Language of Emotions

8/3/2019 Language of Emotions

1/43

COGNI TION AN D EM OTI ON, 1989,3 (2), 81-123

The Language of Emotions:An Analysis of a Semantic Field

P. N. Johnson-LairdMRC Applied Psychology Unit, 15 Chaucer Road, Cambridge, U .K .

Keith OatleyDepartment of Psychology, Glasgow University, Glasgow, U. .

This paper uses a theory of the emotions to motivate a sem antic analysis ofEnglish words referring to emotions. The theory assumes that emotionshave a two-fold communicative function, both externally amongst membersof the species, and internally within the brain so as to bypass complexinferences. It implies that there is a small number of basic signals that canset up characteristic emotional modes within the organism, roughly corre-sponding to happiness, sadness, fear, anger, and disgust. In human beings,these modes can be modulated by the propositional content of the cognitiveevaluation that caused the emotion signal, or else, if this content fails toimpinge on consciousness, these m odes can be experienced as emotion s thathave occurred for no apparent reason . According to this communicativetheory, there should be a set of terms that refer to basic emotions, andthese terms should have no internal semantics, since they cannot be ana-lysed into anything more basic, such as a prototype or a set of semanticfeatures. Other terms should refer to states that combine a basic emotionwith a propositional content. Finally, the theory implies that any emotionalterm should devolve upon one of the five basic emotion modes, or somesubset of them, an d that there will be no need to invoke any other emotion alstates. These predictions were borne out by the semantic analysis of 590emotion words.

Requests for reprints should be sent to P. N. Johnson-Laird, MRC Applied PsychologyUnit, 15 Chaucer Road, Cambridge, U.K. We are grateful to various colleagues for readingearlier drafts of this paper and for criticising ou r ideas, and we thank Ruth Byrne, MartinConway, Gerry Clore, Valentina dUrso, Barbara Dritschel, Nico Frijda, Andrew Ortony,and an anonymous referee for their help. We are also grateful to Tony Anderson forcollecting various data and t o Richard Beckw ith for discussions on WordNE T and emotionalterms. Part of this work was camed out while Johnson-Laird was visiting the CognitiveScience Laboratory, Princeton University, with the support of the James S. McDonnellFoundation.= 3 / 7 4 0 1989 Lawrence Erlbaum Associates Limited


2/43

82 JOHNSON-LAIRD AND OATLEYINTRODUCTION

William James (1890, p. 485) argued that no single coherent outcome islikely to be produced by the analysis of words referring to emotions:If on e should seek to name each particular one of [the emotions] of which thehum an hea rt is the seat, it is plain that the limit to their num ber would lie inthe introspective vocabulary of the seeker, each race of men having foundnames fo r some shade of feeling which othe r races have left u ndiscriminated.If we should seek to break the emotions, thus enumerated, into groups,according to their affinities, it is again plain th at all sorts of groupings wouldbe possible, according as we chose this characte r or that as a basis, and thatall groupings would be equally real and true.We will advance an alternative proposal: Not only are there naturallyoccurring groups of emotions, but these naturally occurring groups formthe basis of th e m eanings of English emotion terms. They a re the subjec-tive experiences that emotion terms denote.We ca n discern two different kinds of alternative to our conjecture. Thefirst is well represented by the quotation above from James, and in morerecent times by Mandler (1962, 1984). On this view, emotion partlydepen ds on a heterogeneous s et of events that occur to the person havingthe experience. In Jamess theory, these ev ents are internal. In Mandlerstheory, they ar e meaningful external events that a re used to label arou sal.

As, in either ca se, such events are heterogeneo us, and often idiosyncratic,there is no reason to suppose that analyses of emotion terms will tell usanything substantial. This view fits well with the ethnographic thesis thatemotions are culturally variable and reflect the rudimentary theories andtaxonomies of folk psychology. Ordinary language refers to a wide varietyof different ideas about an assortment of experiences, and provides nobasis for a unified theory . These naive acc ounts will be ultimately replace dby a scientific understanding that will retain little or nothing of folkintuitions or terminology.The second view is that an emotion is a sequence that includes aneliciting cond ition, a cog nitive evalu ation, ph ysiological activatio n, a changeof action readiness, and finally an action (e.g. Frijda, 1986). Emotionwords may refer to all or any part of this sequence of events-which part ofthe sequence they do refer to may well be culturally and individuallyidiosyncratic. The conclusion is, again, that emotion terms are largelyheterogeneous, and that no coherent classification of them is possible.Nevertheless, some recent work analysing the emo tion lexicon from thesequential standpoint has been ca m ed out by Fehr and R ussell (1984), andby Shaver et al. (in press). These investigators argue that the concept of


3/43

THE LANGUAGE OF EMOTIONS 83emot ion depends on a proto type, ra ther than a se t of necessary andsufficient conditions. Different emotions a re more or less good exemplar sof the core c oncept, much as different exemplars of categories such as birdor fruit a re m ore o r less good exemplars of their respective categories (seee.g. Rosch, 1973). Because semantic analysis primarily concerns the tacitgrasp of concep ts, not an understanding available to introspection, i t is noteasy to d emo nstra te which particular concepts ar e prototypical-preciselybecause w e lack conscious access to the m. Prototypicality is oft en claimedto have been demo nstrated by showing that subjects rate instances of acon cept as varying in typicality an d are faster to verify good exe mp lars thanpoor exemplars . Armstrong, G lei tman, and Glei tman (1983) poin t out thatsuch phenomena are not decisive, because they may occur, as theseauthors show, with concepts that do have necessary and sufficient con-ditions.

