear translation into Lao of English disgusted, namely khi-
diat. 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 ef-
fluent material and waste products (khi-muuk [muuk = nose] 'mucus'; khi-
thaw [thaw = 'ash'] 'ashes'; khi-luaj [luaj = 'mill'] 'sawdust'), as well as
negative personal characterisations (khi-khuj [khuj = 'chat'] 'snobby'; khi-
lak [lak= 'steal'] 'thief; khi-qaaj [qaaj= 'shy'] Overly shy'). (The mor-
pheme -dial has no independent meaning outside of the expression khi-diat
'disgusted'.) The usage of khi-diat is more restricted than that of the Eng-
lish 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' some-
one, or 'that' something has happened.2
1.4. 'Happy'
The English expression happy may be translated into Lao as dii-caj 'good-
heart', 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 Eng-
lish 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 en-
tail 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 lead-
ing to the feeling. This prior cognitive evaluation is the hallmark of emo-
tion (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 com-
fort which sabaaj 'content' allows. Rather, it is a profound state of well-
being 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 un-
expected, 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 distinc-
tion in Malay.) Tok-caj, however, can include the idea of a cognitive evalu-
ation 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-
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 re-
spect to tok-caj, some specific event must accompany the evaluation lead-
ing to the feeling of 'surprise'.
1.6. 'Afraid'
The most obvious translation of afraid or scared into Lao is jaan. It is un-
clear 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 sugges-
tive of immediate and present danger. Jaan is also used in Lao as a com-
plement-taking predicate expressing what the speaker merely suspects
rather than literally fears (cf. English I'm afraid she's not coming), and un-
like 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. Eng-
lish ?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', oc-
curs 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 grammati-
cal 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 dis-
cussion).
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', caj-
kaa [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 ex-
tensive study.
1.8. Comment
We began with the question of correlations across languages between emo-

tions and facial expressions. We are yet to consider facial expressions, but
it is already clear that particular facial expressions cannot be said to uni-
versally express particular emotions. This is because when eliciting and/or
describing Lao speakers' attribution of inner states to people making vari-
ous 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, con-
centrating on lexical items and idioms which can be used to describe par-
ticular facial expressions (or features thereof). I will note the kinds of at-
tributions 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 cor-
relation, 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.
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 poly-
morphemic expressions, such as 'rotten face', 'smelly face', 'stuck eyes',
and 'clenched mouth'. In between these extremes are a number of 'cran-
berry' expressions (i.e. in which one component does not independently
occur), such as naa-beq 'ready-to-cry face' and taa-loo 'exposed-whites-
above-iris eyes', in which the modifying elements (beq and loo, respect-
ively) have no independent meaning.
Syntactically, these facial expression descriptions involving head nom-
inals 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 con-
struction ( 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 func-
tions, 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) in-
volves 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 poss-
ible. 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 ab-
stract "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-
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 physi-
cally disgusted person only.
2.2.3. naa-khm '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 ex-
pressions 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 sa-
baaj(-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 [pro-
gressive-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 qaa-
paak (where in English one would say open wide or say 'ah'). The qaa-
paak 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 Open-
mouthed', 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 open-
mouth] '?S/he spoke open-mouthed'.
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 nom-
paak '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 sugges-
tive 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. In-
formants 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 'exposed-
whites-above-iris eyes' with some incongruous pose, such as a smiling
mouth, Lao informants attribute to the person a state of insanity (baa 'in-
sane'), 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 ex-
pose 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', al-
though it may also mean that they are merely son-caj 'interested' in some-
thing.
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 surpris-
ing 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 hap-
pening 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 fa-
cial 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 mo-
mentaneous nor extended time profile. Thus, taa-khaang 'stuck-eyes' is
not associated simply with being tok-caj 'surprised', but with this plus
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-whites-
above-iris eyes', this is associated with being jaan 'afraid', as well as tok-
caj 'surprised'.
2.4.6. hetkhiwson 'making-eyebrows-collide'

Lao does not have a term equivalent to English/rovwi, and the idea of bring-
ing one's eyebrows together is expressed as het khiw son 'make the eye-
brows collide'. (The verb son is used elsewhere most notably for the but-
ting of buffaloes' heads when they clash.) Inner states of son-caj 'inter-
ested', or kamlang-khit [progressive-think] 'thinking' are attributed to
someone displaying this facial expression.
162 N.J.Enfield
2.4.7. het khiw ct-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. Simi-
larly, 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 de-
scriptions of facial expressions make specific reference to components of
the face, and I suggest that in further research, it would pay to consider fa-
cial expression as having a greater componentiality than is sometimes as-
sumed. (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 "emo-
tions". What are denoted by words for 'emotions' are not actual phenom-
ena, 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 emo-
tion, we cannot escape our reliance on language. Scientific inquiry is
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 ex-
pressions 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 Eng-
lish, and not all societies know and/or use English.
It is interesting to consider the parallel between research on facial ex-
pression 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 pro-
vided 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 ethologi-
cally 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 fa-
cial 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-un-
rounded, low-back, and low-front vowels, respectively. All data are from fieldnotes,
and from field interviews conducted in Sisavat and Saphang Mo villages in Vien-
tiane, Lao PDR, January-May, and July-August 1998. Syban Khoukham's assist-
ance is gratefully acknowledged. Pitsana Vayaphanh also provided generous con-
sultation, 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 as-
pect 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 physi-
cally 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. '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] woke up at six o'clock', khooj tuun caw [I awaken you] (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 sub-
ject khooj T.
4. Even so, colour research has resulted in interesting findings, such as the various im-
plicational 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 ex-
pressions? 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 ques-
tion to ask (cf. Lucy 1996 on colour).
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 Ptz and Marjolijn H. Verspoor (eds.),
Explorations in Linguistic Relativity, 125-157. Amsterdam: John Ben-
jamins.
2001 Semantic analysis of body parts in emotion terminology: avoiding the
exoticisms of "obstinate monosemy" and "online extension". Prag-
matics 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. Lin-
guistics of the Tibeto-Burman Area 13(1): 141-158.
Russell, James A.
1995 Facial expressions of emotion: What lies beyond minimal universal-
ity? 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
Zajonc, R. B.
1980 Feeling and thinking. Preferences need no inferences. American Psy-
chologist 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 emotional-
moral 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 dis-
course 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 'deter-
mined', 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 can-
not approach an understanding of Malay attitudes about human nature and
about social life without understanding this quintessentially Malay con-
cept.
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 contrast-
ing it with English heart, second, to advance and argue for an explicit sem-
antic 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, of-
fended', 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 com-
parable 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 isjan-
tung), 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, per-
haps, but not one's heart). On the same point, it is noteworthy that certain
fixed expressions involving hati are more felicitously translated into Eng-
lish using mind than heart. For example, senang hati (senang 'relaxed,
easy') as 'peace of mind'.
(1) ... seribu satu masalah memenuhi hatinya. Ingatannya pada ibu-
nya 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 ex-
pressions such as those in (2) and (3), in which the hati is depicted as ter-
tanya 'spontaneously asking' an emotionally loaded question, or as bisik
'whispering' an emotional reaction. In general, the hati is capable of "ex-
pressing" itself in ways which can be reported with a wide range of speech-
act 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 ,
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 at-

tributed directly to a person's hati, as though the hati itself were the "ex-
periencer". That is, rather than saying the equivalent of 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 diiris-
iris, 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 de-
sire 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 some-
thing. 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 hen-
dak 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 im-
portant 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, in-
volve 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'. Conse-
quently, 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 differ-
ence between the expressions di hati (lit. 'at hati') and (di) dalam hati (lit.
'inside hati'}, the latter expression conveying a greater sense of conceal-
ment, 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 ex-
pression 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 at-
traction. For example, Hatinya terpikat kepada Jasni 'her heart was snared
by Jasni', i.e. she was secretly in love with him. Sundry other hati ex-
pressions 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 concen-
tration 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. jaga hati 'to be protective of (someone's) feelings' (lit.
'watch over hati')
b. ambil hati 'to charm (someone), win over' (lit. 'get hati')
c. suka hati 'do as you please, do whatever you feel like' (lit.
'please hati')
d. tidak sampai hati 'can't stand to (do something)' (lit. 'not
reach hati')
e. cuba hati 'test (someone's) honesty or commitment' (lit. 'try
hati')
f. 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'. In-
deed, jaga hati is often paraphrased as memelihara perasaan 'tend to feel-
ings' or bertimbang rasa 'take account (lit. weigh) feelings'; and the com-
bination hati dan perasaan 'hati and feelings' is not uncommon.
In the expressions in (14b)-(14d), however, we find hati being em-
ployed 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 re-
garded, in Malay culture, as a risky and immature way to act), and the ex-
pression 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.
The same broad range of semantic effects is found among fixed nominal
expressions involving hati. Most of these have a two-part structure, con-
sisting of hati and an adjective. They can be grouped formally according to
whether the adjective follows or precedes the noun hati. Attributive adjec-
tives normally follow their head nouns in Malay, but hati + adjective is the
less frequent of the two patterns. Abdullah's (1990) Kamus Simpulan Ba-
hasa [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 + ad-
jective 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 pre-
posed 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 simi-
lar in construction to English phrases such as sick at heart). The locus con-
struction with hati also seems somehow to "personalise" the state being de-
scribed. 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 ex-
plicitly 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: b. Aspects of moral character:
besar hati 'proud, thrilled with' baik hati 'kind, good-hearted' (baik 'good')
(besar 'big')
kecil hati' feeling hurt' (kecil 'small') busuk hati 'vicious, mean' (busuk 'putrid')
geli hati 'amused, tickled' (geli tulus hati 'sincere, honest' (tulus 'pure, un-
'ticklish') adulterated')
iri hati 'envious, spiteful' murah hati 'generous hearted', (murah
'generous')
panas hati 'angry, worked up' (panas rendah hati 'humble, modest' (rendah
'hot') 'low')
sejuk hati 'calmed, reassured' tinggi hati 'proud' (tinggi 'high')
(sejuk 'cool, cold')
susah hati 'troubled' (susah 'difficult') putih hati 'sincere, honest' (putih 'white')
senang hati 'contented, relaxed' suci hati 'pure of motive' (suci 'morally
(senang 'easy') pure')
hancur hati 'devastated' (hancur tabah hati 'determined' (tabah 'resolute,
'crushed') persevere')
puas hati 'satisfied' (puas 'sated') lembut hati 'tender-hearted' (lembut
'tender')
sakit hati 'annoyed, resentful' tawar hati 'lost interest' (tawar 'bland')
(sakit 'sick, pain')
putus hati 'despairing' (putus keras hati 'stubborn' (keras 'hard')
'broken asunder')
lapang hati 'content' (lapang
'vacant, open')
(16) hati + Adjective

a. Moral states (often in religious b. States of mind:
contexts):
hati bersih (lit. 'clean') hati keras 'determined' (keras 'hard')
hati kotor (lit. 'dirty') hati terbuka '(with) an open mind' (terbuka
Open')
hati nurani 'enlightened, virtuous' hati waja 'resolute, determined' (waja
(nurani 'bright') 'steel')
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'
To round out this quick picture of the uses ofhati, it is instructive to look
at someperibahasa 'traditional sayings' andpepatah 'maxims'. Tradition-
ally, 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 har-
dest 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. Per-
haps 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 Iain-
lain '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. Sehati sejiwa.
One hati, one soul.'
b. 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 poss-
ible, 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 lan-
guages, 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, accord-
ing 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 de-
scribe 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 refer-
ence to the concepts of 'good' and 'bad' (one can 'feel' surprise and inter-
est, 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 re-
flects the fact that, so far as feelings directed towards other people are con-
cerned, the heart is specifically linked with the potential for positive feel-
ings (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 re-
cognised 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 pro-
pose 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 perma-
nently their psyche and personality in both a psychological (zahir) and
spiritual (batin) sense."
The bulk of the explication attempts to articulate the rich cultural con-
strual 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 com-
ponent (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. compo-
nents (c)-(e), spells out the role of the hati as the wellspring of interper-
sonal 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 monitor-
ing and regulation of the hati, as mentioned in the earlier discussion).
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 'say-
ing 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 'troubled, worried, preoccupied' (susah 'difficult')

senang hati 'relaxed, comfortable, happy' (senang 'easy')
sakit hati 'annoyed, offended, resentful' (sakit 'sick, pain')
puas hati 'satisfied, pleased' (puas 'satisfied, sated')
kecil hati 'feeling hurt, disappointed' (kecil 'small')
Each of these expressions is commonly verbalised by means of the pro-
ductive 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 me-
muaskan, and corrresponding to senang we have menyenangkan. Thus:
(26) menyusahkan hati 'make troubled, preoccupied'

menyenangkan hati 'make comfortable, please'
menyakitkan hati 'annoy, offend, embitter'
memuaskan hati 'satisfy, please'
mengecilkan hati '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 menca-
ri 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, com-
ponent (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 in-
clude 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
man was definitely ruminating on what was going to happen to him. His ill-
fated life was evidence for him that there was worse to come.)
(29) Lelaki tua semakin susah hau. Dia sentiasa mengeluh mengen-
angkan 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 un-
wanted event befalling someone, component (d) spells out the consequent
sense of responsibility for finding a solution or resolution ( have to do
something'). Component (e) states that, for the time being, at least, the ex-
periencer is at a loss about what to do ( 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 compo-
nents (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 me-

nyusahkan 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 cop-
ing 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 con-
cerned about possible misfortune (thinking 'something bad can happen to
someone; I don't want this') and feels under pressure to act (thinking
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 com-
ponent (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
Some features of this explication can be justified by reference to the fol-

lowing 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 some-
one else one feels loyal toward or protective of).
(35) Anak rusa berasa senang hati kerana tuannya sentiasa menjag-
anya 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 ar-
rived at his cousin's house safely.'
As suggested by these examples, and also by the following one, the ex-
pression 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 re-
ferred 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 re-
sponsibilities 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 address-
ed 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 memer-
angkap 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' (kun-
ing 'urine'), so on the basis of their English counterparts one might im-
agine 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 brood-
ing over wrongs or supposed wrongs'.
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 diction-
aries tend to combine a reference to marah 'angry, offended' with tersing-
gung 'have one's feelings hurt'. For example: berasa marah atau tersing-
gung kerana dihina atau ditipu 'feel offended or hurt because being ma-
ligned or cheated' (Abdullah 1996: 265); berasa tersinggung atau marah
kerana sesuatu sebab atau perbuatan orang lain 'feel hurt or angry be-
cause some reason or action of another person' (Zakaria 1988: 118). No-
tice that these two sources differ on whether the cause of sakit hau is typi-
cally or predominantly verbal (as implied by Abdullah's reference to
being maligned or cheated), or whether the cause might be an action of an-
other 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 experi-
encer. 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 aman-
ah. 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-ver-
bal events - by a social event such as divorce and its attendant compli-
cations, 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.'
Puas hati
The expression puas hati designates a good feeling resulting from some-
one'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 man-
aged 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 mat-
any 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.
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 Cliff Goddard
(52) 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 necess-
arily, says something) and this makes us realise that this person doesn't ac-
tually 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 dis-
missive 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 over-
heard Razi talking about her like that, she would feel kecil hati. These con-
siderations 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 ker-
ana setiap kali berjumpa mereka akan mengeluarkan perkataan
yang bukan-bukan tentu juga kita rasa kecil hati.
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.'
(55) "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 kata-
kata 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 re-
gard after all; hence, component (d) 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 benar-
benar 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 problemati-

cal 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 gen-
eral 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 cross-
linguistic semantics. Numerous studies have shown that seemingly "basic"
English words such happiness, anger, fear, love, and pride do not have pre-
cise equivalents in other languages, and, conversely, that the apparently
"basic" emotion vocabulary of other languages often resists straightfor-
ward 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 as-
sume 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 mis-
taken. 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,
what is needed is a series of detailed semantic studies of secondary lex-

emes for emotions (and related concepts) in a range of languages.
A second lesson is that to arrive at an authentic picture of the Malay vo-
cabulary of emotion, it is not enough to stop at the one or two dozen mono-
lexemic 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 elab-
orate vocabulary of emotion; for example, Thai, cf. Diller and Juntanama-
laga (1990). If we identify these languages and study these elaborated sys-
tems, 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. Ethnopsychol-
ogy is much broader than this. Wierzbicka (1992) has shown that studying
the Russian key concept of duSa 'soul' opens up a uniquely Russian per-
spective on human nature. I have no doubt that studying the Malay key con-
cept of hati can open up a uniquely Malay perspective on human nature,
and, in particular, on the role of interpersonal feelings in the human condi-
tion.
APPENDIX:
Semantic Primes - English and Malay Exponents
Substantives Time
I-AKU, - WHEN/TIME-BILA/MASA(WAKTU)
SOMEOWE-SESEORANG NOW-SEKARANG
SOMETHING-SESi/ArC/ BEFORE-SEBELUM, AFTER-SELEPAS
PEOPLE-ORANG A LONG TIME-LAMA, A SHORT TIME-
SEKEJAP
FOR SOME -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, HEAR-DENGAR ABOVE-(>/) 5, BELOW-(/?/)
BAWAH
Speech INSIDE-O/) DALAM
SAY- 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, HAPPEN-TERJADI IF-KALAU, -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, SMALL-KECIL
Determiners and Quantifiers Intensifier and augmentor

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 4G7A7V, KIND-/BN/5
SOME-BEBERAPA, A.LL-SEMUA LIKE-MACAM
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 Aus-
tralian 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 fac-
tor is that one may form any number of causative expressions which have hati as ob-
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 de-
scribe 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 ex-
pressions like menyenangkan hati and memuaskan hati (for example) are based di-
rectly 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 sug-
gests the transient state of the experiencer's hati, and would be appropriate in repor-
ting 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 tepat-
tepat. 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 pertama-
petang tu.
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
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 Benja-
mins.
Goddard, Cliff and Wierzbicka, Anna (eds.)
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 In-
donesia. Cambridge: Cambridge University Press.
Karim, Wazir Jahan (ed.)
1990 Emotions of Culture. A Malay Perspective. Singapore: Oxford Univer-
sity Press.
Maeda, Narifumi
1975 Family Circle, Community, and Nation in Malaysia. Current Anthro-
pology 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): 557.
Oey, Eric M.
1990 'Psycho-collocations' in Malay: A Southeast Asian areal feature. Lin-
guistics 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.
Wierzbicka, Anna
1992 Semantics, Culture, and Cognition: Universal human concepts in cul-
ture-specific configurations. New York: Oxford University Press.
1999 Emotions Across Languages and Cultures: Diversity and universals.
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 is-
sues 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 cat-
egory of things that people would want to know more about, and to read
about in this and other volumes, arises from a particular cultural perspec-
tive or perspectives. While methods of study and investigation vary across
academic disciplines, the activities of collecting data, observing, experi-
menting, 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 lan-
guage-independent and culture-neutral metalanguage, as most of the
studies in this volume do, we need to consider that this is a culturally situ-
ated activity. Even if our metalanguage can be culture-free, when we en-
gage in constructing definitions or explications, trying to arrive at precise
statements of meaning that represent the necessary and sufficient condi-
tions for the use of a linguistic expression, formulating and discussing sem-
antic 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 some-
what 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 car-
ried out in a wide range of linguistic and cultural contexts, by people of di-
verse 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 underesti-
mated. 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 in-
terest 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 illuminat-
ing, when investigating emotional or other meanings, to cultivate aware-
ness of the cultural and intercultural dimensions of the enterprise, and to
see both the processes and outcomes of the investigation in this light. Aim-
ing for complete scientific objectivity in a culturally situated study would
be not only foolish, but less informative than taking into account the cul-
tural 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 cat-

egory, 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 con-
cepts 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 formulat-
ing 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 (Hender-
son 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, with-
out 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;
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 pro-
cess of translating definitions into the language under investigation and
back to English, often more than once, very helpful in clarifying and refin-
ing 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 ap-
propriate 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 proto-
type 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 re-
lated 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 first-
language Arrernte speakers living in Alice Springs, all with some experi-
ence in bilingual education and Arrernte language work: Margaret Heifer-
nan, 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 attract-
iveness of the image of listening to one's own inward state (cf. also God-
dard 1989: 54). Participants referred to the topic of our discussions as ang-
ketye 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. Ay enge akurne awelhe-me.

IsS bad feel-NPP
feel (something) bad.'
b. Ayenge mwarre awelhe-me.
IsS good feel-NPP
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 be-
cause you won't talk to me" the physical or emotional interpretation is un-
ambiguous. 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 re-
lated 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 con-
sidered whether the expressions collected could be arranged into sub-
groups of related feelings, and found that it was easy for speakers to ar-
range them intuitively into seven groups. These are shown below, with
some typical members (from various word-classes) of each subgroup. The
rough glosses give some indication of the meanings, but for many of these
words there is no adequate English equivalent.
(2) a. Feeling good (mwarre awelheme)

akangkeme 'happy, having a good time'
ilkelheme 'enjoying'
atnewe-atnewe 'relaxed, calm'
atneratye 'trust, confidence in someone'
b. Wanting something (ahentye-aneme)
ahentye 'wanting, liking'
alhirreme 'admiring, fancying'
kwakwe 'keen'
ingketye 'desirable'
c. Feeling bad (akurne awelheme) because something happened
alhwarrpe 'sadness, sorrow, grief
irrare 'loneliness, yearning'
irrareltheme 'depressing'
d. Feeling bad towards someone else
ahe 'anger'
ahakngerre 'aggressive'
akiwarre 'infuriated'
arnkelye 'resentful'
ingkerte-ingkerte 'jealous'
e. Feeling bad about something
akurne awelheme 'regretting'
apurrke-apurrke 'bored, exasperated'
ilkelhetyakenhe 'uncomfortable, fed up'
f. Feeling bad because something bad could happen
akngerre anthurre itirreme 'worrying'
atere 'fear'
apure 'shame'
g. 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. Wil-
kins (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 speak-
ers. These feelings arise from transgressions of social and cultural norms
and relationship boundaries, and from severe pressures imposed by in-
vasion of and damage to Arrernte country by outsiders. English lacks pre-
cise equivalents or ways of expressing most of them, with the dual out-
comes of frequent intercultural misunderstanding in difficult situations,
and significant semantic differences between some Aboriginal English
words and their non-Aboriginal English counterparts. Although this analy-
sis deals mainly with Arrernte, Aboriginal English and other local lan-
guages 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 be-
tween these two closely linked languages and cultural groups, despite fre-
quent interaction and kinship bonds between them.
As a first example, consider the following examples from Arrernte nar-
rator Willie Rice (1986: 28-31), who provided the English glosses in con-
sultation 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.'
(4) ...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) trans-
lated using the English words angry and cheeky respectively. This reflects
the Aboriginal English usage, where hostile could be a closer non-Aborig-
inal 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 Ar-
rernte, 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, instrumen-
tal and other functions in addition to its function here as a manner adverb
formative (Wilkins 1989: 176). This is followed by the inchoative deri-
vational suffix -irre, whose occurrence with many emotion predicates (in-
cluding 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 sem-
antic comparison and analysis. These are listed below, with English glosses
from Henderson and Dobson (1994):
(5) a. ahele-irreme
'get angry, be angry at someone'
b. arnkelye
'not talking to people because you're unhappy
about something; sulking, in a sulk'
c. akiwarre
'showing anger; in a tantrum, sulking,
having lost your temper, pissed off,
flying off the handle'
204 Jean Harkins
d. ahakngerre
'aggressive, angry, violent, tough, mean,
savage, vicious, dangerous, touchy. In local
Aboriginal English, "cheeky" has this meaning,
and not its standard English meaning.'
e. 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 trans-
lated 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 sulk-
ing 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 of-
fence, 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 twen-
ty examples of each in order to observe its range of use as broadly as poss-
ible, 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 Ar-
rernte texts, but most of the available writings were factual or event-
oriented 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, indicat-
ing 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
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 emo-
tional response - whether a person felt akiwarre or arnkelye or ahele-ir-
reme 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 rea-
sons of anonymity. In a close-knit community, it could be undesirably ob-
vious 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 of other
NSM explications came into play, though for reasons that had more to do
with social realities than with issues of indexicality in semantic represen-
tation (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
did something very bad'
c. the akurne ile-tyenhenge renhe ikwere-nge
IsA bad do-SBSQ 3sO SSDAT-ABL
will do something bad to because of this'
d. X alakenhe itirre-me-le, X akurne awelhe-me Y-ke
X thus think-NPP-ss X bad feel- 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 feel-
ings is far from incompatible with the cognitive perspective of NSM expli-
cations for other languages: component (a) expresses a concept of the ex-
periencer's feeling something as the result of her own thoughts. This com-
ponent 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 ex-
periencing arnkelye: a strongly negative evaluation of the offender's ac-
tion, 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'
b. Y-le akenge mpware-ke-nge atyinhe-ke
Y-ERG rotten make/do-pc-os ISPOSS-DAT
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
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
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- 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 im-
pulses (d) can persist inwardly or find expression in a show of temper. In-
terestingly, 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 uncommunicative-
ness 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
English sulky might be closer to these Arrernte meanings than to the non-
Aboriginal 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 af-
fects 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'
b. Y-le atyinhe akurne ile-ke
Y-ERG ISPOSS bad do-pc
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
don't want 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
-0 renhe ikwere-nge
Y-ABS 3SO 3SDAT-ABL
want to do something bad to because of this'
e. 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 in-
herently interpersonal, directed at some specific offender. Participants in-
sisted that Y-ke -DAT' was a necessary element in the first and final com-
ponents 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 cate-
gorisation 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 in-
terpersonal emphasis (Y-ku -DAT'), and the desire for retaliation in mirp-
anarinyi but not kuyaringanyi.
Space does not permit a full analysis here of ahakngerre or atnerte am-
peme, but we can observe that ahakngerre, somewhat like the Western De-
sert pikaringanyi (Goddard 1991,1998), seems not to be an emotional feel-
ing but a general propensity to do something bad to someone. The fact that
ahakngerre, like Aboriginal English cheeky, can be predicated of non-hu-
mans and inanimates, such as snakes or poisonous plants, raises an inter-
esting problem of whether or not this should be represented as a metaphor-
like "wanting to do something bad to someone", or in terms of the possi-
bility 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,
against a rich background of culturally situated experiences, beliefs, valu-

es, 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 Kvecses' (1987: 213) anger scenario or Frijda's (1986) elicit-
ing 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 re-
solved, or escalated (cf. examples 6-12 and 15 above). Some of these cul-
tural 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 se-
mantics 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 practi-
tioners 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 fa-
cilitating 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 cul-
tural setting - a far cry from some popular notions of semantics as an ab-
stract, 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.) pro-
posed 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 practi-
cal 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 unman-
ageably time-consuming way to produce Arrernte language definitions for
more than a handful of Arrernte words. Nonetheless, the intercultural in-
sights gained through the use of this method of analysing meanings, par-
ticularly for important social and cultural concepts, were valued by all con-
cerned. These outcomes are at least as important, indeed more so for most
practical puposes, as the artefacts of semantic formulae that were pro-
duced. The fact that a group of interested speakers could gain some satisfy-
ing 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 com-
munication 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 represen-
tation of meanings. While some aspects of the NSM approach remain con-
troversial, 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
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 emo-
tional 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 Abo-
riginal Development, the Arandic Dictionaries Project, and Pamela Ditton for prac-
tical assistance that made this work possible.
2. Abbreviations used in morphemic glosses (mostly following Wilkins' 1989 analy-
sis) are: 1, 2, 3[person]s[ingular]/p[lural]; ABLative; ABSolutive; ADVerb; Associ-
ative; Afgent]; CONTinuous; DATive; DS different subject; DUal; EMPHatic; ER-
Gative; FREQuent; IMPerative; INCHoative; INSTrumental; LOCative;
Neg[ative]IMPerative; NMZ nominaliser; Nom[inal]NEGator; NPC non-past com-
pletive; 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 Ptz and M. H. Verspoor (eds.), Explo-
rations in Linguistic Relativity, 125-157. Amsterdam: John Benja-
mins.
Fillmore, Charles H. and Beryl T. Atkins
1992 Toward a frame-based lexicon: The semantics of RISK and its neigh-
bours. 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 Semanti-
ca 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 Univer-
sity Press.
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 Under-
standing. St Lucia: University of Queensland Press.
1995 Desire in language and thought. PhD thesis, Australian National Uni-
versity.
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
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 (.), 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
Table of the Centre for Applied Language and Literacy Research,

Edith Cowan University, Claremont, Western Australia, 1 July.
Wierzbicka, Anna
1972 Semantic Primitives. Frankfurt: Athenum.
1996a Japanese cultural scripts: Cultural psychology and "cultural gram-
mar". Ethos 24(3): 527-555.
1996 b Semantics, Primes and Universals. Oxford: 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.
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 Gram-
mar. 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-sym-
bolic words, which form a conspicuous group in the Japanese lexicon
(cf. Shibatani 1990: 153). Japanese onomatopoeic words are generally re-
ferred 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 mi-
metic words than sound-onomatopoeic words (cf. Tamamura 1982).2 It has
also been noted that Japanese is rich in psychomimes which describe vari-
ous emotion/sensation states.
While non-native speakers of the Japanese language will certainly face
difficulties in acquiring psychomimetic words, they need to master the cor-
rect use of them if they wish to communicate effectively with Japanese
people. Psychomimetic words play a vital role in Japanese everyday lan-
guage life, and thus they are an indispensable key for outsiders to under-
stand 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 in-
tangible 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 Japanese-
English 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) HaQI*gyoQ to shita shunkan meian ga ukanda.

good idea occurred to me when I felt haQ/*gyoQ.'
b) *HaQ/gyoQ to shite sono ba ni tachisukunda.
'Feeling *haQ/gyoQ, I stood there paralysed.'
(2) a) Uso tsuita no ga baresoo na n de. hiya-hiyal?biku-biku na
n da.
feel hiya-hiy al?biku-biku since it seems my lie will be
found out.'
b) Uso tsuita no ga bareta node. * hiya-hiy a/biku-biku na n
da.
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 se-
lected psychomimetic words, so that learners can actually use them cor-
rectly. Space permits the examination of only seven psychomimes here.
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 concep-
tualising emotions. The semantic metalanguage developed by Wierzbicka
and colleagues is employed as a tool to define these words. This metalan-
guage 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 compo-
nents 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 inter-
views), 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 JapaneseEnglish
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 ab-
breviations are given in the list of sources at the end of this paper. Unrefer-
enced 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. Fur-
thermore, the differences between emotions, for example, between hiya-
hiya 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 Jap-
anese 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 as-
pects of a word, they do not usually succeed in doing this. Many Japanese-
English 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 onomato-

poeia dictionary] (Mito and Kakehi 1981) gives "feel nervous, be afraid"
for the meaning of hara-hara. However, 'feel nervous' or 'be afraid' can-
not 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 sup-
posed 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 dis-
junction 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 sem-
antic situation in which it is used. In fact, most Japanese-English diction-
aries 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 pro-
vides a shortcut to a purely intuitive knowledge acquired by lengthy 'im-
mersion in a language', a clear and general explanatory formula may pro-
vide a shortcut to a semi-intuitive knowledge acquired through lengthy
study of examples, supported by time-consuming exercises".
Neither vague analogous words from another language, nor lengthy ex-
planations, 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 lan-
guage. These words are "maximally clear, maximally simple, and maxi-
mally universal" in the sense of having absolute equivalents in all the lan-
guages of the world, and thus are intuitively understandable (non-techni-
cal). 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 psycho-
mimes by using this metalanguage. In my explication of Japanese psycho-
mimes, 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 (specifi-
able) 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 dis-
cussed in section 2. They are gyoQ, haQ, dokiQ for something like 'sur-
prise' feelings. Although there is a standard verb odoroku or a quasi-sound-
symbolic 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 antici-
pated bad happening.
Hara-hara, hiya-hiya, biku-biku, and odo-odo, are often translated or
glossed as "nervous" in Japanese-English dictionaries. According to Mori-
ta, 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, hiya-
hiya, 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 mean-
ings, in fact they do mean different things and function differently in Jap-
anese. When they are translated into the universal semantic metalanguage,
the similarities and the differences between those concepts become appar-
ent, and the meaning of each psychomime becomes comprehensible to out-
siders.
2. Momentary emotions
HaQ, gyoQ, dokiQ are psychomimes denoting something like the feeling of
'surprise' about an unexpected happening. The quality of immediate reac-
tion 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 kiite haQ to shitari... 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 dic-
tionary 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
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 intu-
itively understandable words.
For the discussion of each psychomime, examples in context are given
first, then my proposed definition, followed by amplification or justifi-
cation of particular elements in the definition.
2.1. HQ
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 iida-
su 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 katasu-
mi o kasumete toon, haQ to omoimashita. Moshimo ano Dosuto-
shi 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 omoida-

shita. 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, Kupu-
kupu 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 shito-
shito 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: feel haQ. hearing the music of old times.'
(10) Densha orita totan, kasa ya kaimonobukuro o shanai ni okiwa-
sureta 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 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. pull my-
self 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.'
(13) Tomodachi ga hootai de gurugurumaki ni shite, doa no tokoro ni

arawareta no de, watashi wa haQ to shita. [PG]
: was startled for a moment [felt haQ] when my friend appear-
ed 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: 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) X feels something for a very short time
(b) because X thinks something
(c) sometimes a person thinks something like this:
(d) "something happened now
(e) I didn't think before that this would happen
(f) I know something now because of this"
(g) because of this, this person feels something for a very short time
(h) X feels something like this
(i) because X thinks something like this
According to Morita (1988: 83), the Japanese word odoroku, a non-
sound-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
This feeling occurs suddenly when one reacts to some happening im-
mediately. There is an unexpected feeling in haQ, although its degree is not
as strong as that of gyoQ ( didn't know before that something like this
226 Rie Hasada
could happen') and dokiQ ( didn't think before that this could happen').
The trigger of this haQ feeling can be anything. It does not have to be re-
markable, 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 sen-
tence (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 ordi-
nary 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.

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 sen-
tence (7) and (10); we recognise something anew, such as someone's
beauty as seen in sentence (3) and (14). Therefore, haQ is often used to-
gether 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
torinaosu 'pull oneself together' in (11).
There are a lot of sound-symbolic words which refer to the state of ab-
sent-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.
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 some-
thing 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 mitsumekaketa ga, ude wa monosugoi atsu-
sa datta. [T: 159]
RH: 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. : 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?
(18) Sukoshi zutsu, jigazoo no seisaku ni torikakatte mimashita. Jibun
demo gyoQ to shita hodo. insan na e ga dekiagarimashita. Shika-
shi, 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 my-
self feel gyoQ]. Here was the true self I had so desperately hidden.'
[NLH: 55]
(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]

: 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 com-
mit 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 tashika-
meyoo to, gurutto ushiro no hoo 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 for-
got 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
(d) "something happened now
(e) I know: this is bad
(f) I did not know before that something like this could happen
(g) I can't think now
(h) I can't do anything now because of this"
(i) because of this, this person feels something bad for a very short
time
(j) X feels something like this

(k) 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 some-
thing 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...
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 discover-
ing something "bad".
The psychomime gyoQ came to be used much more widely and com-
monly 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 Ko-
rean 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 occur-

rence 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 inter-
jection 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 de-
scribed 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 osshatta. Watashi wa dokiQ to shite...
[Dazai "Shayoo (setting sun)" in JEOMTD]
J-E: 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: feel dokiQ to see the demand for payment of my credit
card.'
(27) Tetsuo ga itta. "Chigau, nigeru na'.' dokiri6 to shite kare o mit-
sumeruto... [KY: 158]
RH: Tetsuo said, "No, don't run away (from your trouble)." I felt
dokiri. then I gazed at him.'
(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 mit-
suketa 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
(c) sometimes a person thinks something like:
(d) "something happened to me now
(e) I didn't think before now that this could happen
(f) something happened to my heart now because of this
(g) I could hear something like this: dokiQ"
(h) because of this, this person feels something for a very short time
(i) X feels something like this
(j) 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 at-
tached to the trigger of this emotion. Sometimes the trigger could be some-
thing 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.
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 some-
one'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-
tence (29). Other parts of the body would not feel dokiQ: *atama/hara/ka-
rada ga dokiQ to sum '*My mind/belly/body feels dokiQ'.
The degree of unexpectedness of dokiQ is strong enough to make the ex-
periencer think he/she had a strong heart beat. The intensity of this unex-
pectedness is stronger than that ofhaQ, where no particular physical mani-
festation 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 doki-
doki 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 doki-
doki.
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 sen-
tence (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 de-
rived 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: hara-

hara, hiya-hiya, biku-biku, odo-odo.
As usual, existing bilingual dictionaries merely list possibly related Eng-
lish 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 experi-
encer'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 ner-
vous [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 audi-
ence 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]
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 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 kur-
ashikata 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 katta hito wa, ichinichi ichinichi no kabu no neugoki ni
hara-hara shiteiru soo da. [JF: 60]
RH: hear that the people who bought the stocks are feeling hara-
hara about the movement of the price of stock day by day.'
Meaning of hara-hara
(a) X feels something
(d) "I see: something is happening to someone/something (Y)
now
(e) something bad could happen to in a very short time because
of this
(f) I don't want this to happen
(g) because of this, I want to do something
(h) I don't know what I can do"
(i) because of this, this person feels something bad
(j) this person thinks:
(k) "I will see (something/someone) is falling down [hara-hara]
into something bad"
236 Rie Hasada
(1) this person feels this for some time

(m) X feels like this
I see: something is happening to someone/something (Y) now
something bad could happen to 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 some-
thing, as seen in the following sentences.
(42) *Takai tokoro ni noboru to ketsui shita hito mite hara-hara sum.
*'I feel hara-hara seeing a person who decided to climb the high
mountain.'
(43) *Takai tokoro ni nobotta hito mite hara-hara sum.
*'I feel hara-hara seeing a person who climbed the high moun-
tain.
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 (ne-
macystis 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 housekeep-
ing, 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/ob-
ject (Y) because of something in the on-going process. They feel this poss-
ible danger in the immediate context ("now").
A person who feels hara-hara for does not want something bad to
happen to Y. A person who does not care if something bad happens to
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 Ya-
kuruto so that their favourite team will win.
The experiencer feels like doing something about the possibility of
something bad happening to Y. However, since the experiencer of the hara-
hara 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 ex-
periencer of hara-hara feeling cannot do anything about the on-going situ-
ation, the psychomime hara-hara often collocates with phrases like ki
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 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 Rie Hasada
(45) ... 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
(someone/ something) falling into a dangerous or frightening situation
("I will see is falling down into something bad").
3.2. Hiya-hiya
Examples:
(46) Oya-kyoodai-tachi wa, byoonin kinodokugam yon mo, sono toki

no surudoi sakebigoe o kikasareru no ga yarikirenai no de, hiya-
hiya 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 hiya-
hiya shita yo. [FSB: 58]
B: 'Lying just isn't worth it, after all. I was scared stiff [felt hiva-
hiyq] that I'd get caught any minute.'
(48) Hahaoya no saifu kam ichiman'ensatsu 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
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]
: 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: was feeling hiya-hiya inwardly, thinking it might be painful
for you.'
Meaning of hiya-hiya
(c) sometimes a person thinks something like:
(d) "something bad can happen
(e) I don't want this
(f) because of this, I want to do something
(g) I don't know what I can do"
(h) because of this, this person feels something bad
(i) this person thinks:
(j) "if I were in a cold place, I would feel like this [hiya-hiya]"
(k) this person feels this for some time
(1) X feels something like this
(m) 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 ob-
serving 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 experi-
encer. 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 happen-
ing to oneself (e.g. one's mother's scolding or accusation) after the reveal-
ing 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 happen-
ing to them to be realised.
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 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 hiya-

hiya 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 suw-
atta. 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 -> : " wondered if I was going to be scolded severely and I was
nervous 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. [Mishi-
ma "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 hairima-
shita 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 qui-
vered.' [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 yosoo-shite, byooin e iku no wa, mattaku biku-
biku-mono datta. [GGJ]
RH: was really in a state of feeling biku-biku when I went to the
hospital, anticipating a bad result of the examination of my stom-
ach.'
(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 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 prob-
ably come down on him like a ton of bricks.'
(66) 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.'
(67) 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
(d) "something bad will happen to me
(f) because of this, I want to do something
(g) I cannot do anything"
(h) because of this, this person feels something very bad
(i) this person thinks:
(j) "I am trembling/shaking [biku-biku]"
(k) this person feels this for some time
(1) X feels something like this
(m) 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 sen-
tence (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 feel-
ing 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 experi-
encer. 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 tsuita no ga baresoo na n de. ?biku-bikulhiya-hiya na n da.

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, hiya-
hiya 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 sen-
tence (60), the speaker's conversation partner feels biku-biku while being
close to the speaker who usually does something bad after drinking. In sen-
tence (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 situ-
ation 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.
The experiencer does not want this to happen.
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 situ-
ation 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
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 shak-
ing movement. People do not necessarily actually tremble when they ex-
perience the emotional biku-biku feeling, like biku-biku in the phrase nai-
shin biku-biku shinagara 'feeling an inward tremor' [the inward biku-biku}
of sentence (56): this trembling is metaphorical, and not a physical move-
ment. 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, 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 per-
son 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 shin-

nyuusei 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: 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 it-
sumo inochigake de, odo-odo shite aruite iru. [PG]
: want to go to a place where there are no cars. We pedestrians
are always nervous [feeling ofto-oifo]. risking our lives when walk-
ing.'
(72) Hitomae ni deru to odo-odo suru. [cf. KNJED]
: 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 odo-
odo 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 yo-
wasa ga am.
J-*E: 'But his eyes, in spite of this well-built body had an ex-
pression 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 odo-
odo}. and wasn't in a state to answer any questions.'
(78) 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
(d) "something bad can happen to me at any time
(e) because I can't do much
(f) I don't want this
(g) because of this, I want to do something
(h) I don't know what I can do"
(i) because of this, this person feels something bad
(j) this person feels this for some time
(k) X feels something like this
(1) 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 pre-
suppose 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, anticipat-
ing 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 sen-
tence (74), the girl is feeling odo-odo after failing the examinations, since
she feels herself incapable and without confidence, and imagines some-
body might do something bad (sneering, backbiting, etc.) to her. In sen-
tence (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
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.
The resulting feeling is "something bad".
4. Conclusion
The major goal of this paper was to explicate the meaning of selected psy-
chomimes to an extent that can give non-native speakers a greater under-
standing 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 anal-
ogous words from other languages, nor lengthy explanations, nor accumu-
lations 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 se-
lected psychomimes can be clearly shown and made comprehensible to
outsiders, when they are translated into the Natural Semantic Metalan-
guage based on universal human concepts and represented in the frame-
work 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 expli-
cation with a language-independent semantic metalanguage. The extent of
similarities and dissimilarities in labelling and the conceptualisation en-
coded 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.
(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 lexi-
cal group in Japanese, and are continually acquired and used throughout life
by Japanese speakers. Understanding the correct use of Japanese sound-
symbolic 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 Jap-
anese 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 pre-
pared with the financial assistance of the Toyota Foundation.
2. The direction of transfer is always from phonomimes to mimetic words (phenom-
imes/psychomimes), and not vice versa (Kakehi 1992: 18).
3. The Romanisation used here follows the Hepburn system. The Hepburn system in-
dicates 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 necess-
ary 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.
Izard, Carroll E.
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 Dai-
gaku bungakubu kiyoo, 1-12.
Kimizuka, Sumako
1967 A Contrastive study of the Japanese Structural Device Onomato-
poeia + 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]
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 So-
ciety 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 Spe-
cificity in Cognition and Culture, 431-452. Cambridge: Cambridge
University Press.
SOURCES CITED - JAPANESE
AS Asahi Shimbun: Shukusatsuban [Asahi Newspaper: compact edition].

Tokyo: Asahi Shimbunsha.
BP Beyond Polite Japanese: A Dictionary of Japanese Slang and Collo-
quialisms. 1993. Yoneyama, Akiko. Tokyo: Koodansha.
252 Rie Hasada
CS Chippsu Sensei Sayoonara [Good-bye, Mr. Chips]. 1956. Hilton,

James. Translated from English by Jyuuzaburoo Kikuchi. Revised ed.;
Tokyo: Shinchoosha, 1966.
FK Fujin Kooron [Women's Public Opinions]. Tokyo: Chuuoo Kooron-
sha.
FKB Funanori Kupu-Kupu no booken [The Adventures of Kupukupu the
Sailor]. 1975. Kita, Mono. Tokyo: Shinchoosha
FSB Flip, Slither, and Bang. 1993. Fukuoka, Hiroko. Tokyo: Koodansha
GG Giongo-Gitaigo Jiten [Onomatopoeia and Mimesis Dictionary]. 1974.
Yasushi Amanuma (ed.). Tokyo: Tokyoo-doo.
GGJ Giongo-Gitaigo Jiten [Onomatopoeia and Mimesis Dictionary]. 1978.
Asano, Tsuruko and Haruhiko Kindaichi (eds.). Tokyo: Koodansha.
GGT Giseigo-Gitaigo Tsukai-kata Jiten [Usage Guide to Japanese Ono-
matopoeias]. 1993. Atooda, Toshiko and Kazuko Hoshino (eds.).
Tokyo: Sootakusha.
HM Hashire, Merosu [Run, Melos]. 1967. Dazai, Osamu. Tokyo: Shin-
choosha. "Kokyoo" [Homecoming]. 207-224.
JEOMTD Wa-Ei Giongo-Gitaigo Honyaku Jiten [Japanese-English Onomato-
poeia and Mimesis Translation Dictionary]. 1984. Fjita, Takashi and
Shinichi Akiho, (eds.). Tokyo: Kinseidoo.
JF Japanese for Foreigners: adverbs. 1987. Kayaho, N. et al Tokyo: Ara-
take Shuppan.
KKA Kagami no Kuni no Arisu. 1959. Carroll, Lewis. (Translated from Eng-
lish by Tadataka Okada.) Tokyo: Kadokawa Shoten.
KNJED Kenkyuusha's New Japanese-English Dictionary (Fourth edition/
27th Impression) 1993. Masuda, Koh (ed.). Tokyo: Kenkyuusha.
KY Kanashii Yokan.[Melancholic Premonition]. 1988. Yoshimoto, Ba-
nana. Tokyo: Kadokawa.
NS Ningenshikkaku [No Longer Human]. 1952. Dazai, Osamu. Tokyo:
Shinchoosha.
PG Nichi-Ei Giongo-Gitaigo Katsuyoo Jiten [A Practical Guide to Jap-
anese-English Onomatopoeia and Mimesis]. 1989. Ono, Shuuichi
(ed.). Tokyo: Hokuseidoo.
T Tsugumi. 1989. Yoshimoto, Banana. Tokyo: Chuuoo Kooronsha.
TB THINK BIG! 1988. Terasawa, Yoshio. Tokyo: Shufunotomosha.
SOURCES CITED - ENGLISH

AKS The Adventures of Kupukupu the Sailor. 1985. Kita, Morio. (Trans-
lated from Japanese edition by Ralph F. McCarthy.) Tokyo: Koodan-
sha International.
DO 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.
FC Floating Clouds. 1965. Hayashi, Fumiko. (Translated from the Japan-
ese edition by Y. Koitabashi. and M. C. Collcutt.) Tokyo: Harashoboo.
GBC Good-Bye, Mr. Chips. 1937. Hilton, James. 8th ed. London: St. Paul's
House.
NLH No Longer Human. 1958. Dazai, Osamu (Translated from Japanese by
Donald Keene.) New York: New Directions Publishing Corporation.
RM Run, Melos!: and other stories. 1988. Dazai, Osamu. (Translated from
Japanese by Ralph F. McCarthy.) Tokyo: Koodansha International,
Ltd. "School Girl". 43-101.
TLG "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 re-
membered my teacher's story and suppressed my anger. (Li Lu Moving the moun-
tain, 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 elabor-
ations, 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 psy-
chologists 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 examin-
ation of the Chinese lexical data might contribute to this debate.
2. Nu
Tian zi zhi nu, fu s hi
Heaven son of nu lie corpse
bai wan, liu xue qian li
hundred 10 thousand flow 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 be-
comes 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. Al-
though 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 Classi-
cal Chinese writings and - even though nu is certainly not the most com-
mon 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 Char-
acters, 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 mani-
festations, symptoms and effects of nu. Before proceeding with the expli-
cation 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).
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 mian
nu hold full/whole face
'The face is full of nu:
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 qie chi
nu ASSOC grit tooth gnash tooth
'gnash one's teeth in nu'
258 Pawel Kornacki
(7) nu ya tu zui
nu tooth protrude mouth
'showing teeth in nu'
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 vio-
lence' - 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 lei
nu ASSOC roar roar like thunder
'to roar like thunder with nu'
(13) nu de da fa lei ting
nu ASSOC big send thunder- clap
'to thunder with nu'
However, nu can also be communicated with the eyes only, and the si-
lence which ensues then, heightens the tension of the whole situation:
(14) 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 fol-

lowing 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
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
violent wind is howling with nu.' (?anger/rage)
(20) ji feng nu huo
wild wind nu 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 hy-
pothesis that the meaning of nu contains a component referring to the im-
possibility 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 other-
wise "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 men-
tioned 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 can-
not 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 con-
nected 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-), 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".
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 be-
come m/')> whereas the following sentence, which would be a literal trans-
lation of the English one, appears to be unacceptable:
(24) *Ta yin na-jian shi er fa -zhe nu
he because that-CL thing CONJ send -ASP nu
'*He is getting nu about it.'
Thus, nu is also a very sudden feeling - a conjecture which may be sup-
ported 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 Pawel Kornacki
(26) 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 huo
nu ASSOC two eye send fire
'so nu that one's eyes send 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 medi-

cine (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 com-
ments 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:
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 re-
sistance 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 re-
vealed by others, one's rights infringed, observing certain striking physical charac-
teristics of someone whom we habitually cannot stand (taoyan) (for example, es-
pecially 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 swal-
lowing, 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 nu

person fatigue/stressed therefore much nu
'Much nu results from fatigue.'
Gathering the above clues together, and in order to proceed with further
discussion, I would propose the following explication:
nu
(b) X thinks something like this:
(c) something happened
(d) this is bad
(f) because of this, I want to do something very bad to someone
now
(g) I can't not do this
(h) X feels something very bad because of this
(i) because of this, X can't think now
(j) something is happening inside X because of this
(k) something bad can happen to X because of this
Arguably, the prototypical case of 'anger' evokes the idea of "bad ac-
tion" (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 incompat-

ible 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 recon-
sider 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 some-
thing hateful), or a protracted period of toil.
Nu can be a powerful and uncontrollable emotion. The metaphorical ref-
erences 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
directed at a person - rather than at an object - as it is spelled out in com-

ponent (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 char-
acter (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 Chi-

nese medicine and philosophy (cf. Porkert 1976: 167-8). As far as its con-
temporary colloquial usage is concerned, Han-Ying Cidian (The Chinese-
English 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/en-
raged, (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 Pawel Kornacki
(31) 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, ex-
claims:
(32) Erzi! Ta qi de hun shen de

son he qi ASSOC whole body ASSOC
xue yi re, erzi gan shenme, ni!
blood one hot 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.'
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
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 con-

trol. According to Ots (1990:40), 'When a Chinese says wo sheng qi le (lit-
erally: 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 dif-
ferent 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 "physio-
logical" 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 re-
sentment', 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 'stor-
age', 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 sug-
gest 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. '<?i'-die'), 'being half-dead
with qi' (qi de ban si - '^'-Assoc-half-dead'), or 'being so qi that one is al-
most dead' (yi qi yi-ge si, lit. One qi one-CL die'), as in:
(40) Jin Qi de hing, benlai you qi se, chu-le

Jin Qi ASSOC illness before have rise health happen-ASP
zhe yi-jian shi, ta yi qi yi-ge si
this one-CL thing he one qi one-CL die
'Jin Qi has already begun to feel better, but when this thing hap-
pened, he was almost dead with qi.'
Being shenglqi suggests bodily tension, which could be interpreted as a
sign of the internal conflict between the will to strike out and the thought of
restraint, so that one speaks of:
(41) qi de zhi duosuo

qi ASSOC keep tremble
'trembling with qi'
And the singular aggressive impulse might be sought in:
(42) qi de zhi duo jiao
qi ASSOC keep stamp foot
'to stamp one's feet with qi'
Similarly to nu, qi is also seen in Chinese as harmful to the experiencer,
as expressed in the following folk comment:
(43) qi da shang s hen, shu da s hang gen

qi big harm body tree big harm root
'Big qi harms the body/health, big tree crushes the roots.' (Chu
1970: 168)
and in another one, which also refers to the related concept of nao (cf. the
next section):
(44) bu yao qi, bu yao nao, qi
not want qi not want nao qi
qi, nao nao ren kuai lao
qi nao nao person quickly old
'It's no use being qi, it's no use being nao, people who are qi and
nao will grow old quickly.' (Chu 1970: 168)
I would propose the following explication for sheng/qi:
sheng/qi
(d) I don't want this
(e) I want to do something bad because of this
(f) X feels something bad for a long time because of this
(g) something is happening inside X because of this
(h) something bad can happen to X because of this
270 Pawel Kornacki
The prototypical situation which makes one shenglqi involves disagree-

ment ("diswant") when things do not happen the way one wants them to
happen - a long queue separates one from the ticket office in the railway
station, someone refuses to move over in a crowded bus, a pupil gives a stu-
pid answer, or parents refuse to buy a toy. These are not necessarily "big"
(important), or inherently "good/bad" (moral) things - one could hardly be
shenglqi because of, e.g., a war breaking out, or a friend's terminal illness
(although one possibly could speak of nu in such contexts). The emphasis
seems to be not so much on the condemning judgement "this is bad" (as
was the case with nu), as, more simply, on the assumption that these are the
things that no one would normally want to happen. An alternative to the
component (c) might, however, be a component like "someone did some-
thing" - the problematic cases under such an interpretation would be the
situations like queuing, or bike crashes which, strictly speaking, don't have
to involve any very specific agency, or intentions, but are rather seen as
eminently undesirable events. An even stronger case for the possible irrel-
evance of the human agency to this concept might be found in that one can
be shenglqi as a result of, for instance, unexpected rain which spoils one's
plans to go for a trip.
In contrast with nu, shenglqi involves only a fairly general wish to "do
something bad" (cf. component e). References to overtly aggressive be-
haviour are minimal, and there seems to be more stress on physical discom-
fort and tension (suppression, "hot blood", trembling), than on specific
"active" retaliation. It might also be the case that what is involved in this
concept is a wish to make someone "feel something bad" on account of
what happened (which comes to the foreground especially in the sentences
in the durative aspect). Thus, a person who is being shenglqi for several
days, might, for instance, try to refuse to talk, see, or share things with the
people around, or, if the qi is being suppressed, he or she might certainly
cherish such thoughts. A relevant expression here is also chulqi (lit. 'put
out', or 'exit' qi), which to some extent suggests the "externalisation" of
one's "bad feelings". Interestingly, however, this phrase can be used in the
collocation na moumou shilren chulqi ('take-certain-thing/person-put out-
<7/'), which often suggests that one's qi was vented not on the actual wrong-
doer, but rather that it was displaced. Thus, e.g., in one informant's
example, after a marital quarrel, the husband can na haizi chulqi ('take-
child-put out-^f), by shouting at, or hitting him/her. It seems, therefore,
appropriate to suggest a component like (e), which simply spells out the
unquestionable will to do something bad but does not prejudge the likeli-
hood of "venting" one's qi (since both restraint and retaliation are possible,
as well as lexicalised, although there seems to be more evidence in favour
of 'restraint', such as, for example, the fact that one speaks of "dying"
rather than "killing" with/from qi).
As suggested before, there seem to be fairly good reasons to think of
shenglqi as an emotion which may extend over an indefinite period of time.
It is also potentially harmful to the experiencer, but again, it seems, in a dif-
ferent way from nu. Shenglqi can be rather seen as something lingering
('smouldering' rather than 'exploding') and hidden ('kept in the stomach'),
something that perhaps happens as much TO a person, as it does INSIDE him
or her.
4. Nao (huo)
Another colloquial word often perceived as closely related to shenglqi is

nao (which in the spoken language is usually expanded to the two-syllabic
naohuo, lit. '/zao-fire'). Shaver et al. (1992:192) translate it as 'annoyance/
irritation', Bucher (1986: 423) glosses it as 'angry, unhappy', whereas the
standard Han-Ying Cidian (The Chinese-English Dictionary, 1988: 491)
provides two separate meanings: (1) 'angry, irritated, annoyed', and
(2) 'vexed, worried'. In one informant's succinct description, naohuo
refers to the situation when:
(45) yudao bu ruyi de shi,

come.across not as wished ASSOC thing
renshou bu liao er shenglqi
endure not able CONJ 'angry'
One comes across something one doesn't want, something that
can't be endured any longer, and because of that one becomes
shenglqi'
Consider the following example taken from a story set among music stu-
dents (Liu 1987: 215):
(46) Na-ci gangqin kaoshi ta you de-le si-fen

that-time piano exam he again get-ASP four-point
272 Paw et Kornacki
dagai you shi yinwei bei- bu-xia Baha

perhaps again be because memorise not-down Bach
Ta da wei naohuo, wen Li Ming wei
he big for naohuo ask Li Ming for
shenme ta de-le si-fen er Li Ming bu
what he get-ASP four-point CONJ Li Ming not
chang lian gang que neng de wu-fen
often practice piano but can get five-point
'He scored four points again at the piano exam - probably because
he didn't memorise Bach. He got very naohuo, and asked Li Ming
why he got four, whereas Li Ming, who didn't practice much, got
five.'
The hero of this passage had been portrayed before as a very ambitious
and gifted, if perhaps a bit overzealous student. Although shenglqi would
be quite acceptable in this context, naohuo focuses here exactly on the 'un-
bearability' of the whole case. This situation is not only different from the
one which the student had hoped for, but also, from his perspective, un-
wanted and strangely inexplicable. The idea of something one wants not
happening is also present in the following example, taken from an inform-
ant's story about how his attempts to secure a recommendation from his
work unit, which would enable him to take up an extra job, were almost
frustrated:
(47) Wo yudao zheme-ge daomei lingdao
I meet SO-CL unlucky superior
xin-li zhen you dian naohuo
heart-in really have bit naohuo
'It was bad luck that I met this superior, I felt really a bit naohuo.'
Meeting an "unlucky" (i.e., for the speaker) superior was something he
certainly wished didn't happen - as when in the process of application he
met someone who only saw problems with his idea and was unwilling to
agree.
Encountering an obstacle which repeatedly frustrates one's efforts brings
forth the feeling of naohuo. The following example is also taken from Liu's
(1987: 214) story, in which another student, who tries to write music in a
fairly idiosyncratic style, is in many small but palpable ways disregarded
by some people in the school, so that:
(48) Shi Bai dui zhe-xie ren yu Jia jiaoshou

Shi Bai towards these person and Jia professor
wuxing de duikang you qi you nao
invisible ASSOC opposition both qi both nao
'Shi Bai was both qi and nao with the invisible opposition of pro-
fessor Jia and other people.'
I would suggest the following explication of this concept:
naohuo (X wei mou jian shi er naohuo - 'X is naohuo because of

something')
(c) something bad happened to me
(f) something bad can happen to X because of this
Very much like shenglqi, naohuo is provoked by subjectively diswanted

events, which accounts for the wide-spread mutual exchangeability of the
two. Yet naohuo seems to be focused on a slightly different mental attitude
of the experiencer, which, in addition to the disagreement, involves a bad
feeling caused by the subjectively bad situation. Such an interpretation ac-
counts also for the following example, quoted in Han-Ying Cidian (1988:
491):
(49) Duiyu ta na-zhong ting-bu-jin piping

towards he that-kind listen-not-enter criticism
de taidu, women gandao naohuo
ASSOC attitude we feel naohuo
'We are naohuo at his not listening to criticism.'
Someone refuses to pay attention to the criticism. The words simply 'do
not enter' into his head - although we would like this person to behave in
one way, he keeps doing the opposite. This aspect of the meaning of nao-
huo probably accounts for why this word can sometimes, as Bucher (1986:
274 Pawet Kornacki
423) suggests, be translated as 'unhappy', although 'frustrated' might be

perhaps more fitting.
Like nu and shenglqi, naohuo is also a "harmful" emotion. The follow-
ing saying, interestingly, suggests laughing (as opposed to, say, some kind
of retaliation) as an antidote to naohuo, which again suggests a relatively
"experiencer-oriented" perspective on this feeling:
(50) nao yi nao, lao yi lao

nao one nao old one old
xiao yi xiao shao yi shao
laugh one laugh young one young
'Nao makes you older, and laughter younger.' (Chu 1970: 169)
5. Fen
The next Chinese concept to be analysed in this paper is fen. Its various
(morphemic) collocations point, first of all, to the strongly moral, evalu-
ative, and, in a sense, "public" quality of this emotion, which seems to be
absent from, or relatively downplayed in, the notions of nu, sheng/qi, or
nao(huo).
Thus, for instance, one can speak of such politically significant senti-
ments as gongfen (lit. 'public-/?'), which Metzger (1981: 19) describes as
"the feelings of 'righteous anger' aroused by authority figures perceived as
morally deficient", yifen (lit. 'righteousness-/?') glossed inHan-Ying Ci-
dian (1988: 821) as 'righteous/moral indignation', and exemplified in the
expression geming yifen (lit. 'revolutionary yifen', i.e., 'revolutionary in-
dignation'). According to one native speaker, people labelled as "traitors"
(maiguozei) or "spies" (tewu) were during the Cultural Revolution often
singled out and persecuted in the name ofminfen (lit. 'people-/?'), and re-
ferred to as 'arousing' (jiqi) or 'facing' (mianzhong) minfen.
Fen is typically stirred in the situations which can be summed up with the
phrase bu gongping (roughly, 'not fair/just'). This association is lexicalised
in another collocation -fenfen bu ping (lit. '/<?/<? not just'), which could
describe the feelings aroused in a situation in which, for example, a crimi-
nal escapes or somehow evades the proper punishment, or a student is un-
deservedly (in the eyes of others) praised or criticised by a teacher.
Wang (1993: 74), in his quite revealing paper on different ways of sub-
duing "bad feelings", speaks about/*? associating it mainly with two kinds
of circumstances, induced by specific 'ways of thinking' (sixiang moshi).
Firstly, fen is associated with what Wang calls "Yinggai" huo "Bu ying-
gai" sixiang ('"Should" or "Should not" thinking'), which is implied in
situations where something that we think 'should' (yinggai) happen in fact
doesn't happen, and/or something that we think 'should not' (bu yinggai)
happen does happen. Wang gives here the following example: "When the
girlfriend with whom you were together for several years breaks off with
you, you can become fenfenran (lit. '/<?/z/<?w-like'): 'This should not be so!
(Zhe tai bu yinggai le - this-too-not-should-) There is 'good feeling'
(ganging) between us, why should we split?'" According to Wang,
thoughts of revenge or of getting even with the partner may be the outcome
of this situation.
The second type of context relevant to fen is called by Wang "Bu gongp-
ing" sixiang (' "Not fair/just" thinking'), where the bad feelings are gener-
ated because a person has evaluated a certain situation as bu gongping. In-
terestingly, Wang provides here not only an example of a relevant situation,
but also the hypothesised thoughts of a person involved in it. The key word
in this passage is the bi-morphemic qifen, the most common and colloquial
of the/? compounds:
For example, one day you go window-shopping, and find something which could be
a beautiful gift for your wife. You immediately want to buy it and give it to her, but
unfortunately you run into an 'abominable' (taidu elie) shop-assistant who not only
'neglects you' (daiman ni) but also shouts abuse at you. Obviously, in such a situ-
ation there appear a lot of thoughts in your mind which can make you extremely qifen
(feichang qifen), so that you wish to hit that person, and when you go back home you
can hardly calm yourself.
Subsequently, as a way of resolving such feelings, Wang advises to write

down ones 'agitated' (//) thoughts - the ones which he finds plausible in
such a situation, are as follows, given below in a word-for-word translation
(with Wang's original numbers):
(I) Na-ge hundan tai met you jiaoyang le

that-CL stupid.egg too not have culture PRT
jianzhi mei you yi dian qiye daode
simply not have a bit profession morality
276 Pawel Kornacki
'That bastard has no "culture" at all, not even a bit of professional

morale.'
(II) Shangchang jingli yingga kaichu zhe-ge jiahuo
shop manager should drive.expel this-CL fellow
'The shop-manager should throw him out.'
(III) Pohuai-le wo mai liwu song ren de
ruin-Asp I buy gift send person ASSOC
hao xinqing, na jiahuo gai si
good feeling that fellow should die
'He spoiled my good "intention/feeling" of sending someone a
gift, that rascal should die.'
(IV) Wo shizai ren-bu-liao zhe-kou qi
I really bear-not-able this-CL insult
can't possibly stand such treatment.'
As a first approximation to the explication of (qi)fen one might observe

that it involves not so much a "disagreement with some bad state of affairs"
(as nu), or a certain 'frustration' ("things do not happen the way I want
them to happen", as was the case with shenglqi and nao(huo)), as the feel-
ing of being wronged, and perhaps somehow 'hurt'. The two above
examples ("dejected boyfriend" and "slighted customer") might evoke a
fairly specific cause of qifen - such as, for instance, that "someone did
something bad to me". This, however, although certainly possible, is too
specific. One can become qifen not only as a result of a direct personal ac-
tion, but also, for example, upon hearing a disturbing piece of news, or in-
formation. Thus, for example, one Chinese native speaker mentioned be-
coming qifen upon hearing about a recent pronouncement of a Japanese
politician who, according to the informant, tried to falsify and diminish
Japan's involvement in the Chinese-Japanese war. In this case it probably
wouldn't be appropriate to speak about the event directly in terms of
"someone doing something bad to someone else" - a more plausible hy-
pothesis could be rather formulated along the lines of "someone did some-
thing bad, when someone does something like this people feel something
bad because of this".
fen
(c) some people did something bad
(d) it is very bad if people do things like this
(e) when someone does something like this people feel some-
thing bad because of this
(f) I would want to do something because of this if could
(g) X feels something bad because of this
(h) X thinks something bad because of this
(i) sometimes it is good if people feel like this
The basic stimulus to fen is the morally reprehensible ("bad") human ac-
tion. It is, however, not necessarily the case that the agent provoking this
emotion must be a "singular" person ("someONE") - in fact, the "social"
sentiments such as gongfen ('public-/?'), or minfen ('people-/?'), could
easily be provoked by the behaviour of some groups of people, labelled,
e.g., as 'bureaucrats' (guanliao), 'bad elements' (huaifenzi), etc., hence
the reference to "some people" in component (c).
Components (d) and (e) represent the negative moral and social judge-
ments attached to fen. The behaviour causing/? breaks expected social
standards ("it is very bad if people do things like this"), and consequently
hurts the interests and feelings of other people (component e). From the
perspective offen, people like an 'abominable' shop-assistant, 'traitors', an
inconsiderate lover, or a lying politician appear, first of all, to share a dis-
regard for others.
Arguably, the feelings and thoughts induced by such offending attitudes
are not inherently accompanied by the overt aggression, or a certain vio-
lence peculiar to, e.g., nu. The unfortunate lover, according to Wang
(1993), can 'hate/resent' (yuanhen) his partner, or think of revenge
(baofu). The customer thinks (but - at least in Wang's scenario - only
thinks) of hitting the nasty assistant, and his supposed ideas (I-IV) do not
seem to involve anything as specific and immediate as, for instance, "I
want to do something bad to someone/this person now". What is involved
is rather more like a painful mental recycling of the offending event
(points III and IV), the idea that "someone should do something about
this" (point II), and bad thoughts about the moral qualities of the assistant
278 Paw et Kornacki
(point I) - referred to in component (h). A compatible scenario of fen is

also evoked in the following chengyu:
(51) fen shi ji su

fen world hate ways/customs
'to fen the world and hate its ways'
In ancient China, this expression could be used to describe an honest of-

ficial who, for example, upon realising the degree of corruption, or some
other outrageous practice at the court, decides to give up his office and to
seek refuge in a solitary, far-away place, ruminating over the depravity of
the world. If anything, such a feeling tended to be, first of all, self-destruc-
tive, and if serious enough, might have led to a suicide which was under-
stood as an act of moral protest (Prof. Yin Banyan, personal communi-
cation).
It is also interesting to notice the lack of figurative associations
which - as with nu - would focus on the uncontrollable or aggressive as-
pects offen. There are no lexicalised expressions which would speak of
'*being/<?n beyond self control' (cf. nu), or '*protruding teeth m fen'
(cf. nu). Furthermore, fen can't qualify natural phenomena - something
which puts this emotion firmly in the "properly human" context, since the
natural phenomena do not provide any inherent "moral qualification" - the
sea doesn't '*roar with/<?' (cf. nu), and the wind doesn't '*howl with fen1
(cf. nu). Fen, however, can be spoken of as a sort of a burden, something
which, according to another chengyu, can fill one's chest:
(52) Yi fen tian ying

righteous fen fill chest
'righteous fen fills one's chest'
But unlike in the case of nu, the morphemic compounds of fen suggest
that whereas it is probably not a "pleasant", or a "good" feeling it can
sometimes be the "right" or "proper" feeling carrying an attitude of a cer-
tain moral superiority. Thus, there are no expressions like *gongnu ('pub-
C-/IM', cf. gongferi), *nunu buping ('nunu not just', cf.fenfen buping), or,
perhaps most tellingly, *yinu ('righteous-m*', cf. yifen). All these semantic
clues are expressed as the last component (i) of the proposed explication.
An interesting conceptual parallel might perhaps be sought here in the
much-discussed Ifaluk concept of song, extensively studied in Lutz (1988)
and re-analysed from the NSM perspective by Wierzbicka (1992 a: 147-8).

Originally translated by Lutz as 'justifiable anger', song is an emotion
which is not necessarily characterised by direct physical violence, can be
self-destructive (as in an attempted suicide), and is usually felt by people in
a higher/authority position when a wrongdoing has taken place. A possible
difference - at least in the case of expressions like 'public' fen, 'people's'
fen, or 'righteous' fen (cf. Metzger 1981) - might be that it is rather the
people in the authority positions, or an otherwise salient social group, who
would be targeted as an object of fen by the rest of the society.
6. Taoyan
Chinese-English dictionaries usually associate a quite impressive number

of glosses with this concept. Han-Ying Cidian (1988: 669), for example,
which considers taoyan to have three meanings, gives the following:
(1) 'disagreeable, disgusting, repugnant', (2) 'hard to handle, trouble-
some', and (3), as a transitive verb, 'dislike, loathe, be disgusted with'.
Bucher (1986: 440) adds 'to dislike, be sick of, be disgusted with'.
There seems to be remarkably little restriction on the possible range of
things which may cause this feeling. People, behaviour, events, illnesses,
weather, animals, insects, plants and things - all can be qualified as
taoyan taoyan de ren (person), taoyan de xingwei (behaviour), taoyan de
shiqing (matter), taoyan de hing (illness), taoyan de tianqi (weather),
taoyan de dongwu (animal), taoyan de chong (insect), taoyan de zhiwu
(plant), taoyan de dongxi (thing). All of them may also become objects of
one's taoyan. Firstly, taoyan seems to be connected with things that we dis-
approve of, i.e., that we think are bad things (to do):
(53) Wo hen taoyan ta-de fangjian luanqibazao

I very taoyan he-ASSOC room mess
taoyan the mess in his room.'
(54) Wo hen taoyan ta
I very taoyan he
zai nude mianqian chuiniu
there.be female in front of boast
taoyan his boasting before the girls.'
280 Pawel Kornacki
(55) Wo tebie taoyan ta

I especially taoyan he
mei tian dao wan kan dianshi
every day till late watch TV
especially taoyan that he watches TV till late every day.'
So far, taoyan appears to involve the following judgement: "this/some-

thing is bad, I don't want this". Consider now the following fragment of a
conversation taken from Wang Meng's short story (Renmin Wenxue
1987/4: 33). The husband, who has recently returned to his small native
town in China from a scholarship in Canada, keeps comparing China to the
luxuries of Canada. His wife flares up, and reminds him of her poor life
while he was comfortably away:
(56) taoyan\ shuo de na hua jiu taoyan\

taoyan say-ASSOC that words just taoyan
You shi Jianada! Gou le, ni na-ge
again be Canada enough PRT you that-CL
gai si de Jianada
should die ASSOC Canada
What you are saying is just taoyanl Canada again! That's
enough, that damn Canada of yours...'
The wife's response needs little comment. She doesn't want to hear the
story all over again, and she certainly "feels something bad because of it".
A similarly structured reaction can be found in a more "physical" response
of a music student to the new, rather unusual composition of his friend, en-
titled "The Fourth Pattern of Primitive Tension" (Xiaoshuo Lin 1984/6:22):
(57) Ni xie de dongxi dou xiang gou

you write ASSOC thing all like dog
shi! Wo yi beizi mei ting-guo zheme taoyan
shit! I all life not listen-ASP so taoyan
de yinxiang, jianzhi taoyan tou le
ASSOC sounds simply taoyan extremely PRT
'The stuff you wrote is just like stinking dog-shit! I haven't heard
such taoyan sounds in all my life, it's simply taoyan beyond
words.'
Although, as this example shows, metaphorical associations of taoyan

with excrements are possible, one should not, arguably, look for a proto-
type of taoyan in a natural reaction to such things. Similarly, spitting out
something that tastes bad - the prototype of 'disgust' - is probably better
reserved for the expressions zuo'ou ('vomit, nauseating'), or exin ('feel
like vomiting, nauseating'), which apart from describing reactions to the
physical stimuli (smells, tastes) can be also extended to extremely despic-
able human behaviour, so that one informant, for example, spoke only of
someone who is a traitor selling out his friends to the enemy as zuo'ou de
ren ('a nauseating person'). An interesting collocation is the expression 'a
taoyan illness' (taoyan de hing), which means an illness which is hard to
cure, troublesome and 'nasty' - and which probably could be considered as
a more apt "experiential prototype" of taoyan - something bad, diswanted,
towards which we feel something bad. To account for such an interpre-
tation, I would suggest the following explication:
taoyan
(c) this is bad
(d) I feel something bad because of this
(f) X feels something bad because of this
The scope of component (c) is very broad. It must account for quite var-
ied things that can be called taoyan, such as, for instance, people, weather,
or sounds, hence also the simplicity of the judgement "this is bad". Taoyan
things are unwanted things (of which 'nasty' illnesses are a good example).
Obnoxious' behaviour certainly qualifies for being called taoyan. An in-
formant who watched the excessively "sociable" behaviour of a student
who got drunk at a disco said:
(58) Ta ganma na-yang ling ren taoyan

he why on earth that-way make person taoyan
'Why on earth does he make everyone taoyanT
This appears to conform quite well to component (e) of the explication.
Arguably, taoyan implies also a conscious judgement about one's feelings
282 Pawel Kornacki
(expressed in the explication by embedding the "subjective feeling" com-

ponent (d) under the "cognitive" component (b)), so much so that coming
to have this emotion can be presented along "quasi-inferential" lines, as in
another informant's comment about one of her acquaintances' constant vi-
sits:
(59) zuotian lai-le, jintian lai-le, mingtian

he yesterday come-ASP today come-ASP tomorrow
ye yao lai, zhen taoyan
too want come really taoyan
'He came yesterday, he came today, he also wants to come tomor-
row, it's really taoyan'
7. Conclusions
The two Chinese concepts which would merit an especially detailed com-
parison with the English 'angry/anger', are, arguably, nu and shenglqi - the
first one on account of being actually considered a natural "human feeling"
(ren qing) in the traditional Chinese psychology (see Yan Guocai 1982),
the other on account of its social and experiential salience, confirmed by its
early acquisition and widespread use in everyday language.
As discussed in the Sections 2.1-3 of this paper, nu is involved in a
number of expressions which bring to mind concepts like 'rage', 'fury', or
'aggression', rather than simply 'anger'. It is perhaps interesting to note
here that Darwin, while claiming that "anger, indignation - these states of
mind differ from rage only in degree, and there is no marked distinction in
their characteristic signs" (1872: 246), still decided to analyse them in sep-
arate sections. The bodily symptoms of rage which he enumerated, come,
in fact, remarkably close to the lexicalised Chinese 'formulae' (chengyu)
describing nu, so that, for example:
(a) "the face reddens or becomes purple, with the veins on the fore-
head and neck distended" (1872: 240)
- nu de lian hong bozi CM ('so nu that the face is red, and the neck
bulges')
(b) "the teeth are clenched or ground together" (1872: 241)
- nu de yao ya qie chi ('so nu that one grinds one's teeth')
(c) "such gestures as the raising of the arms, with the fists clenched, as
if to strike the offender, are common" (1872: 241)
- nubi ('-raise arms'); nu de wo quan ton zhao ('so nu that one
clenches the fists and shows the fangs')
(d) "the hair sometimes bristles" (1872: 242)
- nu fa chong guan (One is so nu that the hair raises one's cap')
(e) "the lips, however, are much more commonly retracted, the grin-
ning or clenched teeth being thus exposed" (1872: 243)
- nu ya tu zui ('so nu that the teeth protrude from one's mouth')
(f) "the eyes are always bright, or may, as Homer expresses it, glisten
with fire" (1872: 242)
- nu de muguang ru ju ('so nu that the eyes are blazing like the
torches')
(g) "the paralysed lips then refuse to obey the will, and the voice sticks
in the throat" (1872: 242)
- nu bu neng hui ('so nu that one can't reply')
Ekman, who in one of his papers (1993: 386) speaks about "the anger
family" which "would include variations in intensity stretching from an-
noyance to rage", suggested in an earlier article that "in all members of the
anger family the brows are lowered and drawn together, the upper eyelid is
raised and the muscle in the lips is tightened" (cf. Ekman 1992: 172). In-
terestingly, the Chinese expression which comes closest to Ekman's con-
figuration includes a reference to nu - rather than to some other, less "in-
tense" 'anger'-concept (e.g., nao) - so that one speaks of heng mei nu mu
('horizontal eyebrows, nu eyes'). This might suggest that the universal fa-
cial expression of "anger", as envisaged by Ekman, is more naturally
spoken of as applicable only to some members of the "family". It is rel-
evant to note that qi - when compared with the strikingly graphic "descrip-
tions" of nu - shows in the face to a much less specific degree (so that one
generally describes the face as becoming red or white with qi, rather than,
e.g., baring the teeth, or frowning with qi).
Van Brakel (1994: 2) in his critique of the psychological and anthropo-
logical studies on "emotions" lists twelve different definitions of anger put
forward by various scholars. Consider, however, the following statement,
not included in Van Brakel's list, but taken from Izard's Human Emotions
(1977: 329-30) "a very common stimulus to anger is the feeling of being
284 Pawel Kornacki
either physically or psychologically restrained from doing what one in-

tensely desires to do".
When spelled out in an NSM paraphrase, the main line of Izard's idea
might run as follows:
I want to do something
I can't do this
This, however, appears to be quite different from the "stimulus" compo-
nents of nu\
something happened
this is bad
I don't want this
as well as shenglqi:
something happened
I don't want this
"Anger", in Izard's view, would seem to be first of all a response to the
frustration of personal intentions. N u and shenglqi, however, appear to be
focused more on the explicit appraisal of the relevant situations - in some-
what 'moral' terms, as far as nu is concerned ("this is bad"), or as a person-
al rejection in shenglqi ("I don't want this") - and the frustrations of per-
sonal aims, although certainly possible as triggers of these emotions, are
presented from a somewhat different angle. It is "things" that happen, and
it is their "badness", or our "diswant" of them, which cause these feelings;
and what matters here is reactions rather than intentions. This "reactive",
or, as it were, "situation-(rather than goal)-oriented", character of a con-
cept like nu - and even more so shenglqi, is perhaps relevant to a certain
paradox in the contemporary Chinese "emotional style" (cf. Middleton
1989; Bond 1991), as noted in Ots's article (1990: 40):
Chinese culture emphasises inhibition of strong emotional expressions. Little
children are scolded by their mothers for aggressive behaviour. This traditional value
system is in some degree of conflict with the finding of this study that anger was the
only emotion that patients referred to on their own and talked about with little in-
hibition. (...) We do not have sufficient data about the psychosocial situations which
account for day-to-day frustrations and aggressiveness in present-day China to com-
prehend this obvious contradiction.
Summary of explications
nu
(d) this is bad
(f) because of this, I want to do something very bad to someone
now
(g) I can't not do this
(h) X feels something very bad because of this
(i) because of this, X can't think now
(j) something is happening inside X because of this
(k) something bad can happen to X because of this
shenglqi
(e) I want to do something bad because of this
(f) X feels something bad for a long time because of this
(g) something is happening inside X because of this
(h) something bad can happen to X because of this
naohuo
(c) something bad happened to me
(f) something bad can happen to X because of this
fen
286 Pawel Kornacki
(c) some people did something bad

(d) it is very bad if people do things like this
(e) when someone does something like this people feel some-
thing bad because of this
(f) I would want to do something because of this if I could
(g) X feels something bad because of this
(h) X thinks something bad because of this
(i) sometimes it is good if people feel like this
taoyan
(c) this is bad
(d) I feel something bad because of this
(f) X feels something bad because of this
References
Bond, Michael H.
1991 Beyond the Chinese Face: Insights from Psychology. Hong Kong: Ox-
ford University Press.
Bucher, Urs
1986 Vocabulary of Modern Chinese. Felsberg: Tobun.
Chu, Chieh-fan
1970 A study of Chinese proverbs on health and hygiene. Bulletin of the In-
stitute of Ethnology Academia Sinica Taipei, 30(2): 165-237.
Darwin, Charles
1872 The Expression of Emotion in Man and Animals. London: Murray.
(Reprinted 1979, London: Julian Friedmann.)
Ekman, Paul
1993 Facial Expression and Emotion. American Psychologist 48 (4):
384-392.
Goddard, Cliff
1991 Anger in the Western Desert: Semantics, culture and emotion. Man 26:
602-619.
Hanyu Chengyu Fenlei Da Cidian [A Big Dictionary of the Classified chengyu].

1987 Huhehaote: Neimenggu Renmin Chubanshe.
Han-Ying Cidian [A Chinese-English Dictionary].
1988 Beijing: Shangwu Yinshuguan.
Hanyu Changyongzi Zidian. [A Dictionary of Frequently Used Characters].
1977 Beijing: Renmin Chubanshe.
Izard, Carroll
Li, Lu
1990 Moving the Mountain. London: Pan Books.
Lin, Keh-ming
1981 Traditional Chinese medical beliefs and their relevance for mental ill-
ness and psychiatry. In: Arthur Kleinman and Lin Tsung-Yi (eds.),
Normal and Abnormal Behaviour in Chinese Culture, 95-111. Dor-
drecht: T. Reidel.
Lin, Yutang
1972 A Chinese-English Dictionary. Taipei: Gwei Gwan Tu Shu.
Liu, Suola
1987 Tansuo xiaoshuoji [Exploratory short stories - a collection]. Shang-
hai: Wenyi Chubanshe.
Lutz, Catherine
1988 Unnatural Emotions: Everyday sentiments on a Micronesian atoll and
their challenge to Western theory. Chicago: University of Chicago
Press.
Matthews Chinese-English Dictionary
1960 Hong Kong: Oxford University Press.
Metzger, T.
1981 Selfhood and Authority in Chinese Political Culture. In Arthur Klein-
man and Lin Tsung-Yi (eds.), Normal and Abnormal Behaviour in
Chinese Culture, 7-27. Dordrecht: T. Reidel.
Middleton, D. R.
1989 Emotional style: Cultural ordering of emotions. Ethos 17 (2):
187-201.
Morohashi, T.
1960 Dai kan-wajiten [A Big Chinese-Japanese Dictionary]. Tokyo: Tai-
shukan Shoten.
Oatley, Keith
1992 Best laid schemes: the psychology of emotions. Cambridge: Cam-
bridge University Press.
288 Pawei Kornacki
Ots, T
1990 The angry liver, the anxious heart, and the melancholy spleen: The
phenomenology of perceptions in Chinese culture. Culture, Medicine,
and Psychiatry 14(1): 21-58.
Porkert, Manfred
1976 Theoretical Foundations of Chinese Medicine. Cambridge, Mass.:
MIT Press.
Renmin Wenxue [People's Literature] (various issues). Beijing: Zuojia Chubanshe.
Roth, H. D.
1996 The inner cultivation tradition of early Daoism. In Donald Lopez (ed.),
Religions of China in Practice, 123-148. Princeton: Princeton Univer-
sity Press.
Russell, J.
1991 Culture and the categorization of emotions. Psychological Bulletin 10:
426-450.
1994 Is there universal recognition of emotion from facial expression? A re-
view of cross-cultural studies. Psychological Bulletin 115(1):
102-141.
Shaver, P., S. Wu, and J. C. Schwartz
1992 Cross-cultural similarities and differences in emotion and its represen-
tation. In P.Clark (ed.), Emotion, 175-211. London: SAGE Publi-
cations.
Shweder R.
1993 The cultural psychology of emotions. In: Michael Lewis and Jeanette
Haviland (eds.), The Handbook of Emotions, 417-431. New York:
Guilford Publications.
Van Brakel, J.
1994 Emotions: a cross-cultural perspective on forms of life. Prepared for
W. M. Wentworth and J. Ryan (eds), The Handbook of Social Psychol-
ogy, vol. 2.
Wang, Ling
1993 Kongzhifennu de xinlixue fangfa [Psychological methods of control-
ling anger]. Fuyin Baokan Ziliao, Xinlixue, B4(l): 73-76.
Wierzbicka, Anna
1992 a Semantics, Culture, and Cognition: Universal Human Concepts in
1992b Defining emotion concepts. Cognitive Science 16: 539-581.
Xiandai Hanyu Cidian [A Dictionary of Modern Chinese]
1988 Beijing: Shangwu Yinshuguan.
Xiaoshuo Lin [Literary Forest]
1984-87 Ha'erbin: Wenyi Zazhishe (various issues).
Xinbian Xiaoxuesheng Zidian [A New Character Dictionary for Primary School