Part of the pu rpose of this pap er is to test whe ther an ord erly semanticsof English e mo tion term s is possible. If so, it questions these two alterna-t ive positions, an d ma kes our theory of emotion m ore plausible. T wo linesof work are close to our own in that they too analyse the meaning ofemotion terms and specify their contribution to the truth condit ions ofsentences in which they occur. One analysis is due to Wierzbicka (e.g.1972, 1987). She goes further than we do , and proposes tha t al l emotionterms can in principle be analysed. We have benefitted from her w ork , butwe shall argue that only some emotion terms have a semantic analysis,wherea s other s de note unanalysable primitives.Th e o ther work closely related to ours is that of Ortony, Clore , and Foss(1987) and Clore, Ortony, and Foss (1987). In their initial work, theygathered together a large corpus of mental and affective words, and c am edout a componential analysis of them in order, in part, to distinguishbetween those words that referred to emotions and those that did not. Th eanalysis was based o n as few assum ptions as possible, in a way th at was notspecific to a ny particular theo ry of emotion including their own. Th eir firstdist inction was between internal and external condit ions (Ortony et al . ,1987). External conditions include descriptions of behaviour or objectivestates of affairs, such as Moses was abandone d in the bullrushes. T heterms that occur in these cases may have emotional connotations, but theydo not necessarily refer to emotional states. Within the terms denotinginternal condit ions, Ortony et al . dist inguish mental condit ions from non-mental co nditio ns, such as hungry and thirsty. Nex t, they dividemental terms up into those that focus on affects, those that focus onbehaviour, and those that focus on cognit ion. They propose, as we do,that em otions are mental states (not se quences that include elici tingconditions, actions, etc.), but these states, they claim, are valenced in thatthey imply moving towards or away from something.


4/43

84 JOHNSON-LAIRD AND OATLEYClore et al. (1987) have discovered that not just they, but also their

undergraduate subjects, could distinguish between emotional and .non-emotional states on the basis of a simple linguistic test. Their test for agenuine emotional term, such as happy, is that subjects rate bothfeeling happy and being happy as emotions. A term such as ignore d,however, is not a genuine emotional term , because subjec ts rate feelingignored as an emotion, but not being ignored. Thus, Clore et al.provide us with a helpful initial clarification concerning which terms aretruly part of the emotional lexicon.The feature that differentiates our analysis from those of Wierzbicka,and Ortony, Clore, and their colleagues, is that their approaches did notstart with any strong commitment to a particular theory of emotions.Wierzbicka views her work as part of a general semantic analysis thatranges over the entire lexicon, and so she is equally concerned with themeaning of red and cup. Ortony, Clore, and their colleagues havecomm itted themselves to the idea that emotions are mental sta tes, but theirmain aim is to derive a com putationally tractable calculus of the kind fromwhich a language understanding program me would be a bl e to deriveinferences. In contrast, we shall examine the consequences of our com-municative theory of emotions for the semantics of emotion words.

A COMMUNICATIVE THEORY OF EMOTIONSThe theory of emotions that motivates our semantic analysis has beenpresented in detail elsewhere (see Oatley & Johnson-Laird, 1987). It scentral assum ption is that emotions serve a communicative function bo thwithin the brain and within the social group. In both cases, em otions aresimple signals that propagate pervasively within the system. Unlike thesignals of a natural language, these signals do not have a propositionalstructure. T he m eaning of a propositional signal depen ds on combining themeanings of its parts according to its syntactic structure, whereas themeaning of a non-propo sitional signal is not comp osed ou t of th e meaningsof its parts. Its parts have no m eaning in themselves. A good example of anon-propositional signal is an alarm call: I t has a structure that enables it tobe easily recognised, but its significance does not d epen d on com bining themeanings of its parts according to the overall structure of the signal.Th e theory assumes that there is a small set of non-propositional signalsthat arise at recognisable junctures in plans. The signals govern themanagem ent of plans, particularly those that are concerned with multiplegoals. They enable an organism to react in a general preparatory wayrather th an eith er with the stereotyped response of a fixed action pa tternor with intricate behaviours that depend on costly inferential processing.He nce , eac h em otiona l signal is associated with a specific physiological


5/43

THE LANGUAGE OF EMOTIONS 85pattern, which perhaps has its own neurochemical basis. It acts within thecognitive system to set the modules of the system into a co-operative modeappropriate to the juncture in the plan. The organism is thus prepared toact in certain ways and to communicate emotional signals to others. Anawareness of this action readiness can contribute to the experience of anemotion, but one can experience some emotions without any consciousnessof a propensity towards certain actions. Those actions that can communi-cate emotions include facial expressions, movements and gestures, andtone of voice. As with the internal system, the external signals can set thecommunity of individuals into an appropriate emotional mode.

A major assumption of the theory is that mental architecture consists ina hierarchy of separate processors, or modules, that carry out computa-tions in parallel, and that an emotion can be set up by a cognitiveevaluation occurring at any level in this hierarchy. The evaluation can setthe processing modules into one of only a small number of emotion modes.These modes constitute the fundamental elements out of which all subjec-tive experiences of emotion are constructed, and, from our analysis of thejunctures at which they arise, w e take them to correspond to thoseexperiences that have in English as their closest labels: happiness, sadness,anger, fea r, and disgust. Around each mode, there may cluster a family ofrelated emotional experiences, e.g. if the mode of happiness has an object,then it constitutes a feeling of attachment. Consciousness depends on theprocessing module at the top of the computational hierarchy (see Johnson-Laird, 1983). In human beings, the normal subjective experience of anemotion accordingly depends on consciousness receiving both the emo-tional signal and a propositional message encoding the cognitive evaluationthat caused the signal to propagate in the first place. The theory allows,however, that the emotion can be consciously experienced in the absenceof a propositional message. The system as a whole may be in a particularmode for a relatively short period of time, in which case the state isordinarily referred to as an emotion, or it may remain in a particular modefor some time and often in the absence of any propositional informationabout its cause, in which case the state is ordinarily referred to as a mood.Certain personalities may even be constitutionally biassed towards onemode rather than others.