Students]
1986 Beijing: Renmin Chubanshe.
Yan, Guocai
1982 Woguo guanyu qinggan de ji zhong xueshuo [Some theories concern-
ing emotions in our culture], Xinlixue tongsu 1982(6): 34-38.
You, Zhiren et al.
1986 Xiandai Hanyu tongyici bianxi [Discrimination and Analysis of Syn-
onyms in Modern Chinese]. Ningxia: Renmin Chubanshe.
Zhang, Yaoxiang
1947 Qingxu Xinli [Psychology of emotions]. Beijing: Shangwu Yinshu-
guan.
Human emotions viewed through the Russian
language
Irina B. Levontina, Anna A. Zalizniak
1. Introduction1
Linguistic representation of human emotional life is one of the most im-

portant pieces of the linguistic picture of the world. The notion of a lin-
guistic (or naive) picture of the world comprises two different ideas: that
the picture of the world offered by the language differs from the "scien-
tific" one, and that every language paints its own picture. The majority of
linguistic studies of emotion words in Russian focus on the first idea, i.e. on
universal properties of emotions which may be identified through lin-
guistic analysis of particular words (cf. lordanskaja 1970, 1972, 1984;
Arutjunova 1976: 93-111; Anna Zalizniak 1983, 1992; lordanskaja and
Mel'Cuk 1990; Apresjan and Apresjan 1995; Ju.D. Apresjan 1995; Uryson
1995 et al.). Recent research has seen a shift toward interest in language-
specific features in designating emotions, leading to several works focuss-
ing on "Russian emotions" (Wierzbicka 1990, 1992 a; Pen'kovskij 1991;
V. Apresjan 1995; Apresjan 1997 a, b; Levontina 1997 a, b; Uryson 1997;
cf. also investigations of other specific concepts revealing features of Rus-
sian mentality: Zalizniak and Levontina 1996, 1999; Levontina and Sme-
lev 1996 a, b; Smelev 1997 a, 1997b; Levontina 1999; Arutjunova 1996,
1997, 1999).
"Russian emotions" can be studied in two ways. First, by searching for
"specifically Russian" words (i.e. words comprising conceptual configur-
ations peculiar to the Russian language and missing in other languages:
such words as toska, udal', rodnoj or nadryv). Second, by dealing with
words that refer to universal human categories and can be translated into
other languages, but have some language-specific aspects of meaning
(such words as soast'e, radost', obida, razluka, drug - as opposed to hap-
piness, joy, offence, separation, friend). Our article will analyse words of
both types (leaving them without translation in the first case and indicating
approximate translation equivalents in brackets in the second). We do not
292 Irina B. Levontina, Anna A. Zalizniak
aim to provide a complete description of the world of feelings in Russian,

in the manner of Ju. D. Apresjan (1995) or Wierzbicka (1992 a). Ju. D.
Apresjan has examined emotions in the context of other "systems" regulat-
ing human existence; A. Wierzbicka reveals certain tendencies in the Rus-
sian language significant for the analysis of emotional words (such as
"emotionality", "irrationality", and "moral passion"). Here we shall focus
on those concepts which are not mentioned in the literature or are not de-
scribed in detail. In so doing, we shall try to uncover various aspects of the
emotional life of a person who speaks Russian.
Specific features of how Russian presents emotions are related to the
conception of the dusa ('soul'), which is considered not only as the most
important part of a human being (the receptacle of feelings) but also as a
physical object, a part of the human body (see Uryson 1995; Smelev
1997b). That is probably the reason for the most characteristic feature of
Russian emotional life - a paradoxical combination of predilection for
"high" or "idealised" feelings and an almost physical perceptibility accom-
panying them.
While studying "Russian emotions" we often encounter the notions of a
"national character", "Russian soul" (cf. Pesmen 2000), etc. Naturally, we
have in mind only particular fragments of a linguistic picture of the world,
i.e. a system of culturally significant stereotypes reflected in linguistic en-
tities. Language is both the source and the target of our analysis: we make
no statements about the nature of the Russians.
Our methodology is based on A. Wierzbicka's conception of interre-
lations between words and concepts which she formulates as follows: "I do
not claim, needless to say, that the absence of a word from a language
proves the absence of the corresponding concept, or the inability to form
this concept. But the presence of a word proves the presence of the concept,
and moreover, its salience to a given culture.... English has no word for the
feeling encoded in the Polish word t^sknic. Does this mean that native
speakers of English do not know (never experience) the feeling in ques-
tion? Not necessarily. Individual speakers of English have no doubt experi-
enced this feeling. But the Anglo-Saxon culture as a whole has not found
this feeling worthy of a special name" (1992 a: 21, 123).
Linguists in general, and researchers studying the linguistic picture of
the world in particular, deal with phenomena of three types: words, con-
cepts they incorporate and fragments of reality they refer to. Usually the
entities of the first type are designated in italics, the second type in inverted
commas, and the third type in Roman characters. However, due to the syn-
cretism commonly found in real word usage, we can use italics for all three
types of entities without special comment.
2. What do Russians feel?
2.1. Feeling good things
Radost' (roughly, 'joy') and udovol'stvie (roughly, 'pleasure')2
As noted by many specialists (Tolstoj 1995: 314; Smelev 1997 a: 481), the
Russian linguistic picture of the world is characterised by the opposition of
the "elevated" and the "earthbound", the "world of the sublime" and the
"terrestrial world", with a marked preference for the former. (This dualism
would seem to be rooted in the peculiar character of Russian Orthodoxy,
which determines the features of Russian culture in general: the polar-
isation of value judgments and the absence of an axiologically neutral
zone; see Lotman and Uspenskij 1994). Many important concepts exist in
Russian in two variants corresponding to these two levels, and are some-
times denoted by different words - cf. the following pairs of words which
are contrasted according to the principle of "high low": istina (roughly,
'truth') andpravda (roughly, 'truth'), dolg 'duty' and objazannost' Obli-
gation', dobro 'good' and blago 'good, welfare'. The pair radost' (roughly,
'joy') and udovol'stvie (roughly, 'pleasure') provides an excellent example
of such a polarisation of the Russian linguistic consciousness.
The words radost' and udovol'stvie have been analysed by Pen'kovskij
(1991). Of the differences noted in that article there are, we suggest, two
main ones which determine all the others. The first is that radost' is a feel-
ing, whereas udovol'stvie is merely a "positive sensual-physiological reac-
tion" (linguistic evidence for this statement can be found in the peculiar-
ities of its metaphorisation and conceptualisation in general). The second,
and in a sense main one, is that radost' belongs to the "elevated", spiritual
world, whereas udovol'stvie belongs to the "low", profane, corporeal.
These fundamental differences involve other, more particular ones
(cf. Pen'kovskij 1991). Thus, the axiological polarisation within the pair
radost' - udovol'stvie is conditioned by the fact that radost' is associated
294 Irina B. Levonna, Anna A. Zalizniak
with the abilities of the soul, whereas udovol'stvie is an attribute of the

body. So, one can say in Russian: duSa raduetsja (lit. '[one's] soul re-
joices' = 'it's a joy'), radovat'sja dusOj (lit. 'to rejoice with [one's]
soul' = 'to be delighted') and duSevno rad ('delighted'), and, on the other
hand, we say plotskie udolvoVstvija ('carnal pleasures'), but not *plotskie
radosti. Moreover, since the opposition duSa - telo already belongs to the
system of other axiologically significant oppositions (high - low,
heavenly - earthly, sacral - profane, inner - outer, etc.), a corresponding
distribution also takes place in the radost' - udovol'stvie pair.
All this is not, however, specific to the Russian linguistic consciousness:
the opposition of soul and body as "high" and "low" is a constant of Chris-
tian culture in general. Meanwhile, one more essential human attribute is
lacking here: a person's mental capacities, intellectual activity. What place
does this third element occupy in the system of binary oppositions? Eng-
lish has the word mind (which, according to Wierzbicka, is as essential for
the Anglo-Saxon linguistic consciousness as dusa for the Russian,
cf. Wierzbicka 1992 a: 46, 63), and which, including as it does the intellec-
tual sphere, is in opposition to the body. Cf. the trichotomy of duSa
'soul' - telo 'body' - urn 'mind' in Satunovskij (1996: 295), who differ-
entiates between desires of the body (physiological), desires of the soul
(for example, the desire to conquer, the desire for the happiness of someone
you love) and of the intellect (the desire to know, understand, find the
answer).
With regard to the place of intellect in the Russian linguistic picture of
the world, the following can be said. The very absence of a concept which
is comparable to dusa in importance is indicative (the importance of a con-
cept manifests itself, in particular, in the combinability of the correspond-
ing word, in wealth of metaphors and idioms). This role cannot be claimed
by um 'intellect', razum 'reason', rassudok 'sense', or evengolova 'head',
which is of the richest combinability. The main thing, however, is that in
the Russian linguistic consciousness urn 'intellect' is of relatively little
value. Thus the evaluation contained in the words bezumnyj (lit. "without
intellect" = 'mad') and bezrassudnyj (lit. "without sense" = 'reckless') is
by no means uniquivocally negative. Bezrassudstvo is similar to udal'
(roughly, 'daring'), which, axiologically ambiguous, is rather positive (see
Smelev 1997 a: 489); bezumie ('madness') "removes man from the normal
world and in a sense places him above it" (Plungjan and Raxilina 1993:
121). Yet the word bezduSnyi (lit. 'soulless') implies a definitive condem-
nation. Tjutchev's famous poem Umom Rossiju neponjat'... 'Russia is not

to be understood with the intellect' contains not only a clear statement, but
also a hidden implication (which follows from the next line arsinom ob-
scim ne izmerit' 'nor can be measured with a common yardstick', that
"true" knowledge is not attained by means of the intellect. The same idea is
expressed directly further on: v Rossiju mozno tol'ko verit' One can only
believe in Russia'. In other words, the only true, valuable knowledge is lo-
calised in the soul or heart, but not in the head.
This idea, which is undoubtedly specific to the Russian linguistic con-
sciousness, is also confirmed by the use of the words radost' and udo-
vol'stvie, from which it follows that, when confronted by the need to fit the
intellect into the framework of the binary "soul - body" opposition, the
Russian language places it in the "low" sphere, linking the intellectual with
the "corporeal" as opposed to the "spiritual". According to the Russian lan-
guage, an elegant proof of a theorem or a witty joke gives us udovol'stvie,
not radost'. In Russian, intellectual pleasure (intellektuell'noe udo-
vol'stvie) is on the same level as physiological or motor pleasures, and does
not attain the "higher" level of joy (radost').
So, pleasure, while being axiologically neutral or even positive, in the
Russian linguistic picture of the world shows a clear tendency to slide into
the sphere of negative assessment: a person who has zazdu udovol'stvij
('thirsts for pleasure') and spends his life v pogone za udovol'stvijami ('in
the pursuit of pleasure') is pathetic, and one who is vsegda dovol'nyj sam
soboj, svoim obedom i zenoj ('always pleased with himself, his dinner and
his wife', Pushkin, Evgenij Onegin 1: XII) is ridiculous. The words
dovol'stvo soboj ('being pleased with oneself) and samodovolstvo ('self-
contentment') clearly contain a negative assessment.
It is indeed hard not to detect a manifestation of the notorious Russian as-
ceticism here.
Scast'e (roughly, 'happiness') and naslazdenie (roughly, 'enjoyment')

A comparison of the radost' - udovol'stvie pair with the sast' e - naslazde-
nie pair suggests that they indicate a "proportion": scast'e is great radost',
and naslazdenie great udovol'stvie.
That naslazdenie is great udovol'stvie seems intuitively obvious: for
example, in S.I. Ozegov's dictionary of Russian language naslazdenie is
defined as "the highest degree of pleasure". And, indeed, the respective
words show considerable similarity in meaning and combinability. Thus,

naslazdenie and udovol'stvie have one and the same source, which lies in
the sphere of the physiological and intellectual (unlike radost' and scast'e
that belong to the sphere of the soul). One can experience udovol'stvie or
naslazdenie from bathing in cold water, walking or from stimulating con-
versation. Aesthetic impressions are more complicated: in this area the dif-
ference between both sensations lies not only in the strength of the feeling,
but also in its quality. If we get pleasure from a work of art (for example,
from an exhibition or theatrical performance), this feeling is more rational,
conditioned by a positive assessment of the object - whereas the enjoyment
we get (for example, from music and poetry) is an immediate impression,
which has no evaluative basis. And such enjoyment, unlike pleasure, be-
longs to the "high" sphere - cf. the word-combination istinnoe naslazdenie
'true enjoyment', whereas *istinnoe udovol'stvie 'true pleasure' is imposs-
ible in Russian, the very word istinnyj 'true' belonging to the "high" sphere
(see Smelev 1997 a). We could cite also other examples of "high" naslazde-
nie (which co-occurs with sast'e):
Slava
Mne ulybnulas'; ja v serdcax ljudej
Nase l sozvuija svoim sozdan'jam.
Ja scasfliv byl: ja naslazdalsja mirno
Svoim trudom, uspexom, slavoj.
(Pushkin, "Mozart and Salieri")
... at last fame deigned
To smile at me; and in the hearts of men
I found an echo to my own creation.
Then I was happy, and enjoyed in peace
My labors, my success, my fame
(Pushkin 1936: 429)
The grounds for the rapprochement of scast'e with radost' are the fol-
lowing. Firstly, both are "high" feelings. The words radost' and scast'e
often appear together in a text, e.g.:
The cliffs and sea, and slanting rays of the setting sun -1 seemed to see all this still
when I awoke I remember feeling joyful [= rod]. A sensation of happiness
[=scast'e], till then unknown to me, passed through my heart (Dostoevsky, "The
Adolescent").
Both states may lack any concrete reason, as in the foregoing example
and the following one:
So death has brushed against me, yet I still cannot lose a sense of constant happiness
[= sast'e]. God only knows where it has come from and how it will end... There
seems to have been so much in my life - and I am happy [=soastliva] with life itself!
(N. Berberova. "The Accompanist").
Apart from this there is an obvious similarity in metaphorical associ-

ations: scast'e, like radost', can perepolnjat' someone ('fill a person to
overflowing'); can be nezamutnennym ('unsullied') or bezoblacnym
('cloudless'), or on the contrary, omraat'sja by something ('to be dark-
ened by something'): it can b&priliv scast'a ('surge of happiness'); a per-
son can svetit'sja scast'em ('shine with happiness'), etc. (On the meta-
phorisation of radost' see Arutjunova 1976: 100; Pen'kovskij 1991: 151.)
Finally the closeness of antonyms should be mentioned: the opposite to
sast'e is gore ('grief, woe') and to radost' is ogorcenie ('grief, pain').
Nevertheless it would be wrong to say that scast'je is simply great ra-
dost' . Thus, for example, a person may have several radosti (cf. te nemno-
gie radosti, kotorye u nego v zizni ostalis' 'the few joys which remained to
him in life'), but only one scast'e (if at all). Indeed, scast'e is an indepen-
dent concept and so important for the Russian linguistic consciousness,
that we shall now consider it separately.
In her article (1992b) A. Wierzbicka quotes the following passage from
Baranczak (1990): "Incidentally, it is also interesting that Slavic languages
don't have an exact equivalent for the verb 'to enjoy'. I don't mean to say
that Americans are a nation of superficial, backslapping enjoyers and
happy-makers, as opposed to our suffering Slavic souls. What I am trying
to point out is only one example of the semantic incompatibilities which
are so firmly ingrained in languages and cultures that they sometimes make
mutual communication impossible - or, rather, they turn it into a ritual ex-
change of meaningless grunts and purrs. 'Are you happy?', an Eastern
European is asked by his cordial host. 'Yes, I am.' 'Are you enjoying your-
self?' - 'Sure, I am.' What else can be said? What would be the point in try-
ing to explain that his Eastern European mind does not necessarily mean
what his American vocabulary communicates?"
Indeed, the difference between the Russian scastliv, scast' e and the Eng-
lish happy, happiness is so great that it makes one doubt whether it is right
to regard these words as translation equivalents. The word happy is an
everyday word in English, happiness denotes an emotion associated with a
"genuine" smile (Wierzbicka 1992c: 297-298), and is one of the five or six
"basic emotions" picked out on the basis of facial expressions which cor-
298 Irma B. Levontina, Anna A. Zalizniak
respond to them and which seem to be universal (see, for example, Izard
1977; Johnson-Laird & Oatley 1989; for the criticism of the notion "basic
emotion" see Wierzbicka 1992a: 119 ff.).
First, the Russian word scast'e is by no means an "everyday word". It
definitely belongs to the "high" category and has a very strong emotional
charge, which produces two opposite tendencies in its usage corresponding
to the two extremes of the "Russian soul". The first one is the tendency to-
wards asceticism, anti-hedonism and a certain modesty (a kind of "pudeur"
that is sometimes ascribed to the ideology of Bolshevism, but has of course
a much longer history), which makes the speakers avoid using "high" and
"strong" words and class them as almost "improper", unmentionable ones.
The other, opposite, tendency corresponds to the Russian urge to talk
"about the main things" and vyvoraivat' dusu naiznanku (lit. 'turn the soul
inside-out' = 'pour one's heart out') which explains why, in spite of the
existence of the first tendency, the word scast'e is fairly frequent and char-
acteristic of Russian discourse (cf. data from the abovementioned article
Ufimceva 1996).
Second, sast'e in Russian is certainly not a "basic emotion", in any
sense of that term. It belongs to the sphere of that which is ideal and unat-
tainable in reality (cf. Pushkin's Na svete sfast'ja net... There is no hap-
piness in the world...') alongside the "meaning of life", "ideal love" and
other fundamental categories of being.
English happiness is evidently closer to dovolen 'content' and udo-
vol'stvie 'pleasure'. According to A. Wierzbicka, in contrast to the "weak-
er" English happiness, the French bonheur, German Glck and Italian fel-
icita express "a common European concept" of an emotion which "fills a
human being to overflowing, leaving no room for any further desires or
wishes" (Wierzbicka 1992 a: 299). The abovementioned words from Euro-
pean languages are indeed closer in meaning and in usage to the Russian
sast'e than English happiness, but there are also some important diver-
gencies here. One of them is the absence of a strong emotional charge and
its consequences (see above).
We must also mention the differences in the usage of French heureux and
Russian sfastliv, which derive from the fact that French has no equivalent
for Russian rad (lit. joyful = 'glad'). The meaning of Russian quasi-adjec-
tive rad in general corresponds to that of radost', but a bit "weakened" by
the fact of its use in etiquette formulas like Ja rad tebja videt' ('I'm glad to
see you'). The Russian word dovolen 'content', cognate with udovol'stvie,
discussed below, but also with udovletvorenie 'satisfaction' is more distant

from udovol'stvie: it is more "rational". Being used in a construction with a
subordinate sentence (Ja radldovolen, oto...} the meaning of both is shifted
to the "mental area" (cf. Wierzbicka 1988: 312; Zalizniak 1992: 41) and
their opposition is less evident, as in the pair radost' - udovol'stvie. But
rad is nevertheless more "altruistic" and "high", whereas dovolen is more
"pragmatic" and "low". There is also a third member of the opposition,
scastliv, which differs from rad in "degree" of feeling and its emotional
colouring. This difference is ignored in French. Since the sphere served in
Russian by three words (dovolen - rad - sastliv) is divided between two
(content - heureux), the meaning of both of them is somewhat shifted. So,
sentences such as Ja rad, cto u tebja vsS v porjadke, to tvoj syn postupil v
universitet, fto ty k nampriedes etc. ( am glad, that all is well with you,
that your son has got into university, that you are coming to see us'), etc.
can be conveyed in French either by using the word content, which corre-
sponds to the Russian dovolert, and therefore is a bad equivalent, or by the
word heureux in a somewhat "reduced", "weaker" meaning.
2.2. Feeling bad things
Toska (roughly, 'yearning')

Among the untranslatable emotion words in the Russian language which
could be classified as denoting "negative emotions" there is, first and fore-
most, the much discussed toska, which contains a specifically Russian con-
cept (see Wierzbicka 1990,1992 a: 169-174; Uryson 1997; Smelev 1997 a;
Stepanov 1997). It has been noted in particular that the ability to feel toska
is not only a characteristic feature of the "Russian soul", but also is one of
the noblest abilities of the soul (Wierzbicka 1992 a: 171). The connection
between the Russian toska and the Russian geographical expanses has also
been discussed (Smelev 1997 a: 490).
In the present article we shall not consider the word toska in detail, but
confine ourselves to indicating the connection of toska with the words
tesno 'cramped', tSietno 'in vain' and tos no 'sickening'. With two of them
the word toska would appear to be linked also etymologically (see Fasmer
1996, IV: 90; Cernyx 1994, II: 253), but we are interested here mainly in
the synchronic relations between these concepts.
The connection of toska with the idea of 'crampedness' is the most per-
sistent (although between toska and tesno there is no etymological link): it
has obvious inner justifications and is also confirmed by a multitude of
semantic parallels. The semantic connection here is that the sensation of
crampedness, compression, is evidently a "physical analogue" for the feel-
ing of toska, e.g.:
...trevozit
Ee~ revnivaja toska,
Kak budto xladnaja ruka
Ej serdce zmet.
(Pushkin, "Evgenij Onegin", 6: III).
...Disturbs
her the ache of jealousy,
as if a cold hand
compressed her heart (Pushkin 1975: 228)
The association of toska with tesno is supported also by numerous

examples of similar semantic development: old Russian tuga 'sorrow' and
Russian tuzit' 'to grieve' derive from tugoj 'cramped, pressing', French an-
goisse 'anxiety' and German Angst 'fear' from Latin angustia 'cramped-
ness', Latin word angor has both meanings 'compression' and 'anxiety'
(cf. Stepanov 1997: 675-676). (Cf. the formula for interpreting emotion
words proposed in Apresjan and Apresjan 1995: "a person's soul experi-
ences something similar to what is experienced by the body when...")
The connection between toska and tscetno consists in the following.4 The
feeling denoted by the Russian word toska has three components, as it
were, three facets, which may be combined in different proportions. One of
these facets brings toska close to fear (cf. Stepanov 1997: 670-684), an-
other to melancholy or nostalgia and a third to boredom and revulsion. The
toska that is close to fear is smertnaja or smertel'naja ('mortal') toska.
Note that the adjective here - as in the phrase smertnyjlsmertel'nyj
strax - is not only an intensifier (the lexical function Magn5, cf. lordanska-
ja 1984: 807): it may also indicate the source of the feeling. Originally
smertnaja toska is toska aroused by the thought of death, its inevitability or
proximity, and thus about the vanity of life. (Cf. the analogous meaning of
the adjective in the phrases smertnyj grex 'mortal sin', originally: 'sin pun-
ishable by death', and also smertel'naja obida 'mortal offence', originally:
Offence, which can be expiated only by death' in opposition to krovnaja
obida lit.: 'bloody offence', originally: Offence which can be expiated

only by bloodshed', cf. Lotman 1994: 165.) So toska is, inter alia, a sense
ott$etnost' 'futility'. This original connection, which seems to have been
lost in the current linguistic consciousness, may, as is often the case, be re-
stored in poetic language - cf. the following example from Andrei Plato-
nov's novel "The Foundation Pit":
The exciting sounds of the sudden music gave a feeling of conscience, they proposed
that the time of the life be thriftily preserved, that the far distance of hope be walked
to the very end and attained, so as to find there the source ofthat rousing song and not
to weep in the face of death from the melancholy of futility [= ot toski tscetnosti].
(Platonov 1978b: 17)
Finally, the connection of toska with tosno is as follows. The word toSno
(a fairly specific one in itself) is a derivative from the impersonal verb tos-
nit' 'to feel sick', but, unlike the verb which has both a literal and figurative
meaning, is used only figuratively, i.e., to indicate a certain state of mind
(soul). The connection of physical sickness with the state of the soul de-
noted in Russian by the word tosno and in general the association of tosno-
ta 'sickness' with dusa (cf. expressions such as s dusi vorotit, dusa ne
prinimaet, i.e. 'it makes me sick', lit.: 'the soul doesn't admit it') would ap-
pear to be explained by the simple fact that a sensation of physical sickness
is located in that part of the body (i.e., in the chest) where, according to
naive anatomy, the soul is. This aforementioned state of mind is close to
toska - when toska shows its least "poetic" and most sombre side. This
semantic similarity, reinforced by the phonetic and etymological ones, ex-
plains why the word tosno takes the place of the impersonal predicative:
i.e., similar to the triple sets of the type obidet'sja - obida - obidno ('to
take offence' - Offence' - 'it's offensive'), bojat'sja - strax - strasno ('to
fear' - 'fear' - 'it's frightening'), skucat' - skuka - skucno ('to be
bored' - 'boredom' - 'it's boring'), etc. we have toskovat' - toska - tosno,
cf:
Ax njanja, njanja, ja toskuju,
Mne tosno, milaja moja
(Pushkin, "Evgenij Onegin", 3: XIX)
Oh, nurse, nurse, I feel dismal,
I'm sick at heart, my dear (Pushkin 1975: 158)
There was an unbearable anguish in his heart [= toSno na duse], and the familiar,
well-read books could not drive this oppressive anguish [= toska] away (M. Gorky).
Of all the Russian writers Dostoevsky has always been regarded as the
most striking exponent of the "Russian soul": the Russian passion for suf-
fering, soul-searching, "clarifying relations", Russian nadryv, rushing into
extremes of humiliating oneself and insulting others, of the most foul self-
abasement and the most unrestrained arrogance, between adoration and
profanation - all this forms the world of Dostoevsky's characters, through
which the inhabitants of Western Europe, beginning from the school
bench, try to penetrate the enigmatic "Russian soul". On the other hand,
many studies have been devoted to Dostoevsky's language, certain features
of which have long since become commonplace in courses of Russian lit-
erature. However, the idea of linking Dostoevsky's language directly with
the Russian linguistic picture of the world did not appear until quite re-
cently and belongs to N. D. Arutjunova (see Arutjunova 1996, 1999).
In this connection we shall examine the typically "Dostoevskian" word
nadryv which expresses a characteristic extreme state of the "Russian soul"
and is quite untranslatable, and also the words obidno and sovestno, which
are less obviously but no less indisputably specifically Russian concepts
(and, incidentally, hold a place of considerable importance in Dostoevsky's
world).
Nadryv
The word nadryv is one of the most capacious, expressive and ingrained
words in Russian culture, which is very hard to translate into English. The
Russian poet Lev Losev enumerates in a poem the repellent signs of the
mad Russian life. Among other things the poem says:
And he also said in excitation:
I don't like those drunken nights,
Penitential frankness of drunkards,
The Dostoevski] nadryv of the informers
It is interesting that this word, similarly to the words avos', toska, leri,
etc. is frequently associated with the epithet "Russian". (On these combi-
nations see Plungjan and Raxilina 1996.)
This word comprises a poignant psychological sketch. Nadryv presup-
poses an emotional state when a person cannot handle his or her feelings.
The person is either overcome by emotions making him or her forget about
limits, manners and even decency, or they come out ponderous and disin-
genuous. In addition, nadryv presupposes a certain masochistic self-reflec-
tion and hysterical confession which sounds jarring to the people around.
Cf.:
Bely suffered desperately, driven from humiliated meekness to furious arrogance - he
cried that it was a blasphemy to reject his love.... He spent several months abroad and
came back still suffering and with "Cup of the Snowstorms" - the weakest of his
symphonies because it was written in nadryv (V. XodaseviC).
It is noteworthy that the root -ryv- (-rv-) 'to tear' forms many words de-
scribing the emotional life of a human being: poryv 'impulse', razryv 'rup-
ture', sryv 'collapse' - as well as the modern slang otryv 'enjoying doing
something a lot'. The action of tearing is perceived as something closely
related to an emotional state. Therefore, the gesture of tearing open one's
shirt on one's breast (rvanuf rubaxu na grudi), which is considered char-
acteristically Russian, is very important.
The noun nadryv is mainly connected with the verb nadryvat'sja which
means 'to do something with a great effort'. But more important than the
context of work is that of shouting, crying and, especially, coughing. Con-
sider the expression nadryvat'sja ot kaslja 'to have a whooping cough', ad-
jective nadryvnyj, adverb nadryvno and the usage of the noun nadryv in the
word combination kasljat' s nadryvom 'to cough distressingly'. When we
say that a man nadryvajetsja ot kaslja, kaSljajet nadryvno or s nadryvom,
we mean a specific type of cough. It is not one where the throat aches, but
one that comes out of the depths of one's lungs and shakes the entire body.
This cough is not necessarily loud but hollow, as if a man were trying to
push out of his depths the things that hinder him to breathe and to live. Such
a cough takes a person's energy, and the person still coughs uncontrollably.
Similarly, to say something with nadryv does not mean to cry it out. One
can speak with nadryv in a hardly audible manner but in a way it is clear
that the words come out from the bottom of the soul. It is torturous to say
them but it is impossible not to say them.
Thus, it is clear why the word nadryv describes two opposite situations:
an uncontrolled emotional outburst or an expression of forced and inauth-
entic emotions. In the first case one shows one's deeply hidden intimate
feelings with a frightening frankness, that exposes the things which are
supposed to be most personal (cf. the abovementioned example about An-
drey Bely). In the second case one goes even further. One is so deeply in-
volved in introspection that one finds in one's soul something that does not
exist. That is why with nadryv extremely exaggerated, distorted, or even
imaginary feelings are expressed. In this case nadryv verges on falseness or
grotesque. Consider another example from V. F. Xodasevic where he dis-

cusses a rumor that before his death the Russian philosopher V. V. Rozanov
became so poor that he started gathering cigarette butts:
It is very probable that he started gathering them, but maybe there was a certain nad-
ryv there and even some stylization? Because to show oneself poor and humiliated is
in the "style" of Rozanov, and it was all the more natural in light of such a tragedy.
Often a semantic motivation appears through figurative, metaphorical

meanings of words. So, probably, duSevnyj nadryv, a 'soul-nadryv' is re-
lated also to the literal understanding of nadryv as a gap. An external shell
opens, the interior comes out through a hole, and the surrounding people
look aside in embarrassment.
In discussing the dostojevskij nadryv Lev Losev is right not only with re-
spect to the fact that a drunken confession of a snitch is a thing that can be
naturally called dostojevscina ('Dostoevsky-stuff'). Furthermore, the word
nadryv with this meaning began to penetrate the Russian language just
after Dostoevsky.
In general, the world of Dostoevsky, which comprises characters who
display hypertrophied feelings, pathological personalities, and broken des-
tinies, is a continual nadryv. The behavior and speeches of Marmeladov
from Crime and Punishment can be a good illustration of this concept. The
psychological complex imbedded in the word nadryv is the most irritating
thing in Dostoevsky - for Nabokov, for example. Cf.: "Dostoevsky's lack
of taste, his monotonous dealings with people suffering pre-Freudian com-
plexes, the way he has of wallowing in the tragic misadventures of human
dignity - all this is difficult to admire" (Nabokov 1981: 104).
However, the word nadryv comes from his last novel The Brothers Ka-
ramazov. It is one of the most important words in the novel. One of the
novel's parts is entitled "Nadryvy" 'Crack-ups'. The same word is re-
peated in the titles of chapters 5, 6, and 7: "Nadryv in the drawing room",
"Crack up in the izba", "And out in the fresh air." The translator, David
McDuff, comments: "In Russian nadryvy can be translated approximately
as 'cracks', 'ruptures', 'harrowings', but also as 'hysterias'. Nadryv like its
French parent dechirement connotes a breaking, tearing and straining be-
neath an intolerable weight of mental, emotional and spiritual suffering. It
is a constantly recurring theme through the novel. 'Crack up' is offered as a
near equivalent". V. Nabokov drew attention to the cryptic character of
these titles. The heroes of the novel repeatedly say the word nadryv as if
tasting it and trying to understand the world through this word. The major-
ity of cases of its usage relate to Katerina Ivanovna and her feelings to-
wards Dmitri. First Dmitri humiliated her in promising to give her money,
in order to save her father, only if she comes alone to take it from him. Then
he gave her the money and demanded nothing in exchange. Soon after that
she wrote him and told him she loved him and wished to be his wife. From
the feeling of insulted human dignity, sacrificial ecstasy, and hatred grow
love and passion (which are also accompanied by a certain desire for sac-
rifice). Katerina Ivanovna says:
"Even if he marries that... brute creature", she began majestically, "...From this day
forth I shall never, never give him up! she uttered in a kind of hysteria [= nadryv in the
Russian text-/.L., A.Z.} of pale, forced ecstasy." (Dostoevsky 1983: 215)
The situation is aggravated by the fact that due to artificially excited

passion for Dmitri, Katerina Ivanovna rejects the love of Ivan.
She has taken her revenge on me for all the insults she has constantly and every mo-
ment endured throughout all this time from Dmitry insults dating from their first en-
counter... Because even their first encounter stayed in her heart as an insult. ...I shall
now depart, but let me tell you, Katerina Ivanovna, that in reality he is the only one
you love. And you love him increasingly the more he insults you. That is the form
your hysteria [nadryv in the Russian text-/.L., AZ.] takes. (Dostoevsky 1983: 219)
Although the word nadryv is so complicated in its meaning, it penetrated

the language in a free and natural manner - because it corresponds per-
fectly to the general character of the Russian language and culture. And
even after its adoption into the Russian lexicon, the original meaning of the
word nadryv was not dissolved. Instead it was kept in the Russian mind
through its genealogy (it is reminiscent of the sestidesjatniki lit. 'people of
the 1860s' and subsequent cultural reflection, similarly to the wordporyv
'impulse' which also keeps an imprint of the previous epoch, see Vinogra-
dov 1994: 791-792.). The origin of the notion of nadryv in Dostoevsky's
later works is a remarkable fact. By the use of it Dostoevsky showed one of
the fundamental categories of behaviour and thinking of the Russian raz-
nocintsy (mid 19th century intellectuals united by their radical political be-
liefs; literally 'from different social origins').
The culture of the gentry valued external lustre, good manners, comme il
faut, high life, or dandyism. The lack of external lustre was called vulgarity
or bad style (see Lotman 1994). In his "Youth" Tolstoy described the world
view of a young nobleman for whom it was better to die but not to be not
comme ilfaut. The raznoftntsy saw in the culture of the gentry falsity and
superficiality. Therefore, they rejected external lustre. Instead, they sug-
gested the cult of frankness and a deep understanding of things. The raz-
notintsy had a peculiar mode of behavior, which was a thundering blend of
reckless, unrestrained frankness, romantic melodrama, and a propensity to
the depraved. This blend is familiar to readers of Belinsky's letters to Ba-
kunin, Cernystevsky's diaries or almost any personal document from the
people in this social group or psychological type. (On the aesthetics of the
everyday behavior of the raznointsy see Paperno 1996.)
Returning to the theme of coughing, let us remember consumption - the
disease of the raznoZintsy. This consumptive whooping (nadryvnyj) cough
accompanies "the cough of the soul." This is an anguishing frank (nad-
ryvnyj) confession where the shameful and ugly truth about a person is re-
vealed.
Manifest drinking is a strikingly peculiar feature of the nadryv aesthetics
cultivated by the raznocintsy. It is said that What a sober person has on his
mind, a drunk has on his tongue: alcohol intoxication is a perfect moti-
vation for refusal to refrain from expressing one's feelings. This refusal be-
came culturally significant fact, and the abyss of alcoholism a synonym of
spiritual depth. Since then this type of behavior has never had such a high
cultural status. It degenerated into self-parody.
The 20th century gave a new aesthetic dimension to the word nadryv. For
Dostoevsky nadryv was still interesting and aesthetically attractive though
it was fraught with lies. Now it is usually assessed as tastelessness: "Do
you like Andreev? - No, states S. Dovlatov. He is flamboyant and with
nadryv". And Brodsky says about Pushkin in admiration: "Still Pushkin
belongs to the gentry, don't forget it. If you like, he is an Englishman - a
member of the English Club - in his attitude towards reality he is reserved.
He has not got the thing that is called nadryv".
Obida (roughly, Offence')

The emotion of obida is specific to the Russian language. In English,
French and German there is no single translation equivalent: to the Russian
obida, obidet'sja correspond words either with the meaning more close to
the Russian oskorblenie, oskorbit'6 (English offence, to offend, French of-
fense, offenser, German beleidigen, Beleidigung) or words with a figurative
meaning (English to hurt, to wound [someone's feelings], Fr. blesser, Ger-
man krnken, cf. also Russian ranit'). The corresponding emotion is ob-
viously peripheral for English and German (at least neither word occurs in
the list of thirty-six common English emotion words in Wierzbicka 1972
and in the list of the forty most frequent German emotion words in Mel'ouk
and Wanner 1996). By contrast, for the Russian language the words obida
and obidet'sja are extremely frequent, and the emotion denoted by them is
one of the most important for the Russian linguistic consciousness (the
word obizat'sja is included in the list of forty emotion words described in
lordanskaja 1970).
The verb obidet'sja denotes a certain emotion which arises as a reaction
to an action by another person. One cannot obidet'sja on oneself or on a
natural phenomenon - although one can, for example, be angry with one-
self or despair of the bad weather.
What exactly is the emotion of obidal Obida consists in feeling pity for
oneself linked with a certain disillusionment in another person, caused by
the fact that this person by his or her words or acts has shown, as it were,
less affection or regard for us than what we had expected from him or her
(the disillusion must be a local one, which has not yet become global: in the
latter case there is no room left for obida). Obida is perceived as experi-
encing something unfair towards yourself (cf. the inclusion of the idea of
'unfairness' in the interpretation of the word obizat'sja in lordanskaja
1970: 19). The conception of unfairness arises from the idea that the other
person owes to oneself something good (good feelings or respect), often
without any definite grounds. So, obida is an aggressive feeling: it means a
reluctance to accept the situation that seems unfair, a demand for compen-
sation. Obida is opposed to a resignation, on the one hand, and to the state
of forgiveness on the other.
We propose the following explanation for the Russian obida:

(a) this other person did something
(b) because of this I felt something bad
(c) I think that this person should not do this
(d) I feel something bad about me because of this
(e) I feel something bad toward this person because of this
The components (a)-(b) constitute a presupposition, (c)-(e) are assert-
ive: Ja ne obidelsja did not take offence' means "I don't find that this per-
son ought not do it", "I don't feel anything bad about me because of this"
and "I don't feel anything bad toward this person". Note that Ja prostil
forgave' differs from Ja ne obidelsja first and foremost by the fact that it
does not include the negation of the component (e), meaning don't feel
anything bad toward this person (even though I think the action was un-
fair)'.
The Russian obida can be compared to the Polish word przykro, de-
scribed by Anna Wierzbicka (this volume), who writes: "It is clear that the
concept of Offended' has a good deal in common with the Polish concept
of przykro. Roughly speaking, what the two share is some perceived 'defi-
ciency of regard'. But first, in offended, this 'deficiency' concerns an ap-
parent absence of 'good thoughts about someone'; second, in offended this
appears to be intentionally displayed, whereas in przykro it can be (and
typically is) betrayed or implied unintentionally; and third, in the case of
offended the experiencer must be different from the agent (I cannot be of-
fended by my own action), whereas in the case of przykro the experiencer
can coincide with the agent: I can feel przykro because of something that I
have done myself (causing a bad feeling for someone else)".
The Russian obida shares with the Polish przykro the first two features
(the third making przykro close to the Russian sovestno, see below), and is
opposed to the English offended more or less in the same way.
So, in accord with the second feature (common to przykro and obida)
obida often arises as a result of an inference made by the "offended" per-
son. Different people are more or less inclined to such inferences. Hence
obida is an egocentric feeling, the source of which is rooted, in the final
analysis, in ourselves. A disposition to obida is the characteristic of the
"offended" person: therefore an obidiivyj person is someone who is in-
clined to take offence (and not to give it - which, from the point of view of
Russian morphology is just as possible, cf. obmanivyj = 'which deceives'
and not 'which is deceived').
The meaning of the impersonal obidno is even more specific. The emo-
tion expressed by the word obidno is somewhat different from that con-
tained in the words obida and obidet'sja. Thus we can say Mne obidno, to
nikto ne pritronulsja k moemy salatu 'I'm offended (lit.: to.me it's.a.pity,
ADV.) that no one has tried my salad', but it would hardly be correct in this
situation to obidet'sja 'take offence' at the guests or to harbour an obida
against them (provided, of course, that there are no grounds for suspecting
deliberate malicious intent in this act). On the other hand, the phrase Mne
obidno, to Ivan ne pozdravil menja s dnem rozdenija am offended that
Ivan did not wish me a happy birthday' sounds rather strange. It would be
more natural to say Ja obidelsja na Ivana za to, cto on ne pozdravil menja s
dnem rozdenija or Mne obidno, cto nikto iz moix druzej ne pozdravil menja
s dnem rozdenija am offended that none of my friends wished me a
happy birthday': the former sentence describes my attitude towards Ivan,
the latter, my feeling pity for myself. In the impersonal obidno, the obidik
(the person who causes the offence) disappears: there is no syntactic op-
portunity for mentioning this person. Thus, whereas obidet'sja and obida
describe an attitude of the offended person to the offending one, in obidno
the stress shifts to the offended person himself. We feel obidno when we
perceive, as A. Wierzbicka says, a "message of indifference" from the
outer world.
The feeling expressed by the word obidno arises when something dear to
the person in question is abased, ridiculed or simply underrated. This for-
mula can serve as a generalisation for all cases: the "something dear" is
either the person him/herself, or another person, or thing that is dear to
him; it can also be the sum total of one's own wasted efforts: it is obidno if
they are spent in vain (cf. the abovementioned example with salad).
Sometimes obidno comes close to zalko, cf.:
(1) Obidno/'zalko byloby upustit' takuju

hurtful/pitiful (ADV.) it.would.be to.miss (INF.) such
vozmoznost'.
opportunity
'It would be a pity to miss such an opportunity.'
Other examples: obidno /zalko ujezzaf ran'se sroka 'it's a pity to leave
before the time is up'; obidno I zalko, cto ne udalos vstretit'sja 'it's a pity
we did not manage to meet'. Obidno and zalko in this case describe the
feeling that arises at the thought that circumstances which have taken a
certain turn might have developed differently and that this "differently"
would have been better. Both words express the idea of some positive al-
ternative. It is rooted in the very word zalko (see Zalizniak 1991); in obid-
no it appears by the transformation of the idea of a "just" state of affairs
which is being violated. The difference between impersonal zalko and
obidno result from the fact that the "raison d'etre" of sozalenie (regret) is
"it could be otherwise", whereas the "main idea" of obida is "it must be
otherwise".
So, we can say zalko, ito sejas leto: nel'zja pokatat'sja na lyzax 'What a
pity it is summer now and we can't go skiing', but we cannot say *Obidno,
to sejias leto. Obidno is appropriate only when the positive alternative is a
more "normal" version of the development of events, cf.: Kak obidno, ito
isportilas' pogoda 'What a pity the weather has turned bad'; obidno, to iz-
za kakogo-to pustjaka vs sryvaetsja 'What a pity that it has been called off
because of a trifle'; tak obidno zabolet' v pervyj den' kanikul 'What a pity
to get ill on the first day of the holidays'. In a construction with a subordi-
nate perfective infinitive, obidno is close in meaning to ugorazdilo (see Za-
lizniak and Levontina 1996), in the distribution of the communicative
status of the components. Both zalko and obidno combine with the subor-
dinate infinitive form, a construction specific to the Russian language, and
are without any satisfactory equivalent in European languages.
The peculiarity of the Russian obida, obidno can also be seen by com-
paring it with another Russian word: dosada, dosadno (roughly, 'annoy-
ance', 'annoyed'). Whereas obida is always felt in relation to someone
else, i.e. it is a feeling towards someone, dosada can only be felt about
something, some circumstances which prevent the fulfillment of one's
wishes: this could also be another person, but in dosada this person ap-
pears only as a mechanical obstacle. Therefore one can dosadovat' 'be an-
noyed' with oneself, for example, at one's own stupidity or bad luck. The
source of dosada is a single event from which no consequences are
drawn - unlike obida, where what is important for us is usually not the act
itself, but what "lies behind it" (namely, an attitude towards us which we
do not like). Obida is akin to grief and tears, dosada to anger and aggres-
sion.
To conclude we will say that the inner state described by the Russian
word obidet'sja is fairly close to that presented by the English to be hurt
(see the explanation of the latter in Wierzbicka, this volume), the main dif-
ference between the respective concepts being the degree of their salience
in the linguistic picture of the world. The Russian obida is an elaborated
concept of a high cultural significance, while the same cannot be said about
the English to be hurt. Linguistic evidence of this is seen in the existence of
a substantive and other derivatives (lacking for the English to be hurt}. The
feeling called obida possesses a certain independence in Russian, in the
sense that one can do different things with obida: derzat' 'to hold' or kopit'
'to accumulate\pomnit' 'to remember' or zabyvat' 'to forget' them; obida
itself can rasti v dushe 'to grow up in the soul' orproxodit', uletuiivat'sja
'to evaporate' and so on. Another important difference: feeling hurt is a

spontaneous reaction, whereas obida is more rational, it often results from
a certain intellectual processing of what was said or done. (For an extensive
discussion of obida see also Pesmen 2000.)
Sovestno
The concept of sovest' 'conscience' plays an important role in Russian cul-
ture. Sovest' performs a double function: on the one hand, it enables us to
distinguish between good and evil, and on the other, guides our actions, di-
recting them towards good. As Ju. D. Apresjan puts it, in the Russian lin-
guistic picture of the world sovest' is seen as "a strict internal judge" and as
"a moral restraint which blocks the realisation of amoral desires" (Ju. D.
Apresjan 1995: 353).
The fact that sovest' can gryzt' 'gnaw', terzat' 'tear to pieces', mudt'
'torment' and in general arouse various unpleasant sensations postfactum,
when the act has already been committed (in the case when sovest' has per-
formed its first function and failed in its second) is common to many lan-
guages. The sensations aroused by sovest' are connected usually with the
idea of "gnawing, biting" - cf. Russian ugryzenija sovesti, German Gewis-
sensbisse, French remords, English remorse having similar etymology (no
longer alive in the last two). Specific to Russian is the fact that it has an
idiomatic way of representing the role of sovest' in taking a decision: the
predicative adverb sovestno, whose meaning differs considerably from that
of the substantive sovest'. Let us take a closer look at this.
(2) Mne bylo sovestno ego ob . prosit'.

to.me it.was shameful(ADV.) him of that to.ask (INF.,
IMPERF)
was ashamed to ask him to do that.'
The word sovestno is close to stydno, on the one hand, and to neudobno,
on the other. Sovest' 'conscience' and styd 'shame' in Russian form a pair
of related and complementary concepts, whose proximity is reinforced by
expressions such as ni styda, ni sovesti' [someone has] no shame, nor con-
science' parallelisms such as Kak tebe ne stydno? and Kak tebe ne sovest-
no? (both: 'you should be ashamed of yourself), by the similarity in the
model of government of the words stydno and sovestno. And yet the word
stydno denotes a feeling of shame, one of the more fundamental and al-
legedly universal (see Izard 1977) human emotions, whereas the word so-
vestno contains a specifically Russian concept and may be translated into
other European languages only insofar as it is synonymous with stydno
(cf. the dictionary translations offered for Russian sovestno: English to be
ashamed, German sich sch men, French avoir honte). (On the concepts
styd and sovest' see also Arutjunova 1996, 1999.)
The point of intersection for the spheres of sovestno and stydno are bad
actions, i.e. actions condemned by moral norms (cf. Apresjan 1997 b: 417).
However, each of these concepts also has its own spheres. First and fore-
most, unlike sovestno, which characterises actions (i.e., intentional acts
controlled by the subject), the sphere of stydno covers the most varied cir-
cumstances, including some that do not depend on the subject at all (as, for
example, one's origin or physical defect); stydno may be for actions com-
mitted by someone else. Thus, it can be both stydno and sovestno to mis-
lead other people (because a lie is a morally reprehensible, i.e. bad action),
whereas to believe cock-and-bull stories can be only stydno, but not sovest-
no - because the stupidity you thereby reveal doesn't belong to the cat-
egory of "bad actions".
On the other hand, stydno is not necessarily for something bad. It may be
simply something intimate: something which should be private, but has be-
come public. This includes various circumstances concerning the body and
sex, and also, for example, the manifestation of strong feelings or excessive
exposure. Sovestno cannot apply to any of this.
The emotion of shame can be rendered in Russian by a verb: stydit'sja 'to
be ashamed'. But there is no *sovestit'sja, because there is no such emo-
tion except for the construction with a subordinate infinitive where it ap-
pears in the perfective form posovestit'sja: X posovestilsja to do = 'X
hasn't done because X felt sovestno about it' (cf. section 5).
The word neudobno is also specifically Russian. The nearest English
equivalent inconvenient differs from neudobno in frequency and stylistic
register, neudobno being a common and everyday word. The original
meaning of the pair udobnolneudobno is physical comfort. Neudobno re-
flects a sense of the possible violation of some norms of decency or rules of
conduct, which are usually rather complex and subtle and are determined,
inter alia, by relations between the people in question (cf. Mne neudobno
ego ob tom prosit', a ty mozes' eto sdelat' 'It's inconvenient for me to ask
him, but you could'). In certain contexts sovestno seems almost synony-
mous with this neudobno, cf. Mne sovestno/'neudobno, to ty vse vremja
cto-to delajes", a ja tol'ko sizu i razgovarivaju Tm ashamed/embarrassed

that you are busy doing something all the time, while I just sit and talk'. In
modern Russian neudobno in this sense is gradually extending its use,
driving out sovestno. (Apart from this "ethical" meaning the word neudob-
no, and its antonym udobno, has a "pragmatic" meaning, which strangely
enough is also hard to express in other European languages - this fact was
brought to the attention of one of the authors by E. N. Savvina - cf. Kogda
tebe bylo by udobno so mnoj vstretit'sja? 'When would it be convenient for
you to meet me?' Mne neudobno vesti zanjatija vpervojpolovine dnja 'It's
inconvenient for me to take classes in the morning').
The proximity of sovestno and neudobno is based on a common compo-
nent connected with the idea of causing someone else an unpleasant feel-
ing. This idea is absent in stydno, where "the other" performs an important,
but quite different function (see glosses below; cf. also Arutjunova 1997).
In general it can be said that it is sovestno to do something that harms
someone else (and in this respect there is a shift of accent in sovestno com-
pared with sovest' where one's relations with oneself or with God are at the
centre). The action may not even be bad or unfair in itself, and the "harm"
caused may be the result of a perfectly neutral action (for example, exac-
ting rent or a fine): but if the other person is poor, to take this person's last
few pennies is sovestno, nevertheless. A person inclined to feel sovestno is
called sovestlivyj celovek, and this is a moral category - whereas stydlivyj
'inclined to feel shame' refers only to a person's behaviour.
The words discussed in this section can be glossed as follows:
stydno
(a) someone else can think something bad about me
(a) because of this I can think something bad about me
(a) because of this I feel something bad
(d) like I do when I know that I have done something bad
sovestno
(a) I have done (want to do) something
(a) someone else can feel something bad because of this
(d) like I do when I know that I have done something bad
neudobno
(a) I want to do something
(a) someone else can feel something bad because of this
3. What do Russians feel toward one another?
3.1. (don't) want to see you'

We should like to draw attention to the fact that many designations of feel-
ings in the Russian language are connected with the idea of 'seeing'.
Let us begin with love. The connection of love with the desire to see the
beloved person is very widespread: examples of this from European poetry
from Sappho onward are so plentiful that there can hardly be any need to
quote them.7 The special feature of Russian is that this connection is re-
flected in the inner form (semantic motivation, mode of nomination) of
some words. Thus, one of the Russian words for addressing the beloved,
which has no translation equivalent, is nenagljadnyj , literally, '[so beauti-
ful that] it can never be enough looking at it' (cf. section 3.3). The connec-
tion of love with looking is also reflected in the word ljubovat'sja 'to ad-
mire', from the root meaning 'love'.
At the opposite pole (cf. Axmatova's Ne ljubis", ne xoies' smotret' ...
'you don't love me, you don't want to look at me') the main word, nenav-
ist' 'hatred', is derivative of the idea "to see", unlike the West European
languages where the words meaning 'hatred' are undecomposable
(cf. Eng. to hate, Ger. hassen, Fr. hair, Ital. odiare). According to Fasmer
(1996, III: 63), the church-slavonic word nenavideti is formed by negation
of navideti 'to look gladly at somthing or sombody'. The same idea is
found in the Russian expressions videt' ne mogu\ glaza by moi [na nego]
ne gljadeli can't stand the sight of. The desire not to see is reflected in
the inner form of the word otvrasZenie 'revulsion', lit. 'turning away
from'; cf. also the expression otvernut'sja [ot kogo-to] 'to avert one's eyes
from somebody'.
There are also some emotion words where the original connection with
the idea "to see" is no longer alive. According to Fasmer (1996, III: 100)
the word obidet' 'to offend', discussed in section 2.2., derives from ob-
videt', where the preposition ob- has the meaning 'around, skirting, by-
passing', cf. obnesti [kogo-to ugo$eniem] 'to leave someone out when
serving food', obdelit' 'do someone out of his share', obvesit' 'cheat in
weighing', obojti vnimaniem 'neglect, disregard'. Thus, the Russian
obida has grown out of the idea of obdelit' vzgljadom 'not to look at
someone'.
A similar idea is included in the word prezrenie 'contempt'; prezirat'
[kogo-to] literally means 'to look over someone', i.e., not to see the person
(being somehow above him/her).
3.2. want to talk to you'
In this section we shall consider the words drug 'friend', obs enie 'contact'
and otnosenie 'relation, attitude'. The word drug (and also its synonyms in
Russian and equivalents in other languages, in particular, English friend) is
analysed in detail in Wierzbicka (1997, ch. 2). Moj drug in Russian is a per-
son I want to be with and talk to, and to talk po duSam 'intimately, sin-
cerely' - cf. Wierzbicka's components want to be with this person',
want to talk to this person' and' I can say anything to this person' (1997:63).
All that can be added to Wierzbicka's analysis is perhaps that the Rus-
sian word drug seems about to share the same destiny as English friend.
This word in English, as Wierzbicka points out, has undergone a semantic
evolution consisting of a "weakening" in its meaning: from the meaning,
close to that of the Russian drug, with the main semantic component
want to do GOOD THINGS FOR this person' to the meaning centered round
the idea want to do THINGS WITH this person' (Wierzbicka 1997: 40). In
modern Russian discourse (particularly in the speech of people who have
lived abroad for a long time) the use of the word drug increasingly appears
as a nomination for people with whom the speaker goes on Sundays to a
barbecue.
The words obS enie (roughly, 'contact, communication') and obscat'sja
'to have contact with' have not yet attracted the attention of linguists. Yet
these words are very specific and difficult to translate. There are at least
two features which distinguish them from their translation equivalents. The
first is that the words obScenie and particularly obscat'sja belong to every-
day language and contain the idea of something "warm" and "friendly";
the second is that obscat'sja in the modern language refers to a concrete
process.
Obscat'sja in Russian means 'to talk with someone for a while in order to
preserve dusevnyj kontakt with that person' (dusevnyj kontakt, close con-
tact, literally 'contact between the souls of the interlocutors'). The fact that
the generic seme here is 'to talk' is demonstrated by the existence of the
valency of content: na kakuju temu On what subject'. The talk which forms
the content of obscenie need not necessarily be about "highbrow things". It
can be about mere trifles as well, because the main thing in obSenie is to
strengthen friendly contact. The word obsat'sja also contains the idea of a
certain purposelessness in this activity and the radost' or udovoVstvie ob-
tained from it (the former being due to the abovementioned dusevnyj kon-
takt, the latter, to the intellectual interaction, cf. section 2.1). This meaning
of the word obscat'sja appeared relatively recently: it is a colloquial new
formation and by older and/or more conservative speakers of the Russian
language is regarded as vulgar. Not long ago this verb used to be used only
to denote an "generic state" (using Z. Vendler's terms), like pitat'sja 'to
feed' or pravit' 'to rule', cf. Deti v etom vozraste obscajutsja v osnovnom
so svoimi sverstnikami 'Children of this age mix mainly with their peers'.
It is characteristic that from the verb obSoat'sja in modern colloquial Rus-
sian different manners of verbal action are easily formed: delimitative
(poobscaf sja polcasa 'to do it for half an hour'), perdurative
(proobscat'sja ves' veier 'to do it all evening'), saturative (naobscat'sja
vdovol' 'to do it to one's heart's content') and even semelfactive (pb-
scnut'sja 'to do it within a very short time, but intensively'). The last verb
is regarded as a vulgarism not only by the purists.
In order to do something, including poobsiat'sja s druz'jami, Russians
must do three difficult things: sobrat'sja ('to make up one's mind', see Za-
lizniak and Levontina 1996), vybrat'sja (more or less the same as so-
brat'sja, but when speaking about actions one does outside, e.g. visiting
people) and dobrat'sa (reach the place). All three words (derived from the
same root meaning 'to take') encode language-specific concepts and are
untranslatable: the first two of them reflect properties of the "Russian
soul", and the third, those of the Russian landscape (and mode of life in
general). Indeed, the five hundred years that, according to Pushkin, Russia
needed to improve its roads, are not yet over.
Another purely Russian occupation, which is, in a sense, something quite
opposite to the obScenie, is called vyjasnenie otnosenij (lit. 'clarifying re-
lations'). This language-specific expression means: to express resentment
to each other concerning not actions or statements (i.e., something "tan-
gible"), but considerations of a purely ideal, unobservable and unverifiable

nature, which in Russian are denoted by the untranslatable word otnosenie
(something like 'attitude'). The otnosenie (of one person to another) is his
inner state, which may manifest itself in some way or other or may remain
unexpressed, without losing its existential status, however. In the Russian
linguistic picture of the world this mysterious otnosenie (of the other per-
son to me) is the only thing that really matters in interpersonal relations - at
least it is much more important than anything this other person could do to
me. The phrase Kak ty ko mne otnosis'sja? (roughly, 'How do you feel
about me?') is not just a question used by teenagers to make the other per-
son say they love them: it is also an explicit or implicit subject of a most
significant part of Russian dialogues (and particularly between Dos-
toevsky's characters).
3.3. *I feel the same thing you feel'
Zalost' (roughly, 'pity')

One of the properties that makes people human is the ability to feel pain in
response to the suffering of others. In Russian, there are several words de-
scribing this type of feeling: nouns zalost' 'pity', socuvstvie 'sympathy',
sostradanie 'compassion', u astie 'compassionate help' and a number of
corresponding adverbs (, zalko}\ verbs (zalet', soauvstvovat', sostra-
dat'); and adjectives (zalkij, zalostlivyj, ucastlivyj). The meanings of these
Russian words and the difference between them and their English and
French equivalents have been described in detail (see Zalizniak 1991;
V. Apresian 1995; Levontina 1997 a). The particular significance of the
feeling of zalost' in Russian culture has often been pointed out (see
Wierzbicka 1992 a: 169). The propensity to this emotion (zalostlivosf) is
associated with an image of a typical Russian woman.
V
Zalost' in the Russian language is both one of the most spontaneous and
one of the most culture-mediated emotions. It is an uncontrollable and im-
mediate response of the soul to the suffering of others. Zalost' may em-
brace, cover, squeeze the heart, etc. Cf.:
An unexplainable zalost' always grew in me when I spoke to him (N. Berberova,
"Italics are mine").
v
Zalost' implies a state of mind closely associated with physical sen-

sations. The feeling of zalost' is often accompanied by sensations of dis-
comfort, wearisome anxiety, and spasmodic contractions inside the body.
This feeling may be painful and cause tears. In this respect zalost' re-
sembles tenderness on the one hand, and disgust on the other hand, and fre-
quently the two occur in combinations: there can be neznaja zalost'
('tender pity') and brezglivaja zalost' ('fastidious pity'). Cf.:
Xodasevic, with some pity mixed with disgust, recalled how those lousy girls in rags
clung to him trying to undress him directly at the stairs and raised their torn skirts
over their heads crying obscenities" (N. Berberova, "Italics are mine")
Zalost' differs from similar emotions in the following way. Soouvstvie,

sostradanie and uiastie imply a partial identification of oneself with an-
other person, and a wish to share his or her sorrow, while zalost' denotes a
certain feeling that does not depend on the feelings of another person. That
is why one can feel zalost' towards oneself, or towards the dead. One can
feel zalost' at the sight of a simple small or weak creature, which is not
necessarily suffering. V. V. Rozanov said: "Pity is in the small things. That
is why I love everything small" .
Additionally, zalost', unlike sotuvstvie, presupposes awareness of a cer-
tain distance between the subject and object of this emotion; moreover,
zalost' is often accompanied by a slightly protective attitude towards the
object. So, the adjective zalkij means causing zalost' and, as a result, dis-
dain. Cf.:
Those traditions had a tinge of both blissful sophistication and haughtiness. There
was haughtiness in the desire of the nobles to live the way humble folk live, in the
acute zalost' they felt (always as misunderstanding of the one on the top or aside),
arid in the shame for the advantages they had.. - Russian literature of the 19th century
abounds with zalost', while people who created it, with a few exceptions, hardly
knew any hardships in their life (L. Ya. Ginzburg).
During the Soviet epoch, due to the cult of force fostered by official So-
viet ideology, the mere word zalost' became associated with something hu-
miliating - cf. Gorky's expression which became a pithy saying Zalost'
unizaet eloveka 'Pity humiliates man'. Soviet books and movies are full
of similar rhetoric. Even V. Vysotsky sang: "When I see broken wings/I do
not feel any zalost' and there's a reason for it: I hate violence and impo-
tence".
The appearance of such connotations associated with the feeling of pity

is typical of European culture of the 20th century in general.
People around were becoming more and more pitiless [= bezzalostnyj], and that was
the call of the times, not a fashion, it lasted for a century, not a day. Pitilessness in the
people of our times can be traced back to the 80's and 90's of the last century when
Strindberg wrote his "Confession of a Fool". ... "Feel pity for me!" - but nobody
could or wanted to feel pity. The word pity was then on the verge of extinction. It is
not accidental that nowadays this word is used only in an offensive humiliating sense:
with the overtone of disdain in French, of vexation in German and of ironic malevo-
lence in English" (N. Berberova, "Italics are mine").
When understood this way, zalost' is different from love: one can often
hear it said in Russian something like this: Ne ljubit, a tol'ko zaleet 'He
does not love her, he simply has some pity for her'. But this concept of
zalost' is alien to traditional Russian culture, where "to feel zalost'" means
"to love", and this concept is specific to Russian (cf. Wierzbicka 1992 a:
168-169; Stepanov 1997: 281). In Russian culture zalost' was always
understood as one of the supreme values, as a feeling very close to love and
not offensive at all. Cf.: "Sometimes people experience a genuine and pro-
found feeling on Earth. There is always a trace of zalost' in it. The more we
love, the more we perceive the object of our admiration as a victim" (B.
Pasternak, "Doktor Zivago"). V. V. Rozanov wrote about his sick wife:
"Her face was all in tears. I froze. Both in admiration and zalost'". This
philosopher believed zalost' to be one of the fundamental values of life; he
wrote: "Is there any zalost' in the world? Beauty - yes, sense - yes. But
zalost")... Do stars feel zalost'l The mother does: and let her be above the
stars".
A. Platonov usually names the feeling of maternal love by the word
zalost', cf.:
It turned out that Simon was alive because he felt his mother's zalost' for him, and he
could keep the peace of her mind just remaining safe in this world. (Platonov 1978 a:
299)
In this world, only his sister would live on, but she then would give birth to a child,
and zalost' for it would become stronger than her grief for her dead, crushed down
brother." (Platonov 1978b: 60)
Platonov in general believes that zalost' in man is a proof of his being

alive; when a person stops feeling pity, he or she might as well be dead, cf.:
The mother half opened her eyes, they were suspicious, prepared for any kind of mis-
fortune in life, and already whitish with indifference - and she said in her own de-
fense: "I don't feel sorry for you [= nine tebja ne zalko] now and I don't need any-
one -1 have become like stone; please put out the lamp and turn me on my side, I
want to die." (Platonov 1978b: 60)
A similar concept of 'love-zalost' is imprinted in the word rodnoj (lit-

erally, 'native', but closer in meaning to 'beloved')
Rodnoj
There are three ideas that form the basis of terms of endearment. Some of
them directly indicate the respective feeling: ljubimyj, zelannaja, ljubov'
moja, radost' moja, siast'e moe, nenagljadnaja ('beloved', 'wished', 'my
love', 'my joy', 'my happiness', 'the person I can't stop looking at admir-
ingly'), etc. Others express the value and uniqueness of the object of the
feeling: dorogoj, edinstvennaja, bescennyj, zolotoj, sokrovisie moe, zo-
lotce ('dear', 'the only one', 'priceless', 'golden', 'my treasure', 'my
gold'). And lastly, many words indicate the pleasant features or prettiness
of the object of the feeling (and other more sophisticated attributes): xoro-
Saja 'good', sladkij 'sweet' etc.; cf. in English: honey, sweetypie and
sweetheart. This idea is often expressed through comparison with small
and pretty animals - zajka, kiska ('bunny', 'pussy'), etc.: a creature of
small size and young age is seen as always nice.
Among the numerous Russian terms of endearment, however, there is
one which stands apart. It seems to be one of the most important and, no
doubt, the most peculiarly Russian term of address - rodnoj (literally,
'native'). It has a diminutive form rodn'enkij and a number of other de-
rivatives. The obsolete addresses rodimyj and krovinoska 'a small drop of
blood' are close to it in their meaning. The hero of Ju. Malecky comments
on the address rodnaja in the following way: "Rodn'en'kaja. Closer than of
the same [Russian] tribe, belonging to a very peculiar kin of my own rather
than to a masculine or feminine or neutral gender, she is the only one of
just-my-own kind."
The address rodnoj means: I think about you as if you were a (beloved)
blood relative. This type of address is rare even in other Slavic languages. It
differs from others by its special emotional coloring. Rodnoj, rodnaja ex-
presses profound tenderness, confidence, feeling of mutual understanding
and spiritual intimacy rather than romantic love and passion. "When this
adjective is used as a form of address outside the family, and outside the
domestic circle, its great emotional strength is also linked with the impli-
cation of belonging, a total lack of distance and the existence of ties felt to
be absolute and unseverable" (Wierzbicka 1997: 84). Although rodnoj is
related to erotica to a lesser extent than other words of love, the degree of
intimacy of that word is higher than of the words milyj (darling) and even
ljubimyj (beloved). One would hardly call a person rodnoj at the beginning
of a love affair, even when the feeling of love has been already expressed
through other words of endearment. Some people do not have the word
rodnoj in their love lexicon at all, as it seems to them shockingly open.
The Russian adjective rodnoj primarily states the existence of blood ties.
However, rodnoj is not quite the same as rodstvennik (relative); cf.:
Nastja, her daughter and the only rodnoj person, lived far away, in Leningrad (K.
Paustovskij).
If in this phrase we replace the word combination "the only rodnoj per-
son" by the combination "the only relative", the meaning will change. The
original message - human loneliness will be replaced by data from a
questionnaire. The word rodnoj has some special warmth - as if blood re-
lations were not only an objective fact, but something one feels in one's
heart:
Suddenly Nadja felt in her heart that aunt Frosja was there her only rodnoj person.
She saw the familiar shape of her fingers similar to that of her mother's and the fam-
iliar slightly high cheekbones. And she felt a sudden tenderness for aunt Frosja as
never before (Yu. Trifonov).
Just because rodnoj is more than simply a relative, this word freely com-
bines with such words as such and forms degrees of comparison (rodnee,
samyj rodnoj):
"Mother! How could it happen? I have no one else in my life who is or will be rod-
nee" (K. Paustovski).
Since the word rodnoj means primarily not "of the same blood", but
"someone organically linked to the speaker", it is widely used to describe
the relationship of people who are not blood relatives. In Russian it is poss-
ible to "become rodnoj", cf.:
Once Katja even burst into tears - though she never cried - and suddenly those tears
made her extremely rodnoj to him, made him feel acute pity and a sort of guilt with
respect to her" (I. Bunin).
322 Irina B. Levonna, Anna A. Zalizniak
Blood relationship is a great value in all cultures, being a part of the basic
opposition "own vs alien". But for the Russian culture the feeling of kin-
ship is especially important. Moreover, in Russian the power of love can
transform a stranger into a sort of blood relative: rodnoj and even rodneje
rodnogo (closer than rodnoj}. The feeling of unity with the beloved person
can be so strong that this unity seems to be as indissoluble as blood ties.
Naturally, this feeling can be expressed, one way or another, in any other
language, but the Russian language has canonised it through providing a
special word for it and putting this word into the mouths of people speaking
of love.
The metaphor of blood relations is widely used in Russian colloquial and
common forms of addressing strangers. In other than Slavic languages one
can hardly find such an abundance of this type of address forms: otec, pa-
pasa, mat', mamasa, synok, doka, sestrenka, brat, bratok, bratcy, tetka,
djadja, ded, babuSka, babulja, vnudka ('father', 'dad', 'mother', 'mum',
'son', 'daughter', 'little sister', 'littlebrother', 'brother', 'brothers', 'aunt',
'uncle', 'grandfather', 'grandmother', 'granny', 'my dear granny', 'grand-
daughter') (cf. Lixaoev 1987; Kronhaus 1996). The form of address is de-
termined by the speaker who correlates his or her own age with that of the
addressee and figures out what relative that person could be for him or her.
It should be noted that religious addresses brat (brother) and sestra (sister)
are different in principle: they have nothing personal in them and just indi-
cate equality of all people before God.
So, for the Russian culture kinship is a strong value that has a strong
emotional coloring to it. Love for one's own brethren however does not
imply indifference or hostility to others. On the contrary, the warmth of
familial or congenial relations serves as an example of a kind attitude to
people in general. The linguistic facts that have been discussed here seem
to confirm the traditional conception of "breadth" (= Sirota) of the "Rus-
sian soul", its ability to embrace the entire world with a kind of parental,
fraternal or filial love.
4. What do Russians tend to feel when they part from their

beloved?
Russian has at its disposal a number of words describing feelings that
people experience when separated from the ones who are dear to them:
skuiat', toskovat', nexvatat', nedostavat', stoskovat'sja, istoskovat'sja,

soskuit'sja ('to miss', 'to long for', 'to miss somebody', 'miss badly',
'yearn for', 'to start missing'). The most peculiar combination of ideas can
be found in the word soskudit'sja 'to start missing'. To designate the state
of separation with someone dear Russian uses the word razluka, which is
also language-specific and hard to translate. In this section we will discuss
two specific Russian words: soskuit'sja and razluka.
The phrase {someone} soskudilsja means that this person has begun
missing someone from whom he or she is separated, the feeling becoming
ever more acute with the passage of time until the next meeting takes place.
While razluka (the state of being in separation) starts at the moment of sep-
aration in space, a person can soskuiit'sja only after some time has passed
since the separation.
How much time is to elapse before a person can soskuit'sjal As a rule,
several days, weeks or months. When people are in close contact (for
example, lovers or a mother and a child), it could be hours. Subjectively,
people close to each other perceive this time interval as a long one; cf.:
From school Nikitin would go to his coaching and when he at last reached home
about half past five he would feel as happy and excited as if he had been away a whole
year. He would run panting upstairs, find Masha, take her in his arms, kiss her, swear
he loved her and couldn't live without her, claiming to have missed [=soskuoilsja] her
terribly, asking in panic whether she was well and why she looked so solemn. (Che-
khov 1984:197)
The word soskuiit'sja implies that the emotion felt becomes ever
stronger, the longer the separation; it presupposes a certain "accumulation"
of feeling. It is correct to say "In a fortnight he soskuiilsja even more." On
the contrary, the sentence *Snaala on soskuailsja, no ierez dve nedeli os-
voilsja i perestal skudat' (At first he soskuftlsja, but in a fortnight he got
accustomed to the situation and this feeling disappeared) sounds incor-
rect - although in real life such a situation would be quite normal. For in-
stance, if a young man has left for another city to study there it will be cor-
rect to say: "At first he was missing home a lot, but then he becomes com-
pletely carried away with student life, and he absolutely forgot his home
and mother". But in this case we must use the verb skuat', and not sosku-
Zit'sja because soskuit'sja doesn't allow for the possibility of reverse
development of the feeling: it presupposes that the longing will persist till
the desired resumption of the contact.
Another important feature of the word soskuit'sja which distinguishes it

from the word razluka 'the state of separation' and from skuat' 'to miss
somebody' is its peculiar emotional colouring. It would hardly be an exag-
geration to say that the word soskuit'sja describes a positive emotion: it is
constituted by happy and impatient expectation, an anticipation of the joy
of meeting. (The same idea of pleasant anticipation can also be found in the
meaning of the Russian word progolodats'ja 'to get hungry', which is
similar to soskuiit'sja in its inner form as well. The state described by the
verb progolodat'sja - contrary to feeling real hunger - is, to some extent, a
pleasant state, because it promises pleasure in the near future.) This word
would be quite appropriate in a situation when the pleasant reunion is still
distant, but it will be more appropriate when the period of separation is
nearly over and waiting for someone dear becomes a pleasure rather than a
torture, cf.:
Dymov bought snacks and candies and went to his wife who lived in their country-
side cottage. He had not seen her for a fortnight and started missing [= soskuiilsja]
her badly. When sitting in the railway carriage and later when he wandered through
the large grove trying to find their cottage, he felt hungry and tried to imagine how
he would leisurely dine with his wife and then flop into bed. And it did his heart
good to look at his package with caviar, cheese and white fish. (Chekhov, "The
Fidget")
Moreover, a person often can blissfully keep saying "I soskuiilsja so

much!" when the separation is already over. This exclamation expresses
first and foremost the joy of the long-awaited reunion, while the phrase "I
skual so much!" sounds rather as a complaint (stating that the period of
separation was very unpleasant) and when repeated several times can be
boring.
So, in Russian there is a word describing the emotions of a person sep-
arated from the one who is dear to him or her in an optimistic rather than
gloomy way. It is significant that this word not only exists in the language,
but it is frequently used in everyday speech. This fact runs counter to the
widespread idea that the vast expanses of Russia impose an imprint of
gloom upon the life of the Russians; for example, "Russians are almost un-
able to feel joy. ...The Russian soul is suppressed by the vast Russian fields
and abundance of Russian snow, it gets drowned and dissolved in that vast-
ness", from an essay by N. Berdjaev "On the power of space over the Rus-
sian soul". Such a picture is more in line with another very peculiar Rus-
sian word toska 'anguish, yearning', see section 2.2.
Whereas the verb soskuit''sja describes a certain inner state of a person,

the noun razluka, at first sight, refers to certain external circumstances (i.e.
separation in space). Nevertheless, its emotional colouring is even stronger
than that of the word soskucit'sja. Although these words differ stylistically
(soskuit'sja is an everyday word while razluka is rather poetic) and in
terms of mood (soskuif sja is optimistic, razluka is sad), they have much
in common8.
Like soskuit'sja, razluka has no precise equivalents in West European
languages (cf. Beyer 1995). The main difference between the Russian
word razluka and its equivalents in the dictionaries (e.g. English separ-
ation) consists in the fact that one can be in razluka only with somebody or
something close, dear (a beloved person, as well as home, fatherland, etc.).
Similarly, one can soskudt'sja only of somebody/something very dear and
close. Therefore, if the relations between two persons are not intimate
enough, utterances like "I soskuilsja" or "We live in razluka" would be
inappropriate, because they presuppose more intimate relations than there
really are. In this respect razluka is even more exigent than soskuit'sja:
razluka presupposes love, while for soskucit'sja a pleasant habit would be
sufficient. One can soskucit'sja 'start missing' one's computer or rye
bread, but it would be an exaggeration to speak of razluka in this situation.
If we use the word razluka with respect to the relations between two
friends, we mean a special type of romantic friendship, which is not the
case when we use the word soskuit'sja. "Razluka with a favorite cat"
sounds ironical, whereas the word soskuit'sja is quite appropriate in this
situation. Compared to soskudt'sja, the word razluka is much more dra-
matic. A person can hardly soskucit'sja so badly as to go crazy or commit
suicide, while it is possible to die in razluka (cf. ne vynes razluki 'could not
withstand razluka' - or Axmatova's Kljalas', ito pogibnes" v razluke 'You
swore you would perish in razluka').
Razluka presupposes that for the two given persons it is normal to be to-
gether (cf. the strange phrase "to live in razluka from her niece"). Lovers
or parents with young children are the people for whom the constant con-
tact is habitual and its cessation is painful. That is why the word razluka is
associated mainly with these two prototype situations, cf.:
"Maria Nikolaevna, I love my bride, and the razluka with her is not so easy for me."
(I. S. Turgenev)
"I am unhappy because of razluka with my son. I beg you to allow me to see him just
one time before I leave." (L. N. Tolstoy, "Anna Karenina")
Anna Karenina has become in Russian culture a symbol of a mother suf-

fering from razluka with her child; cf.:
Ty tak ze vual' opravljaes' v prixozej.
Ty Anna nad spjaSiim Serezej - razluka
(M. Tsvetaeva, "Razluka")
Like her, you adjust your veil in the hall,

razluka, you are Anna bent over sleeping Seriozha!
So, razluka is a particular state that has both internal and external as-
pects. To be in razluka means to be separated physically but remain to-
gether emotionally. This idea distinguishes the Russian razluka from ras-
stavanie as well as of the English separation, French separation, German
Trennung, Italian separazione etc. All these words can designate both
physical and emotional distancing in various combination, but they lack
that sense of contradiction between the internal and external aspects which
is inherent in the Russian word razluka. Moreover, the word separation
(and its equivalents in other West European languages) is used as a legal
term designating a particular marital status: a state normally preceding di-
vorce, when the spouses live separately. This kind of separation can take
place even within the same flat. In this case the word separation (German
Trennung etc.) becomes an opposite to Russian razluka, because it desig-
nates a situation when two persons are physically together while emotion-
ally separated from each other.
Normally, razluka is a temporary absence of contact. This word wouldn't
be used in the situation where no reunion is planned. It would be strange to
say about a person who is leaving for good (without the possibility of re-
turn, as it was e.g. for Soviet emigration in the '70s), that he is facing raz-
luka from his relatives. It would be as strange as to speak of making a stop
when one has no intention to travel onward. When lovers think of their sep-
aration in terms of razluka, they hope that their feelings will withstand the
razluka, that the beloved person will not stop loving and waiting. In the
Russian lexicon of love there is a special term for it: dozdat'sjalne
dozdat'sja ('not stop/stop loving', literally, a resultative mode of action
from the verb meaning 'to wait'), cf.:
Razluka bystro proneslas'
Ona menja ne dozdalas',
No ej proSoaju
(V. Vysotskij. "Toj no'ju ja ne pil ne el")
Razluka passed quickly,

She did not wait for me [ = loved me no more, had another lover],
But I forgive it...
There is one more semantic feature of razluka that distinguishes this

word from rasstavanie and its equivalents in West European languages
(separation, etc.): the idea of independence from the human will. This is a
rather characteristic element of the Russian conceptualisation of the world
that can be found in a large number of words and impersonal constructions
(see Wierzbicka 1992 a: 413-417; Zalizniak and Levontina 1996). Razluka
is not always caused by an intervention of external forces: as when a man is
drafted to serve in the army or sent to jail, etc. Even when a person leaves
voluntarily, the word razluka presents the situation in such a way as if the
separation has taken place by itself, following a certain course of events
and in a sense independent of the human will.
Feelings experienced in razluka are one of the best explored domains in
Russian culture. Razluka is the main theme of a great number of Russian
songs and poems. For the Russian lyrical consciousness razluka is the re-
verse side of love. Love and razluka are a standard poetic couple, though
their disposition can be different, ranging from cheerful, "Love in the
world/is stronger than separation [= razluka]" in the Soviet soldier's song
"Don't Cry, Girl!", to desperate.
Ljubov' sil'nej razluki, no razluka
dlinnej ljubvi
(I. Brodskij, "Dvadcat' sonetov k Marii Stjuart")
Love is more powerful than razluka,
the latter is more lasting
(I. Brodsky, "Twenty sonnets to Mary Queen of Scots", translated by Peter France
and the author)
Ljubov' - luk
Natjanutyj - luk: razluka
(M. Tsvetaeva. "Poema konca")
Love - that means a bow/The drawn bow: razluka
(M. Tsvetaeva, "The Poem of the End")
The words razluka, razlu it' derive from the root luk and are etymologi-
cally connected with the word luk 'bow' (see Fasmer 1996, II: 537);
cf. words with other prefixes: otlucit' (ot cerkvi) 'excommuniate', otlu-
cit'sja 'to be absent', slucit', slucka 'coupling', zaluiit' [kogo-to k sehe v

gosti] 'get someone to come to one's place'.
In Russian poetry razluka is a value in itself, being not only an obstacle
to love, but a form of love, and, in a sense more "sublime" form (cf. sec-
tion 2.1.). The Russian poet of razluka par excellence is Josef Brodsky. He
found a concrete image for it when he placed himself and his beloved at the
opposite coasts of the ocean and showed a way to meet each other - to ob-
serve a non-existing star located at the line perpendicular to the section
joining the two points, the two lovers as seen from celestial height.
Razluka
est' summa nasix trex uglov, a vyzvannaja eju muka
est' forma tjagoten'ja ix
drug k drug u.
So separation
means our three sharp angels' sum,
while all its agony (let's rationalize)
is just a product of
their mutual gravitation
(I. Brodsky. "A Song to No Music", translated by David Rigsbee and the author)
All that geometry and physics were invented by Brodsky when he lived
in Russia. Then his beloved was an American. But when Brodsky crossed
the ocean, he retained this happily found lyrical construction and started
addressing it to a Russian love. Having changed his address, he changed
also the addressee, so that both could be, as before, separated by the ocean.
His life became a realisation of his own metaphor.
5. Conclusion
The Russian language creates language-specific ways of viewing emo-
tional states first of all owing to the words which contain unique (i.e. absent
in other languages in a ready-made form) conceptual configurations, the
analysis of which was the subject of the present article. However, a no less
important part of the picture of the world seen through the Russian lan-
guage can be found in the unique conceptual configurations contained in
the meaning of syntactic constructions.
The Russian language possesses a formal device which makes it possible
to express idiomatically the connection between the inner states of subjects
and their actions, namely, a construction with an infinitive subordinate to

an inner-state predicate. One of the most characteristic of these is the con-
struction with a personal verb in the perfective, the meaning of which may
be described by the general formula didn't do so because (of an inner
state)': postesnjalsja sprosit' (lit. 'was too embarrassed to as\i'),polenilsja
sxodit' (lit. 'was too lazy to go'),pozalel otdat' (lit. 'was too attached to it
to give it away') etc. The meaning of this construction type cannot be con-
veyed in Western European languages except by explanatory paraphrase.
The Russian impersonal constructions have also been studied exten-
sively. A. Wierzbicka has made a study of a number of them, relating their
meanings to more general features of Russian which they embody
(Wierzbicka 1992 a). The reflection of the "Russian soul" in the syntax and
morphology of the Russian language is a subject for separate study; here
we shall mention only one construction where, thanks to the combined ef-
fects of all levels of the language, a unique conceptual configuration is cre-
ated which embodies a characteristic Russian attitude to the action itself. It
is the combination of an impersonal predicate with a subject in the dative
and a subordinate infinitive:
(3) nam bylo zalko rasstavat'sja.
to.us it.was sorry(ADv) to.part (INF. IMPERF.)
'We were sorry to be parting.'
Cf. also: mne obidno slysat' 'that sounds offensive to me', grex tebe
zalovat'sja 'you ought not to complain', slusat' tebja boleznenno 'it is
painful to listen to you', etc. This construction provides the speaker of Rus-
sian with a ready-made form to convey a certain culturally salient type of
meaning, and from this point of view, the productivity of this construction
in Russian is not surprising.
Briefly, the nature of the conceptual configuration contained in this con-
struction is as follows. The subject here is in the semantic role of "recipi-
ent". He responds to a state of affairs produced by someone else: he does
something, but does not take responsibility for it. Whereas in the nomi-
native model the subject acts himself and takes responsibility for his ac-
tions, and in the accusative one he does not take responsibility, nor does he
do anything either, but is merely subjected to something that happens to
him, this dative-subject construction expresses a situation dear to the "Rus-
sian soul": I am doing something, but, at the same time it is kak by ('as if)
it were not me (on the Russian kak by see Arutjunova 1996). As a Russian
proverb says: M o ja xata s kraju, ja nicego ne znaju (and, we can add, ni

zacto ne otveaju) 'My house is on the edge of the village, I don't know
anything' (and am not responsible for anything'). The subject is on the
sidelines, as it were, on the edge of what is taking place, outside the sphere
of responsibility for it (cf. the feature "peripheral" assigned to the dative
case in Jakobson 1971).
Thus in Russian syntax the desire to renounce responsibility for one's
own actions acquires, if not a justification, at least a codified means of ex-
pression.
Notes
1. This work was supported by the Research Support Scheme of the Open Society In-
stitute, grant # 797/1997. We express gratitude to those who have assisted with this
paper: Konstantin Bogatyrev, Jean Harkins, Steven Hassman, Alexej Smelev, Eliot
M. Tretter, Anna Wierzbicka, Andrej Zalizniak.
2. Some interesting conclusions concerning the present state of the Russian linguistic
consciousness have been obtained by processing data from the new Associative Dic-
tionary of Russian (see Ufimceva 1996). On the basis of the frequency with which a
word occurs as a response to a given verbal stimulus the author comes to a con-
clusion about the relative significance of this or that concept for the Russian lin-
guistic consciousness. The results obtained by N. V. Ufimceva confirm both certain
common ideas about the "Russian soul" and also the results of linguistic analysis by
other authors. In particular, words with the highest rate (and higher than the corre-
sponding words in English) in modern Russian include: radost' 'joy', stast'e 'hap-
piness', drug 'friend' and razgovor 'conversation'.
3. Cf. the explanation given in Dictionnaire dufrangais contemporain, by J. Dubois et
al. (Paris: Larousse, 1971): "Seditd'unetreanimedontlesdesirs, les gouts sontsat-
isfait, qui a ce qui lui plait". And the difference between rad and dovolen in general
corresponds to the opposition radost' - udovoljstvie, which we have considered
here, both in their syntactic properties and the presence of a mental component
(cf. Wierzbicka 1988: 312, Zalizniak 1992: 41).
4. Etymologically both words go back to a root meaning 'empty'. It is noteworthy that
the Russian v/ordpustoj 'empty' reproduces the same semantic derivation 'empty'
-'vain', as in: pustye ugrozy 'empty threats',pustaja trata vremeni 'time-wasting'
vpustuju 'in vain'.
5. Lexical function, a notion introduced by I. Mel'Cuk (e.g. Mel'cSuk 1995: 26), associ-
ates with a given lexical item a lexeme or phrase which expresses a certain general
meaning, such as 'very' or 'great' (LF Magn), 'begin' (LF Incep) etc.
6. The differences between obida and oskorblenie (note that the latter doesn't describe
an emotional feeling) consist in the fact that oskorblenie is caused when socially de-
termined norms are violated, whereas obida arises on individual, personal grounds:
the set ofoskorbitel'nye (causing oskorblenie) utterances is a part of the cultural text
of the given social and cultural community, whereas the set of obidnye (causing
obida) utterances cannot be determined. One can say also that oskorblenie arises
when honour is involved, in obida - feelings. The Russian language reacts more
sharply to the latter (cf. Ufimceva 1996 on the relatively slight significance of the
concept of cest' 'honour' for the Russian linguistic consciousness).
Cf. the famous translation by Catullus (Carmen LI):
Hie mi par esse deo videtur
llle, si fas est, superare divos
Qui sedens adversus identidem te
Spectat et audit
There is a third item here - the word skuCat' mentioned above. We do not treat it in
detail here, as it is less specific than soskuiit'sja and razluka (cf. English to miss,
French s'ennuyer [de quelquun]. This word is related to the word soskuiit'sja, but
its tone renders it fairly compatible with the word razluka. Expression skuat' v raz-
luke 'to miss somebody in separation' is a very frequent word combination in Rus-
sian. These two words could be similarly explained through each other: one can say
that razluka is such a type of separation when people skuajut 'miss each other',
and, vice versa, skuat' means to have such a feeling which one normally experi-
ences in razluka. If a person does not skucaet v razluke, that means there is no raz-
luka at all.
References
Apresjan, Jurij D.
1995 Obraz Celoveka po dannym jazyka [The model of a person based on
linguistic evidence]. In Jurij D. Apresjan, Izbrannye trudy, vol. II,
348-388. Moscow: Jazyki russkoj kul'tury.
1997 a Radovat'sja, likovat', torzestvovat' [dictionary entries for Russian
verbs comparable to rejoice]. Novyj ob"jasnitel'nyj slovar' sinonimov
russkogo jazyka, 312-315. Moscow: Skola Jazyki Russkoj Kul'tury.
1997 b Stydit'sja, smu$at'sja, stesnjat'sja, konfuzit'sja [dictionary entries for
Russian words comparable to shame]. Novyj ob"jasnitel'nyj slovar'
sinonimov russkogo jazyka, 417-423. Moscow: Skola Jazyki Russkoj
Kul'tury.
Apresjan, Valentina
1995 Emotions: Linguistic representation and cultural conceptualization.
PhD dissertation. Los Angeles, USC.
Apresjan, Valentina and Jurij D. Apresjan
1995 Metafora v semantieskom predstavlenii emocij [Metaphor in the
semantic representation of emotions]. In: Jurij D. Apresjan, Izbrannye
trudy, vol. II, 453-465. Moscow: Jazyki russkoj kul'tury.
332 Irma B. Levonna, Anna A. Zalizniak
Arutjunova, N. D.
1996 Stil' Dostoevskogo v ramke russkoj kartiny mira [The style of Dos-
toevsky and the Russian model of the world]. Poetika, Stilistika, Jazyk
i kul'tura. Pamjati T. G. Vinokur, 61-90.
1976 Predloienie i ego smysl [The sentence and its meaning]. Moscow:
Nauka.
1997 sty de i stuzhe [ On the relatedness of the Russian words for shame
and cold]. Voprosy jazykoznanija 2: 59-70.
1999 Ponjatie sty da i sovesti v tekstax Dostoevskogo [The concepts of
shame and conscience in the works of Dostoevsky]. In N. D. Arutju-
nova and I. B. Levontina (eds.) Logieskij analizjazyka. Obraz ielove-
ka vjazyke i kul'ture, 320-345. Moscow: Indrik.
Beyer, Th. R., Jr.
1995 Marina Tsvetaeva and Andrej Belyj: Razluka and Posle razluki.
Wiener zionistischer Almanack 35: 97-132.
Baranczak, Stanislaw
1990 Breathing under water and other East European Essays. Cambridge,
MA: Harvard University Press.
Cernyx, P. Ja.
1994 Istoriko-etimologiieskij slovar' sovremennogo russkogo jazyka [His-
torico-etymological Dictionary of Contemporary Russian], vols. 1-2.
Moscow: Russkij jazyk.
Chekhov, Anton
1984 The Russian Master. Oxford: Oxford University Press.
Dostoevsky, Feodor M.
1983 The Brothers Karamazov Translated by David McDuff. Harmonds-
worth: Penguin.
Fasmer, M.
1996 Etimologioeskij slovar' russkogo jazyka [Etymological Dictionary of
the Russian Language], vols. 1-4. St. Petersburg: Azbuka.
lordanskaja, Lidija
1970 Popytka leksikografieskogo tolkovanija gruppy russkix slov so
znaceniem Cuvstva [An attempt at a lexicographic analysis of a group
of Russian emotion terms]. MaSinnyjperevod iprikladnaja lingvistika
13: 3-26.
1972 Leksikografioeskoe opisanie russkix vyrazenij, oboznaajuScix fizie-
skie simptomy cuvstv [A lexicographic description of Russian ex-
pressions for physical symptoms of emotion terms]. Masinnyjperevod
iprikladnaja lingvistika 16: 3-30.
1984 Slovarnye stat'i bojat'sja, vostorg, vosxiSoat', gnev, strax. Tolkovo-
kombinatornyj slovar' russkogo jazyka [Dictionary entries for Russian
words comparable to fear, delight and anger]. Wiener Slavistischer Al-
manach 14.
lordanskaja, Lidija and Igor Mel'cuk

1990 Semantics of two emotion verbs in Russian: bojat'sja 'to be afraid' and
nadejat'sja 'to hope'. Australian Journal of Linguistics 10(2):
307-357.
Izard, Carroll E.
Jakobson, Roman
1971 [1958] Morfologiceskie nabljudenija nad slavjanskim skloneniem [Morpho-
logical observations concerning Slavic declension]. In: Roman Jakob-
son, Selected Writings, vol. 2, 154-183. The Hague/Paris: Mouton.
Johnson-Laird, Philip N. and Keith Oatley
1989 The Language of Emotions: an Analysis of a Semantic Field. Cogni-
tion and Emotion 3 (2): 81-123.
Kronhaus, M.
1996 Russian forms of address in communication and culture. International
Journal of Psycholinguistics 13(1): 67-79.
Levontina, Irina B.
1997 a Zalost', soCuvstvie, sostradanie, uoastie [Dictionary entries for Rus-
sian words comparable to compassion]. Novyj ob"jasnitel'nyj slovar'
sinonimov russkogo jazyka, 107-112. Moscow: Skola Jazyki Russkoj
Kul'tury.
1997 b Milyj, dorogoj, ljubimyj [The Russian words comparable to dear and
beloved}. Russkaja rec' 5: 55-62.
1999 Homo piger. In: N. D. Arutjunova and I. B. Levontina (eds.), Logie-
skij analiz jazyka. Obraz celoveka vjazyke i kul'ture, 105-113.
Levontina, Irina B. and A. D. Smelev
1996 a Russkoe "zaodno" kak vyrazenie ziznennoj pozicii [The Russian ex-
pression zaodno as a manifestation of an attitude to life]. Russkaja rec'
2: 53-57.
1996b "Poprecnyj kus". Russkaja rec' 5: 55-60.
Lixaev, D. S.
1987 Zametki russkom [Remarks on "Russianness"]. In: D. S. LixaCev. 7z-
brannye raboty, vol. 2, 418-494. Moscow:
Lotman, Ju. M.
1994 Besedy o russkoj kul'ture [Conversation about Russian culture]. St. Pe-
tersburg: Iskusstvo-SPB.
Lotman, Ju. M. and B. A. Uspenskij
1994 Rol' dual'nyx modelej v dinamike russkoj kul'tury (do konca XVIII
veka) [The role of dual models in the dynamics of Russian culture]. In
B. A. Uspenskij, Izbrannye trudy, vol. 1, 219-253. Moscow: Jazyki
russkoj kul'tury.
Mel'cuk, Igor.
1995 The Russian language in the meaning - text perspective. Wiener Sla-
wistischer Almanack 39.
Mel'uk, Igor and L. Wanner