Bodily sensations are another form of mode, but they are distinct fromemotions in their causation, termination, and communicative conse-quences. Bodily sensations have physical causes, e.g. deprived of food onefeels hungry. They can be terminated by other physical causes, which inturn produce further bodily states. They have bodily and behaviouralconsequences that have a direct purpose. Emotions, however, have psy-chological causes. They are created by cognitive evaluations, e.g. theperception of a predator makes one fearful, its disappearance reduces the


6/43

86 JOHNSON-LAIRD AND OATLEYfear. T hey have consequences that include ritualised, or symbolic, behav-iours that no longer serve any function other than the communication ofthe em otion , e.g. a n alarm cry, or a laugh. These behaviours can commu ni-cate an em otional state to other members of the species, and sometimes tomembers of other species.Basic emotions often have bodily sensations accompanying them.According to our theory, a distinctive physiological state is associatedwith each emotion m ode, and on e can be aware of its bodily and somaticconsequence s-sweating, and a racing pulse, say, as a result of fear. Yet,these conseque nces are dissociable from the e motion ; one can experiencethe bodily sensations without the emo tion; one can experience the em otionwithout the bodily sensations. Of course, the cognitive evaluation of suchsensations may in turn lead to emotional consequences. L ust is an interest-ing case in point, because it depends on both a bodily state-sexualarousal-and an emotion of desire. The sensation can be produced byphysical stimuli, and pleasure can b e experienced in this way without thenormal concom itant emotion. When one is touched o n the skin, it makes agreat difference who is doing the touching. Different evaluations mayinduce love, fear, or repulsion. Pain is another bodily sensation that isintimately associated with emotion, and indeed the word pain and itscognates can be used to denote either the sensation or the emotion ofsadness. Emotions may be both aroused by, and m odify, the experience ofbodily sensations.Finally, the theory allows that there is a special category of complexemotions. They emanate from consciousness, because they arise fromcognitive evaluations that depend o n access to the mod el of the self, e.g.embarrassment, jealousy, and regret. These emotions are inextricablybound up with the propositional message that captures their cause: Theycannot be exp erienced without some aw areness of th e circumstances thatoccasioned them.

THE SCOPE OF THE PAPEROu r theory of emotions receives empirical support from a range of empiri-cal observations concerning both somatic and be havioural phenom ena (seeOatley & Johnson-Laird, 1987). It also yields many predictions that wehave yet to test, e.g. the possibility of a dissociation between feeling andpropositional content for only certain classes of emotion. Our aim in thispaper is to follow-up the consequences of the theory for the semantics ofemotion words. If the theory is correct in its essentials, then three mainpredictions follow. First, emotional terminology should be analysable intocoherent categories. Second, all terms denoting emotions ultimatelydepen d o n just the five basic families of emotion m odes, roughly speaking:


7/43

THE LANGUAGEOF EMOTIONS 87happiness , sadness, an ger, fea r, and disgust. This prediction allows that aword ma y deno te disjunctively mor e than on e of thes e modes; its essentialclaim is that the emotional component of any words meaning never goesoutside t hem . Third, words referring to em otions will reflect the str uctur eof emotional experience as posited by the theory, and so some words maybe used to re fer to basic emo tions, which can be experienced w ithout theindividual experiencing them being aware of their cause or their object ,whereas o ther words will designate emotions that can be experience d onlywith a known cause or a known object . There should also be wordsdesignating complex emotions that have a highly specific propositionalcontent tha t cannot be divorced from their subjective experience .It may turn out that there is no coherent or useful classification ofemotion words. This eventuali ty would show that the posit ion adop ted byJames , and o ther m ore recent psychologists, such as Mandler an d Frijd a, isthe correct one. I t may turn o ut tha t emot ion words refer to subjectivestates other than those that de pend on the five basic modes. If th ere is anysuch word-as evinced by, say, the judg em ents of subjects or the entries tobe fou nd in dictionaries, then th e theo ry is false in its curr ent form ulatio n.It may turn out that e motional words cannot be analysed in terms of thesorts of experience postulated by our theory. If so, then the theory isradically false. We have either proposed an erroneo us account of emot ionsor the terminology of daily life is wholly remote from the real nature ofemot ions (or both) .The remainder of our paper is organised in five main sections. First, wedescribe the corpus of emotion words that we have collected, and clarifycertain conceptual , morphological , and syntactic matters. Second, weconsider the vexed question of whether or not basic emotion words have asemantic analysis. Th ird, we outl ine the so rt of structure that is typicallyto be found in any semantic field as a guide to what we might expect foremotional words. Fourth, we describe each of the different classes ofemotional words and summarise their semantic analyses. The co rpus i tself,along with ou r semantic analyses for each word, is presented in App end ix 1,and a set of related terms that d o not deno te emotions is presented inAppendix 2. Finally, we draw some conclusions about our analyses.

THE CORPUS OF EMOTION WORDSWe collected a representative sample of words denoting emotions (seeAppendix 1) by inspecting a nu mb er of sources. First, we exam ined the se to f 196words which Fehr and Russell (1984) had obtained when they aske d200 subjects to write down instances of emotions. Many of the subjectsresponses denoted, not emotions p e r se, but expressions of emot ions (e .g .laughter, smiling, crying, tears, frown), bodily states associated with


8/43

88 JOHNSON-LAIRD AND OATLEYemotions (e.g. strong, tiredness), properties of emotion (e.g. deep, posi-tive, negative, expressive, mixed, disturbed, uncontrollable, turbulent),characteristics of behaviour motivated by emotion (e.g. sincerity, giving,helping, sharing, violence), personality traits related to emotion (e.g.outgoingness, gentleness, sensitive, stubbornness, hardness, vulnerability,hyperactive), states of mind associated with emotions (e.g. confusion,uncertainty, arousal, control, conflict, thinking, meditating, alert), andcognates and superordinates of emotion (e.g. reactions, responsive, state,communication, expression). None of these words refer to emotions ac-cording to the feeling X and being X test devised by Clore et al.(1987). Hence, we excluded them from our sample (but see Appendix 2,which lists many such words). Second, we included in our sample all thewords that occurred in the Clore et al. (1987) corpus that the experimentersor the subjects (or both) considered to contain an affective component.Third, we included the words of Tillers (1988) corpus. These threecorpora provided us with a total of 327 words. Finally, as a result ofscouring thesauruses, dictionaries of synonyms, previous psychologicalstudies based on emotional terminology (e.g. Davitz, 1969, 1970; deRivera , 197 7), and an unpub lished list devised by Richa rd Beckw ith as partof George Millers WordNET project, we were able to add a further 263words. They either passed the feeling and being test, though they werenot included in either the Fehr and R ussell or Clore et al. corpora , o r elsethey denoted causes of emotions, which we included for reasons that willbecome clear presently. Almost certainly, we have inadvertently omittedsome English words denoting emotions, but our sample of 590 words iscertainly representative an d extends previous corpora.Syntax a nd MorphologyIn order to present an analysis of emotion words as economically aspossible, we shall try to avoid analysing all the different morphologicalvariants of the same underlying root. The vocabulary of emotions doesindeed contain words from all the main open-class categories: nouns,verbs, adjectives, and adverbs. Many root morphemes take appropriatesuffixes to allow them to serve in all four categories. For example, fear isboth a noun and a transitive verb, but it is also the root of certainadjectives, fearful, fearless, and fearsom e, and their correspondingadverbs, fearfully, fearlessly, and fearsomely. These adverbs havealso been turne d into nouns: fearfulness, fearlessness, and fearsom e-ness. A t the ro ot of all of these words is the same morphem e den oting thesame basic emotion . Th e interpretation of the suffixes is straightforward.They attribute the emotion o r its denial to an individual, or they attributethe power of causing it to an individual; they map these notions into a