1996 Lexical functions and lexical inheritance for emotion lexemes in Ger-
man. In: L. Wanner (ed.), Lexical Functions in Lexicography and
Natural Language Processing, 209-278. (Studies in Language Com-
panion Series, vol. 31.) Amsterdam: John Benjamins.
Nabokov, Vladimir
1981 Lectures on Russian Literature. New York: Harcourt Brace Jovan-
ovich.
Papemo, I.
1996 Semiotika povedenija: Nikolaj Cernyaevskij - oelovek epoxi realizma
[The semiotics of behaviour - Nikolaj CernySevskij as a represen-
tative of the epoch of realism]. Moscow: Novoje Literaturnoje oboz-
renije.
Pen'kovskij, A. B.
1991 Radost' i udovol'stvie v predstavlenii russkogo jazyka [Attitudes to
joy and pleasure reflected in the Russian language]. Logiieskij analiz
jazyka. Kul'turnye koncepty, 148-155.
Pesmen, Dale
2000 Russia and Soul: An Exploration. Ithaca, NY: Cornell University Press.
Platonov, Andrei.
1978 a Chevengur. Translated by Anthony Olcott. Ann Arbor: Ardis.
1978 b Collected Works. Translated by Thomas P. Whitney. Ann Arbor: Ardis.
Plungjan, V. A. and E. V. Raxilina
1993 "Madness" as a Lexicographic Problem. Logiceskij analiz jazyka.
Mental'nyje dejstvija, 120-127.
1996 "S chisto russkoj akkuratnost'ju ..." [A Russian expression meaning
"with a purely Russian precision"]. Moskovskij lingvistieskij Zurnal,
vol.2, 340-351.
Pushkin, Alexander
1936 The Poems, Prose and Plays of Alexander Pushkin. Translated by
A. F. B. Clark. New York: Random House.
1975 Eugene Onegin. Translated from the Russian, with a commentary, by
Vladimir Nabokov. Princeton: Princeton University Press.
Satunovskij, I. B.
1996 Semantika predlozenija i nereferentnye slova [Sentence semantics and
non-referential words]. Moscow: Jazyki russkoj kul'tury.
Simcuk, E. G.
1991 Iz istorii leksiki, svjazannoj s duxovnym mirom eloveka (dr.-russk.
toska i ego okruzenie). Istoriko-kul'turnyj aspekt leksikologioeskogo
opisanija russkogo jazyka, 75-85.
Smelev, A. D.
1997 a Leksioeskij sostav russkogo jazyka kak otrazenie "russkoj dui" [The
Russian lexicon as a mirror of the "Russian soul"]. In: T. V. Bulygina
and A. D. Smelev (eds.), Jazykovaja konceptualizacija mira (na ma-

teriale russkoj grammatiki), 481-495. Moscow: Jazyki russkoj
kul'tury.
1997 b Dux, dua i telo v svete dannyx russkogo jazyka [Spirit, soul and body
in the light of the Russian language]. In: T. V. Bulygina and A. D. Sme-
lev (eds.), Jazykovaja konceptualizacija mira (na materiale russkoj
grammatiki), 523-539. Moscow: Jazyki russkoj kul'tury.
Stepanov, Ju. S.
1997 Konstanty. Slovar' russkoj kul'tury [A Dictionary of Russian Culture].
Moscow: Jazyki russkoj kul'tury.
Tolstoj, N. I.
1995 Jazyk i narodnaja kul'tura. Ocerki po slavjanskoj mifologii i tnoling-
vistike [The language and national culture: Essays on Slavic mythol-
ogy and ethnolingustics]. Moscow: Jazyki russkoj kul'tury.
Ufimceva, N. V.
1996 Russkie: opyt eSfe odnogo samopoznanija. Etnokul'turnaja specifika
jazykovogo soznanija, 139-162.
Uryson, E. V.
1995 Fundamental'nye sposobnosti Celoveka i naivnaja "anatomija" [Basic
human abilities and "naive" anatomy]. Voprosy jazykoznanija 3: 3-16.
1997 Toska, unynie, peCal', grust' [Dictionary entries for words comparable
to sadness]. Novyj ob" jasnitel'nyj slovar' sinonimov russkogo jazyka,
441-445. Moscow: Skola Jazyki Russkoj Kul'tury.
Vinogradov, Viktor V.
1994 Istorija slov [A history of words]. Moscow: Tolk.
Wierzbicka, Anna
1972 Semantic Primitives. Frankfurt: Athenum.
1988 The Semantics of Grammar. Amsterdam: John Benjamins.
1990 DuSa 'soul', toska 'yearning', sud'ba 'fate: Three key concepts in Rus-
sian language and Russian culture. In: Zygmunt Saloni (ed.) Metody
formalne w opisie je_zykow stowianskich [Formal methods in the de-
scription of Slavic languages]. Bialystok: Bialystok University Press.
1992 a Semantics, Culture, and Cognition. Universal Human Concepts in
1992 b Defining emotion concepts. Cognitive Science 16: 539-581.
1992c Talking about emotions: semantics, culture, and cognition. Cognition
and Emotion (special issue on basic emotions) 6 (3/4): 285-319.
1997 Undestanding Cultures through their Key Words. New York: Oxford
University Press
Zalizniak, Anna A.
1983 Semantika glagola bojat'sja v russkom jazyke [The semantics of the
Russian verb bojat'sja ('to fear')]. Isvestija AN SSSR, ser. lit. i
jaz. 42(1): 59-66.
1991 De la nature du regret. Bulletin de linguistique generate et appli-

quee 17: 61-75.
1992 Investigations in the semantics of inner state predicates. (Slavistische
Beitrge, vol. 298.) Munich: Otto Sagner Verlag.
Zalizniak, Anna A. and Irina B. Levontina
1996 Otrazenie nacional'nogo xaraktera v leksike russkogo jazyka. (Raz-
myslenija po povodu knigi: Anna Wierzbicka. Semantics, Culture and
Cognition) [The Russian lexicon as a reflection of the Russian national
character. (Thoughts arising from reading Anna Wierzbicka's book Se-
mantics, Culture and Cognition)]. Russian Linguistics 20: 237-246.
1999 S ljubimymi ne rasstavajtes' ... [The Russian phrase meaning 'Don't
part from your loved ones']. In: N. D Arutjunova and I. B Levontina
(eds.), Logioeskij analiz jazyka. Obraz ieloveka v jazyke i kul'ture,
52-63.
A culturally salient Polish emotion: Przykro
(pron. pshickro)
Anna Wierzbicka
1. Introduction: emotions, words, and cultures

Languages differ widely in the nature of their vocabularies. Distinctions which seem
inevitable to us may be utterly ignored in languages which reflect an entirely differ-
ent type of culture, while these in turn insist on distinctions which are all but unin-
telligible to us.
Such differences of vocabulary go far beyond the names of cultural objects such as
arrow point, coat of armor, or gunboat. They apply just as well to the mental world.
(Sapir 1949: 27)
Despite Sapir's warning, many scholars continue to search for universals

on the basis of language-specific (usually English) concepts which have no
exact equivalent in other languages. The field of emotions well illustrates
the trap involved in the attempt to reach for human universals on the basis
of one's native language alone. As I pointed out in my earlier publications
on the subject (cf. Wierzbicka 1997: 9), a typical debate - often, between a
psychologist ("P") and a linguist ("L") - runs as follows:
P: Fear and anger are universal human emotions.
L: Fear and anger are English words, which don't have exact equiv-
alents in all other languages. Why should these English words -
rather than words from language X, for instance, for which English
has no equivalents - be peculiarly well equipped to encapsulate
certain emotional universals?
P: It doesn't matter whether other languages have words for fear and
anger or not. Let's not deify words. I am talking about emotions,
not about words.
L: Yes, but in talking about these emotions you are using culture-spe-
cific English words, and thus you are introducing an Anglo per-
spective on emotions into your discussion.
338 Anna Wierzbicka
P: I don't think so. I am sure that people in those other cultures also ex-
perience fear and anger, even if they don't have words for them.
L: Maybe they do experience fear and anger, but their categorisation
of emotions is different from that reflected in the English lexicon.
Why should the English taxonomy of emotions be a better guide to
emotional universale than that embodied in some other language?
P: Let's not exaggerate the importance of language.
But, by neglecting the semantic differences between words from differ-
ent languages, scholars who take this position end up doing precisely what
they wished to avoid, that is, "deifying" some words from their own native
language and reifying culture-specific concepts which are encapsulated in
them. Thus, unwittingly, they illustrate once again how powerful the grip
of our native language on our thinking habits can be.
As William James noted a hundred years ago, we know from introspec-
tion that, on the one hand, we are capable of a great variety of feelings, and
on the other, that these different feelings are not clearly separated from one
another and can not be counted. Furthermore, as James also noted, every
language imposes its own interpretive grid upon this largely nebulous
world of feelings.
... if one should seek to name each particular one of [the emotions] of which the
human heart is the seat, it is plain that the limit to their number would lie in the in-
trospective vocabulary of the seeker, each race of men having found names for some
shade of feeling which other races have left undiscriminated. If we should seek to
break the emotions, thus enumerated, into groups, according to their affinities, it is
again plain that all sorts of groupings would be possible, according as we chose this
character or that as a basis, and that all groupings would be equally real and true.
(James 1890:485)
Thus, the way people interpret their own emotions depends, to some ex-
tent at least, on the lexical grid provided by their native language. Trying to
develop these ideas in an earlier study I wrote:
... the set of emotion terms available in any given language is unique and reflects a
culture's unique perspective on people's ways of feeling. It also reflects the links be-
tween feelings, cognition, moral norms, and social interaction.
To draw on my personal experience (as a bilingual and bicultural person), I would
say that I tend to perceive my daily emotions in terms of lexical categories provided
by Polish (my native language), such as those illustrated in the following sentences:
A culturally salient Polish emotion: Przykro 339
Przykro mi. 'To me, (it is) unpleasant/hurtful.'

Denerwuj? sig. am making myself upset/nervous/on edge.'
Jestem zachwycona. am euphoric with admiration/delight.'
Jestem zta (lit. am bad') am displeased/angry/furious.'
WSciekla jestem. am enraged/furious/mad.'
Strasznie sig cieszg, ze... am rejoicing terribly that...'
Mam do ciebie zal. (lit. have regret to you') hold something against you, and I am
sad/sorry.'
Tesknig do ciebie. miss you/I am homesick for you/I ache for you.'
Bardzo mnie to boli. 'This hurts me very much, I ache because of this.'
Serce mi sie_ kraje. 'My heart is cutting itself.'
None of the categories illustrated in these sentences has exact equivalents in English,
and the sentences themselves cannot be adequately translated into English.
(Wierzbicka 1994: 135)
This paper is devoted to the first and perhaps the most salient of all these
culture-specific emotion concepts.
To give the reader some idea of how salient the concept of przykro is in
Polish, we can compare the frequency of przykro with that ofsmutno 'sad'
given in a Polish frequency dictionary (Kurcz et al. 1990) based on a cor-
pus of 500,000 words. POT przykro the figure is 15, and for smutno, 8. If we
include the figures for the nouns przykrofo and smutek (8 and 11), and for
the adjectives przykry and smutny (23 and 27), the joint frequencies of the
two groups will be the same (46). Thus, although "sadness" is supposedly a
"basic human emotion", innate and universal, the concept of przykro ap-
pears to be equally, if not more salient in Polish culture. Figures of this
kind, based on a relatively small corpus of half a million words, should not
be pushed too far, nonetheless they do give us some idea of the salience of
the emotion concept przykro in Polish culture.
2. Przykro
The adverb przykro is one of the most important Polish emotion words. It is
normally used in a dative construction, as in the following sentence quoted
in the Dictionary of the Polish Language (Slownik J^zyka Polskiego
1958-69):
340 Anna Wierzbicka
(1) Przykro mi, ie mimo woli obrazitem kobiet%. (Orzeszkowa)1

'I'm sorry that I accidentally offended that woman.'
(lit. '[It is] przykro[A.O\] to me')
The phrase przykro mi (literally, 'it is painful to me') could be loosely

translated into English in a number of ways, such as am sorry', feel
bad', have an unpleasant feeling' or have a painful feeling'. None of
these would-be translations, however, captures the concept conveyed by
the Polish phrase. As a first approximation, we could say that przykro con-
veys a painful feeling caused by a person; and that to that extent it is similar
to the English word hurt. Indeed, English sentences with the word hurt
could often be translated into Polish with the help of the word przykro: in
most contexts, I felt hurt would be naturally translated as Bylo mi przykro.
But hurt implies an action done by another person, whereas przykro - like
the English sorry - can also apply to some actions of the experiencer him/
herself (as in sentence 1 above). Furthermore, hurt (discussed in more de-
tail later) implies an unfulfilled expectation: I cannot feel "hurt" (in an
emotional sense) by something that a stranger or an "enemy" does, I can
only feel "hurt" as a result of something done by someone from whom I
would expect "something good" rather than "something bad" or nothing at
all. Compare:
(2) a. She was hurt by her sister'si husband's/friend's words.

b. IShe was hurt by the dentist's/bus-driver's words.
Przykro doesn't have such implications. In particular, something a den-

tist or a bus-driver (with whom I have no special relation) says to me, can
well cause me to feel przykro or to cause me a przykroSc (noun from
przykro):
(3) Siowa dentysty sprawiiy mi przykrooc.

'The dentist's words were hurtful.'
(lit. they 'caused me aprzykrosc')
(4) Kiedy dentysta to powiedziai, zrobito mi sig przykro.
'When the dentist said this, I felt a painful feeling (I felt upset).'
(lit. 'it became przykro to me')
But przykro differs also from sorry, insofar as it refers necessarily to in-
terpersonal relations (like hurts). For example, I can be sorry when I hear
that someone else lost their keys. Przykro, however, could not be used in
such a context:
(5) ?Kiedy usfyszatam, ze mo ja sqsiadka zgubila klucze,

zrobito mi si^ przykro.
'When I heard that my neighbour had lost her keys, I felt bad.'
If someone loses their keys this means that "something bad happened to
that person", not that "someone did something (that caused a bad feeling in
someone)". Sorry is applicable in the case of "bad things happening to
someone", but przykro is not: it requires a human causer ("someone did
something").
Not all human actions can cause a feeling of przykro. Typically, this feel-
ing is caused by somebody's words, as in the following example (with the
noun przykrosc):
(6) Widocznie slowa mo je sprawify jej przykrosc wielkq, bo nerwowy,
bolesny usmiech wykrzywit chwilowojej usta. (Zapolska)
One could see that my words caused her a great przykrofc,
because for a moment a painful, nervous smile twisted her mouth.'
But the feeling of przykro can also be caused by physical actions (other
than words), and even by omissions - provided that these actions or
omissions can be interpreted as conveying a particular message: a message
of indifference, an absence of "good feelings" for someone else. This
message can be formulated in universal semantic primes (see the Introduc-
tion) as follows:
[someone can think:]

"this person doesn't feel anything good towards me"
And more precisely:
[someone can think:]

"when this person thinks about me
this person doesn't feel anything good"
342 Anna Wierzbicka
For example:
(7) Ju2 mi kilka razy mowit, ze ja za duzo pracujg; widac wigc,

ze ijemu przykro, ze mnie tak malo widzi. (Dabrowska)
'He has already told me a few times that I work too much; so it's
clear that he, too, feels upset (przykro) that he sees so little of me.'
The person spoken of clearly interprets the speaker's behaviour as send-

ing a message of "indifference" (that is, as indicating a lack of any good
feelings towards the other person). Were the word przykro to be replaced
with some other word of negative emotion (for example zal 'regret/sorry'
or smutno 'sad') the sentence would no longer include any reference to an
imputed message of indifference. The same applies to the following sen-
tence:
(8) Piotr poczul, ze jest tutaj zupelnie niepotrzebny...

Byto mu nieprzyjemnie, a nawei przykro. (Zeromski).
'Peter felt that nobody needed him there...
He had an unpleasant, and even a painful (przykro) feeling.'
In this sentence, the adverb nieprzyjemnie (also used in a dative con-

struction) implies little more than an unpleasant feeling, but przykro, intro-
duced in an "and even" frame, is clearly presented as more than that (i.e.,
worse than that). More specifically, przykro implies here that Peter feels
not only useless but also, so to speak, unloved; or rather that he interprets
the behaviour of other people, who are not making any demands on him, as
sending a message of "indifference", of absence of "good feelings".
In trying to identify the invariant cognitive scenario associated with the
word przykro, it is helpful to consider some examples without any wider
context and without any additional clues. For example:
(9) Byio mi bardzo przykro i glupio z powodu tego, co sie_ stalo. (Pie-
tak)
had a painful and embarrassed feeling because of what had hap-
pened.'
Here, the sentence (adduced by Stownik Jgzyka Polskiego without any
context) does not refer to any preceding human action. Yet the word
przykro (as well as glupio, to be discussed below) makes it quite clear that
the event referred to must have involved some human action: one could feel
neither przykro nor gtupio because of some misfortune unrelated to any

human causer. Even without the context, the sentence makes it clear that
someone had done something as a result of which someone else could feel
neglected, unloved, rejected, offended.
This brings us back to the first example, repeated here for the reader's
convenience:
(10) Przykro mi, ie mimo woli obrazitem t%

4
1 have a painful feeling that I unintentionally offended that
woman.'
In this case, the speaker has done something because of which someone
else ("the woman") could think: "when this person thinks about me this
person doesn't feel anything good". In this case, the missing "good feel-
ings" can be interpreted as respect - and as a result the woman feels of-
fended. The speaker himself, thinking about what happened, feels przykro.
Thus, all the examples discussed are consistent with the following expli-
cation:2
(11) Byto mi przykro

(a) I felt something bad
(b) because I thought:
(c) someone did something
(d) because of this someone else could think:
(e) "this person doesn't feel anything good towards me"
Most commonly, the "someone" mentioned in component (c) is different
from the experiencer mentioned in component (a), whereas the "someone
else" in component (d) is identical with that experiencer (in compo-
nent (a)); it is also possible, however, that the experiencer (in compo-
nent (a)) is identical with the causer (in component (c)).
This is in fact one of the most striking features of the concept of przykro:
that it encodes a bad feeling caused either by someone else's apparent emo-
tional rejection of us (or indifference towards us), or by our own action
which may appear to someone else to indicate our rejection of them (or in-
difference towards them).
It is noteworthy in this context that the noun przykrosc is commonly used
as a direct object in bi-transitive sentences referring to interpersonal sen-
tences, such as
344 Anna Wierzbicka
(12) X zrobif (/sprawii) Y-owi przykrosc

X.NOM did (/did [caused]) Y-DAT pain/hurt.ACC
'Person X did something painful to person Y.'
The syntax of the noun przykrosc is revealing, for it highlights the inter-
personal character of the emotional experience. The noun referring to the
causer of przykrosc can be inanimate, but it has to refer to some aspect of
somebody's behaviour, for example:
(13) Jego slowa (zachowanie) sprawiiy jej przykrosc.

'His words (behaviour) did[caused] her a painful feeling.'
I cannot think of any English emotion terms which would have similar
syntax, and consequently similar semantic implications. One can say in
English, of course, things like He (X) did her (Y) a bad turn, which refer to
something person X is "doing to" person Y, but in this case no emotional
experience on the part of Y is implied. One can also say things like He (X)
gave her (Y) a black look, which do imply a feeling, but in this case what is
referred to is a feeling of the agent, not the target person (Y). Finally, one
can also say things like He (X) caused her (Y) a lot of pain, but a sentence
like this would have to refer to a long-term effect, not to a brief interper-
sonal episode. For example, normally one would not say: IHis words
caused her (a lot of) pain.
Furthermore, a sentence like "X caused Y a lot of pain" does not have to
have a human subject and can refer to physical pain: Her inflamed knee
caused her a lot of pain. All this is different from a Polish sentence like:
(14) X zrobil Y-owi przykrosc
X.NOM did Y-DAT pain/hurt.ACC
which implies quite unambiguously that X is a person and Y, another per-
son, and that it is an action of person X which causes a painful, and
thought-based, feeling in person Y.
The syntax of the dative construction, as in
(15) X-owi bylo przykro
X-DAT was painfully[ADV]
'X felt (for a while) a painful feeling.'
does not show the interpersonal character of the painful event and in this
case the agent can be identical with the experiencer. Yet in this case, too,
the meaning of the word przykro guarantees that the sentence refers to
some interpersonal act: either X felt something bad (przykro) because
someone else (Y) had done something that suggested coldness, indiffer-
ence, or lack of regard for X, or X felt something bad (przykro) because X
thought that someone else (Y) may have perceived (construed) some action
of X as suggesting coldness, indifference, or lack of regard towards that
other person (Y).
The painful effect of someone's perceived "coldness" or indifference is
often experienced as extremely strong, so much so that the word przykro
often occurs in Polish in combination with the intensifier strasznie 'ter-
ribly':
(16) Strasznie mi przykro.
terribly to.me painfully
am terribly hurt/sorry.'
(17) Byio mu strasznie przykro.
it.was to.him terribly painfully
'He was terribly hurt/sorry.'
This frequent co-occurrence with the intensifier strasznie highlights the
depth of the hurt: even when there is no implication that anyone has done
"anything bad" to another person, or that the two persons in question have a
close emotional relationship, the "slight" of "coldness" or "indifference"
can be perceived as very painful.
3. Semantic implications of the dative construction

There are many emotion terms in Polish which can appear in a dative con-
struction and which - when used in a dative construction - appear to carry
the same semantic component: can't not feel like this'. For example:
(18) a. Jestem smutna[A.O3]

am sad'
b. Jest Am.DAT smutno[\OV]
I feel'sad'
I can't not feel like this
346 Anna Wierzbicka
(19) a. Jestem wesola [ADJ]

I am 'cheerful'
b. Jest W/.DAT wesoto [ADV]
I feel 'cheerful'
I can't not feel like this
The word przykro used in a dative construction appears to carry (in ad-
dition to all the components specified earlier) the same "involuntary" com-
ponent can't not feel like this'.
In fact, in Polish there are many words which outside the dative construc-
tion do not refer to feelings at all but which when used with the dative carry
the "involuntary" component can't not feel like this'. This applies, for
example, to glupi 'stupid', cigzki 'heavy', lekki 'light', and zly 'bad'. One
example:
(20) Bylo mi glupio
(lit. 'It was stupid to me.')
(b) because I thought:
(c) I did something stupid
(d) other people could see this
(e) other people could think something bad about me be-
cause of this
(This is only a very rough explication, but sufficient for the present pur-
poses.) The reference to other people stems not from the meaning of the
word glupio itself, but only from its use in the dative construction. Signifi-
cantly, it is only in the dative construction that one can combine glupio
with the phrase pried ludzmi 'in front of the people':
(21) Byio mi glupio przed ludzmi.

(lit. 'It was stupid to me in front of the people.')
felt embarrassed in front of the people'
4. Some English words comparable with przykro

As mentioned earlier, the English words comparable to przykro include
hurt, offended, and sorry (as well as the expression feel bad). One might
think that the concepts encoded in these words would be elucidated, to

some extent at least, by reputable English dictionaries. In fact, however,
most dictionaries find it hard to deal with such words. For example, Collins
(1993) defines the (relevant) meaning of the word sorry as follows: "feel-
ing or expressing pity, sympathy, remorse, grief, or regret", and the (rel-
evant) meaning of the verb to hurt as "to cause emotional pain or distress to
(someone)". Similarly, Longman Dictionary of the English Language
(1984) defines sorry as "feeling regret, penitence or pity" and hurt as "to
cause mental pain or anguish to; offend". Offend is defined in turn as "to
cause displeasure, anger, or vexation" (Longman Dictionary of the English
Language), and "to hurt the feelings, sense of dignity, etc. of (a person)"
(Collins English Dictionary and Thesaurus). What is characteristic of such
definitions is, above all, their complete failure to capture the semantic in-
variant of any given concept and to pinpoint the differences between all the
concepts mentioned. Thus, having read all the definitions adduced above
one would still have no idea as to what the differences between hurt and of-
fended or hurt and sorry really are.
The only way to do this is to identify the prototypical cognitive scenario
associated with each word, and present these scenarios in terms of the same
set of conceptual primitives. If these conceptual primitives also represent
lexical universals, the emotion concepts involved can be rigorously and in-
sightfully compared not only within one language but also across lan-
guages and cultures.
4.1. Offended
The Oxford English Dictionary offers the following (unattributed)

examples of the word offended (in its emotional sense):
(22) a. He was highly offended at being passed over.

b. You are offended with me.
c. / assure you that I am not in the least offended.
All these examples are useful in highlighting the interpersonal character

of the event involved: they all suggest that one person is "offended" be-
cause of something another person did. In addition, sentence 1 offers an
example of an action which may cause someone to feel "offended".
348 Anna Wierzbicka
What matters here is not the action as such but its apparent meaning: I
feel "offended" by your action if this action suggests to me that you "don't
think good things about me", and that you want me to know it. Schemati-
cally, this can be represented as follows:
(23) / was offended

(b) because I thought about someone else (Y):
(c) "Y did something (Z)
(d) because of this I know now:
(e) Y doesn't think good things about me
(f) Y wants me to know this"
The word "dignity" mentioned in the Collins dictionary's definition of
offended provides a helpful clue: "being offended" does have something to
do with the notion of "dignity", and with our need to have our personal
worth acknowledged by other people. I am offended if I think this basic
need is flouted: not only does the other person fail to think good things
about me, but doesn't even seem to mind letting me know it (for example,
by "passing me over").
If I accidentally discover that someone doesn't think "good things" about
me I may be disappointed, upset, distressed, and so on - but not "of-
fended" - unless this lack of regard appears to be displayed openly.
Smith's (1970: 589) dictionary of synonyms offers the following charac-
terisation:
OFFEND (Lat. offendgre, to strike against) relates always to the conduct of one person
towards another, and implies, therefore, conscious agents on both sides, and a con-
dition of real or supposed slight on one side. It belongs to superiors and equals rather
than to inferiors to be offended. In the case of equals, it still implies an alleged defi-
ciency of regard or consideration.
This is apt and insightful and it corresponds, I think, to the proposed

cognitive scenario.
It is clear that the concept of offended has a good deal in common with
the Polish concept of przykro. Roughly speaking, what the two share is
some perceived "deficiency of regard". But first, in the case of offended,
this "deficiency" concerns an apparent absence of "good thoughts about
someone", whereas in the case of przykro it concerns an apparent absence
of "good feelings"; second, in the case of offended the "deficiency" in
question appears to be openly and intentionally displayed, whereas in the

case przykro it can be (and typically is) betrayed or implied unintention-
ally; and third, in the case of offended the experiencer must be different
from the agent (I cannot be offended by my own action), whereas in the
case of przykro the experiencer can coincide with the agent: I can feel
przykro because of something that I have done myself (causing a bad feel-
ing for someone else).
4.2. Hurt
Hurt is often treated as virtually interchangeable with offended. Recall, for

example, the Collins dictionary's definition of offend: "to hurt the feelings,
sense of dignity etc. of (a person)". Smith's (1970: 491) definition of hurt,
while helpful, is also very similar to its definition of offended:
HURT ... is used of wounded feelings, and denotes the sense of having been treated un-
fairly, inconsiderately, or without due respect. We are sorry for circumstances. We
are grieved at acts and conduct. We are hurt by treatment or behaviour.
But the range of use of the two words (hurt and offended) is clearly not
the same, and clearly hurt has more to do with feelings than offended; it is
not an accident that while one can either hurt or offend a person, one can
only "hurt someone's feelings" (not "offend feelings"). A few examples
(from Bernard Shaw's plays):
(24) a. / can' t fall in love and I can' t hurt a woman 's feel ings by tell-
ing her so.
b. Mrs. Warren winces, deeply hurt by the implied indifference
towards her affectionate intention.
c. You can marry her, if you hurt her feelings by refusing...
d. How deep it would hurt her to think it was because I didn't
care for her.
e. Oh, do stay, Rejjy, I shall really be hurt if you desert us.
What all these examples suggest is that one person "feels something
bad" as a result of something that another person does, and that the "hurt-
ful" action is unexpected and apparently inconsistent with the affected per-
son's expectations. Schematically, this can be represented as follows:
350 Anna Wierzbicka
(25) was hurt (by what X had done)

(a) Y felt something bad
(b) because thought:
(c) "X did
(d) X could know that I would feel
something bad because of this
(e) I didn't think X would do something like this
(f) because I thought X wouldn't want me to feel something
bad"
Smith (1970: 491) observes that "when used of mind or feelings, hurt is
employed analogously [to the physical hurt], in the sense of receiving a
rude shock", and I think this idea of a "rude shock" is consistent with the
component of unexpectedness (d) in the scenario proposed above. Accord-
ing to the proposed explication, hurt does not imply that the experiencer
feels "unloved": doesn't think "X doesn't feel anything good towards
me," he/she only thinks "I didn't think X would do something like this be-
cause I thought X wouldn't want me to feel something bad".
How then does hurt differ from the Polish przykrol The most obvious
difference concerns, as mentioned earlier, the "other person" requirement:
I cannot feel "hurt" (in an emotional sense) by anything that I myself do,
whereas I can feel przykro because of something that I have done myself if
someone else may feel "unloved" or "uncared for" as a result of my action.
A second clear difference has to do with the "unexpectedness" of the
"hurtful" action (component (e) of hurt). An action which causes przykroso
is not similarly unexpected. Furthermore, in the case of przykro it is an im-
plied "indifference" (or lack of regard) which causes a "bad feeling." If I
am "hurt" this implies that I would not have expected the other person to
act the way they did because I would not have expected them to want me to
feel something bad. The fairly wide use of this word in modern English
may reflect the cultural norm that in social interaction it is important not to
"hurt other people's feelings" and the assumption that normally people
would not want to make other people feel something bad (and would try to
avoid doing so). The cultural assumption behind the Polish v/ord przykro is
different: what matters most is not so much that we should not make other
people "feel something bad" as that we should feel, and show, good feel-
ings for other people.
4.3. Sorry
The adjective sorry, often assumed to be identical in meaning with the

noun sorrow, in fact means something very different, and is, as Smith
(1970: 695) put it, "used in a lighter sense and of more ordinary matters".
Smith adds that "we are commonly sorry for what is on our own account",
and offers the example:
(26) / am sorry that I was not at home when you called.
He also adduces, however, an example from Shakespeare which illus-
trates a different use of sorry (not "on our own account"):
(27) / am sorry for thee, friend, 'tis the Duke's pleasure.
When sorry is combined with for, it seems clear that its meaning can be
represented as follows:
(28) X was sorry for
(a) X felt something bad
(b) because X thought:
(c) "something bad happened to person Y"
This scenario is similar to przykro in so far as it involves two people. In
this case, however, no one has to do anything: the feeling is presented as
caused by a thought about something that "happened" to someone, not a
thought about something that someone "did".
When used without the preposition for, as in "I am sorry that X", sorry
does not seem to necessarily imply that "something bad happened to an-
other person" or even that "something bad happened to somebody". For
example, if I say I'm sorry that I couldn't come to your party, I do not mean
to suggest that something bad happened to either you or me. I do wish to
imply, however, that what happened was "bad for someone" (presumably,
me, but possibly, you). This suggests the following semantic formula for
sorry that:
(29) X was sorry that happened
(b) because X thought:
(c) "something (Z) happened
(d) this was bad for someone"
352 Anna Wierzbicka
Since, however, the frame: "emotion-word THAT" suggests a feeling re-

stricted in time to the moment of thinking about the proposition introduced
with the word that, a more accurate formula would read:
(29') X was sorry that Z happened

(b) when X thought:
(c) "something (Z) happened
(d) this was bad for someone"
Or, alternatively:
(29") X was sorry that Z happened

(a) when X thought about , thought:
(b) "this was bad for someone"
(c) when X thought about this, X felt something bad
Often, what happened ("Z") is due to some action of the experiencer. I do
not think, however, that it is necessary or appropriate to posit for sorry that
an additional, separate meaning along the lines of "apology" or "remorse"
(as suggested by some dictionaries). But the word sorry used on its own,
without an overt subject, does have a distinct meaning, referring to some-
thing that the speaker him- or herself has done and that was bad for the ad-
dressee:
(30) Sorry
(a) I feel something bad (now)
(b) because I think:
(c) "I did something
(d) it was bad for you"
Of the three uses of sorry considered here and linked with three different
grammatical frames (X was sorry for Y, X was sorry that Z, and Sorry) it is
the second use which comes closest to the Polish przykro and could some-
times be used as its approximate translation equivalent. For example:
(31) Byto mi przykro, ze niechcqcy obrazUem kobiet%.

was sorry that I had inadvertently offended that woman.'
But here, too, there are considerable differences in meaning: przykro

implies that someone did something, whereas sorry that (on its own) does
not; przykro implies a perceived lack of good feelings for somebody else,
whereas sorry that does not; and sorry that implies that the event referred
to was "bad for somebody", whereas przykro has no such implications.
4.4. Feel bad
The phrase feel bad looks so simple and straightforward that it is seldom
recognised for what it really is: a language-specific English idiom, with a
meaning different, and not predictable, from the meanings of its two parts.
For example, a question like "How are you feeling today?" is more likely to
be answered with "I feel good", or "I feel wonderful", or "I feel awful",
than with "I feel bad". In contrast to the other three phrases, "I feel bad"
(often, "I feel bad about it") suggests a cognitive scenario which is irrel-
evant to a question about a person's well-being.
Broadly speaking, "I feel bad" can be compared to "I feel guilty", al-
though the two are by no means identical in meaning. First of all, "I feel
bad" refers to not just any "bad feeling (or feelings)" that I may have, but to
a bad feeling resulting from something that I have done or that I have failed
to do. Secondly, "I feel bad" refers, in an oblique way, to "other people". If
my conscience is telling me that I have done something bad which doesn't,
however, involve other people, I could readily say that "I feel guilty", but
not that "I feel bad". A person who is on a diet and who has gone on a
binge, could say, half in jest, that they "feel guilty", but probably not that
they "feel bad about it". At the same time, if something really terrible
happens as a result of my actions, I could say that "I feel guilty", even "ter-
ribly guilty", but not that "I feel bad about it". In fact, saying that "I feel
bad about it" is almost like saying that "I didn't really do anything
bad" - while at the same time admitting that there was something about my
behaviour that was not quite right, and that was bad for someone else.
"Feeling bad" seems to be particularly appropriate for situations where I
could have done something good for someone else and didn't. I don't have
to "feel guilty" (I tell myself), for I don't think that I have done "anything
bad"; at the same time I can't deny to myself that what I have done was bad
for someone else. An example of such a scenario:
354 Anna Wierzbicka
Polish teachers don't like cheating any better than the Americans. One Polish in-
structor, who is infuriated by cheating, caught a university student red-handed and
sent her out of the exam room. "I felt really bad", he said, "I was wet with sweat, she
was crying". He thinks that unclear policies on cheating and a reluctance among
teachers to act as executioners result in some degree of tolerance. (Klos Sokol 1997:
194)
In this vignette, the teacher consciously thinks: "I didn't do anything

bad", and yet seeing the student crying, he can't help thinking that what he
did was bad for the student, and he feels something bad because of that.
These observations suggest an explication along the following lines:
(32) I feel bad

(a) I feel something bad
(b) because I think:
(c) "I did something"
(d) I can think that I didn't do anything bad
(e) I can't not think that it was bad for someone else
More often than not, "feeling bad" would be linked with having caused
someone else to "feel something bad", but this does not seem to be invari-
ably so. For example, if I give another person some of my food (hiding,
however, the best of it and keeping it for myself), and if this other person
thanks me profusely and praises my generosity, I could "feel very bad" - al-
though the other person wouldn't feel anything bad at all. Nonetheless, the
proposed explication would still be consistent with the situation: I could be
telling myself that "I didn't do anything bad", but perhaps I would be un-
able not to think that what I did was bad for that other person.
Assuming, then, that this is what "feeling bad" means, how does it differ
from the Polish przykrol The two are similar insofar as they both refer to a
"bad feeling" caused by a thought about someone else, and about some-
thing that someone has done ("something", but not "something bad"). In
the case of "feel bad", however, the experiencer of the "bad feeling" is
identical with an agent who has done something, whereas in the case of
przykro this may or may not be the case. Furthermore, "feel bad" is linked
with a thought that the action was bad for someone else, whereas przykro
implies no such thought; on the other hand, przykro implies that the action
could be interpreted as showing an absence of "good feelings" towards an-
other person, whereas feel bad does not imply that.
5. Conclusion
The English words and expressions discussed are probably closer to
przykro than any others. It can be safely concluded, therefore, that English
doesn't have a word or expression corresponding exactly to this Polish con-
cept.
This is not to say that speakers of English never experience the emotion
associated in Polish with the word przykro; only that they do not think ha-
bitually about their experiences in these terms. The situation conceptual-
ised by a Polish speaker in terms of przykro would be likely to be concep-
tualised in English in a different way - for example, as a situation linked
with the experience of being "hurt", "offended", "sorry", or "feeling bad".
Each language provides its speakers with a different set of interpretive
categories and encourages them to conceptualise their emotional experi-
ence in terms of a different set of scenarios. It is likely that the set of scen-
arios linked with frequently used language-specific categories (lexical and
lexico-grammatical), points to a culture's central concerns and values. The
Polish word przykro and the grammatical construction associated with it is
a good case in point.
In an earlier study of the norms relating to emotion in Polish culture
(Wierzbicka 1994) I argued that Polish culture places a great deal of em-
phasis on the expression of interpersonal "good feelings", noting that "the
Polish cultural emphasis on warmth or affection must be distinguished
from the Anglo emphasis on consideration and tact", and that "Polish cul-
ture encourages the showing of good feelings toward the addressee rather
than efforts not to hurt or offend the other person" (1994: 163; for fuller
and more recent discussion, cf. Wierzbicka 1999). I also noted that in
Polish culture, behaviour that displays emotion is seen as the norm, not as a
departure from the norm, and that "perhaps no feeling is more valued and
more expected in Polish discourse than a good feeling towards the ad-
dressee". In support of these observations, I wrote, inter alia:
The Polish cultural value of expressing good feelings toward the addressee is... re-
flected in its rich store of terms of endearment, especially metaphorical endearments
(usually used in the vocative case), such as ptaszkul (bird-DlM), sioneczko\ (sun-
DIM), kotku\ (kitten-DiM) robaczku\ (worm-DiM), zabko\ (frog-DiM), skarbiel (treas-
ure), or zlotko! (gold-DiM). Words of this kind may be used not only as parenthetical
terms of address (like honey or sugar in American English), but also as exclamative
utterances in their own right (like darling\ or sweetheart*, in English). Frequently,
356 Anna Wierzbicka
terms of this kind are used with the possessive pronoun moj ('my'), which under-
scores the personal character of the relationship: moje ztoto (my gold), moj skar-
bie (my treasure), or moje sioneczko (my dear-little-sun; cf. English My darlingl
vs. *My honeyl). Metaphorical endearments are particularly common in address-
ing children, but are not restricted to adult-child interaction. (Wierzbicka 1994:
161)
The culture-specific Polish concept of przykro, which focusses on

painful effects of a perceived lack of expected interpersonal "good feel-
ings", points to the same cultural values and expectations and provides
additional evidence for their reality.
Notes
1. This and subsequent examples from leading Polish authors are as cited in Stow-
nik Jezyka Polskiego (1958-69).
2. The format used in this explication and in the other explications of emotion con-
cepts included in this paper is a simplified one. A full explication of an emotion
concept should, I believe, be based on a prototype, and should attempt to de-
scribe not only the cause but also the quality of the emotion. This can be done'as
follows (cf. Wierzbicka 1973, 1992, 1999):
I felt joy
(a) I felt something good
(b) because I thought something
(c) sometimes a person feels something good
(d) because this person thinks:
(e) something very good is happening now
(f) I want this to happen now
(h) I felt something like this
(i) because I thought something like this
In the case of przykro a full explication of this kind would read as follows:
Byto mi przykro
(b) because I thought something
(c) sometimes a person feels something bad
(d) because this person thinks:
(e) someone did something
(f) because of this someone else could think:
(g) "this person doesn't feel anything good toward me"
(h) I felt something like this
(i) because I thought something like this
References
Collins English Dictionary and Thesaurus
1993 John M. Sinclair, general consultant. London: Harper Collins.
James, William
1890 The Principles of Psychology. London: Macmillan.
Klos Sokol, Laura
1997 Shortcuts to Poland. Warsaw: Wydawnictwo IPS.
Kurcz, Ida, Andrzej Lewicki, Jadwiga Sambor, Krysztof Szafran and Jerzy Woronczak
1990 Stownik Frekwencyjny Polszczyzny Wspotczesnej [A Frequency Dic-
tionary of Contemporary Polish] Krakow: Polska Akademia Nauk.
Longman Dictionary of the English Language
1984 London: Longman.
The Oxford English Dictionary
1933 12 vols. Oxford: Clarendon Press.
Sapir, Edward
1949 Selected Writings of Edward Sapir in Language, Culture and Person-
ality. David Mandelbaum (ed.). Berkeley: University of California
Press.
Stownik Jgzyka Polskiego [Dictionary of the Polish Language]
1958-69 Witold Doroszewski (ed.). 11 vols. Warsaw: Panstwowe Wydawnict-
wo Naukowe.
Smith, Charles John. (1970 [1903]). Synonyms Discriminated: A Dictionary of
Synonymous Words in the English Language. H. Percy Smith (ed.).
Detroit: Gale Research Co.
Wierzbicka, Anna
1973 The semantic structure of words for emotions. In: Roman Jakobson,
C. H. van Schooneveld and Dean S. Worth (eds.), Slavic Poetics: Es-
says in Honor ofKiril Taranovsky, 499-505. The Hague: Mouton.
1992 Defining emotion concepts. Cognitive Science 16: 539-581.
1994 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 As-
sociation.
1999 Emotions Across Languages and Cultures: Diversity and Universals.
An inquiry into "sadness" in Chinese1
Zhengdao Ye
What we have to translate, in effect, is not a word but a whole culture, to see how an
emotion and the name for that emotion fits into the systematic worldview, lan-
guage, and the way of life of the society. (Robert C. Solomon 1995: 256)
No inquiry into the emotions (affections) can afford to overlook poetry and narrative,
since the writers of such works are more perceptive than anyone else on these
matters. (Zhongshu Qian 1998: 84)
1. Introduction: a fundamental question of what