9/43

THE LANGUAGE OF EMOTIONS 89manner of performance; and finally they convert these manners of per-form ance into abstract properties. He nce , in general , we shall treat onlyon e or two forms of a w ord, and we shall not at tempt to d eal with all theother morphological variants into which the same root enters.In som e cases, however, there ar e changes in the interpre tation of wordsformed fro m the sam e underlying root. I n i ts emotional sense, for exam ple,the verb affect a nd its participle affecting d en ote the power of movingthe emotions, but the noun affection denotes the narrower concept of anattachment towards someone or something. The re are o ther sh i f ts of thissort . Compare the following pairs for instances of the phenomenon:lovable-lovely, dread-dreadful, awe-awful . The secondmember of each pair does not den ote an em otional state. Likewise, ther e isno guarantee of the productivity of a particular suffix, e.g. hate yieldshateful but not hatesome. Where a root yields different words withdifferent emotional meanings, we will analyse both of them.Th e etymology of emotional terms is complicated, and l ies outside thescope of this pa per . We shall instead app roac h the seman tic field synchron-ically, justifying this strategy on the following grounds. The words thatcontinue to be used frequently serv e a useful semantic function , an d henceour task should be to show that their meanings can be elucidated by ou rtheory.Experience, Concept, WordBef ore undertaking any semantic analysis, it is important to be clear ab outcerta in funda men tal distinctions, which can be illustrated by th e followingpred icam ent: You can be in the grip of a particular emotio n, but it may behard f or you to conceptualise your experience and thus to describe i t inwords. Th is situation enab les us to distinguish three im porta nt entities: a nemotion, a concept of an emotion, and a description of an emot ion . Anemot ion such as embarrassment is what you feel; a concept is a mentalconstruct that en ables you to categorise your experience as one of em bar-rassment; and a description is a way of putting your experienc e, presum -ably by way of its categorisation, into words. The meanings of words areconcepts-those concepts that have been dignified by a word for thepurposes of communication. Hence, when words refer to things in theworld, such as clouds or cuckoos, they do so by way of their meanings-theconcepts that people enter tain abou t those things. But, because em otionsare experienced directly, the l inkages between experience, concept, andword, a re diffe rent, as we shall see.The meanings of emotional words are not immediately available toconscious inspection, and their analysis is complicated by several factors.In particular, emotional vocabulary is not the result of parsimonious


10/43

90 JOHNSON-LAIRDAND OATLEYplanning. Many words referring to emotions have other meanings too;many words are near synonyms an d differ only in their conn otation s andusage; and in English and other languages, some emotional words aresystematically ambiguous because they can be used to refer either to animmediate subjective feeling or to a general predisposition. Fo r exam ple,you can assert I am frightened of her either to refer to a feeling thatcurrently grips yo u or else to refer to your general attitude towards therelevant individual, i.e. how you ar e disposed towards her eve n though youare not actually feeling frightened at the moment of your utterance.Ano ther kind of ambiguity, a s Clore et al. (1987) have pointed out, arisesin the use of words that d o not, strictly speaking, refer to em otions but thatcan be used to convey an emotional state, e.g. feeling ignored. As wehave mentioned, we have tried to exclude the latter sort of words from ou rcorpus.

DO BASIC EMOTIO N WORDS HAVE ANANALYSABLE MEANING?According to our theory, there is a set of basic emotion modes thatcorrespond to internal signals that can impinge on consciousness. Thesemodes-happiness, sadness, ang er, fea r, disgust-should be universallyaccepted as discriminable categories of direct experience. Basic emotionsignals have n o internal structu re that is parsed an d interprete d within thesystem. Hence, it follows from our theory that there is no way in whichwords that refer to the subjective experiences corresponding to thesemodes can be analysed semantically: The modes are primitive subjectiveexperiences that the words denote. They are, as philosophers say, unana-lysable qualia.If you were emotion-blind and unable to experience emotions, thenyou would have no ide a what it was like to feel, say, sadness. Word s thatcan be analysed semantically, whether based on a prototy pe or on necessaryand sufficient conditions, can be communicated to people who are notfamiliar with them. Although there are studies that have explored theanalysis of emotions in terms of prototypes (see e.g. Fehr and Russell,1984; Shaver et al. , 1987), there are no concepts which, if taken forgranted, will enable us to communicate the contribution that the wordsad makes to the truth conditions of sentences. If you are emotionallyblind, we cannot convey the meaning of the word sadness to you. Onebrave attempt in the literature to define its meaning (Wierzbicka, 1972,p. 61) offers the following analysis:

X feels sad = X feels as one does when one thinks that what one has desiredto happen has not happened and will not happen.


11/43

M E ANGUAGE OF EMOTION S 91But, as Wierzbicka (p. 59) herself remarks:

Tho ugh ts have a structure which can be rendere d in words, but feelings, likesensations, do not. All we can do, therefore, is to describe in words theexter nal situations or thoughts which are associated in our memory or in ourimagination with the feeling in question and to trust that our reader orlistener will grasp what particular feelings are m eant.This procedure may be the best we can do , but i t is not good enough fo rsomeone who cannot experience the emotion of sadness. Such an indi-vidual will not know how one feels when one thinks that what one hasdesired to happe n has not happene d. Likewise, the following sentence:

John feels sad even though he d oes not think that what he has desiredto happen has not happened and will not happen.uses the term in a way that explicitly violates the definition. Yet theassertion is entirely sensible, because a basic emot ion such as sadness canbe felt for no known reason-a phenom enon that our theory elucidates.We agree with Wierzbicka (1987) that one of the aims of semant icanalysis is to uncover the set of universal semantic primitives, a nd t hat it ispossible to analyse the meanings of certain em otion terms, such as the onesdenoting complex emotions, into more basic components. Thus, for ex-ample, the meaning of reg ret can be analysed along the following lines:

Regret: sadness as a result of evaluating ones past action as harmfulWh ere we disagree is over the status of words tha t denote basic emotionmod es, such as sadness. Wierzbickas strategy is reminiscent of the accountof em otion in Frijda (1986): She defines an emotio n by recounting a briefscenario of a possible cause of the feeling. B ut, this analysis conveys onlythe conditions in which someone is likely to feel sad; it does not conveyanything about what it feels like to be sad.Why, then, not accept that the meaning of sad is given by a universalsemantic primitive corresponding to on e of the five basic emotio n mo des?One answer is that there is a l ink between sadness, anger, anddisgust-they all de no te negative emotions-and so they can not corre-spond to sem antic primitives (see Wierzbicka, 1972; Frijd a, 1987). As wehave argued elsewhere (see Johnson-Laird & Oatley, 1988), it may befallacious to assume that because several words fall into a superordinatecategory, their meanings necessarily decompose into components, at leastone of which is common to all . We would argue, for example, that themeanings of red , green , b lue, etc. do not each contain a componentequivalent to C O L O U R , plus some other comp onent that dist inguishes the

or wrong in relat ion to ones current stand ards.


12/43


13/43

THE LANGUAGE OF EMOTIONS 93experience that goes along with these observable eliciting conditions andconcomitants is called sadness. Hence, when you observe other peoplein similar situations displaying similar signs, you can attribute the samesubjective experience to them. Your attribution may be wrong: They maybe feigning the emotion, or they may be emotionally blind and lack thesubjective experience, but in general you will be right.

Although observable eliciting causes and concomitants are necessary foryou to learn how to use emotional terms (cf. Wittgenstein, 1953), they arenot part of the meaning of basic emotion words. If they were, it would beanomalous to make such assertions as:

I feel sad but I dont know why.I feel sad even though I dont show it in any way.Are causes and concomitants part of a prototype of the emotion? This is

a difficult question, but one that can be answered by considering somefurther examples. If a speaker asserts:The person I love has left me.

it is reasonable to infer the speaker feels sad. Indeed, the apparent absenceof the feeling in the case of patients suffering from encephalitis lethargica isremarked upon by clinicians (e.g. Meyer-Gross, Slater, & Roth, 1960).Likewise, the inference is even stronger if someone asserts:

The reason I am weeping is because the person I love has left me.Conversely, when a speaker asserts:

I am sadone can infer by default that if the remark is true, then something hashappened to cause the feeling of sadness-there is a variety of possibleexplanations. One can also infer that the sadness is likely to be expressed inthe speakers demeanour and behaviour. However, the reader should notehow we described these default inferences: We said that something hashappened to cause the sadness and that some behaviour will express thesadness. We did not say that the eliciting condition is part of the sadness orthat the concomitant expression is part of the sadness. In short, themembers of a culture have a prototype for the sorts of events that cause anemotion such as sadness, and for the sorts of events that ensue; but they donot have a prototype for the subjective feeling itself. It is an unanalysableprimitive experience. Hence, we conclude that a basic emotion, such assadness, has causes and consequences, but is itself only a part of aprototypical sequence. Complex emotions, as we shall see, are ratherdifferent in this respect.


14/43

94 JOHNSON-LAIRD AND OATLEYTHE STRUCTURE OF THE SEM AN TIC FIELD

Given a basic concept, such as the notion of movement, Miller andJohnson-Laird (1976, sec. 7.5.1) show how th e concept underlies an entiresemantic field, that is, a set of words with meanings that depend on theconcept. They also describe four principal ways in which the concept canbe elaborated in the meaning of a particular word in the sem antic field:1. The w ord may presuppose a m ore restricted range of ap plication ofthe basic concept, e.g. leave presupposes that its subject is a t aparticular location.2. The word may depend on a particular modification of the concept,such as an adverbial modification or manner, e.g. lurch is a

particular way of moving.3. Th e word may signify a causal relation into which the basic conceptenters, e.g. to shift something is to cause it to move.4. The word may introduce an intentional component into which thebasic concept enters, e.g. chase. You can accidentally shift some-thing, but you cannot accidentally chase it.Emotional words ought likewise to d enote different elaborations of th ebasic emotion modes. In order to outline the general structure of thissemantic field, we will follow Miller and Johnson -Lairds proc edu re (1976 ,sec. 6.3.1) and use a series of diagnostic questions in which t o captu re ou rintuitions, and, we hope, those of the reader. A similar procedure is

employed by K eil(l9 79) to delineate the ontology of physical objects. Th euse of intuitions can, of course, be backed up by data obtained frominformants, but the first stage must be to formulate a theory based onintuitions about th e clear cases (see Chomsky, 1965).Th e questions that organise our taxonomy arise directly from our theoryof the emotions. The first question to be asked about any abstract term iswhether it can be used to refer to a feeling. If it can, then one can askwhether it is a purely bodily sensation, such as an itch or thirst, or asubjective feeling, such as happiness or fear, which may include somesomatic elements-see also Clore et al. (1987), who use the same initialquestion. There is, of course, a small class of generic terms, such asemotion and passion, which can be used to denote emotions ingeneral. If a word d enotes an em otion, then one can ask whether o r not itis possible to experience the feeling without knowing its cause or object.Terms denoting feelings that can be experienced without knowing theircause or their object correspond to ou r category of basic emotions. (Theseterms can also be used to describe feelings with known causes; the point is


15/43

THE LANGUAGE OF EMOTIONS 95that they need not be used in this way.) Of course , an emot ion can beexp erien ced in different ways and in differing degr ees of intensity, an d sowe can ask whether a term deno tes such a modification, e.g. e lationrefers to an intense form of happiness, and so i t counts as a simplemodification of a basic emotion.If a word denotes a feeling that must have a known cau se or object , thenits analysis calls for a c ombination of a basic em otion mo de with a cognitiveevaluation. T he questions that distinguish thes e words concern the natureof that evaluation. We can ask whether a word concerns the object orsource of an em otion. Thus, if Ja mes fears Joan , then sh e is the object orsource of his fear. He is in a part icular em otional relat ion to her , and wecategorise these words as denoting emotional relations. The relat ion canoften be ex perienced for no known reason, e.g. love.Another question we can ask is whether the word denotes an emotionthat must have a known cause, e.g. a person can be glad because a friendhas recovered from an illness. A word such as glad does not specifyanything ab out the part icular nature of the ev ent el ici ting the em otion, buti t does dema nd some event causing happiness. Hen ce, on e cannot sensiblyassert , I feel glad but I dont know why. W e shall refer to these words asdenoting caused emotions.A n impo rtant class of words in many se mantic domains (see Miller &Johnson-Lai rd , 1976) are so-called causatives. Such verbs exist for emo-tions, and their use in the passive voice provides another way of r e f e m n gto caused emot ions, e .g .