There is a Chinese idiomatic expression jingdiziwa 'frog in the well': a
frog that can never jump out of the well can see only the little circle of sky
above his well, imagining it to be the whole world.2 Students of emotions
from different disciplines - anthropologists, psychologists, and sociol-
ogists - can often unconsciously run the risk of being a 'frog in the well' if
they take their own language as the universal norm. Researchers may not
be aware that their own language, along with its built-in culture and con-
cepts, constitutes a well, and that the deeper the well, the more limited their
view of the world will be.
Part of this trap lies in answering the fundamental question of what -
what is a particular emotion that is particular to its culture? Only after they
are assured of the universal truth of this what can students of emotions
freely embark on the search for why and how in "ethno" studies. To quote
Kirkpatrick and White:
Given the central place of conceptual structures and interpretive routines in our ap-
proach to ethnopsychology, issues of linguistic meaning and translation are of par-
ticular importance in the task of ethnographic description... so understanding what is
being said - is a minimal requirement of ethnopsychological research. (Kirkpatrick
and White 1985: 13; emphasis mine)
Unfortunately, however, in writing about emotions and emotion con-
cepts, researchers often rely exclusively on translation, and exclusively on
English - a linguistic "Trojan horse" (Tollefson's 1995 term). This practice
360 Zhengdao Ye
carries an assumption that the world above the "well" of the Anglo lan-
guage is the whole world outside the well. They forget that the high wall of
the well is built up by the bricks of Anglo values and judgments. It is not
that all researchers of emotions are unaware of the problem of absolutising
their own values. Lutz (cf. 1985, 1987, 1988, 1995), for example, has re-
peatedly emphasised semantic differences between cultures as a methodo-
logical problem in her study of the indigenous psychology on Ifaluk Atoll
(Micronesia). She stated that "in the present study, the question of the
meaning and translation of those terms has been taken as the primary and
logically prior question" (Lutz, 1985: 39). She emphasised this point a dec-
ade later by warning herself to guard against her own preconceptions: "My
sense of what people were saying - requiring an effort to disentangle my
own native emotional understandings from theirs" (Lutz, 1995: 253). Rus-
sell made the same point from a more global perspective: "I take it for
granted that psychologists are interested in the emotions of all people, not
just those who speak English. We must therefore understand the vocabu-
laries of all people" (Russell 1991: 428).
The same problem comes up in the study of Chinese emotions (as well as
in the study of Chinese indigenous concepts in general, cf. Bond 1996;
Yang 1988). Russell and Yik (1996), in emphasising that analysis of the
Chinese lexicon can help reveal the concepts available to Chinese speakers
through which they categorise their emotional experiences, also raised the
fundamental question of what. In their comprehensive and revealing re-
view of the research on Chinese emotion, Russell and Yik tabulated
95 Chinese emotion-describing words that were translated to or from Eng-
lish from 18 studies (Table 12.2: Translation of Emotion Terms, in Russell
and Yik 1996: 177-179). They were listed with their "English equivalents".
This list is illuminating as well as alarming. Russell and Yik pointed out
that among the aforementioned 95 words, 66 were found in only one study
(where the translation has not been replicated). Among the remaining
29 cases, which allowed researchers to examine the replicability of trans-
lation across studies, the one-to-one correspondence between Chinese and
English was replicated in just five translations. The researchers supposedly
investigated the same concepts, but in fact they did not.
Some researchers do recognise the disparity and inconsistency between
concepts across cultures (cf. e.g. Schwartz, White, and Lutz 1992;
Shweder and LeVine 1984). The question is how to make explicit the
meaning of emotion concepts without imposing researchers' own cultural
values on them. In other words, there is an urgent need to develop adequate

means in "ethno" studies through which researchers can articulate, from a
neutral position, the meaning of seemingly "inexplicable" and "undefin-
able" emotion concepts found in other languages. Undoubtedly, linguistic
theory, especially lexicographic theory, can make a contribution to answer-
ing the question of "what is the meaning".
This paper will attempt to overcome the methodological problems in
emotion studies by relying on the Natural Semantic Metalanguage (NSM)
(cf. e.g. Goddard, 1998; Goddard and Wierzbicka 1994, In press;
Wierzbicka 1972, 1996). By explicating three so-called "sadness-like"
Chinese emotion terms - bei, ai, and chou* - in NSM, this paper will show
how the NSM approach, with a set of empirically grounded and universally
applicable semantic primitives such as SAY, WANT, FEEL, THINK, KNOW, i,
YOU, GOOD, BAD, DO, HAPPEN (see the table of semantic primitives in 5),
can provide a neutral comparative grid for further inquiries into the mean-
ing of emotion concepts across languages and cultures.
The paper will first provide a very general discussion of the Chinese
emotion lexicon from a morphological point of view, followed by an in-
depth semantic analysis of bei, ai, and chou in NSM. The discussion will
draw on linguistic evidence, including well-known textual examples, lexi-
calised items and conventionalised phrases and idioms (in particular
chengyu - usually four character long idioms) that are familiar to the Chi-
nese ear and that serve as lexical entries in such authoritative dictionaries
as Xiandai Hanyu Cidian (XHC).
"There is no richer source of 'data' for reflection on emotion than litera-
ture" (Marks 1995: 7). The literary approach is used to provide concrete
contexts and illustrate the range of use of the emotion terms under dis-
cussion. In the context of China, where literature has always been part and
parcel of the national culture and the best embodiment of life-experience
and cultural spirit (Chen 1977; Qian 1998), it is indispensable to examine
these resources where emotions are described and expressed. Linguistic
tests, especially collocation tests, will also be used in the discussion. As an
empirical experiment, the definitions in NSM will be translated into Man-
darin Chinese. By relying on the exponents of semantic primitives ident-
ified in Mandarin Chinese we can further verify the universal applicability
of NSM. Methodological advantages of NSM will be summarised at the
end of the paper.
362 Zhengdao Ye
2. A word on emotion-describing terms in Chinese
Because of the fuzziness of word boundaries in Chinese due to the writing

system (cf. Chao 1968; Li and Tompson 1981; Lu 1964; Pan, Yip and Han
1993; Zhang 1999), a better way to refer to emotion words in Chinese is
emotion terms or emotion-describing terms. Emotion terms in Chinese ac-
tually consist of two tiers of concepts (due to the nature of the Chinese lan-
guage): qinggan zi ('monosyllabic emotion-describing zf) and qinggan ci
('polysyllabic emotion-describing cf). Both are semantically self-con-
tained. This is not to say that all emotion-describing ci have emotion-de-
scribing zi as their components. Some do, and some don't. Since emotion-
describing zi has a very strong and basic psychological reality among Chi-
nese people, this should be taken as a starting point or first step to investi-
gate Chinese emotion lexicon (see Ye 2000 a for detailed morphological
discussion of the Chinese emotion lexicon, its categorisation, and inner re-
lationships).4 To have a full grasp of the Chinese emotion lexicon, both
emotion-describing zi and emotion-describing ci should be explicated. The
three emotion terms to be analysed in this paper - bei, ai, and chou - fall
into the category of emotion-describing zi.
3. Bei, ai, and chou
The three emotion-describing terms bei, ai, and chou, are often glossed in-
discriminately in various studies and in dictionaries as either sadness, sor-
row or grief. Bei and ai appear in different versions of "basic emotions" in
classical Chinese texts. Russell and Yik (1996) labeled bei and ai as sorrow
in discussing five versions of the "basic emotions" in Chinese. When 50
educated Hong Kong Chinese adults were asked to list the seven basic
emotions, as in the proverb ren you qiqing liuyu (lit. 'people have seven
emotions and six desires'), bei, ai, and chou were mentioned, with an oc-
currence of 66 percent for ai, 14 percent for bei, and eight percent for
chou.5 The English translations were as follows: ai - 'sorrow/grief,
bei 'sorrow/sadness', and chou 'sadness'. This seems to suggest that ai
is perceived as one of the more basic emotions; and that these three Chi-
nese emotion terms share the same conceptualisation with each other, and
also with the English sadness. Is this the case? The discussion will start
with bei.
3.1. bei 'a tragic/fatalistic sadness'6
(1) Qiu Ye Du Zuo1

autumn night alone sit
Du zuo bei shuang hing,
alone sit bei both hair on the temples
kong tang yu er geng.
empty hall will second two hours in the night
Yu zhong shan guo luo,
rain in mountain fruit fall
deng xia cao chong ming,
lamp under grass insect the cry of birds and insects
Bai fa zhong nan bian,
white hair after all difficult change
Huang jin bu ke cheng.
yellow metal not can succeed
Sorrow over these two locks of hair, I sit alone -

In the empty hall, almost the second watch
In the rain a fruit of the mountain falls,
Insects sing in the grasses beneath the lamp
This white hair will never change its color,
The golden elixir cannot be made...
(From Wang Wei: "Sitting Alone on an Autumn Night", in Owen
1985: 134)
Alone at night, sitting in an empty hall, the poet Wang Wei (701-761 AD)
reflects on his state of isolation in the darkness of night, and on his ageing
marked by the universal, unavoidable life change - black hair turning
white. As the temporal emptiness progresses towards the darkest part of the
night (ergeng), the poet can sense the decline of his life. He is conscious
that what awaits him is death. Ageing, the cycle of life, are inevitable and
irreversible: his white hair will never again turn black and no elixir will
ever bring back his youth. He has to submit to the "law of nature". This
"naturalness" of the cycle of life is heard through the thud of autumn fruit
falling from the tree in the silence of the night mountain in the rain, and
through the wailing of the insects, whose imminent death approaches with
every passage of autumn.
364 Zhengdao Ye
When black hair starts to whiten, this indicates that a person is entering
shengming zhi qiu ('the autumn of life'), in other words something bad is
happening (to somebody). White hair is a sign of the decline of life, with
the worst yet to come - death. What makes one bei is not just a bad hap-
pening at that moment, but an uncontrollable force pulling towards a sub-
sequent "tragic" happening determined by nature. Furthermore, bei lies in
the consciousness of the experiencer, who is aware that the expected course
of events is irreversible. One can do nothing to "beat the odds" and reverse
the happening from worse to worst. Thus, the meaning of bei can be expli-
cated along the following lines:8
bei
(c) "something bad happened now
(d) I know that after this good things will not happen any more
(e) I do not want things like this to happen
(f) I cannot do anything
(g) because I know no one can do anything when things like this
happen"
(h) when this person thinks like this this person feels something very
bad
(i) X felt something like this
(j) because X thought like this
Components (a), (b), (i) and (j) set up a prototypical cognitive scenario
referring to a standard situation in which X feels like a person does when
they have certain specifiable thoughts. Facing the same antecedents, emo-
tional stimuli or situational antecedents, different people may feel differ-
ently because of the different thoughts they have, but it is the prototypical
thoughts that single out a particular situation (cf. Fehr and Russell 1984;
Shaver et al. 1987; Wierzbicka 1972).
Component (c) ("something bad happened now") suggests that the
thoughts of bei are a response to a recent event, whose effects are still evi-
dent. The event may happen to somebody, or it may just happen as a natural
process like autumn. In any case, this event triggers a further pessimistic
view of the situation. Component (d) points to the experiencer's conscious-
ness of the on-going deteriorating situation: the very source of bei. The no-
tion that the future happening lies towards the bad end of the scale ("I know
after this good things will not happen any more") is reflected in such com-
monly used compounds as (2 a) and (b):
(2) a. bei ju
bei play/drama
'tragedy'
b. bei guan
bei view
'pessimistic'
"Something good will not happen any more" proposed in component (d)
indicates thoughts about the absence of a good outcome. Compared with a
component like "something bad will happen", which refers to the presence
of a bad result, "something good will not happen" brings out a pessimistic,
rather than a "fearful" view. This pessimism is one of the most important
aspects of the meaning oibei. The following examples provide further evi-
dence for this key conceptual component. In example (3), a reading of the
thoughts that go through Jia Zheng's mind as he listens to all the riddles
from the young members of the family sheds light on the pessimistic char-
acteristics of bei:
(3) Jia Zheng xinnei chensi dao: 'niangniang suo zuo baozu, ci nai yi
xiang er san zhi wu. Ying-chun suo zuo suanpan, shi dadong luan
ru ma. Tan-chun suo zuo fengzheng, nai piaopiao fudang zhi wu.
Xi-chun suo zuo haideng, yifa qingjing gudu. Jin nai shangyuan
jiajie, ruhe jie zuo ci buxiang zhi wu wei xi ye?' Xinnei yu si yu
meng, yin zai Jia mu zhi qian, bugan xing yu se, zhide mianqiang
kan xiaqu... Jia Zheng kanwan, xinnei zi cun dao: 'ci wu dao hai
youxian. Zhishi xiaoxiao zhiren zuo ci ciju, gengjue buxiang, jie fei
yongyuan fushou zhi bei'. Xiangdao cichu, da you bei qi zhi
zhuang. (HLM: v. 1:304)
Her Royal highness wrote about a firecracker which disintegrates
after a single explosion. Ying-chun's subject, the abacus, is in con-
stant commotion; Tan-chun's kite is something which drifts away
with the wind; Xi-chun's temple lamp is even more lonely and ne-
glected. What ill-omened subjects for all of them to choose so soon
after the New Year!' The more he reflected, the deeper his dismay.
366 Zhengdao Ye
But in his mother's presence he dared not disclose it and forced

himself to look at the other riddles.... After reading this Jia Zheng
reflected with dismay, The object itself isn't ill-omened, but what
inauspicious lines for a young girl to write. It doesn't look as if any
of these girls will have good fortune or long life.' Sunk in gloom he
looked the picture of grief as he lowered his head in thought
(DRM:v. 1:325).
Lost in contemplation of a tree, Bao-yu in example (4) thought of Xiu-

yan and her recent engagement as a "tragic" event. With the following
pessimistic thoughts, his heart was filled with bei:
(4) (Bao-yu yangwang xingzi bushe. You xiangqi Xing Xiu-yan yi zele
fuxu yishi.) Sui shuo shi nann da shi, bu ke bu xiang, dan weimian
you shaole yige hao emu. Buguo liangnian, bian ye yao 'lu ye
cheng yin zi man zhi' le. Zai guoji ri, zhe xing shu zi luo zhi kong,
zaiji nian, Xiu-yan weimian wufa ru yin, hong yan si gao le, yinci
bu mian shangxin, zhi guan dui xing liulei tanxi. Zheng beitan
shi... (HLM:v. 1:800)
Though marriage was something everyone must go through with,
this would mean one good girl the less; in just a couple of years she
would be burdened with children as this tree was with fruit; and
just as the apricots would soon be gone, leaving the branches bare,
in a few more years Xiu-yan's hair would be turning silver and she
would lose her beauty too. He could not help shedding tears as he
gazed at the tree. But as he was sighing... (DRM: v. 2:295)9
It is interesting to note that the translator was unable to render the mean-
ing of bei. Nonetheless, tan ('to sigh') conveys, to a certain extent, the
powerlessness and resignation on the part of the experiencer, which is an
integral part of the meaning of bei. Although the experiencer rejects the
outcome (component (e) "I don't want this bad thing to be happening"),
they cannot prevent it from happening. They merely subscribe to the ac-
ceptance of, and the subsequent resignation before the irreversible out-
come of the event. Component (f) ("I cannot do anything") shows one's in-
ability to reverse the course of the event. This is because of the fatalistic
realisation that the event happens independently of one's will. Compo-
nent (g), "because I know that no one can do anything when things like this
happen" reflects this resignation. This powerless feeling bears some resem-
blance to sorrow, where "I cannot do anything" is part of the meaning.
The powerlessness of the individual before the "law of nature" causes a
poignant bitterness, pain and sombreness, almost to the point of "chill and
cruelty" as reflected in the component "this person feels something very
bad", supported by the following compounds:
(5) a. bei Hang

bei cool/cold
'sad and dreary; sombre; dismal'
b. bei tong
bei pain
'painfully sad'
c. bei can
bei miserable/tragic/cruel
'tragic; tragical; pathetic'(ZSHC: 333)
Compared with the cognitive structures of sadness and words related to
it, such as grief and sorrow (see Wierzbicka 1998, 1999 for definitions of
sadness, grief, and sorrow), it is quite clear that none of these English emo-
tion terms includes such a fatalistic view as a component of its meaning.
The emotion bei, then, is tinted with the Chinese view of the world and of
the nature of life. In defining the common expression bei huan U he (6),
Liang Shiqiu, an eminent Chinese scholar of letters and a lexicographer,
offered an insightful definition, which provides valuable evidence for the
meaning of bei:
(6) bei huan U he

bei joys part/separate reunion
'The sorrows of partings and the joys of unions that life has to
offer'. (ZSHC: 333; emphasis mine)
The above definition of 'what life has to offer' echoes components (f)
and (g).
This reminds us of the following line uttered by Su Shi (1036-1101
A.D.), the founder, and a leading representative, of the dynamic school of
poetry in the Ci genre:
368 Zhengdao Ye
(7) Ren you bei huan li he,

man have bei joy parting reunion
yue you yin qing yuan que,
moon have cloudy sunny round lack/absent
ci shi gu nan quan.
This matter ancient difficult whole/perfect
'Mankind is sad at parting; Happy at reunion again.
The moon has weather that changes fine or foul: it waxes and
wanes
From most ancient times, down to our own every days,
The imperfection of all things - has for ever been the case.'
(ZGWG: 229)
Even separation and reunion are perceived by Chinese people as being
subject to natural forces: they, too, are caught up in the snare of the inevi-
table. It is obvious from these examples that glossing bei as 'sorrow' or
'sad', as various scholars have done, fails to convey this important shade of
meaning, moulded by the characteristic Chinese Weltanschauung. The
idiom bei tian ming ren bears out most clearly this fatalistic emotion con-
cept of no alternative:
(8) bei tian min ren

bei heaven pity man/people
'Bemoan the state of the universe and pity the fate of mankind'.
(HCYC: 17)
(?ai tian; *chou tian)
Tian ('heaven') has a fatalistic status in the Chinese ethno theory of life
(cf. Feng 1953: 34). This feature makes it most compatible with bei.
Linked with tian ('heaven') is ming ('predestined fate') - an omnipotent
force that governs human life, according to the traditional view of Chinese
people (cf. Ye 1998) - that is intrinsically linked to bei. It is interesting to
note that tian is not so compatible with ai, and chou tian is unacceptable.
When one ponders ming ('predestined fate', which lies towards the bad end
of the scale) and one's powerlessness before it, bei is often the feeling most
associated with it. This point is illustrated by example (9):
(9) (Xiang-yun) you xiangdao ziji mingku, gang peile yige caimao
shuangquan de nanren, xingqing you hao pianpiande dele yuannie
zhenghou, buguo ai rizi bale. Yushi gengjia beitong, zhi kule

banye.(HLM:\.2: 1487)
Xiang-yun recalled the old lady's goodness to her, then thought of
her own wretched fate, she had only married a talented, handsome
husband with a cheerful disposition, but then he had contracted
this fatal illness which might any day carry him off. In distress, she
wept half the night. (DRM: v. 3: 434)
The next example points in the same direction:
(10) Ta he ta de ganging youshi name hao. Ran' er xianzai fuqin que gei
ta xuanqu le ling yige ta bu renshi de guniang, bingqie haijueding
jiuzai niannei jiehun, tade shengxue de xiwang chengle paoying,
er ta shuo yao qu de you bushi suo zhongyi de nage 'ta'. Duiyu ta,
zhe shizai shi yige dade daji. Tade qiancheng duansongle. Tade
meimiao de huanmeng pomie le.... Ta bufankang, ye xiangbudao
fankang, ta zhi bei dao ziji de mingy un. Tan ta renshou le. Ta shun-
congle fuqin de yizhi, meiyou yidian yuanyan, keshi zaixinli ta que
weizhe ziji tongku, weizhe ta shouhou de nage shaon tongku.
(J: 28)
He was deeply in love with Mei, but now his father had chosen an-
other, a girl he had never seen, and said that he must marry within
the year. What's more, his hopes of continuing his studies had
burst like a bubble. It was a terrible shock to Juexin. His future was
finished, his beautiful dreams shattered.... He did not fight back,
he never thought of resisting. He only bemoaned [bei and mourn]
his fate. But he accepted it. (SWBJ: 28)
When Jue-xin finds out about the arranged marriage ("something bad is
happening to me"), he knows that his future is totally destroyed ("I know
after now good things will not happen"). Juexin does not want this arrange-
ment ("I don't want things like this to happen"), for he is in love with Mei,
yet he knows, being the eldest son in such a large and feudal family, that he
cannot do anything but fulfil his filial responsibility ("I know I cannot do
anything", "because when things like this happen, no one can do any-
thing"). Marriage, in such a feudal family, is not dependent on his own will.
The examples cited above, particularly the opening poem, indicate that
bei involves a process of "transfer" and momentary "stirring" (therefore
370 Zhengdao Ye
only "feel very bad" in (h) without reference to a period of time). The ex-
periencer, stirred by some external stimulus, reacts to it. In this regard, bei
is very different from the long-term feeling of sorrow. The following sen-
tence, which serves as a title for a chapter in HLM, typically reflects a natu-
ral change that is intrinsically conducive to bei, or the kind of external
stimulus that the emotion of bei is typically associated with:
(11) Gan qiu shen fu qin bei wang shi.
stir autumn deep touch zither bei past matters
'Moved by autumn, a zither musician mourns the past.'
The music of the zither is desolate and heavy, like the thud of autumn
fruits falling on the ground. It represents the "explosive" tune of bei, which
seems to be symbolised by bilabial plosives /p/. It is interesting to point out
that many scholars have mentioned that bei is a key emotional tone in tradi-
tional Chinese music (cf. Qian 1998).
From ancient times until now, the emotion of bei has typically been
"stirred" by the season of qiu ('autumn'). Bei qiu (lit. 'bei autumn'), which
is a lexicalised item, has been one of the most important and enduring
themes in Chinese literature (e.g. Xin 1993). Song Yu, from the Kingdom
of Chu (around 2nd century BC), believed to have been the first person to re-
flect this folk view of human feelings in Chinese literature, started the lit-
erary tradition of beiqiu:
(12) Bei zai qiu zhi qi ye!

bei PART autumn PART heaven energy PART
Xiaose xi, cao mu yao luo er
rustling of leaves PART grass tree wobble fall PART
bian shuai!
become decline (ZLWZX: v. l: 264)
'Alas for the breath of autumn!
Wane and drear!
Flower and leaf fluttering fall and turn to decay' (ZGWG: 73)
In tracing the cultural origins of the beiqiu complex, Xin (1993) at-
tributes it to the Yin-Yang cosmology that is characteristic of the Chinese
folk view of the universe. This is echoed by Chu Hsi's assertion (cited in
Fang 1957: 56) that "Heaven utilises Yin and Yang and the five ingredients
to transmute all things in the form of life". Autumn is typically linked with
the start of Yin energy ('negative energy'), which initiates the decline of
the process of Nature until it meets Yang energy ('positive energy') again
to generate new life. Bleak and desolate autumn scenes are transplanted
into a kind of reflexive awareness in the individual, who naturally becomes
aware of unavoidable ageing and death, under the Chinese cosmic belief of
tian ren he yi ('the unity of Heaven and Man') (cf. Fang 1957; Feng 1953).
The processes of mutual generation and mutual restraint in nature are in-
dependent of human will.
To some extent, bei crystallises the process from stirring to reflection,
which is construed on the basis of the particular Chinese folk view of the
universe. The explication of the cognitive content of bei seems to suggest a
plausible explanation as to why bei, not ai, is posited as one of the basic
emotions in the Daoist doctrines, such as the Huangdi Nei Jing ('The Yel-
low Emperor's Canon of Medicine', 5th-3rd century BC) and the Dao De
Jing, which are closely associated with Yin-Yang cosmology.
Reflected in beiqiu is a unique aspect of Chinese emotions, that is, the
fact that seasons play an important role in Chinese emotional experiences.
This is well documented by Liu Xie, a sixth century literary critic, in one of
the most important Chinese treatises on literature Wen Xin Diao Long (lit.
'literature heart carve dragon'):
Springs and autumns follow on in succession, with the brooding gloom of dark Yin
and the easeful brightness of Yang. And as the bright countenances of physical things
are impelled in their cycles, so the affective capability of mind (hsin) too is shaken....
And when autumn's skies are high and the animating air takes on a chill clarity, our
thoughts, sunken in the darkness of Yin, touch on far things. The year has its physical
things, and these things have their countenance; by these things our emotions are
shifted, and from emotions language comes. The fall of a solitary leaf finds its place
in our understanding [and we know that autumn is coming], (in Owen 1985: 22; em-
phasis mine).
For Chinese people, to think about autumn is to think about its correlativ-
es in the cycle of human life, in the dynastic cycle, and in all domains of
life (Owen 1985: 22; Xin 1993; Zhang 1989). In the Chinese cultural con-
sciousness, time, major (seasonal) changes in nature, human life, and the
movement of history are seen as unified, operating in a cyclic pattern.
372 Zhengdao Ye
3.2. ai ('ethical and altruistic grief/mourning')
Both bei and ai are considered as being among the cardinal emotions in
traditional Chinese texts. However they are very different in their proto-
typical cognitive structures. Ai is closely associated with someone's death,
as reflected in a cluster of death-related phrases (13). This can be regarded
as the first major difference between bei and ai.
(13) a. ai yue
ai music
'funeral music'
b. ai ge
ai song
'an elegy, a dirge'
c. ai dao
ai mourn
'to mourn for the deceased'
d. mo ai
silent ai
'stand in silent tribute (to someone's death)'
e. ai zi
ai son
'a male bereaved of this father'
f. ai si
ai thought
'sad memories (of the deceased)'
It is interesting to note that in HLM, one of the most popular Chinese
classics (which I have examined in detail in Ye, 2000 a), ai occurs in death-
related situations only. For example, when her mother dies, the character
Dai-yu is said to be shou sang jin ai (14a), and ai tong guo shang (14b):
(14) a. shou sang jin ai

guard death exhaust ai
'(She) went into strict mourning',
b. ai tong guo shang
ai bitter exceed hurt
'Grief had brought about a replace in the health'.
The close link between ai and death seems to suggest grief, which is
prototypically connected with death. However, phrases like ai si (lit. 'ai
thoughts', 'sad memories of the deceased') and ai ge (lit. 'ai song', 'an
elegy') strongly suggest that thoughts of ai are focused not on the loss
event, but on the other person. Thus, ai is very different from grief. Long,
deep and sympathetic thoughts are key elements in the meaning of ai.
Although ai is prototypically linked with death, it can also refer to other
situations where the person is suffering as a result of some disastrous
events. For example, in his famous Ai Jiangnan Fu (On Grieving Jiang-
nan'), Yu Xin (513-581 AD), exiled after the demise of the Liang dynasty,
expresses his ai over the fall of the Liang dynasty not in terms of personal
'loss', but of the suffering of the nation (ZLWZX: v. 2 (1): 224). In exile,
and thinking of his people who are suffering, Qu Yuan (3rd century) utters
the following lines in his famous poem "Li Sao":
(15) Ai minsheng zhi duo nan.

ai populace part many disaster
'Long did I sigh and wipe away my tears
To see my people bowed by griefs and fears'. (ZGWG: 67)
What Qu Yuan grieves about is the misfortune that has befallen the popu-
lation, and their long-term suffering. An exiled person can be sad thinking
about the misfortune of the nation and the suffering of its people, as re-
flected in the following conventionalised phrase:
(16) ai guo ai min

ai nation ai people
'to grieve one's nation and its people'
In (16), the country is personified. The thoughts are very much focused
on the people.
This range of uses of ai can also be found in the first Chinese anthology,
ShiJing ('The Book of Songs') dated as early as 1 lth-5th century BC:
(17) Ai wo zheng fu, du wei feimin

ai my recruited people solely passive inhuman treatment
'Alas, for us soldiers, treated as though we were not fellow-men!'
(from He Cao Bu Huang, in ZGWG)
374 Zhengdao Ye
In (17), what the experiencer feels ai about is the misery of the soldiers
who were sent on expeditionary duties and who were treated very badly.
On the one hand, "death" seems to be regarded as the prototypical situ-
ation when ai is aroused; on the other hand, it can be extended to situations
where an unfortunate event happens to people or where a nation (in this
sense, personified) is suffering. How is this going to be reflected in the
meaning of ail Using the prototype approach and relying on the semantic
prime LIKE, I would explicate the meaning of ai as follows:
ai
(b) sometimes a person thinks about another person like this:
(c) "this person died
(d) this is very bad for this person
(e) I don't want things like this to happen to this person
(f) I want to do something good for this person because of this
(g) I want to think about this person for a long time"
bad for a long time
(k) when something very bad happens to another person
(1) a person can feel something like this
(m) because a person can think something like this
This explication is constructed around two prototypes, with one, formu-

lated by components (b)-(h), embedded in another one, specified by com-
ponents from (k)-(m). The embedded one specifies a prototypical cognitive
scenario of when a person dies, with the overarching one accounting for
situations when very disastrous events befall another person. The thoughts
and the feelings of the experiencer in such situations can be explained via
references to those in a prototypical situation when another person dies.
Components (d) ("this is very bad for this person") and (e) ("I don't want
things like this to happen to this person") reflect the other-oriented charac-
teristics of ai in which the focus is on the misfortune of another person. It is
not counted as a loss to oneself. The personal character of TO ME is part of
the meaning of grief (cf. Wierzbicka, 1999: 68), but is absent from ai. Be-
cause of this, ai can not only be used for people one knows, but also be ex-
tended to one's compatriots.
With regard to component (f) ("I want to do something good for this per-
son"), bei and ai diverge fundamentally - the fatalistic view of bei
contrasts with altruistic thoughts in ai. This marks the second main differ-
ence between the two. Ai encompasses 'sympathy', 'compassion' or even
'love', revealing a wish for bad things not to happen to the other person,
and a desire to do something good for the unfortunate. In this respect, ai is
much more active than grief in that the experiencer of grief does not go
beyond dwelling on the loss ("I want to think about this; I can't think about
other things"), whereas the experiencer in ai not only actively remembers
the other party but wants to do something beneficial for them. Compo-
nent (g) - "I want to think about this person for a long time" - shows this
altruistic aspect of ai. The volition in (g) makes ai highly compatible with
lian ('loving dearly/pity'), as in ai Han:
(18) ai lian
ai pity/commiserate
'to pity; to commiserate; to have sympathy or compassion for'
(ZSHC: 151)
(? bei lian *chou lian)
The difference between ai lian and bei lian (if bei lian is used at all) is
that in the former the emotion of lian arises from the realisation that dis-
aster befalls the other, while, in the latter, it arises from the realisation of
the powerlessness of those affected and the unavoidability of the event.
When a person dies, perhaps the good wish of doing good things for this
person is best carried out by remembrance - want to think about this per-
son for a long time' (component (h)). Words such as ai si (lit. 'ai thoughts',
'sad memories of the deceased') clearly point in this direction. Ai si is a
long and lingering thought that always occupies people's minds.10 People
may commemorate a person by various means such as songs and elegies.
Component (h) does not have any suggestion of resignation, which is pres-
ent in bei. Its 'sympathy' is realised through active thoughts about the other
person.
The ethical aspect of ai, which is totally altruistic, seems to shed some
light on the positing of ai (not bei) as one of the basic emotions in Confu-
cian texts, such as Li Ji (Book of Rites) and Zuo Zhuan (A Commentary of
Zuo). Ai is valued by Confucius and his disciples. This is consistent with
376 Zhengdao Ye
two of the core values that Confucius advocated: ren ('benevolence') and
ai ('love', a different character from the ai in question). The following
examples show when ai was mentioned in Confucius' Lun Yu (The Ana-
lects).
(19) a. Zizhang yue: 'shi jian wei zhi ming, jian

name say gentleman see danger to life see
de si yi, ji si jing,
gain think of righteousness sacrifice think of awe
sang si ai, qi ke yi yi.
mourning think of ai this can PART PART
Tsu-chang [Confucius' disciple] said, "One can, perhaps, be
satisfied with a Gentleman who is ready to lay down his life in
the face of danger, who does not forget what is right at the
sight of gain, and who does not forget reverence during a sac-
rifice nor sorrow while in mourning". (LY: 19: 1)
b. Zi yue: 'ju shang bu kuan. Wei
Confucius say occupy up not magnanimous perform
// bu jing. Lin sang bu ai. Wu he
rite not awe Face mourning not ai I what
yi guan zhi zai.
with observe him PART
The Master said, "What can I find worthy of note in a man
who is lacking in tolerance when in high position, in rever-
ence when performing the rites and in sorrow when in mourn-
ing?". (LY: 3: 26)
In these quotes, ai is considered on a par with yi ('righteous') and jing
('awe/respect'). In other words, ai carries a positive moral value, whereas
sorrow or grief do not. Thus, translating ai as 'sorrow' or 'grief seems in-
appropriate in such value-specific contexts. The general idea in these
quotes is that one should show ai towards the deceased, and that when one
does this, one achieves a certain moral state. The moral connotation of ai is
in line with the use of elaborate rituals for mourning in Chinese culture. A
large proportion of Li Ji (Book of Rites) is devoted to wailing rituals
(cf. Chai and Chai 1967; see Ye 2000b for discussion of cultural rules for
ritual wailing in Chinese). The phrases ai ji andju ai reflect an important
part of mourning rituals.
(20) a. ai ji
ai sacrifice
'mourning rites' (ABC: 2)
b. ju ai
lift/uphold ai
'wail in mourning; go into mourning' (ABC: 322)
An overt display of ai is wailing. This is probably why ai often collocates
with ku ('cry') or da ku ('big cry'), and why ai seems to be often linked
with sound and voices.
The third difference between bei and ai lies in the long-term intense feel-
ing of the experiencer of ai, and the explosive, momentarily intense stir of
bei. Bei erupts like a volcano. It is easier for the experiencer to get over it,
where ai is a lingering emotion, haunting, and penetrating, which is in line
with tunes found in funeral music. It also reminds one of the wailing of
dying birds in an echoing valley, or the residue of the sound persistently
penetrating one's ears. Ai is also cumulative to the point of ai hui gu li:
(21) ai hui gu U
ai destroy bone stand
'emaciated with grief (ABC: 2)
Generally speaking, ai differs from grief in three main aspects: "person-
orientation" vs. "event-orientation" (the loss); "altruistic" ("another per-
son", "for this person") vs. "personal" ("to me"); and lingering feeling
("feel very bad for a long time") vs. not long-term feeling.
In a sense, ai is closer in meaning to mourning, in that both are related to
death and the thoughts are focussed on the other person. However, there are
some major differences. Firstly, ai, though closely related to death situ-
ations, is not exclusively limited to them. It can be extended to other situ-
ations where the targeted person is suffering greatly. Secondly, ai is a much
more intense emotion that can cause harm to the experiencer. Thirdly, the
moral connotation of ai is missing in mourning. Ai is an emotion that is
highly valued in Chinese culture. Its status as a cardinal human emotion
and the elaboration of /-related phrases are strong manifestations of its
importance in Chinese culture.
378 Zhengdao Ye
3.3. Chou ('confused sadness/worry/melancholy')
Both bei and ai can be found as early as the Shi Jing ('The Book of the
Songs', ll lh -6 th century BC) and pre-Qin texts (before 3 century BC),
where chou was an emotion concept hardly mentioned (it did not appear in
The Analects at all). However, it is chou that has become the predominant
emotional message and theme in Chinese poetry from the Qin dynasty until
the present. The very reason behind this, I believe, is meaning-motivated.
Unlike bei and ai, where the experiencer is more or less "stirred" by or di-
rected towards the third party, chou is fundamentally personal. The concept
"TO ME" is so pervasive that almost all Chinese poets have touched upon
and elaborated this theme. Poems have long been regarded as expressing
the true inner self. It comes as no surprise that chou holds such a special
position in the hearts of poets.
Gale (1992, in Myers 1996) treats chou as a key word in Chinese culture
and glosses it as 'sorrow, loneliness, and homesickness'. The most widely
used bilingual dictionary for students studying the Chinese language,
Jingxuan Yinghan H any ing Cidian (JYHC), gives the English translation
of 'worry; be anxious' under the chou entry, yet the same dictionary in its
English-Chinese section makes no mention of chou under the worry entry.
Apparently there is no equivalent or near equivalent in English for chou.
Chou is specific to the Chinese emotional experience, just as fago is to the
people of Ifaluk Atoll (cf. Lutz 1985). So what exactly is choul What is its
conceptual structure? A reading of the last four lines of a famous Chinese
poem by the lauded "immortal poet" Li Bai, "Xuanzhou Xietiaolou Jianbie
Xiaoshu Shuxu" ("At Xie Tao's High Mansion in Xuanzhou: a Parting
Banquet for the Collator Shu-yun"), would be a good starting point for an
inquiry into the meaning of chou.
(22) Chou dao duan shui shui geng liu,n

draw knife cut water water even flow
ju bei jiao chou chou geng chou.
lift goblet water chou chou even chou
Rensheng zai shi bu cheng yi,
life in world not become intention
ming zhao san fa nong bian zhou.
tomorrow morning give-off hair play with small boat
(ZLWZX: v. 3: 89)
Owen, a prominent scholar of Chinese literature, offers the following

translation for these four lines:
But I draw my dirk and cut the waters -
the water keeps flowing on.
I lift my goblet to melt away sorrow,
but sorrow continues in sorrow.
Man's life in this world may never find
what satisfies the mind -
Tomorrow at dawn let your hair flow down,
for delight sail off in your tiny boat. (Owen 1985: 140)
These four lines (with the first two lines frequently used in contemporary
popular songs) constitute an illuminating starting point for discussion in
that several important aspects of chou are embodied in them: sadness at
leaving a friend; the loneliness of setting out on a journey in a tiny boat on
the vast sea; emptiness from being "abandoned" on the vast sea; the yearn-
ing to rise up in a career and reach the heaven, that is, to be recognised (by
the emperor); depression and dissatisfaction caused by the inability to
achieve what one yearns for; confusion in not knowing what to do after
being "abandoned" by one's friends and by the emperor; the inability to put
an end to one's chou. Finally, the protagonist of the poem sets off on an
journey without a destination in a tiny boat drifting on the same water, bur-
dened with chou.
The pervasiveness and continuity of chou is like unstoppable water - the
more one wants to stop it from flowing, the more turbulent it becomes.
Often, this pervasiveness is described as chou hai (lit. 'a sea of chou', 'a sea
of worries' (ZHSC: 336)). Similarly, chou is geng chou (lit. 'even chou')
when one finds that one is unable to extricate oneself from a predicament
no matter how hard one tries. It is a struggle between wanting to find a sol-
ution to change a situation and not knowing what to do, leaving the experi-
encer caught in a dilemma wanting to overcome a difficult situation, yet
not finding a solution.
Some knowledge of the life experience of the poet Li Bai, the "banished
immortal" will help us to understand his thoughts. In his early years, Li Bai
was recommended to the Imperial Court and summoned by the Emperor to
the capital to be awarded an official title - hanlin (scholar with the highest
literary attainment). Yet, his presumptuousness offended the court and
landed him in exile. His later life was one of constant drifting and great
380 Zhengdao Ye
hardship. This poem was written in his later years, when a parting triggered
mixed feelings of loneliness, sadness, depression, confusion, and anguish
(choii). To Li Bai, this chou was concurrent with his existence and unquen-
chable, as again expressed in the last line of another of his poems "Jiang Jin
Jiu" ("Invitation to Wine"):
(23) Yu ni long xiao wan gu chou.

with you together quench ten thousand ancient chou
'To buy good wine that we may drown age-old woes.'
A typical scene or life experience that would evoke the emotion of chou
is usually cast in the setting of parting, separation, or wandering on foreign
soil without any immediate prospect of return. It is small wonder that Gale
glosses chou as 'sadness, loneliness, and homesickness'. A person living in
a foreign country or in exile is likely to have a kind of 'nostalgia' called
xiang chou (lit. 'hometown chou').12 The following lines from "Huang He
Lou" ("Yellow Crane Tower") and "Feng Qiao Ye Bo" ("Mooring at Night
by Maple Bridge") respectively, written by different poets when they saw
friends off or were away from their home, reflect these typical situations
when chou is experienced:
(24) Xi ren yi cheng huang he qu,

ancient people already ride yellow crane leave
ci di kong liu huang he lou.
this place empty leave yellow crane tower
Huang he yi qu bu fan,
yellow crane once leave not agan back
bai yun qian zai kong you you.
white cloud thousand year empty leisurely
Ri mu xiang guan he chu shi,

sun dusk homeland pass where place this
yan bo jiang shang shi ren chou.
mist wave river on make people chou
(from Cui Hao's "Huang He Lou")
The sage on yellow crane was gone amid clouds white.
To what avail is Yellow Crane Tower left here?
Once gone, the yellow crane will ne'er on earth alight,
Only white clouds still float in vain from year to year.

Where is my native land beyond the setting sun?
The mist-veiled waves of River Han make me home-sick.
(25) Yue lou wu ti shuang man tian,

moon descend crow cry frost fill sky
jiang feng yu huo dui chou mian.
river maple fishboat fire towards chou sleep
Gu su cheng wai Han shansi,
name of a city wall outside chill hill temple
ye ban zhong sheng dao ke chuan.
night half bell sound arrive guest boat
(from Zhang Ji's Feng Qiao Ye Bo)
The moon goes down, crows cry under the frosty sky,
Dimly-lit fishing boats 'neath maples sadly lie.
Beyond the Gusu walls the Temple of Cold Hill
Rings bells which reach my boat, breaking the midnight still.
In parting from a friend, chou is diffusive like a white cloud in the vast
sky (24). In drifting on a boat far away from home during the long night,
chou is pervasive like the endless night itself. One can imagine the toss and
turn of the poet in the long, dark night. In both situations, chou is marked
by loneliness and preoccupied thoughts: "I cannot not think of my best
friend and my hometown".
The numerous writers who have touched on the emotion of chou include
an emperor, Li Yu (937-978 AD), the last monarch of the South Tang dyn-
asty. Let us first look at Li Yu's chou in some of his most famous poems, es-
pecially the last one (25), which has been set to a popular song:
(26) a. Jian bu duan, li hai luan,

cut not break order still mess
shi li chou,
this separation chou
bieshi yifan ziwei zai xintou.
especially a kind of taste in heart
(from Xiang Jian Huan)
Such ties of affection cannot be cut.
Reason still disordered; woeful to part.
382 Zhengdao Ye
We said good-by as friends commonly did.

How I have felt is best known to my heart.
b. Bie lei chun ban, chu mu chou chang duan.
apart come spring half, touch eye chou intestine break
(from Qing Pin Yue)
Parted from homeland for half the spring
All that strikes my eye brings only woe.
c. Chun hua qiu yue he shi liao,
spring flower autumn moon what time finish
wang shi zhi duo shao?
past matters know much little
Xiao lou zuo ye you dong feng,
tiny pavilion yesterday night again east wind
gu guo bu kan hui shou ming yue zhongl
old country not bear back head bright moon in
Diao Ian yu qi yin you zai,
carve column jade step should still be
zhi shi zhu yan gai.
only be red complexion change
Wen jun neng you ji duo chou,
ask you can have how much chou
qia si yi Jiang chun shui xiang dong liu.
just like one river spring water towards east flow
(from "Yu Mei Ren")
Too long the autumn moon and spring flowers last,
I wonder how much they've known of my past.
Last night spring breezes through an upper room -
Remind me too much of my present gloom.
With a bright moon, how could I my country recall -
Without a sense of defeat and despair at all.
The Palace should be still there as before -
With its carved railings; jade-like steps galore.
Only here are changes which my plight entail.
My complexion, once ruddy, has become pale.
Should I be asked how much anguish I have found.
Strange! It is like flowing water, eastward bound.
Here Li Yu shows us his most private and troubled feeling: chou. Li Yu's
life went through a series of drastic changes. In an earlier period, he lived
an extremely luxurious court life filled with love and joy. Later, he was a
political prisoner after losing his kingdom, left to live a life of melancholy
and desolation (cf. ZLWZX). Beautiful things like spring flowers and the
autumn moon could only arouse his deep sorrow. The turbulent eastbound
water is an extension of on-going chou, in time and space. Unbearable nos-
talgia for his lost kingdom, rejection of his present situation, and the diffi-
culties of extricating himself from that undesirable situation left Li Yu fad-
ing away in ill health- 'My complexion, once ruddy, has become pale'.
Based on the examples and discussion, the meaning of chou can be ex-
plicated as follows:
chou
(c) "something very bad is happening to me
(d) before this, I did not think this would happen
(e) I don't want things like this to happen to me
(f) because of this, I want to do something if I can
(g) I don't know what to do
(h) I cannot not think about this all the time"
(i) when this person thinks this, this person feels something bad for a
long time
(j) X felt something like this
(k) because X thought something like this
Component (c) ("something very bad is happening to me") reflects the
personal characteristic of chou and the miserable situation that the experi-
encer encounters. The happening is closely related to the experiencer's per-
sonal benefit. This difficult situation is not just momentary, but continuing
and co-existent with the experiencer of chou. It is like an obstacle. Compo-
nent (d) ("before this, I did not think this would happen") indicates that the
event is unexpected, leaving the experiencer unprepared. While rejecting
the adverse situation (component (e) - "I don't want things like this to
happen to me"), the experiencer wants to do something to change the s'tu-
ation (f) ("because of this, I want to do something if I can"), yet he or she is
caught up in confusion, not knowing what can be done (component (g) - "I
384 Zhengdao Ye
don't know what to do"). The struggle between wanting to find a solution
and not finding one is clearly exemplified in the poems cited above,
through, for example, the interplay of thoughts and actions while drinking
wine, tossing through the chilly night and lamenting in a beautiful season.
Maybe the only thing a person can do is to try to forget chou. Li Bai made
the effort, yet the feeling of chou was so haunting that it could not even be
dissolved by wine. Jie jiujiao chou (lit. 'to quench chou with wine') has
become a typical action when people experience chou. Unlike bei, which is
easy for the experiencer to get over, chou stays with the experiencer. How
can a poet forget the honour of the older days; how can one forget one's
hometown; how can an emperor forget his lost kingdom? The persistence
and unforgettableness of chou is embodied in component (h) - "I cannot
not think about this all the time". "Feel something bad for a long time" in
component (i) accounts for the prevailing preoccupied feeling, and the on-
going misery, like eastbound water spreading over time and space.
The meaning of chou is crystallised in the poems of the most outstanding
and influential Chinese poetess, Li Qingzhao (1084-1151 A.D.). Chou ap-
pears in almost every one of her poems composed in later years. Among
them, "Sheng Sheng Man" ("Every Sound Lentamente"), is a condensed
outpouring of chou. The famous opening lines epitomise the thought "I
don't know what to do" (i). The poem goes:
(27) Xun xun mi mi,

seek seek search search
leng leng qing qing,
cold cold lone lone
qi can can qi qi,13
sad sad pain pain moan moan
zha nuan hai han shi hou 9
just warm still chill time

zui nan jiang xi.
most difficult go rest
San bei Hang zhan dan jiu,
three goblet two wine cup weak wine
zeng di ta wan lai feng ji!
how resist its late come wind violent
Yan guo ye,
wild goose pass PART
z he n g s hang xin,
PROG hurt heart
que shi jiu shi xiang shi.
but be old time acquaintance
Man di huang hua dui ji,
fill ground yellow flower pile accumulate
qiaocui sun,
haggard/thin/pallid fade away/damage
rujin you she kan zhai!
these days there be who bear pick
Shou zhe chuangr,
guard DUR window
du zi zen shengde hei?
one self how bear black
wutong geng jian xi yu,
wutong more with fine rain
dao huanghun,
till dusk,
dian dian di di.
dot dot drip drip
Zhe cidi,
this situation
zeng yi ge chou zi liaode!
how one CL chou character PART (say all)
I've a sense of something missing I must seek.
Everything about me looks dismal and bleak.
Nothing that gives me pleasure, I can find.
Even the weather has proved most unkind.
It is warm, but abruptly it turns cold again.
An unbroken rest - most difficult to obtain.
Three cups of thin wine would utterly fail -
To cope with the rising evening gale.
Myself, into woe, a flight of wild geese has thrown.
But with them, very familiar I have grown.
About the ground, chrysanthemums are bestrewn.
Gathering into heaps - bruised - withering soon.
With myself in utter misery and gloom,
Who cared to save them from their approaching doom?
386 Zhengdao Ye
Standing by the window, watching in anguish stark,

Could I bear alone the sight until it is dark?
Against the tung, the plane trees, the wind rises high,
The drizzle becomes trickles, as even draws nigh.
How in the word 'miserable', can one find-
The total effects of all these on the mind!
Like Li Bai and Li Yu, Li Qingzhao also went through drastic and cruel
vicissitudes in her life. Li Qingzhao had a perfect marriage. She and her
husband shared the same ideals and were utterly devoted to each other.
They composed poetry together and spent all their money on books, scrolls
of art and curios. Both were famous connoisseurs at that time. Unfortu-
nately, the invasion of Northern China by a minority group broke up Li's
happy marriage. In fleeing from their country, her husband died of illness
while the bulk of their priceless collections of books, scrolls and curios
were lost in the war. Alone, and away from her country, Li Qingzhao spent
the rest of her life in misery, loneliness and profound melancholy. In
"Sheng Sheng Man", she expresses her nostalgia for her native town, her
old home and beloved husband, recalling the congenial happy life that is no
more.
No other piece of work captures the invariant core of the meaning of
chou more vividly than this poem. The two pairs of monosyllabic words in
the opening line xun xun mi mi ('seek-seek, search-search') highlight the
state of "I want to do something" but "I don't know what to do". The piling
up of the withered flowers and the tempest wind all symbolise the intense
bitterness of chou that reduced her to qiaocui sun ('withered'), just like the
bruised and withered yellow flower.
Component (c) ("something very bad is happening to me now") links
chou with "sadness-like" emotions. The ongoing nature of the event links
chou with "worry-like" emotions. It is not surprising that chou is often
glossed as worry. The future-oriented feature of chou sets it apart from
both bei and ai, which are directed towards events which happened in the
immediate past. Preoccupied thoughts and the constant desire to do some-
thing give chou an earthly and material touch, in which the experiencer
cannot be transcendent, but instead is deeply locked into the situation. Res-
ignation is absent from the thoughts of the experiencer, and this sets it apart
from sorrow. The experiencer of chou never gives up trying to find a sol-
ution. It is probably because of this that chou is also "mundanely" used in
the daily life of ordinary people, particularly when referring to material

concerns. One of the most common contexts in which it is used is as fol-
lows:
(28) chou chi chou chuan
chou eat chou wear
'To have to worry about food and clothing; to suffer from poverty.'
(ZSHC: 337)
The phraseology describing chou in Chinese is rich and vivid. Most no-
ticeable is that chou is often described by reference to the body - both the
inner organs and the external body parts, especially the face. In terms of
inner organs, the intestines are perceived by the Chinese people to be the
seat of this emotion, as is reflected in the lexicalised item chou chang (lit.
'chou intestines') and other chou c/iang-related chengyu such as the fol-
lowing (also recall Li Yu's reference to chou chang duan 'chou intestines
broken' in 26b) (cf. Ye 2000a):
(29) a. chou chang cun duan
chou intestines inch break
'The sorrow is so deep that it seems to have cut the bowels
into pieces'.(ZSHC: 337)
b. chou chang bai jie
chou intestines hundred knot
'It is as if the worries were tied in knots to one another'.
(ZSHC: 337)
The feeling of chou is so intense that it is "leaked" and etched on the
whole face of the experience^ as evidenced by the following conventional-
ised phrases:
(30) a. chou rong man mian
chou look full face
'to wear a sad look; to look distressed' (ZHSC: 337)
*chou se man mian
chou colour/look full face
b. chou mei ku Han
chou eyebrows bitter face
'look worried and miserable; have a worried look' (JYHC:
69)
388 Zhengdao Ye
c. chou mei bu zhan

chou eyebrows not unfold
'knit one's brows' (JYHC: 69)
d. chou mei jin suo
chou eyebrows tightly lock
'chou eyebrows deadly locked'
These phrases provide valuable linguistic evidence, showing the external
manifestation of the emotion of chou. Two observations are particularly
important. One is that the expression of chou involves the whole face, as
evidenced by its collocation with rong ('looks/expressions/appearances'),
not se ('colour/countenance/look'). The main difference between rong and
se is that the former indicates a more intense facial expression, and empha-
sises the presentation of the whole face, whereas the latter suggests a less
intense facial expression that simply brushes over the face, and often in-
volves a change of facial colour (cf. Ye 2000 a).
The other observation is chou mei ('chou eyebrows'). The chou brows
are the marked feature of a chou face. The existence of a facial feature
named after chou shows that this emotion is perceived as having a very dis-
tinct facial expression, so much so that it warrants a description of its own.
As suggested by (30c) and (d), chou eyebrows are brows that are tightly
locked, drawn together for a long time, as if one were searching for a sol-
ution to overcome an obstacle. This, to some extent, iconically mirrors the
inner state of the experiencer of chou: thinking of a way out, while caught
in confusion ("I want to do something; I don't know what to do") (see Ye
Forthcoming).14
The lasting and intense feeling of chou, and the potential physical harm
that chou can inflict on the experiencer, are reflected in the following com-
monly used phrases of resultative complements (Meng et al., in Kornacki,
1995: 207):
(31) a. chou huai le

chou spoil PFV
'to be ruined by chou'
b. chou bing le
chou ill PFV
'to get sick from chou'
c. chou ku le
chou cry PFV
'to cry from chou'
d. chou bai le toufa
chou white PFV headrhair
'to grow white-haired from chou'
e. chou si le
chou die PFV
'to be dying from chou'
Compared to bei and ai, chou is inextricably linked with a personal pre-
dicament that is forced on one by circumstances. It is active both in "want-
ing to do something" and in preoccupied thoughts. What sets it apart from
both bei and ai is its future-oriented thoughts. In this regard, chou is quite
far away from "sadness-like" emotions. What allows it to retain its link
with "sadness-like" emotions is that chou responds to a concurrent event. It
is not surprising, then, that the meaning of chou has been confused, its
glosses ranging from 'sadness', 'sorrow', and 'melancholy' to 'worry' and
'anxiety'.
4. Sadness - Universal?
Sadness has been proposed as one of the basic emotions by different psy-
chologists in the study of emotions (e.g. Ekman 1992; see Plutchik 1994
and Van Brakel 1993 for different lists of basic emotions). In this chapter,
the precise analysis of the cognitive structure of Chinese "sadness-like"
emotion terms, which are considered to be basic emotions in Chinese cul-
ture, provides convincing empirical evidence that what Chinese people
consider to be basic emotions are very different from those proposed in the
English language. If the postulation of basic emotions were based on the
Chinese language, it could include bei ('fatalist/tragic sadness') and ai
('ethical/altruistic grief/mourning'). Neither of these emotion terms, as
discussed in this paper, have exact English equivalents. There is no reason
to assume that bei and ai are not as basic as sadness (cf. Wierzbicka 1992;
1999).
The "basic emotions" proposed by Chinese people single out what the
Chinese people consider to be important in Chinese culture. The basic
390 Zhengdao Ye
emotions of the Chinese people are shaped by Chinese culture. The whole
concept of bei, which highlights a sense of powerlessness due to the laws
of nature and which is coloured by Chinese people's cyclic view of life
and change, may be absent in other cultures. Bei is more tragic and fatal-
ist, involving a momentary transference. Its association with nature and
seasonal changes is characteristic of Chinese people's emotional experi-
ences. Ai, though involving 'loss', is ethical in nature, arousing altruistic
compassion and sympathy. Although chou does not appear on the lists of
"basic emotions" proposed in traditional Chinese texts, its frequency con-
firms its salience in Chinese emotional experience. It is personal, intro-
spective, mundane and ongoing. English does not share the same Gestalt
as the Chinese "sadness" experience. The claim that sadness is a universal
emotion category is biased towards the English language. It is also inter-
esting to note that there is a strong connection between "sadness-like" and
"worry-like" (as well as "love-like") emotions in Chinese (cf. Shaver, Wu
and Schwartz 1992; Ye 2000 a). Intuitively, and as is often the case in clas-
sifying dictionaries, chou is grouped with you ('loving concern'), another
emotion term bordering on "sadness-like" and "worry-like" emotions,
forming a separate category from both bei and ai. To simply gloss chou as
sadness would distort the true picture of the Chinese people's categoriz-
ation of emotion domains. Solomon has succinctly pointed out this pre-
sumption:
The problem, among others, is that we take the category of emotions to be something
settled, and then wonder whether or to what extent the Chinese category corresponds
to our own. But, of course, it does not, not only because we do not have any such
settled category ourselves but because there is no reason whatever to think that the
Chinese concept - which seems to have more to do with what we might call "auth-
enticity" than with a particular realm of "the mind" - bears any interesting relation-
ship to the kind of classificatory category that 'emotions' is supposed to be in Eng-
lish. (Solomon 1995: 179)
This points to a broader issue of whether or not there are universal emo-
tions - a question that has been one of the central concerns in the study of
emotions among psychologists (cf. Ekman and Davidson 1994). Indeed,
there is no reason to think that the question of whether there are basic emo-
tions is valid. As Shweder rightly argues: ' "Are there basic emotions?" is
such a deceptively simple question that it deserves to be deconstructed'"
(Shweder 1994: 32).
To totally reject the quest for the universality of human nature is not a
constructive attitude. The analyses in this paper, on the other hand, show
that the quest for a common core among human beings is not a question of
all or nothing. One can extract the common structure once the cognitive
scenario is explicated in universal human concepts. On the one hand, bei,
ai, chou and sadness have very different cognitive structures, on the other,
they share a semantic invariant core of "something bad happened/is hap-
pening", which may cross-linguistically link these Chinese emotion terms
with other emotion concepts that are constructed upon this cognitive
basis. What is basic is not the emotion terms, but the shared common "re-
sponse elements" (Scherer 1994: 25) that can be couched in lexical uni-
versals. It is the cognitive element found in emotions that is fundamental
to the universality of emotion. The configuration of these elements is
shaped by culture. They are artefacts of the Chinese culture, shaped by
uniquely Chinese experiences and views of life and the universe.
5. NSM in the Chinese context

Central to the idea of framing definitions in NSM is that they can be trans-
lated into other languages with maximum neutrality. This neutrality
requires not only culture-free conceptual primitives, but also universal
syntax. In the last decade, much work has been done to investigate the
combinatory behaviour of the primitives that is universal, and this has
yielded important findings (Goddard and Wierzbicka 1994, In press). The
explications offered in this paper have adopted the findings arrived at in
this work, and have striven for the simplicity and universality of the NSM
syntax. This section, as an empirical experiment, will present the expli-
cations proposed in this paper in the Mandarin Chinese primitive expo-
nents as identified by Chappell (1994, In press). The Chinese semantic
primitives are listed in the following table:
392 Zhengdao Ye
Substantives:
wo ni shei shenme renmen~ren shenti
I YOU SOMEONE SOMETHING PEOPLE-PERSON BODY
Determiners:
zhei(ge) tong ling(wai)
THIS THIS:SAME OTHER
Quantifiers:
yi Hang (ge) dou duo you de
ONE TWO ALL MANY/MUCH SOME
Evaluators:
hao huai ~ bu hao
GOOD BAD
Descriptors:
da xiao
BIG SMALL
Mental Predicates:
xiang zhidao ganjue-gandao yao kan dao ting dao
THINK KNOW FEEL WANT SEE HEAR
Speech:
shuo zi zhen
SAY WORD TRUE
Actions, Events and Movement:
zuo fasheng dong
DO HAPPEN MOVE
Existence and Possession:
(zai PLACE) you X X you
THERE IS/THERE ARE HAVE
Life and Death:
huo si
LIVE DIE
Time:
shihou xianzai yiqian yihou henjiu
TIME/WHEN NOW BEFORE AFTER A LONG TIME
duanqi you yi duan shijian
A SHORT TIME FOR SOME TIME
Space:
shenme difang ~ nar shang xia yuan jin
WHERE/PLACE ABOVE UNDER FAR NEAR
pang(bian) limian zher
SIDE INSIDE HERE
Logical Concepts:
bu keneng hui yinwei ruguo
NOT MAYBE CAN BECAUSE IF
Intensifier and Augmentor:
hen zai... duo
VERY MORE
Metonymy and Taxonomy:
X you ~ bufen zhong
IS A PART OF X A KIND OF
(double subject construction)
Similarity:
xiang... zheyang ~yiyang
zheyang Verb
LIKE
The parallel explications of bei, ai, and chou in English and Chinese primi-
tives look as follows.15
bei
(c) "something bad happened now
(d) I know that after this good things will not happen any more
(e) I do not want things like this to happen
(f) I know I cannot do anything
(g) because I know no one can do anything when things like this
happen"
bad
394 Zhengdao Ye
bei
(a) X ganjuedao le shenme, shi yinwei X xiangdao le xie shenme
(b) youshihou, yi ge ren hui zheyang xiang:
(c) "buxin de shiqing fasheng le
(d) wo zhidao zhi yihou, haoshi buhui zai you
(e) wo buyao fasheng zheyang de shiqing^
(f) wo (dui zhejian shi) zuobuliao shenme
(g) yinwei, wo zhidao zai zhezhong shi fasheng de shihou,
meiyou ren neng zuo xie shenme"
(h) zhe ge ren zheyang xiang de shihou, zhe ge ren xinli ganjue hen
buhao
(i) X you xiang zheyang de ganjue
(j) shi yinwei X you zheyang de xiangfa
ai
(b) sometimes a person thinks about another person like this:
(c) "this person died
(d) this is very bad for this person
(e) I don't want things like this to happen to this person
(f) I want to do something good for this person because of this
(g) I want to think about this person for a long time"
bad for a long time
(k) when something very bad happens to another person
(1) a person can feel something like this
(m) because a person can think something like this
ai
(b) youshihou, yi ge ren hui zheyang xiang lingwai yi ge ren:
(c) "zhe ge ren si le
(d) dui zhe ge ren lai shuo shi henda de buxing
(e) wo buyao zheyang de shi fasheng zai zhe ge ren shenshang

(f) suoyi wo yao wei zhe ge ren zuo dian shenme
(g) wo yao henchang shijian xiang zhe zhe ge ren"
(h) zhe ge ren zheyang xiang de shihou, zhe ge ren henchang shi-
jian xinli ganjue hen huai
(i) X you zheyang de ganjue
(j) shi yinwei X you zheyang de xiangfa
(k) dang hen buxin de shi fasheng zai lingwai yi ge ren shenshang de
shihou
(1) yi ge ren hui you xiang de ganjue
(m) zhe shi yinwei yi ge ren hui you xiang zhenyang de xianfa
chou
(a) X felt something, because X thought something
(c) "something very bad is happening to me
(d) before this, I did not think this would happen
(e) I don't want things like this to happen to me
(f) because of this, I want to do something if I can
(g) I don't know what to do
(h) I cannot not think about this all the time"
(i) when this person thinks this, this person feels something bad for a
long time
(j) X felt something like this
(k) because X thought something like this
chou
(b) youshihou, yi ge ren hui zheyang xiang:
(c) "henbuxin de shiqing zheng fasheng zai wo shenshang
(d) zai zhe yiqian, wo meiyou xiangdao hui fasheng zheyang de
shiqing
(e) wo buyao zheyang de shiqing fasheng zai wo shenshang
(f) suoyi, wo yao jin wo suoneng de zuo dian shenme
(g) wo bu zhidao wo neng zuo shenme
396 Zhengdao Ye
(h) wo bunengbu lao xiang zhe zhejian shi"

(i) zhe ge ren zheyang xiang de shihou, zhe ge ren henchang shijian
ganjue hen buhao
(j) X you xiang zheyang de ganjue
(k) shi yinwei X you zheyang de xiangfa
These Chinese explications via the Chinese primes are, generally speak-
ing, quite natural. This "feasibility study" of the NSM framework in the
Chinese language lends support to the universal applicability of NSM. The
idea of introducing a value-free interpretive and comparative tool into
"ethno" studies in general is not just necessary, but possible.
Blowers (1996) points out that there have been more recent attempts to
study indigenous concepts from indigenous viewpoints (undertaken by
Chinese researchers), but that there has been little enthusiasm for reporting
them to the English-speaking world. Blowers does not mention the reason
for this lack of enthusiasm. Presumably it was because of mismatches be-
tween concepts in other languages. There are indeed no exact equivalents
in English (or in other languages) for Chinese indigenous concepts. Re-
searchers do not wish an incorrect or inaccurate label to be attached to in-
digenous expressions. However, as this paper has demonstrated, the prob-
lem is not insoluble. With the Mandarin Chinese semantic primitive set and
its syntax, Chinese researchers can explicate meanings of indigenous con-
cepts accurately and neutrally and, equally importantly, they can explain
indigenous concepts from an insider's point of view. The explications in
NSM can be safely translated into semantic primitives identified in other
languages. This could facilitate the two-way information flow among re-
searchers across cultures, and contribute to a better understanding of in-
digenous concepts among researchers across different disciplines.
6. Conclusions
This paper has undertaken a textual based semantic analysis of three emo-
tions terms - bei, ai, and chou - that are often translated as equivalents of
sadness, grief or sorrow. Bei and ai are considered basic emotions in Chi-
nese. Using the semantic analytical framework of NSM, the paper has
shown that these three "sadness-like" emotions do not have exact semantic
equivalents in English. The claim that sadness is a universal emotion is
called into question. The case of chou also shows that simply labelling
chou with other "sadness-like" terms hinders a true understanding of the
Chinese categorization of the emotion domain.
As this paper has illustrated, framing definitions in NSM has three main
advantages in the study of emotion. Firstly, it makes possible the expla-
nation of meanings from an insider's perspective. Secondly, the exact dif-
ferences and connections between concepts, within a culture and across cul-
tures, can be clearly identified. This point is most clearly illustrated by com-
paring, on the one hand, ai and bei, and, on the other, ai and grief. Thirdly, it
allows definitions to be translated into different languages while retaining
neutrality. Emotion terms are like the tip of an iceberg with a deep and co-
lossal block of meaning underneath. The NSM approach allows us to ex-
plore this. This paper has also shown that lexical items, conventionalized
expressions and literary examples are all valuable linguistic sources that
should not be overlooked. The lexicalisation of words such as bei qiu (lit.
'bei autumn') bears witness to the Chinese view of life and the universe, and
to the role that seasons play in shaping Chinese emotional experiences.
Once the meaning is unpacked, the "cultural logic" is readily untangled. The
elucidation of meanings for bei and ai sheds light on the "logic" of their dif-
ferent status in Daoist and Confucian versions of "basic Chinese emotions".
Linguistic theory can contribute to the study of emotions, and the study of
emotions can, and will, benefit from an interdisciplinary perspective.
Notes
1. I am deeply indebted to Professor Anna Wierzbicka for the perspective on cross-
cultural semantics that her work has opened for me, and for her encouragement and
support. I also wish to express my sincere thanks to Dr. Jean Harkins for the many
stimulating talks on various issues of emotions and for her valuable suggestions and
continuous support. My thanks also go to Professor Wenguo Pan, Dr. Xiaonong Zhu,
and Ms Changxin Cong for spending time with me discussing the Chinese emotion
universe and the Chinese translations of the explications.
2. This idiom, jingdizhiwa, originated in the writing of the Chinese Taoist philosopher
Chuang Tzu (cf. HCDC: 265; Feng, 1953).
3. Chinese, in this paper, refers to Mandarin Chinese. The romanisation used through-
out that paper is the Pinyin system. For ease of notation, tones are not marked.
4. I use zi not just to refer to the basic graphic or writing unit, but to an integral unit
constructed of characters where sound, form and meaning converge (cf. Xu 1997,
1998).
398 Zhengdao Ye
5. The Chinese frequency dictionary (XHZTB) shows that that the number of occur-
rences of bei, ai, and chou from a computation of 11,873,029 randomly sampled
characters in a corpus of Social and Natural Sciences are 835,328, and 314 respect-
ively, ranking 1525, 2040, and 2199 among 7754 characters. This means that they
are of relatively high frequency.
6. The provision of these closest (albeit idiosyncratic) English glosses is the task of a
separate project headed by Professor Paolo Santangelo of Istituto Universitario
Orientale of Naples. This project aims at compiling a glossary of expressions and
terms for emotions and mental states used in Chinese texts of the late imperial peri-
od (personal communication).
7. Most linguistic examples follow a tripartite notation, with first line romanisation,
second line morpheme to morpheme notation, followed by a literal translation.
Names in translations (except those that are conventionalised names such as
Chuang Tzu), if in the Wade-Giles system, are changed to thepinyin system for the
purpose of consistency. Where the texts are quite lengthy, only English translations
are given. Where sources are unspecified, the translation is literal. The abbrevi-
ations used for morpheme-to-morpheme notation are as follows: CL = classifier;
DUR = durative; PART = particle; PFV = perfective aspect; PROG = progressive.
8. An explication is a definition formulated through a configuration of semantic primi-
tives. It takes the form of a componential analysis, which is in line with the general
view of psychologists that emotions are "multicomponential phenomena" (Scherer
1994; cf. Frijda, Sato, and Wiers 1995). The semantic framework for explicating
emotion concepts adopted here has been developed by Wierzbicka (e.g., 1972,
1986, 1992, 1999). Further detailed explanations of the rationale behind the NSM
framework for emotion words, and how it reflects and concurs with latest develop-
ments and findings in psychological and anthropological studies of emotions, can
be found in Harkins (1990), Harkins and Wierzbicka (1997), and Ye (2000a).
9. The association between apricots and Xiu-yan is based on the homonymic relation-
ship between xingzi ('apricot') and the surname of Xing Xiu-yan.
10. Si ('thoughts/thinking/loving') plays an important role in Chinese emotions. It is
considered one of the "basic Chinese emotions" in traditional texts, and forms its
own distinct category. Liang (1996: 29) points out that, compared to Plutchik's
(1994) multidimensional model of the emotions, the Chinese classification of basic
emotions has an emphatic si category that cannot be contained within Plutchik's
model. The presence of this emotion category seems to indicate that 'thinking' has
an important status in shaping Chinese emotional experiences.
11. The English translations in the following poems are all from ZGWG. All the poems
cited are well-known among Chinese people and some have been set to popular
songs.
12. Xiang chou is a very important theme among Chinese people living abroad. Articles
on xiang chou are common in newspapers such as the overseas version of the
People's Daily.
13. High pitch is usually associated with 'happy' feelings in English (cf. Hinton, Ni-
chols, and Ohala 1994). The Chinese lines here seem to suggest a different inter-
pretation. This matter requires further investigation.
14. Wierzbicka (1999: 201) explains the meaning of brows drawn together as "I am
thinking now; I want to do something now; I am not doing it now". The expression
chou mei supports this interpretation. This suggests that what is universal is the
iconic meaning of some universal facial gestures. But the interpretation of a par-
ticular facial expression may vary from culture to culture.
15. The syntactic aspect of the combinatorial behaviour of Chinese primitives cannot
be discussed here. It requires a separate study to fully investigate various issues
such as valency options (see Chappell 1994 and In press for a detailed treatment of
this issue).
16. For Mandarin speakers, yao ('want') sounds less natural than xiang when it is ne-
gated by bu. For speakers of other dialects (for example Wu speakers), yao sounds
much more natural.
References
Blowers, Geoffrey H.
1996 The prospects for a Chinese psychology. In: Bond (ed.), 1-14.
Bond, Michael Harris (ed.)
1996 The Handbook of Chinese psychology. Hong Kong: Oxford University
Press.
Chai, Ch'u - Winberg Chai (eds.)
1967 Book of Rites: an encyclopedia of ancient ceremonial usages, religious
creeds, and social institutions. New York: University Books.
Chao, Yuan Ren
1968 A Grammar of Spoken Chinese. Berkeley: University of California
Press.
Chappell, Hilary
1994 Mandarin semantic primitives. In: Goddard and Wierzbicka (eds.),
109-47.
In press Universal syntax and Mandarin Chinese grammar. In: Goddard and
Wierzbicka (eds.).
Chen, Lifu
1977 An Introduction to Chinese Culture. Taipei: The Council of Chinese
Cultural Renaissance.
Ekman, Paul
Ekman, Paul and Richard J. Davidson (eds.)
1994 The Nature of Emotion: Fundamental Questions. Oxford: Oxford Uni-
versity Press.
Fang, Tung-mei
1957 The Chinese View of Life: The Philosophy of Comprehensive Har-
mony. Hong Kong: Union Press.
400 Zhengdao Ye
Fehr, B. and James A. Russell

1984 Concept of emotion viewed from a prototype perspective. Journal of
Experimental Psychology: General Section 113: 464-86.
Feng, Yu-lan
1953 A History of Chinese Philosophy (translated by D. Bodde). Princeton:
Princeton University Press [1934].
Frijda, Nico H., Kaori Sato and Reinout Wiers
1995 Emotions and emotion words. In: James A. Russell, Jose Miguel Fern-
andez-Dols, Antony Manstead and J. C. Wellenkamp (eds.), Everyday
Conceptions of Emotion: An Introduction to the Psychology, Anthro-
pology, and Linguistics of Emotion, 121-44. Dordrecht and Boston:
Kluwer Academic.
Goddard, Cliff
1998 Semantic Analysis: A Practical Introduction. Oxford: Oxford Univer-
sity Press.
Harkins, Jean
1990 Review of A. Ortony, G. L. Clore, and A. Collins, The Cognitive
Structure of Emotions (Cambridge University Press, 1988). Australian
Journal of Linguistics 10: 277-87.
1997 Language: a key issue in emotion research. Innovation in Social
Sciences Research 10(4): 319-31.
Hinton, Leanne, Johanna Nichols and John J. Ohala
1994 Sound Symbolism. Cambridge: Cambridge University Press.
Kirkpatrick, John and Geoffrey M. White
1985 Exploring ethnopsychologies. In: Geoffrey M. White and John Kirk-
patrick (eds.), 3-34.
Kornacki, Pawel
1995 Heart & face: semantics of Chinese emotion concepts. Ph.D disser-
tation, Australian National University.
Li, Charles and Sandra Tompson
1981 Mandarin Chinese: A Functional Reference Grammar. Berkeley: Uni-
versity of California Press.
Liang, Chengmou
1996 Qiqing shuo yu xiandai qingxu xinlixue [The theory of seven basic
emotions and modern psychology of emotions]. Nanjing Shifan Daxue
Xuebao: Shehui Kexue [Journal of Nanjing Normal University: Social

Sciences], 4: 64-7.
Lu, Zhiwei
1964 Hanyu de goucifa [Chinese word-formation]. Beijing: Kexue Chub-
anshe.
Lutz, Catherine
1985 Ethnopsychology compared to what? Explaining behavior and con-
sciousness among the Ifaluk. In: Geoffrey M. White and John Kirkpa-
trick (eds.), 35-79.
1987 Goal, event, and understanding in Ifaluk emotion theory. In: Dorothy
Holland and Noami Quinn (eds.), Cultural Models in Language and
1988 Unnatural Emotions: Everyday Sentiments on a Micronesian Atoll
and their Challenge to Western Theory. Chicago: University of Chica-
go Press.
1995 Need, nurturance, and the emotions on a Pacific atoll. In: Joel Marks
and Roger T. Ames, (eds.), 253-252.
Marks, Joel
1995 Emotions in Western thought: some background for a comparative dia-
logue. In: Joel Marks and Roger T. Ames (eds.), 1-38.
Marks, Joel and Roger T. Ames (eds.)
1995 Emotions in Asian Thought: A Dialogue in Comparative Philosophy.
Albany, N.Y.: State University of New York Press.
Myers, Dan
1996 Strategic waiting in China: an ethnosemantic approach. In: Li Wei
(ed.), Chinese Language, Culture and Communication. Special Issue
of Journal of Asian Pacific Communication, 7(3-4): 91-105.
Owen, Stephen
1985 Traditional Chinese Poetry and Poetics: Omen of the World. Madison,
Wisconsin: University of Wisconsin Press.
Pan, Wen-guo, Po-Ching Yip, and Yang Han
1993 Hanyu goucifa yanjiu [Studies of Chinese word-formation:
1898-1990.] Taiwan: Student Book.
Plutchik, Robert
1994 The Psychology and Biology of Emotion. New York: Harper Collins.
Qian, Zhongshu
1998 Guan Zhui Plan [Limited views: essays on ideas and letters] (selected
and translated by Ronald Egan). Massachusetts: Harvard University
Asia Centre.
Russell, James A.
1991 Culture and the categorization of emotion. Psychological Bul-
letin 110:426-50.
402 Zhengdao Ye
Russell, James A. and Michelle Yik

1996 Emotion among the Chinese. In: Michael Harris Bond (ed.), 166-88.
Scherer, Klaus R.
1994 Towards a concept of 'Modal Emotions'. In Paul Ekman and Richard
J. Davidson (eds.), 32-44.
Schwartz, T., Geoffrey M. White and Catherine Lutz (eds.)
1992 New Directions in Psychological Anthropology. Cambridge: Cam-
Shaver, P., J. Schwartz, D. Kirson and C. O'Connor
1987 Emotion knowledge: further exploration of a prototype approach.
Journal of Personality and Social Psychology, 52: 1061-1086.
Shaver, P., Wu, S and J. Schwartz,
1992 Cross-cultural similarities and differences in emotion and its represen-
tation. In Margaret S. Clark (ed.), Emotion. Review of Personality and
Social Psychology 13: 175-212. Cambridge: Cambridge University
Press.
Shweder, Richard
1994 'You're not sick, you're just in love': emotion as an interpretive sys-
tem. In: P. Ekman - Richard J. Davidson (eds.), 32-44.
Shweder, Richard and R. A. LeVine (eds.)
1984 Culture Theory: Essays on Mind, Self, and Emotion. Cambridge: Cam-
Solomon, Robert C.
1995 The cross-cultural comparison of emotion. In: Joel Marks and Roger
T. Ames (eds.), 253-308.
Tollefson, James W.
1995 Power and Inequality in Language Education. Cambridge: Cambridge
University Press.
Van Brakel, Jaap
1993 Emotions: a cross-cultural perspective on forms of life. In: W. M.
Wentworth and J. Ryan (eds.), Social perspectives on emotion, v. II:
179-237. Greenwich: JAI.
White, Geoffrey M. and John Kirkpatrick
1985 Person, Self, and Experience: Exploring Pacific Ethnopsychologies.
Berkeley, California: University of California Press.
Wierzbicka, Anna
1972 Semantic Primitives. (Linguistische Forschungen, 22.) Frankfurt am
Main: Athenum.
1986 Human emotions: universal or culture-specific? American Anthropol-
ogist 82: 7'59-81.
1992 Defining emotion concepts. Cognitive Science 16: 539-81.
1996 Semantics: Primes and Universals. New York: Oxford University
Press.
1998 'Sadness' and 'anger' in Russian: the non-universality of the so-called

'basic human emotions'. In: A. Athanasiadou - E. Tabakowska (eds.),
Speaking of Emotions: Conceptualization and Expression, 3-28. Ber-
lin: Mouton de Gruyter.
1999 Emotion Across Languages and Cultures: Diversity and Universals.
Cambridge University Press.
Xin, Shibiao
1993 'Shangchun Beiqiu' jiqi wenhua tanyuan ['Shangchun Beiqiu' and its
cultural origin]. In: Jianming Chen and Zhiming Tan (eds.), Yuyan yu
wenhua duoxueke yanjiu [Cross-discipline studies of language and
culture]. Beijing: Beijing Yuyan Chubanshe.
Xu, Dong-qiang
1997 Hexin zi he hanyu de yuyi gouci fa [Core characters and Chinese word-
formation based on meaning]. Yuwen Yanjiu [Language Studies], 3:
1-15.
1998 Shuo 'Zi'-fu lun yuyuan jiben jiegou de jianbie biaozun, jiben tezheng
he ta yu yuyan lilun jieshe de guanxi. [Talking about 'Zi' - also on the
criteria for distinguishing basic unit in languages, the basic features of
the basic unit and its relationship with the construction of language
theory]. Yuwen Yanjiu [Language Studies], 3: 1-12.
Yang, Kuo-shu (ed.)
1988 Zhongguoren de xinli [The psychology of the Chinese people]. Taipei:
Guiguan Tushu.
Ye, Zhengdao
1998 'Ming ruo qin xuan'[Life on a string]-Chinese ethno theory of life.
Unpublished ms., Australian National University.
2000 a The language of emotions in Chinese: a study based on Hong Lou
Meng. M. A. thesis. Australian National University.
2000b 'Cultural Scripts': examples from Chinese emotional expression.
Paper presented at Workshop on Ethnopragmatics, Australian Lin-
guistic Institute 2000, University of Melbourne.
In press Different modes of describing emotions in Chinese: bodily change,
sensations, and bodily images. Pragmatics and Cognition 9(2).
Forthcoming The Chinese folk model of facial expressions: a linguistic perspective.
Zhang, Fa
1989 Zhongguo wenhua yu beiju yishi [Chinese culture and the sense of tra-
gedy]. Beijing: Zhongguo Renmin Daxue Chubanshe.
Zhang, Hongming
1999 Word-hood in Chinese. Paper presented at the Eighth Annual Confer-
ence of the International Association of Chinese Linguistics. Univer-
sity of Melbourne.
404 Zhengdao Ye
Dictionaries and original literary references

ABC Alphabetically-Based Computerized Chinese-English Dictionary.
1997 Defrancis (ed.)
NSW: Allen & Unwin.
DRM A Dream of Red Mansions (translated by Hsien-yi Yang - Gladys
1994 Yang)
Beijing: Foreign Languages Press.
HCDC Handa chengyu da cidian [Handa dictionary of idioms].
1997 Shanghai: Hanyu da cidian chubanshe.
HCYC Hanyu chengyu hanying cidian [ChineseEnglish dictionary of
1991 Chinese idioms]
Heifei: Zhongguo kexue jishu daxue chubanshe.
HLM Hong lou meng Yue-qin Cao - E Gao
1998 Beijing: Renmin Wenxue Chubanshe.
J Jia [Family]
1970/[1931] Hong Kong: Nanguo Chubanshe.
JYHC Jingxuan yinghan hanying cidian [Concise English-Chinese
1986 Chinese-English dictionary]
Oxford: Oxford University Press.
LY Lun Yu - The Analects (translated by D. C. Lau)
1983 Hong Kong: The Chinese University Press.
SWBJ Selected Works ofBa Jin. Sydney Shapiro and Mingjie Wang
1988 Beijing: Foreign Languages Press, v. l.
XHC Xiandai hanyu cidian [A dictionary of Modern Chinese].
1998 Beijing: Shangwu Chubanshe.
XHZTB Xiandai hanyu zipping tongji biao [Frequency charts of modern
1992 Chinese characters].
Beijing: National Language Committee.
ZGWG Zhongguo guodian \venxue gaikuang [A survey of classical Chinese
1994 literature with excepts from masterpieces] Dexin Zhang
Beijing: Beijing Language and Culture University Press.
ZLWZX Zhongguo lidao wenxue zuopin xuan [Selections of historical Chinese
1990 literature] Dongrun Zhu (ed.)
Shanghai: Shanghai Guji Chubanshe.
ZSHC Zuixin shiyong hanying cidian [A new practical Chinese-English word
1973 dictionary] Shiqiu Liang
Taipei: The Far East Book.
Subject and name index
Abdullah 172-3, 185, 193 Berdjaev 324

Aboriginal English 54 Berlin 164-5
Address terms 314, 320-2, 355-6 Beyer 235,332
Adelund 135, 145 Bible 145, 148
Aibike 112 Blowers 396,399
Aibul 112 Bodily images 23, 35, 59-60, 70,
Amberber 22 74-112, 153, 157, 167, 179, 192, 203,
Ameka 11,25-6 210, 248, 262, 292, 294,300,387;sen-
Ames 401 sations 37, 44, 58-9, 151, 200, 210,
Amharic 22, 35-67 232-3, 241, 245, 300, 318; symp-
Anger 2-3, 5-8, 23-4, 69-70, 115-44, toms 2, 17, 35, 56-8, 69-70, 115,
184-6,310 163-4, 245, 257-8, 268-9, 282-3, 303,
Apresjan 9, 26, 291-2, 300, 311-2, 317, 306
331 Bogatyrev 330
Arrernte 24, 197-215 Boll 117-8, 123, 131-2, 145
Arutjunova 291, 297, 302, 312-3, 329, Bolshevism 298
332 Bond 284,286,360,399
Asia 23 Boredom 100,201
Athanasiadou 19, 20, 26 Briggs 3,26
Atkins 211,213 Brockhaus 136, 145
Australian English 150 Brodsky 327-8
Austronesian 74, 112 Bucher 256,268,271,273,279,286
Autumn 370-1 Bugenhagen 11, 23, 25-6, 69, 71, 76,
Awang 185, 193 103, 112
Axmatova 314,325 Bunin 321
Burgman 144-5
Backhouse 249,250 Burma 164
Bakunin 306
Baranczak 297,331 Calm 93-5, 174, 182-4, 201, 262, 268
Basic emotions 5-6, 16-17, 69-70, 121, Campe 135, 145
126,149,154,163,190, 255-6, 283-4, Case marking 60-4, 329-30, 339-40,
297-8, 337-8, 362, 371,389-1, 397 345-6
Belinsky 306 Categorisation 16, 111, 119-21, 134,
Bellah 17,26 154, 163-4, 199-201, 338, 355, 390,
Benedict 4, 26 397, 398
Bennett 18,31 Catullus 331
Berberova 318 Cernysevsky 306
406 Subject and name index
Cernyx 299,332 Disgust 5, 69, 108, 112, 121, 149, 151,

Chai 376,399 281
Chao 362,399 Ditton 213
Chappell 11,26,391,399 Djamour 182,194
Chekhov 323-4,332 Dobson 198,203,213-4
Chen 361,399 Doi 4,27
Chinese 11, 17, 24-5, 199, 255-89, Dostoevsky 302, 304-6, 317, 332
359-404 Doubt 72,87-8,161
Chu 286,370 Dubois 330
Clore 31 Duden 136, 145
Cognitive elements 7-8, 35, 207, 212, Durst 23, 115
391 Dutch 142
Cognitive linguistics 19-20
Collins 31,347-9,357 Egypt 66,97
Colour 163-4 Ekman 1, 5, 27, 69, 113, 149, 163, 165,
Conceptual analysis 19-20 255, 283, 286, 389-90, 399
Conceptual primes/primitives, see NSM Ellis 184
Confucius 375-6,397 Emanatian 18
Cong 397 Embarrassment 53-5,72,102-3,128
Contentedness 94, 151-2, 174, 295, Emotion labels 3, 163
298-9 Enfield 11, 23, 27, 149, 153, 163, 165,
Coope 185,193 192,197, 213
Crying 23, 35, 55-6, 69-70, 108, 155, Eskimo 3
157-8, 377 Ethiopia 36
Cultural identity 3-4, 8, 78, 202, 204, Ethnocentrism 8-10, 13, 17-18, 20-1,
291-2, 303, 317; models 16-18,23-4, 25, 163, 197, 338, 360
70, 111-2, 178, 322, 338, 368, 371, Ethnopsychology 9-10, 23, 149, 163,
397; semantics 19, 201 167, 176, 191, 205, 262-3, 275, 282,
291-2, 302, 359-61, 396
Daoism 397 Europe 3, 23, 140, 176
Darwin 54, 66, 69, 113, 255, 282, 286 Ewe 11
Dative 329-30, 339-40, 345-56 Experiencer 60-4, 71-2, 76, 274, 308,
Davidson 1,27,390,399 329, 344, 349, 354
de Tocqueville 17,27 Eyebrows 161-2,387-8
Debus 120, 143, 145 Eyes 23, 54-5, 75-85, 95, 111, 155,
Defining 9, 14-16, 22, 220-1, 347 159-62
Desire 9, 12, 89, 100, 115, 170, 194,
201-3, 399 Facial expression 55-6,82, 85,109,149,
Dictionaries 122, 172-3, 185, 198, 154-63, 230, 257-9, 282-3, 297,
218-22, 234, 256, 265, 271, 279, 387-8, 399
347-8, 361-2, 378, 390 Fang 370-1,399
Diller 153, 165, 168, 191, 193 Farrell 122,145
Dirven 17,20,27,31 Fasmer 299,314,327,332
Discourse 14, 18, 167-8, 175, 191, 199, Fear 3,35,49,69,92,103,105,112,149,
217, 298, 355, 361, 387 153, 222, 234-48
Fehr 364,400 Happiness 13, 17, 38-9, 44-5, 60, 69,

Feng 368,371,400 89, 94-5, 112, 121, 149, 151-2, 154,
Ferber 199,202 158, 184, 190, 201, 291, 295-9, 320,
Fillmore 211,213 330
Foley 163, 165 Harkins 8, 11, 16, 24, 26, 28-9, 35, 54,
French 3, 5, 11, 298-9, 300, 304, 306, 67, 121, 142, 146, 197-8, 202-4, 214,
311-2,314,319,326 330,397-8,400
Frequency 134, 330, 339,398 Harre 8,29,70,113
Friedrich 17,27 Harris 17,29
Friend 150,291,315-6,330 Hasada 11,24,26,29,217
Friesen 69 Hassman 330
Frijda 8,27,211,213,398,400 Hate 61, 64, 73-4, 81, 139, 172, 274-9,
Frisch 117-8, 123, 132, 145 314-5
Fujita 222 Hearing 22, 38-41, 58, 65, 107, 109,
Fury 51, 82, 98, 109-10, 121, 134, 136, 199-200
138-9, 157, 203-6, 208, 275-6 Heart 74-5, 150, 153-4, 167-93, 196,
232-3
Gale 378 Hebrew 74-5,111,136
Geeraerts 19, 27-8, 140, 146 Heffernan 199
Geertz 70, 113 Heider 170, 194
Gelman 67 Henderson 198, 203, 213-4
German 11, 23, 115-48, 176, 298, 300, Hermeneutical circle 8, 21
306,311-2,314,318,326 Hinton 398,400
Gibbs 18,20,28 Hirschfeld 67
Ginzburg 318 Historical change 16, 23, 135, 141,
Goddard 5, 6, 8, 11-12, 18-19, 22-5, 28, 299-301,305-6,378
35,53,65-6,71, 113, 152-3, 165, 167, Hodapp 126
170, 176, 191, 194, 198-200, 202, 210, Hope 15, 85, 108-9
212-4, 250, 263, 287, 361, 391, 400 Hopper 63, 67
Gorky 301,318 Hbler 17,29
Gothic 141 Hurt feelings 185, 188-90, 337-56
Grammar 11, 17, 22; Amharic 36-7,
46-7, 60-5; German 123-4, 127-31; Icelandic 63
Lao 155, 164; Mbula 112; Pol- Ickler 130, 146
ish 344-6; Russian 328-30 Ifaluk 4
Greek 69,75,111,116,136,141 Ilongot 3, 5
Green 213-4 Inagaki 229,250
Grief 25,300,361-2,367,372-8,389-98 IndoEuropean 141,217
Grimes 66 Insides 75,86-95,387
Grimm 135, 141, 144, 146 Inst, fr Deutsche Sprache 143
Grondelaers 19, 28, 140, 146 Interjections 42-4
Guilt 4, 8-9 Intestines 75, 86-95, 387
lordanskaja 17, 29, 291, 300, 307, 332-3
Hamid 192 Israel 66
Han 362,401 Italian 298,314,326
Izard 5, 29, 223, 226, 250, 283-4, 287, Levontina 24, 291, 310, 316-7, 327, 333
298, 312, 333 Lexicalisation 268,271,292,338
Li Bai 378, 379-80, 384, 386
Jger 146 Li, C.N. 362, 400
Jaisser 153, 165 LiLu 255,287
Jakobson 330,333 LiQingzhao 384-6
James 16, 29, 69, 113, 338, 357 LiYu 386-7,381-3
Japanese 4, 11, 18, 24, 217-53 Liang 367,398,400
Jealousy 64,72, 83-4,170, 201 Lin 262,268,280,287
Johnson 133,144-5 Lips 76, 108-9, 157-8, 162
Johnson-Laird 8, 29, 66-7, 298, 333 Literature 13-14, 24-5, 116-8, 302-4,
Joy 5, 75, 101, 177, 291-9, 356, 367-8 361,370-1,378
Juntanamalaga 153, 165, 168, 191 Liu 271-2,287
Liver 23,75,95-9,111-12
Kakehi 220,249-51 Lixacev 322,333
Karim 171, 172, 178, 194 Logan 16-17,30
Kay 164-5 Loneliness 150, 201, 323-8
Keywords 23, 167, 176 Longing 8-9, 72, 108-9, 170, 291-2,
Khan 193 299-302, 323-8, 331, 380-1, 398
Kimizuka 220,250 Losev 302, 304
Kinship 320-2 Lotman 293,301,305,333
Kirkpatrick 359,400,402 Love 61, 64, 75, 88-9, 136, 170-1, 190,
Kita 164 314,319-22,326-7,390,396
Klappenbach 122, 140, 146 Lu 362,401
KlosSokol 354,357 Lucy 163-4, 165
Kondo 4,29 Luther 136, 146
Korean 229 Lutz 4, 30, 60, 114, 278, 287, 360, 378,
Kornacki 24, 26, 29, 255, 388, 400 401-2
Koukham 164
Kvecses 19, 30, 140, 146, 211, 214 MacLaury 32, 148
Kronhaus 322,333 Madsen 26
Kupperbusch 10, 30 Maeda 182, 194
Kurtz 339,357 Malay 6, 7, 11, 13, 17-18, 23-4, 152-3,
167-95
Lakoff 19,52,67,211,214 Malayo-Polynesian 112
Langacker 19, 134, 146 Malcolm, I.G. 211,214
Lao 11,23,149-66 Malcolm, N. 8, 30
Latin 116, 136, 141,300 Malecky 320
Leavitt 69,70, 113 Mandler 10,30
Lebra 222,250 Marks 361,401
Leibniz 71,113,220,250 Martin 217,251
Lerner 116 Marx 119, 146
Lersch 144,146 Matisoff 153, 165, 168, 194
Leslau 42,67 Matsumoto 30
LeVine 360,402 Matthews 256
Maude 14 Offence 6, 24, 72, 153, 291, 301-2,