I was saddened by his death.We have included these verbs (and also some causative adjectives, such aspoignant) in our corpus. They are identified by asking whether a wordconcerns the cause of an emot ion , e .g . to sadden someo ne is to cause themto feel sad .A part icular form of caused emotions that we shall sepa rate from the restare those that concern goals, e.g. desire. We distinguish these words byasking wh ether, given the relevant emotional state , there is something thatone has a s a goal. We refer to these words as denoting emotional goals.Finally, there are words denoting emotions that combine an emotionalmode and a propositional evaluation that conc erns some aspect of the self.These words can be identified by asking whether an emotion results fromevaluating oneself in some situation, e.g. belonging denotes the feelingthat one is happy in relation to others and vice versa. Because thesecomplex em otions depend on the model on e has of oneself-a model thatis accessible only by way of consciousness-they arise in consc iousne ss andSO cannot be experienced without an awareness of the circumstancesoccasioning th em. We refer to w ords denoting these emotions as complex.


16/43

96 JOHNSON-LAIRD AND OATLEYWords denoting emotions therefore fall into seven main categories

according to our sem antic classification. They can denote :0. Generic emotions, e.g. emotions and feelings.1. Basic emotions, e.g. happiness and elation.2. Emotional relations, e.g. love and hate.3. Caused emotions, e.g. gladness and horror.4. Causatives, e.g. im tate and reassure.5. Emotional goals, e.g. desire and avarice.6. Complex emotions, e.g. embarrassment and pity.

We shall say no more about th e generic terms, but turn to a mo re detailedexploration of each of the remaining categories.THE SEMANTIC ANALYSES OF EMOTIONWORDS

1. Basic Emotional TermsOur theory posits five basic emotion modes, and so we can predict thatthere should be words referring directly to these emotions or to simplemodifications of them . These words should accordingly denote emotionsthat can be experienced without the experiencer knowing their cause,though obviously they can also be used to refer emotions experienced for aknown cause.How can we determine that a word is semantically related to one of th efive modes designated by happiness, sadness, fear, anger, anddisgust. One method is to use the so-called but test (Bendix, 1966;Miller & Johns on-L aird, 1976, sec. 6.3.1). If two words have nothin g incommon, they and their negations can be freely combined with theconjunction but:

He was tired, but he was happyand

He was tired, but he was not happy.Thus, there is no semantic component in common to both tired andhapp y. If two words are semantically related, how ever, then the resultsof one or both of the combinations will be odd. For example, apprehen-sive and petrified, which are both caused emotion terms, share asemantic component, because although it is acceptable to assert:He was apprehensive but not petrifiedit is anomalous to assert:

He was petrified but not apprehensive.


17/43

THE LANGUAGE OF EMOTIONS 97Both words, of course, denote fear, but apprehensive denotes mild fearabout poss ible future events whereas petrified denotes intense and para l-ysing fear. Hence, the oddity of the following sentences:

He was apprehensive, but h e felt fear.He was petrified, but he felt fear.

Similarly, there is a semantic relation between distressed and petrified,because it would be odd to say:

He was petrified, but not distressed.According to our analysis, distress has a disjunctive denotation: Sadnessor fear fo r a known reason, and so the two words have fear as a commoncomponent. The but test can,be helpful in exploring close semanticrelations, but, as Miller and Johnson-Laird remark, it should be used withcaution especially for more distant relations.One modification of the basic modes, which the but test helps toreveal, concerns the intensity of the mode. Thus, to be joyful is to feelconsiderable happiness, whereas to be ecstatic is to feel intense happi-ness. In general, the more intense an emotion, the less likely it is to beexperienced acausally, because it verges on the pathological to feel ex-treme emotions without .knowing the reason why. Yet it seems that basicemotions can be experienced intensely for no apparent reason, as in thecase of joy, irritation, or free-floating anxiety.

Another aspect of a basic emotion mode is its temporal duration.Certain terms refer to a currently experienced emotion, others refer to aprolonged state or mood, and still others refer to an even longer-termstate-a disposition of the personality towards feeling that emotion. Thus,an individual can be described as imtable if he or she is currently angry, orin an angry mood, or has a general disposition to be angry. What ourtheory predicts is that basic emotion terms can be used to refer to moods orto emotional types of personality, because moods and personality types,often do not have a discernible cognitive cause. Both depend on emotionmodes, and it follows that the vocabulary of basic emotions should providedescriptions of moods and personalities. Likewise, it should be possible touse generic emotion terms to refer to moods and personalities. They toodo not require a discernible cognitive cause because they can be used torefer to any emotion including basic emotions. Where a word designates acaused emotion, i.e . one where the cause is known but does not have to fita particular propositional recipe, then it will not in general be appropriateto use it to refer to a mood, still less to a personality type. Thesepredictions are corroborated, as the reader can verify by consultingAppendix 1.Thus, for example, it would be unusual to describe a person ashaving a personality that was jubilant, grief-stricken, terrified, furious, ordisgusted. It is only when we come_o the complex emotions, such asCE S / Z - B


18/43

98 JOHNSON-LAIRD AND OATLEYTABLE 1Basic Emotions: A sample of words denoting basic emotions ofdiffering intensities, which can occur in the absence of an yknown propositional content. These words can also be used torefer to m oo ds and to personality types