Mbula 11, 17,23,69-114 306-11, 314-5, 329-31, 346-9, 355
McDuff 304 Ohala 398,400
McElhanon 74, 114 Onishi 11,31
McMillan 199 Ono 251
McVeigh 23,30 Onomatopoeia 217, 249
Mees 126 Ootsubo 251
Mel'cuk 291,307,330,333-4 Ortony 8,31, 126, 147
Meldau 122, 146 Osborne 116, 137
Meng 388 Osmond 17,31
Metaphor 16, 18, 19, 52, 136, 140, 190, Ots 262, 265, 267, 284, 288
261, 264, 281, 293, 297, 304, 322, 328 Owen 363,371,379,401
Metzger 274,279,287 Ozegov 295
Middleton 284,287
Mimetic words 217 Palmer 18,20,31
Mind 111, 176,294 Pan 362,397,401
Mito 220,251 Papemo 17,29,306,334
Modernity 16,306,350 Papua New Guinea 23, 70, 74
Moore 5,30 Papuan 74
Morality 173-4,277-8,292,293,312-3, Pasternak 319
375-6 Paustovskij 321
Morita 222,225,250 Peeters 11,31
Morohashi 256,287 Peletz 178,194
Morphology 36, 179, 362 Pendergrast 16,31
Mostovaja 26, 30 Pen'kovskij 291,293,297,334
Mouth 76, 106, 155, 157-9 Pesetsky 60-1,67
Myers 378,401 Pesmen 292,311,334
Myhill 144, 147 Philippines 3
Pity 8,309-10,317-20,329
Nabokov 304,334 Place 299,316,324
New Guinea 23,70,74,112 Platonov 301,319,334
New Testament 75 Pleasure 13, 15, 172, 201, 293-9, 303,
Nichols 398,400 316,330
Niemeier 19-20,30-1 Plum 120, 146-7
Nonverbal 18, 23, 25, 151, 264, see also Plungjan 294,302,334
Facial expression Plutchik 389,398,401
Norse 137,203 Poertner 133-4, 137
NSM (Natural Semantic Metalanguage) Polish 25, 292, 308, 337-57
9-13, 21-6, 112-3, 121, 176-7, 191-2, Polysemy 13,38-41,44,54,58,65
198-9,221,248,361,391-7 Porkert 262,265,288
Pragmatics 18,95,297,313
Oatley 8, 29, 66-7, 255, 287-8, 333 Priestley 26,31
Occhi 20,24,31 Prototypes 13-16, 22, 74, 111-2, 120-2,
Oceanic 112 140, 212, 221, 263, 270, 281, 356, 364,
Oey 153,165,168, 194 373-4
Psychomimes 24,217-22 249,260,314-5

Pushkin 298,300-1,316,334 Semantic primes/primitives, see NSM
Seneca 115,147
Qian 359,361,370,401 Septuagint 147
Qu 373 Shakespeare 16, 351
Shame 4, 8, 35, 53-4, 102-3, 190, 198,
Rahman 185 201-3,302,308,311-4
Raxilina 294,302,334 Sharifian 211,214
Relief 44 Shaver 265, 271, 288, 364, 390, 402
Resentment 92, 108-9, 119, 201-8, 268, Shibatani 217,251
277 Shweder 16, 32, 256, 288, 360, 390,402
Rice 202,214 Shyness 54-5, 66, 102
Rochecouste 214 Simcuk 334
Rosaldo 3,31,70,114 Slavic 297,320
Rosenberg 1, 27 Smelev 200, 291-4, 296, 330, 333-4
Ross 112, 114 Smell 72,156-7
Roth 265,299 Smiling 13,23,35, 55-6,66,155-6,158
Rozanov 304,318-9 Smith 348-51,357
Rubin 116 Solomon 359,390,402
Ruch 69,114 Sommerfeldt 126, 131, 147
Russell 4, 32, 120-1, 147,149, 163, 165, Sorenson 69
255, 288, 360, 362, 364,400-2 Sorrow/sorry 24, 43, 170, 297, 340-2,
Russian 8, 9, 13, 17, 24, 176, 181, 346-7, 351-5, 361-2, 367, 378-99
291-336 Soul 176, 191, 292, 294-5, 301-2, 304,
316,330
Sadness 5, 10, 13, 25, 45-8, 66, 69, 91, Sound symbolism 24, 217-49, 370
112, 121, 149-50, 154, 158, 169, 177, Southeast Asia 153, 168
201, 359-99 Spanish 11, 13
Santangelo 398 Speech acts 22
Sapir 337,357 Spirit 111,159
Sato 398 Spiro 164, 166
Satunovskij 294,334 Stacey 18,31
Savvina 313 Starke 126, 131, 147
Sayings 175,256 Stearns 16, 17, 32, 121, 147
Scenarios 14-16, 22, 70, 205-6, 211, Steen 20,28
342, 347-8, 353, 355, 364 Steinitz 122, 140, 146
Schema 20,24,205-6,211 Stepanov 299-300,319,335
Scherer 391,398,402 Stomach 23, 75, 104-6, 111
Schmidt-Atzert 119,143,147 Strohm 119
Schreiber 126, 131, 147 Su 367
Schwartz 288,360,390,402 Suffering 85, 101
Schwarz 126, 143, 147 Sullivan 26
Script 121,205-6,211 Surprise 5, 43, 72, 97, 99, 149, 152-3,
Seasons 370-1 158, 160-1, 170, 177, 201, 218-33
Seeing 58-9, 76-85, 109, 111, 160, 238, Swidler 26
Sympathy 8,46-7,65-6,317-8,378 Werth 19

Western Desert 202,210
Tabakowska 19-20,26 White 359,400,402
Tamamura 217,251 Wiers 398
Taylor 8, 32, 148 Wierzbicka 3-4, 6, 8-12, 14, 16-18,
Thai 153, 168, 191 22-3, 25-6, 28-9, 32, 35, 37, 40-3,
Thompson 63,67,362,400 45-6, 48, 50, 53-4, 59, 63, 65-7, 71,
Throat 76, 107-8 74, 114, 118, 121, 134, 142, 144, 146,
Tollefson 359,402 148, 162, 164, 166-7, 176, 178, 191,
Tolstoy 13,32,293,305,325,335 194-5, 198-9, 210-12, 215, 219, 221,
Translation 115-8,149-50,163,218-22, 249, 251, 255, 263, 279, 288, 291-2,
291, 297-9, 306-7, 312, 316, 323, 326, 294, 297-9, 307-10, 315, 317, 319,
347, 352, 355, 359-60,366, 379, 391-7 321, 327, 329-30, 335, 337, 339,
Travis 11,26,32 355-7, 361, 364, 367, 374, 389, 391,
Tretter 330 397-400, 402
Trifonov 321 Wilkins 11, 26, 29, 33, 195, 199-200,
Tsu-chang 376 202-3,213-5
Tsvetaeva 326-7 Wilson 172
Turgenev 325 Winzeler 184, 195
Turner 126, 147 Wittgenstein 8, 120, 144, 148
Worry 63,167-8,179-82,201,379,386,
Tzu 397
390
Writing system 362, 397
Ufimceva 330,335
Wu 288,402
Universals 6, 8-13, 25, 35, 50, 163, 256,
Wundt 134,148
283,291,312,389-91
Upset 50-3, 86, 90-1, 93-4, 100-1, Xie 371
108-9, 174 Xin 370,373,403
Uryson 291,292,299,335 Xodasevie 303,304
Uspenskij 293,333
Xu 397,403
Values 182-3, 211, 284, 293, 295, 328, Van 282,289
319,355,375-7 Yang 360,403
Van Brakel 255, 283, 288, 389, 402 Yankunytjatjara 5,202,210
VanGeert 10,32 Ye 25-6, 33, 199, 359, 362, 368, 372,
Vayaphanh 164 376, 387-8, 398, 403
Vengefulness 82, 85, 109-10, 121, 139, Yik 360,362,402
277 Yin 278
Vinogradov 305, 335 Yip 362,401
Vysotsky 318,326 Yu 34,289
Yusof 185, 194
Wade-Giles 398
Wang 262, 275, 277, 280, 288, 363 Zaenen 63, 67
Wanner 307,334 Zajonc 152, 166
Weber 126 Zakaria 185, 195
Zalizniak 8, 24, 34, 291, 299, 309-10, Zhu 397

316-7, 327, 330, 335-6 Ziegler 126, 143, 147
Zhang 260, 263-4, 289, 362, 371, 403 Zimbardo 69, 114
Words and phrases index
This index lists words and phrases discussed or defined in the book. Definitions are in-
dicated by bold page numbers. All are listed in the order of the English alphabet.
Glosses in brackets are a rough guide, not translation equivalents. Languages are:
Am(haric), Ar(rernte), Ch(inese), Du(tch), Fr(ench), Ge(rman), Gr(eek), He(brew),
If(aluk), Il(ongot), It(alian), Ja(panese), Lao, La(tin), Ma(lay), Mb(ula), Po(lish),
Ru(ssian), Sp(anish), Th(ai), Ya(nkunytjatjara).
adat (customs, Ma) 175 115-44, 149-51, 154, 162, 190,

affara (ashamed, Am) 53-4, 54, 55 201-12, 337-8
afraid 35, 149, 153, see also fear Angst (fear, Ge) 11,300
Aggression (aggression, Ge) 119 anpeme (touch, Ar) 200
aggression 121 apateme (nonplussed, Ar) 201
ahakngerre (aggressive, Ar) 201, 204, apure (shame, Ar) 201-3
210 apurrke-apurrke (bored, Ar) 201
ahe (interj., Am) 43 rger (anger, Ge) 23, 116-9, 122-5,
ahe, ahele, ahele-irreme (anger, angry, 125-6, 127-31, 141, 143-4
Ar) 201-6,209-10 arnkelye (resentful, Ar) 201,204-5,207,
ahentye (want, Ar) 201-3 208
ai (grief, Ch) 25, 361-2, 372-4, 374, arrara (furious, Am) 51
375-8, 386, 389-91, 394-5, 396-8; as- (causative, Am) 49,55,61
~ ge (elegy) 373; ~ji (mourning rites) ashamed 35, 53-4
376-7; ~ lian (commiserate) 375; ~ si atere (fear, Ar) 201-3
(mourn) 373; ju ~ (wail) 376-7 atneratye (trust, Ar) 201
ai (love, Ch) 376 atnerte ampeme (angry/hungry, Ar)
akangkeme (happy, Ar) 201 204-5, 210
akenge (very bad, Ar) 209, 211 atnewe-atnewe (relaxed, Ar) 201
akiwarre (infuriated, Ar) 201, 203-6, atnulkeme (shocked, Ar) 201
208,211 atyinhe (mine, Ar) 208,211
akngerre anthurre itirreme (worry, Ar) aw/(at, Ge) 130-1
201 aweme (hear, Ar) 199-200
akurne (bad, Ar) 209, 211; ~ awelheme awelheme (feel, Ar) 199-201
(feeling bad) 201,210 ay eye (story, Ar) 206-10,212
aid (say/do, Am) 42 ayn (eye, Am) 54-5; aynaffar (shy per-
alak'k'9S9(cry, Am) 55-6 son) 54; aynadarak' (shameless) 55
alhirreme (admire, Ar) 201 ayya (see, Am) 58
alhwarrpe (sad, Ar) 201 azzana (sad, Am) 45-8; azzana-ll- 47;
amae (affection, Ja) 4 azzana-bb- 47-8
ambil hati (charm, Ma) 172
amad (ashen, Am) 57 baa (insane, Lao) 159
amuk (berserk, Ma) 184 bad 12-13; feel bad 24, 41-2, 346,
anger, angry 3, 5-8, 16-17, 35, 75, 353-4, 354, 355
414 Words and phrases index
baik (good, Ma) 13,173-4 chengyu (sayings, Ch) 256, 282

baj (leaf, Lao) 155 chou (sorrow, Ch) 25, 361-2, 378-83,
baofu (revenge, Ch) 277 383, 384-91, 395-6, 396-9; ~ chang
basma (hungry for meat, Mb) 71 (gut-wrenching sorrow) 387; ~ hai (sea
ba, baniddet (manner, Am) 51 of troubles) 379; ~ mei (knotted brow)
bei (sadness, Ch) 25, 361-4, 364, 387-8, 399; jie jiu jiao ~ (drown one's
365-73, 375, 378, 389-91, 393^t, sorrows) 384; ~ rang (sorrowful look)
396-8; ~ can (tragic) 367; ~ dao (be- 387-8; xiang ~ (homesick) 380-1, 398
moan) 369; ~ huan li he (sorrows and ci (polysyllabic word, Ch) 362, 367
joys of life) 367-8; ~ Hang (dismal) cinta (love, Ma) 170
367; ~qiu (autumn melancholy) 370-1; colere (anger, Fr) 3, 5
~ tian (lament) 368; ~ tong (woeful) content 299
367 cry 35, 55-6, 158
Berserkerwut (berserk rage, Ge) 136-7 Cuvstvovat' (feel, Ru) 9
besar (big, Ma) 173-4
bie-zhe qi (resentful, Ch) 268 dadaru (doubt, Mb) 72
bikkuri (surprise, Ja) 220, 229 ddbb 9n 9 (infuriated, Am) 51
biku-biku (fearful, Ja) 218-9, 222, 234, delighted 15
241-3, 243, 244-5, 250 dam-ufalla (blood boiled, Am) 60
bitterness 177 dassita (happiness, Am) 38-9, 45, 60
body 99,102,111,294-5 di hati, di dalam hati (in/at heart, Ma)
bonheur (happiness, Fr) 298 170-1
boo-sua (doubtful, Lao) 161 dii-caj (glad, Lao) 151-2,158
bo-pen-nang (no problem, Lao) 151 diniggat'e (shock, Am) 58
bse (bad/angry, Ge) 117, 144 disappointed 46-7, 65, 75
brain 111 disgust 121, 149, 151,281
bu gongping (unjust, Ch) 274-5 doch (modal, Ge) 129-30
buah hati (sweetheart, Ma) 171 dokiQ (thump, Ja) 24, 218, 221-3, 226,
bueno (good, Sp) 13 230-1,231,232-3,248
buruk (bad, Ma) 13 dosada (annoyed, Ru) 310
busuk (putrid, Ma) 173-4 dovolen (content, Ru) 298-9, 330
buut (rotten, Lao) 156 dovol'stvo soboj (pleased with oneself,
Ru) 295
ca (cry, Lao) 158 dozdat'sja (keep on loving, Ru) 326
caj (heart, Lao) 150, 153-4; dam drug (friend, Ru) 291, 315-6, 330
(ruthless) 153; ~-haaj (angry) 150, dumat' (think, Ru) 9
156, 162; ~-kaa (daring) 153; kuang duSa (soul, Ru) 191, 292, 294, 301-2,
(generous) 153, dii (glad) 151-2, 304,316
158; naak (indecisive) 153-4;
nooj (offended) 153; sia (sad) embarrassed 53-5, 102
149-50,158; son- (interested) 160-1; emotion 8-10, 16, 35, 37, 70, 163,
tok (surprise) 152-3,158,160-1 199-200, 390
cay (heart, Th) 153,168 enjoy, enjoyment 13, 297
cemburu (jealousy, Ma) 170 enthuse 17
cheeky 204,210 ergeren (anger, Du) 142
esli(if, Ru) 9 haaj (tell off, Lao) 150

eyebrows 162 haji (shame, Ja) 4
eyes 162 happy, happiness 13, 17, 121, 149,
151-2, 154, 184, 190, 291,297-8
fagag ah (smile, Am) 55-6 haQ (startle, Ja) 218, 221-5, 225, 226-7,
fago (compassion, If) 8, 378 230, 232-3
fear 15, 190, 337-8 hara-hara (apprehensive, Ja) 220-2,
feel 14, 35-9, 58, 65, 199-200; feel bad 234-5, 235-6, 236-9, 243, 249-50
24, 346, 353-4, 354 Ha (hate, Ge) 139
felicita (happiness, It) 298 hate 314
fen (hate, Ch) 24, 255, 274-7, 277, hati (heart, Ma) 18, 23, 153, 167-77,
278-9, 285; gong- (public) 274, 177-8, 179-93; ambil ~ (charm) 172;
277-8; min~ (people's ~) 274,277; qi~ berdebar ~ (~ throbbing) 192; besar ~
(infuriated) 275-6; yi~ (indignation) (proud) 174; baik~ (good ~) 174;
274, 278 buah ~ (sweetheart) 171; busuk ~ (vi-
farrajirhat (fear, Am) 48-50, 58, 61 cious) 174; cuba ~ (test honesty) 172;
fit-u amadmassala (ashen-faced, Am) 57 dekat di ~ (close to ~) 175; di dalam ~
fret 17 (to oneself) 171; di sudut ~ (in depths
friend 291,315 of~) 170; geli- (amused) 174; han-
frown 161-2 cur ~ (crushed) 174; hasrat - (heart's
fume 17,51 desire) 170; ~ berat (heavy ~) 174;
Furcht (fear, Ge) 128 ~ bersih (clean ~) 174; ~ keras (deter-
furia (fury, La) 136 mined) 167, 174; ~ kecil (innermost ~)
furieux (angry, Fr) 3,5 170; ~ kotor (unclean ~) 174; "Iain-
fury, furious 121, 134, 138-9 lain (not in harmony) 175; ~ luka
(wounded ~) 174; ~ murni (noble ~)
ga (liver, Ch) 262 174; ~ mutu (melancholy) 174; ~ nur-
gijoogo (psychomimes, Ja) 217 ani (enlightened) 174; ~ tawajuh (de-
giongo-gitaigo (onomotopeia, Ja) 217 vout) 174; ~ terbuka (open mind) 174;
giseigolgiongo (phonomimes, Ja) 217 ~ terkilan (upset) 174; ~ waja (resol-
gitaigo (phenomimes, Ja) 217 ute) 174; ~ walang (melancholy) 174;
gleich (just, Ge) 129 in ~ (envious) 174; ikut ~ (follow ~)
Glck (happiness, Ge) 298 175; isi ~ (contents of heart) 170;
glupio (stupid, Po) 342-3, 346 jaga ~ (protect feelings) 172; jatuh ~
good 12-13,41-2 (fall in love) 171; kecil- (feel hurt)
gore (woe, Ru) 297 168, 174, 179, 188, 189-90; kehen-
grief 362, 367, 373-5, 377, 396-7 dak - (heart's wish) 170; keras - (stub-
grimace 155, 162 born) 174; lapang - (content) 174;
Grimm (grimness, Ge) 136 lembut - (tender) 174; murah ~ (gener-
grin 155, 160 ous) 174; niat - (desire in heart) 170;
Groll (rancour, Ge) 119 panas - (angry) 174; puas - (satisfied)
grust' (sad, Ru) 13 168, 174, 179, 187, 188, 193; putih -
guilt, guilty 4, 9 (sincere) 174; putus - (despair) 174;
gyoQ (jolt, Ja) 218, 221-3, 225, 227-8, rendah- (humble) 167, 174; sakit -
228-9, 230, 232-3 (annoyed) 168, 174, 179, 184, 185,
186, 190; se~ (of one mind) 175; irrareltheme (depressing, Ar) 201
sejuk ~ (calm) 174; senang ~ (relaxed) -irre (inchoative, Ar) 203
168, 174, 179,182, 183-4, 193; suci ~
(pure) 174; suka ~ (please oneself) jaak-thaam (questioning, Lao) 161
172; susah- (troubled) 167-8, 174, jaan (afraid, Lao) 153, 159, 161
179, 180, 181-2; tabah ~ (determined) jaga hati (protect feelings, Ma) 172
174; tawar ~ (lost interest) 174; ter- jantung (heart, Ma) 168
pendam di ~ (buried in ~) 170; tidak jatuh hati (fall in love, Ma) 171
sampai ~ (can't stand) 172; tinggi ~ jingdiziwa (frog in the well, Ch) 359,397
(proud) 174; tulus ~ (sincere) 174; jay 75, 177,291,356
ubah ~ (change ~) 172
hear 38-41,58,65 ka- (recipient, Mb) 73
heart 167-8, 170, 173, 176, 177-9, 190; kaipa (gloat, Mb) 71
broken ~ 75, 168; pure of~ 173; sick kak by (as if, Ru) 329-30
at- 173, 184, 190 -kam (do/get, Mb) 72
het khiw sonlcoot-kan (frown, Lao) 161-2 kanamaala (sense, Mb) 72
het taa-naj (make big-eyes, Lao) 160-1 kar kopoono isoloorjgo (collective anxie-
heureux (happy, Fr) 298-9 ty, Mb) 105, 109
hiya-hiya (shivery, Ja) 18, 218-9, 222, kasih (care, Ma) 190
234,238-8,239,240-1,243-5,249-50 kecil hati (feel hurt, Ma) 168, 173-4,
hod-e tambboc'abbacd (stomach lurched, 179,188, 189-90
Am) 59-60 k' anna (jealous, Am) 64
hope 15 k^nakkana (uneasy, Am) 58
hostile 203 kete- (liver, Mb) 23, 75, 95-9, 111-12;
hurt 24,169,306-7,310-11,340,346-7, ipas ~ (enraged) 98, 99, 110; ~ guruk
349-50, 350, 355 (impulsive) 96, 109-10; ~ ibayoulibe-
leu (furious) 98, 109-10; ~ ikam kegl
iffoy (interj., Am) 44 we k (startled) 97; ~ imap (astonished)
-ig (adj. suffix, Ge) 135,144 99; ~ ipakpak (furious) 98; ~ ise (en-
ikarrut zuru- (biting lips in anger, Mb) thusiastic) 97, 110; ~ kembeita lai
108-9 (furious) 98; ~ malmal (angry) 97-8,
ikut hati (follow heart, Ma) 175 108-10; -pitpit (impetuous) 96-7,
ilkelheme (enjoy, Ar) 201 109-10
ilkelhetyakenhe (uncomfortable, Ar) 201 kheew (teeth, Lao) 157; kat (biting
imagine 177 teeth) 159
implication 8 khiat (angry, Lao) 150-1, 156
m (in, Ge) 123 khi-diat (disgusted, Lao) 151
indignation 5 khit (think, Lao) 161
ingkerte-ingkerte (jealous, Ar) 201 khiw (eyebrows, Lao) 161-2; het ~ son/
ingketye (desirable, Ar) 201 coot-kan (frown) 161-2; nik-~ (raise
interest 111 eyebrows) 161
ipas kete- (enraged, Mb) 98, 99, 110 ki o torinaosu (pull oneself together, Ja)
ira (anger, La) 116,136,141 226
irao lele- (satisfied, Mb) 94 -kilaala (recognise, Mb) 72
irrare (lonely, Ar) 201 kilyowt (kidneys, He) 74-5
kizuku (recognise, Ja) 226 (reconciled) 93-4, 110; ~ ipasiksik

know 12 (unable to refrain) 92; ~ ipata (sad) 91,
kon (ghost, Mb) 111 94, 109; ~ ipurulimagga (willing) 89,
koo-luang (create an issue, Lao) 151 110; -isaana (upset) 86, 90-1, 94,
kopo- (stomach, Mb) 23,75,104-6, 111; 108-9; ~ iurpalbe (choose) 87-8, 110;
kar ~ono isoloorjgo (collective anxie- ~ iurpa (suspect) 88; ~ iurur (perplex-
ty) 105,109; ~bok (stomach full) 104; ed) 88; ~iwe ru (doubt) 87-8; ~pa
~ ikam ken (fright) 105; ~ iqgwor (want/love) 88-9, 94, 109; - sa (re-
(stomach growling) 104; ~ isaana sentful) 92,108-9
(nauseous) 105; ~ iyouyou (stomach- -ley (hear, Mb) 72
ache) 105; ~ kutkut (anxious) 105, lev (heart, He) 74-5
109; ~rru (anxious) 74, 105, 109; Liebe (love, Ge) 136
~ suruunu (diarrhoea) 105 liget (passion, II) 3
kotor (dirty, Ma) 173-4 lips 162
ku (cry, Ch) 377 ljubimilyj (beloved, Ru) 321
kuli- (skin, Mb) 23, 76, 99, 101-4; ljubit' (love, Ru) 314, 319-20
~ ibayou (fever) 87; ~ imalai (tired) love 75, 190, 327, 390
101; ~ imet (determined, accustomed) -lulu (vomit, Mb) 72
104; -imoto (afraid) 103, 104, 109;
~ ire youyourjana (suffer) 101; ~ isa- malo (bad, Sp) 13
rakrak (shiver) 101; ~ isu (wrinkled) malu (shame, Ma) 190
101; ~ iswidit (sunburn) 101; ~ ka- marah (angry, Ma) 6-8, 185, 191
miarj (shame) 102-3,109; ~ mburaana -martu (sneeze, Mb) 72
som (not want) 194; ~ muriini (re- mala- (eye, Mb) 23, 75-85, 95, 111;
freshed) 101; ~ puliizi (embarrassed) ~ berber (covet) 82, 85, 108-9; ~ ikam
103, 109 (glimpse) 77; ~ Ha (think about) 80;
kumbu- rjgure- (calf of leg, Mb) 108 ~ imapalpaala (obsessed) 82-3, 109;
kunu- (spirit, Mb) 111 ~ imbeleele (forget) 80; ~ imbol (mor-
kuyaringanyi (angry, Ya) 6-7, 210 ally aware) 78; ~ imiili (think of past)
kwakwe (keen, Ar) 201 80; ~ imun (ignorant) 80-1; ~ irjgal
kwo- (mouth, Mb) 76, 106; ~ itaanda (careful) 79-80, 109; ~ ise (admire)
(agog) 106,109; ~ iyouyou (talked out) 81, 109; ~ iyaryaara (alive) 78; ~ /yo-
106 yo (dislike) 84, 85, 109; ~ kinirjana
(greedy) 83; ~ legalerja (scatter-
laugh 55, 66 brained) 81; -mbeij (spy) 77; ~ mbur-
le- (recipient, Mb) 73 mbur (jealous) 83, 84; ~ mburri (con-
lele- (insides, Mb) 23, 75, 86-95, 99, fused) 78-9; ~ molo (clairvoyant) 79;
110; irao ~ (satisfied) 94; ~ ambai ~ munrjaana (ignorant) 80; ~ pambaa-
(happy) 86, 89, 94-5, 109; ~ be (want) ra (vengeful) 82, 85, 109-10; ~ pasom
89; ~ ibayou (hot) 87; ~ ikam uraata (despise) 81; -pot (reluctant) 84, 110;
(moved) 92, 93, 108; ~ iluumu (calm) -putput (furious) 82, 85, 109-10;
93, 94-5, 110; ~ imarjga 110; ~ imbai ~ sirjiini isu (suffer) 85; ~ sigsirj
(upset with) 93, 109; ~ imbol (stub- (eager) 18, 85, 109-10; -po se ~ (mem-
born) 89; ~ imbukmbuk (anxious) 92, orise) 80; --(hope) 85, 109
108; ~ imet (determined) 94; ~ imiili -mbot (live, Mb) 72
-mbur (jealous, Mb) 72, 84 ni- (being, Mb) 23, 75-6, 99-101, 110;
-meete(die, Mb) 72 ~ ambai (healthy) 99; ~ gesges (bored)
-mel (fall, Mb) 72 100,109; ~ ikilaalalikanamaala (sense)
men (smelly, Lao) 156 99; ~ imbeeze (fit) 99; ~ ipeele (awake)
menmeen (happy, Mb) 71, 95 99; ~ isaana (exhausted) 99; ~ ise (jubi-
marrara (bitter, Am) 58 lant) 101, 108-9; ~ itekteege (shiver)
miar) (shame, Mb) 102 99; ~ iyouyou (ache) 99; ~ pa som (not
mii khuam-suk (happy, Lao) 151-2 want) 100; -puliizi (upset with) 100,
milyj (darling, Ru) 321 109; ~ sapa (resentful) 92,100,108-9;
mind 111, 176,294 ~ sanaana (upset with) 100, 101, 109
mmg(fate, Ch) 368-9 nik-khiw (raise-eyebrows, Lao) 161
miri (thirsty, Mb) 71 Him (smile, Lao) 155, 158
mirpanarinyi (angry, Ya) 6, 210 nom-paak (clench-mouthed, Lao) 155,
miss 331 159
m/Y(with, Ge) 123 nu (rage, Ch) 24, 255-65, 263, 267-70,
mood 10 274,276-8,282-4,285; bao~ (explode
-morsop (startled, Mb) 72 in rage); fa ~ (get angry) 261; kuang-
-moto, motogana (fear, Mb) 72 (mad rage) 260; ~bi (raise hands) 258;
mourn 377 ~ma (spirited horse) 260; xi~ (calm
muan (fun, Lao) 150 one's anger) 262, 268
mune (heart, Ja) 232-3 Ogora (brain, Mb) 111
murah (generous, Ma) 173-4 Qgumtur (embarrassment, Mb) 72
murni (noble, Ma) 173-4 -Ogug (cough, Mb) 72
-murur (surprised, Mb) 72 ggure- (throat, Mb) 76, 107-8; ikaml
muu (friends, Lao) 150 izuuru ~ (savour) 107, 108; ~ bulbul
mwarre awelheme (feeling good, Ar) 201 (disgusted) 108; ~ imamaaza (thirsty)
108; ~ imun (breathless from crying)
-n (accusative, Am) 37 108; ~ isaana (crave) 108,109; kumbu-
naa (face, Lao) 155-7; beqlnuk (ready- ~(calf) 108; nama- ~ (wrist) 108
to-cry face) 155,157-8; buut (scowl) -gu (offended, Mb) 72
156; khum (furious-face) 157; ~-men
(smelly-face) 156-7; pii ~ (next year) obida, obidet' (offence, offend, Ru) 291,
155 301-2, 306-7, 307, 308-11, 314-5,
nadryv (hysteria, Ru) 291,302-6 329-31
nama- ggure- (wrist, Mb) 108 obSfenie, obSoat'sja (communication,
-namut (sweet, Mb) 72 communicate, Ru) 315-6
nao(huo) (annoyed, Ch) 24, 255, 269, odo-odo (timid, Ja) 222, 234, 245-7,
271-3, 273, 274, 276, 283, 285 247, 248
naslazdenie (enjoyment, Ru) 295-9 odoroku (surprise, Ja) 220, 225, 229
ndddada (burn, Am) 51 offence, offended 6, 24, 291, 306-8,
nakka (touch, Am) 38 346-8, 348, 349, 355
nenavist' (hatred, Ru) 314 ogorenie (grief, Ru) 297
neudobno (inconvenient, Ru) 312-4, 314 omou (think, Ja) 226
naw (be, Am) 45 open-mouthed 158
ngaw (lonely, Lao) 150 opthalmos (eye, Gr) 75
orge (fury, Gr) 136 radost' (joy, Ru) 291, 293-9, 316, 330
otnoSenie (attitude, Ru) 315-7 rage 17, 118, 121
otryv (revel, Ru) 303 rasa (feel, Ma) 170
otvraSienie (revulsion, Ru) 314 razgovor (conversation, Ru) 330
outrage 5 razluka (separation, Ru) 291, 323-8,331
raznofintsy (intellectuals, Ru) 305-6
pa (prep., Mb) 72 -re (see, Mb) 72
paak (mouth, Lao) 155, 157-9; nom-~ rebuke 52
(clench-mouthed) 155, 159; qaa rejoice 17
(open-mouthed) 158 remorse 311
pepatah (maxims, Ma) 175 ren (benevolence, Ch) 376
pepe (don't, Mb) 71 ren qing (human feelings, Ch) 256, 282
perasaan (feelings, Ma) 170 rendah hati (humble, Ma) 167
peribahasa (trad, sayings, Ma) 175 revenge 121
petel (hungry, Mb) 71 rindu (longing, Ma) 170
phii (spirit, Lao) 159, baa (lunatic) rodnoj (beloved, Ru) 291, 320-2
159 rodstvennik (relative, Ru) 321
pikaringanyi (angry, Ya) 6-8, 210
pine 17 sabaaj (content, Lao) 151-2,158
pissed off 150, 162 sad, sadness 10, 13, 46-7, 66, 121,
pity, 319 149-50, 154, 169, 177, 359-99
pleased 15, 17 sak' 3 (laugh, Am) 56
-po se mata- (memorise, Mb) 80 sakit hati (annoyed, Ma) 168, 173-4,
-pok (uneasy, Mb) 72 179, 184,185, 186,190
poryv (impulse, Ru) 303, 305 samodovol'stvo (self-content, Ru) 295
prezrenie (contempt, Ru) 315 sbr (break, Am) 36
pride 190 scared 153
progolodats'ja (hunger, Ru) 324 sdast'e (happiness, Ru) 291, 295-9, 320,
przykro (hurt, Po) 25, 308, 337-43, 343, 330
344-56, 356 scastliv (happy, Ru) 298-9
puas hati (satisfied, Ma) 168, 173-4, scowl 155-5
179, 187, 188, 193 sdelat' (do, Ru) 9
pustoj (empty, Ru) 330 sedih (sorrow, Ma) 170
Seele (soul, Ge) 176
qaa-paak (open-mouthed, Lao) 158 sehati (of one mind, Ma) 175
qi (energy, Ch) 265, 268, 270-1, 283; samma (feel, Am) 22, 37-40
bie-zhe ~ (resentful) 268; chu~ (vent senang hati (relaxed, Ma) 168, 179,182,
anger) 270-1; -fen (infuriated) 275-6; 183-4, 193
~ren (irritate) 266; ~si (half-dead with separation 291, 326
anger) 268 serdce (heart, Ru) 196
qiu (autumn, Ch) 370-1 shame 4, 8, 66, 102, 198
shenglqi (anger, Ch) 24, 255, 265-9,
rabbia (fury, It) 134 269, 270-4, 276, 282, 284, 285
Rache (revenge, Ge) 139 shinkeishitsu (hyperanxiety, Ja) 222
rad (glad, Ru) 298 shy 55, 66,102
si (thoughts, Ch) 398 hear) 107; ~ ikam kik (ears ringing)

sia-caj (sad, Lao) 149-50, 158 197; ~ iluumu (not have to hear any-
simmet (feeling, Am) 40-1 more) 107, 110; ~imun (deaf) 107;
skazat' (say, Ru) 9 ~ ituntun (eager to hear) 107;
skuat' (boring, Ru) 301, 323-4, 331 ~ iyouyou (earache) 107; ~ iyouyou pa
slucit'sja (happen, Ru) 9 (tired of listening) 106-7,109
smile 35, 55-6, 66, 155-6, 158 tan (sigh, Ch) 366
smutno (sad, Po) 339, 342 taoyan (loathe, Ch) 24, 255, 263-4,
io(so, Ge) 129 279-81,281,282,286
sofuvstvie (sympathy, Ru) 317-8 te/o(body, Ru) 294
son-caj (interested, Lao) 160-1 tadsssata (happy, Am) 445
song (anger, If) 278-9 19k'at't' ah (livid, Am) 51-2
sook-saw (unhappy, Lao) 150, 158 tak'ot't'a (angry, Am) 22, 52-3
sop (lip, Lao) 157; ~-beq (about-to-cry t'dlla(hate, Am) 61,64
lip) 157-8 tambboi'abbaoa see hod-e
sorry, sorrow 24, 43, 340-2, 346-7, (bitter, Am) 58
351-2, 353, 355, 362, 367, 379, 396-7 (upset, Am) 50-3
soskuit'sja (start missing, Ru) 323-5, tsnbok' abbok' 9 (terrified, Am) 49
331 tanskka (touched, Am) 58
sostradanie (compassion, Ru) 317-8 terkejut (startled, Ma) 170
soul 176,294-5,302,330 tersinggung (hurt feelings, Ma) 185
sovest' (conscience, Ru) 311 (t9)s9mma (feel, Am) 22, 37-40
sovestno (shameful, Ru) 302, 308, 311-3, tfsknic (longing, Po) 292
313,314 tesno (cramped, Ru) 299-300
sozalenie (regret, Ru) 309 thymos (rage, Gr) 136
spirit 111 tidak sampai hati (intolerable, Ma) 172
splangxnon (intestines, Gr) 75 tir tir (pounding, Am) 59
stomach, butterflies in 75 tok-caj (surprise, Lao) 152-3,158,160-1
stray, (fear, Ru) 300-1 Tollwut (rabies, Ge) 135
styd, stydno (shame, ashamed, Ru) 311-3, toorn (anger, Du) 142
313 toska (yearning, Ru) 8-9, 291, 299-302,
suka hati (please oneself, Ma) 172 324; toskovat' (long for) 323
sulk 17, 122, 204, 209 toSno (anguish, Ru) 301-2
surprise 43, 149, 152-3, 177 tsietno (in vain, Ru) 299-301
susah hati (troubled, Ma) 167-8, 179, tulus (pure, Ma) 173-4
180, 181-2 -tuntun (itch, Mb) 72
sympathy 46-7, 65-6 tuun (awaken, Lao) 164
tuzit' (grieve, Ru) 300
t- (vb prefix, Am) 62 -twer (longing, Mb) 72
taa (face, Lao) 155; het ~-naj (make big-
eyes) 160-1;khang (stuck eyes) 155, ber (about, Ge) 123, 130-1, 143
160-1; loo (exposed-whites eyes) ufastie (compassion, Ru) 317-8
155,159,161 ;~-n7 (eyes held open) 160 udal' (daring, Ru) 291,294
takut (fear, Ma) 170 udovol'stvie (pleasure, Ru) 293-9, 316,
talqa- (ear, Mb) 76,106-7; ~ ikam (over- 330
urn (intellect, Ru) 294-5 xotet' (want, Ru) 9

-ur koipa (hate, Mb) 73-4
-ur mata- (hope, Mb) 85, 108-9 -yamaana (feel, Mb) 72
-ute (know, Mb) 72 yang (positive energy, Ch) 370-1
uze (sweaty, Mb) 71 yao (want, Ch) 399
-yes (sting/burn, Mb) 72
vengeance 121 yin (negative energy, Ch) 370-1
Verlegenheit (embarrassment, Ge) 128 -yoozo (smell, Mb) 72
Verzweiflung (despair, Ge) 128 you (loving concern, Ch) 390
yuanhen (resent, Ch) 277
want 12
ware ni kaeru (come to, Ja) 226 za/(sad, Po) 342
waddada (love, Am) 61,64 ialef, zalko, zalost' (pity, Ru) 8, 309-10,
will 111 317-20,329
woede (rage, Du) 142 zi (monosyllabic word, Ch) 362, 397
worry 63,386,390 znat' (know, Ru) 9
wrath 17, 136 Zorn, zornig (anger, angry, Ge) 23,
Wut, wtend (rage, angry, Ge) 23,116-9, 116-9,122,126,127-39,138,141,144
127-41, 137-8, 144 zuo' ou (nauseating, Ch) 281
zum- (lips, Mb) 76, 108; ikarrut ~ (bit-
xiang (think, Ch) 392, 399; ~ chou ing lips in anger) 108-9; ~ imbukmbuk
(homesick) 380-1,398 (pouting) 108-109

Jean Harkins, Anna Wierzbicka - Emotions in Crosslinguistic Perspective

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Jean Harkins, Anna Wierzbicka - Emotions in Crosslinguistic Perspective

Încărcat de

Drepturi de autor:

Formate disponibile

Emotions in Crosslinguistic Perspective

Printed on acid-free paper

Library of Congress Cataloging-in-Publication Data

Emotions in crosslinguistic perspective / edited by Jean Harkins,

Die Deutsche Bibliothek Cataloging-in-Publication Data

Emotions in crosslinguistic perspective / ed. by Jean Harkins ;

Testing emotional universals in Amharic 35

Emotions and the nature of persons in Mbula 69

Why Germans don't feel"anger" 115

Linguistic evidence for a Lao perspective on facial expression

Hati: A key word in the Malay vocabulary of emotion 167

Talking about anger in Central Australia 197

Meanings of Japanese sound-symbolic emotion words 217

Concepts of anger in Chinese 255

Human emotions viewed through the Russian language 291

A culturally salient Polish emotion: Przykro (pron. pshickro) . . . . 337

An inquiry into "sadness" in Chinese 359

Subject and name index 405

from different languages. Before proceeding to the language-specific

1. Emotions and languages

have often been seen as an important key to understanding cultural patterns

2. Linguistic evidence in studying emotions

pects of angry feelings and behaviour: pikaringanyi is associated with ac-

3. Emotion concepts as cultural artefacts

as Frijda (1986), Ortony et al. (1988), Johnson-Laird and Oatley (1989),

be explained in terms of the latter. For example, while there is no word in

4. Natural Semantic Metalanguage (NSM)

The claim that ordinary people's subjective experience can only be

Substantives I, YOU, SOMEONE (PERSON), SOMETHING

To reflect the special status of such words as exponents of universal

5. Describing feelings through prototypes

In literature, feelings are frequently described by means of comparisons:

Pleased (X was pleased)

6. Emotion and culture

on the study of the emotion lexicon, collocations, grammatical devices,

7. Cultural semantics in cognitive linguistics

Conceptual analysis of meaning carried out in universal semantic primes

8. Cognitive and semantic analysis in practice

tural meanings without imposing external interpretive frameworks. These

standing of human nature, and in the role of interpersonal feelings in Malay

of - without making unwarranted generalisations about "national char-

Davidson, Richard J. and Paul Ekman (eds.)

Gibbs, Raymond W. Jr. and Gerard J. Steen (eds)

Harkins, Jean and David P. Wilkins

Niemeier, Susanne & Rene Dirven (eds.)

In press Different modes of describing emotions in Chinese: Bodily

In this paper I present a semantic description of emotion predicates in Am-

(1) a. All languages have a word for FEEL.

j. In all languages, there are alternative grammatical construc-

2. An overview of Amharic grammar

Amharic belongs to the Ethiosemitic sub-branch of the Semitic language

(2) aster wods bet hedged

Amharic is a nominative/accusative language and exhibits a mixture of

3.1. A word for FEEL

(5) *w9nb3r- u t3- samma- fin

(6) a. minim a- y- samma- tint- mm

(8) a. ddssita te- samma- w

b. aster temma- n zafan as- sdmma- i- w

(12) a. simmet 'feeling' (cannot mean'hearing')

(13) a. yd- haz9n simmet

(14) a. aster bird t9- samma- t

It is important to note that physiological experiences such as 'be hun-

(15) a. aster (in) bdrr^d- at