Basic ModesHappiness Sadness Fear Anger Disgut

~~

Light-hearted Wistful Timid GrouchyCarefree Gloomy Tense Touchy QueasyHappy Sad Anxious Irritable NauseaHigh Melancholic Fearful A n g r yEuphoric Depressed Panicky IrascibleEcstatic Wretched Craven Splenetic

jealousy and shyness, that we again encoun ter words that a re suitable todescribe dispositions.As there are 109 words in Appendix 1 hat can be used to denote basicemotions, we cannot consider them all here, but Table 1 presents anillustrative subs et. As the table shows, basic terms gen eralise naturally tomoods and personalities. Th e labels at the head of the table-happiness,sadness, and so on-are not unique names for the five emo tion mo des ,but rath er tho se words of everyday English that seem most closely to referto the m odes in their unm odified forms. Th e table includes words denotingmild, ordinary, and intense emotions. None of these states necessarilydepends on a conscious awareness of its cause.2. Em otiona l RelationsEmotions are typically about some one or som ething; they are m ore likelyto be experienced in relation to individuals or their actions than m erely in avacuum. Hence, one should expect there to be words that refer to therelation between someone who experiences an emotion and its object, e.g.James fears Joan.There is nothing problematical about the idea that fear and anger canhave objects. Love and hate must likewise have objects, and our theoryimplies that they too depen d on a combination of emotional mode-happi-ness an d disgust, respectively-with the cognition identifying the person orentity towards whom, or which, the emotion is felt. This source of theemotion can be treated as its cause, but there are some subtleties to beironed out. One can experience an emotion towards someone withoutknowing why they engender the feeling. Thu s, just as one can be happy for


19/43

THE LANGUAGE OF EMOTIONS 99TABLE 2Emotional Relations:A sample of words that express an emotionalrelation between the subject, who experiences the emotion, andthe object of the emotion

Basic ModesHappiness Sadness Fear Anger DisgustLike Miss Afraid of Aggrieved with DislikeLove Mourn Fear Angry with HateAdore Grieve for Dread Scorn Loathe

no known reason so, too , one can love, hate , o r fear som eone wi thoutknowing the reason why. As theorists, we d o not doub t that the re is a causeof such emotions, but that cause may have l i t tle or nothing t o do with theobject of the emo tion, and may not have ente red the consciousness of theperson experiencing the em otion.Hatred is often expressed in displays of anger, but this connection is acontingent rather than a necessary one: you can hate someone, or some-thing, without feeling anger; you can be angry with som eon e you do nothate. Sadness can also have an object , as when you miss someone fromwhom you ar e separated. Sympathy and pity dep end on the sam e emotion-al mode but they are complex feelings that we will come to presently.Table 2 presents som e examples of th e 76 terms in ou r corpus that can beused to refer to emotional relations.Of th e term s denoting emotional relations, there is only on e subset thatmight be taken not to correspond to a basic mode, namely, bravery,cour age, boldness, and their cogna tes. In ou r view, to be brave i s not tofeel fear-or at least not to manifest it-in circum stance s likely to provo keit . Braver y does not have a particular phenomenology oth er than pe rhaps aslight feeling of being pleased with oneself, or a particular physiology othe rthan, perhaps, that of some conflicting symptoms of fear . I t i s a lack of anemotion rathe r than the posit ive presen ce of one. Similarly, serenity,peacefulness, and their cognates, refer to the mildly pleasurable sensationassociated with a lack of dysphoric emotion in circumstances that mighthave provoked i t .3. Caused EmotionsCe rtain words deno ting emotions normally signify a feeling that h as a causeknown to the individual experiencing it. For instance, if you say, I amglad, then, as the but test shows, you feel happy, but you cannot


20/43

100 JOHNSON-LAIRD AND OATLEYTABLE 3Caused Emotions: A sample of words denoting emotions that havecauses known, in part, to the person experiencing the emotion

Basic ModesHappin ess Sadness Fear Ang er DisgustCheered Dejection Apprehension DisgruntledEnjoym ent Sorrow Consternation MiffedGlad Heart-broken Afraid Indignant DisgustDelight Inconsolable Panic Cross SickOverjoyed Desolate Terror Furious

properly disclaim all knowledge of what has occasioned the feeling. Th us,we can modify the but test to have as its second clause: but I dontknow why, or but I know why. The test shows that words refemng tobasic emotions (see e.g. Table 1) do not necessarily contain a cognitivecom ponen t. It is perfectly sensible to assert, for example:

I am happy but I dont know whyor

I am happy but I know why.However, it would be odd to claim:

I am glad but I dont know whybecause the term is normally used to relate an emotion to a reason orcause, e.g. I feel glad because the winter is over, or to express anattitude towards a proposition (a propositional attitude in philosophicalparlance ), e.g. I am glad that winter is over. Hen ce, the language makesthe distinction predicted by our theory: Some emotions are experiencedwithout knowing their cause or reason, and others-those we refer to ascaused emotions-are experienced for a known reason. Of course, allemotions have a cause, and so ou r label is meant to imply merely that someaspects of it are known to the ex periencer. The five basic emotional modesought to underlie the caused emotions, too , and this prediction is borne outby the analyses of the terms refemng to caused emotions. There are 101words in our corpus that can be used to denote caused emotions, and wepresent a set of typical examples from them in Table 3.4. Causatives and Em otionsOne comm on form of discourse abou t caused emotions relies o n causativeverbs. These verbs, as we have noted, express the relation between thecause of an emotion and the person who experiences i t , e.g. The news


21/43

T H E LANGUAGE OF MOTIONS 101annoyed the President. The passive form of the verb can accordingly beused to refer to a caused emotion: The President was annoyed by thenews. In general, the description of the cause of an emotion is theconverse of the description of a caused emotion: The two run alongtogether in parallel. But there are some exceptions to this principle: somecausative verbs denote the cause of a complex emotion (e.g. humiliate);and some have passive forms that do not denote emotions (e.g. to chafeSomeone is to cause them to feel anger, but speakers do not ordinarily referto being chafed). A few adjectives also denote properties that causeemotions (e.g. poignant, tragic, dreary).

Although we have not marked the distinction in the analyses in Appen-dix 1, it is worth noting that some causatives refer to causes that are not theobjects of the emotion. For example, it may be true that:

Joan frightened Jamesbut Joan as such may not be the object of Jamess fear, because he mayhave been frightened by something that she did. Indeed, he may not evenrealise that it was she who was responsible. Other causatives, however,refer to both the cause and the object of an emotion. For example, if it istrue that:

Joan intimidated Jamesthen something that she did, or something about her, caused him to fearher.

Over 180words in our corpus can be used as causatives, and they divideup into a number of families. As we expected, there are verbs that denotecauses of each of the five main emotion modes, and we present someexamples in Table 4. In addition, however, some denote causes of any

TABLE 4Causatives: A sample of words that can be used to denote thecauses of emotions. Their passive forms can accordingly beused to denote caused emotionsBasic Modes

Happine ss Sadness Fear A nge r DisgrcstConten t Deflate Disquiet Irk PutoffPlease Disillusion Perturb Peeve Ali ena teAmuse Dampen Wony Irritate EstrangeDelight Depress Scare Annoy RepelTransport Sadden Frighten Enrage Nau seateEnthrall Disappoint Terrify Incense SickenExhilarate Desolate Petrify Infuriate Revolt


22/43

102 JOHNSON-LAIRD AND OATLEYemotion (e.g. excite, provoke, stir, overwhelm ), some denote genericallythe causes of negative emotions: sadness, anger, fear, or disgust (e.g.upset, disturb, bother, trouble, distress), and some denote the causes ofcomplex emotions (e.g. humiliate, embarrass, and encourage). Certaincausative verbs denote the restoration of emotional equilibrium aftersadness (e.g. cheer up, console, solace), fear (e.g. hearten ), or anger (e.g.mollify, appease , placate). Only one set of causative verbs appears to falloutside the domain of the basic modes, and these verbs concern surprise(e.g. amaze, astonish, flabbergast). We have argued, however, that sur-prise is not a distinct emotion, but a reaction to an unexpected ev ent thatcan be the precursor to any of th e five emotion modes (Oatley & Johnson-Laird, 1987).

5. Emotional GoalsEmotions often function as motives that lead to characteristic behavioursdesigned to achieve goals. Love may lead to approach, sadness to with-drawal and inaction, fear to flight, anger to aggression, and hatred toavoidance. The achievement of a goal produces happiness, and certainwords denote the state of having a goal (e.g. inclination, desire, need,want). Ot he r terms den ote specific sorts of goals (e.g. avarice, curiosity,greed, lust). Unfulfilled goals may lead to sadness or to anger, and somewords denote these states (e.g. discontent, disappointment, frustration).There are still other verbs that express the sadness that results fromthwarted love or desire (e.g. long for, pine for, lovesick). There are 42words denoting emotional goals in our corpus (see Appendix 1).6. C omp lex EmotionsWords that denote basic emotions can be used to refer to complexemotions too. Thus, someone can say:

I felt anxious because I was aware'that I had made a fool of myself infront of those peopleand then agree that the experience was one of emba rrassm ent, which is amild fear or sham e brought on by a self-conscious assessment of oneself ina social situation. As term s that refer to basic emotions can also be used torefer to complex emotions, the structure of the language must n o t beconfused with the underlying structure of em otions. The words in Table 1can refer to both basic and complex emotions. This possibility is to beexpected given our analysis, because we claim that all complex emotionsdevolve on the basic emotion modes. W here a pa rticular complex emotionoccurs frequently in a culture, and is perhaps of special significance in


23/43

THE LANGUAGE OF EMOTIONS 103social relations, then, as in other domains of discourse, appropriate termsare Iikely to have entered the language to refer to it. However, there is animportant asymmetry: A term refemng explicitly to a complex emotion isrestricted to it, and is not interpretable as refemng solely to the underlyingbasic emotion. Thus, for example, the word embarrassment denotes acomplex emotion, and it cannot be used merely to refer to a basicunderlying emotion. The reason for this asymmetry is that terms whichexplicitly designate complex emotions possess a complex semantic struc-ture. Knowing how to use them properly, that is, knowing their contribu-tion to truth conditions, depends on a grasp of the propositional content ofthe cognitive evaluation that creates the complex emotion. The word couldnot have been coined, or maintained in the language, unless this prop-ositional information is available to members of the language community.It is part of the subjective experience of the complex emotion.

The preceding argument provides us with a powerful linguistic tool foridentifying terms that refer to complex emotions (and not to basic emo-tions). Given that any complex emotion depends on a basic emotion mode,it follows that there will be a term refemng to a basic emotion that renderstrue an implication of the following form:

If you feel complex emotion C, then you feel basic emotion B.Here is an example of such an implication:

If you feel regret then you feel sad.However, the converse implication is not necessarily true:

If you feel basic emotion B , then you feel complex emotion C.Indeed, the implication:

If you feel sad then you feel regretis not generally true, though it may be true on occasion.

The same pattern of inferences can be generated for terms that refer todifferent degrees of a basic emotion, but a further step in the argumenteliminates these cases. Complex emotions depend on a propositionalcontent reflecting the high-level cognitive evaluation giving rise to them.Hence, a paraphrase of an assertion about a complex emotion can alwaysbe provided by using a basic emotion term in a context that captures thiscontent. For example:

If you feel regret then you feel sad as a result of evaluating a pastaction as harmful or wrong in relation to ones current standards.Our theory accordingly implies that a word specifically denoting a

complex emotion should be analysable in terms of a basic emotion and


24/43


25/43

THE LANGUAGE OF EMOTION S 105there a re feelings engendered by a comparison with others that are feelingsabout them, e.g. various forms of sympathy and empathy, and the morebitter feelings of envy and jealousy. The latter is instructive: If you feeljealous, then you judge yourself likely to be supplanted by a third party inan attachment, and in consequence you feel hatred for the third party. Ofcourse, you may be angry, sad, or fearful, too, but you can experience thepangs of jealousy without feeling anything but cold hatred for the thirdparty. And if you have this feeling for the relevant reason, it would bewrong to deny that you felt jealous. A representative set of terms denotingcomplex emotions is summarised in Table 5 , which shows the basic modefor each of them.

There are no cases-in which the meaning of a complex term appears to lieoutside the basic modes. Other highly specialised complex emotions reflecta cultural influence on their propositional content that differs from onesociety to another. They include aesthetic, religious, sexual, and othertranscendental feelings. Examples of words referring to such emotionsinclude piety and accidie (i.e. spiritual torpor), and words that have beenimported into English to make up lexical gaps, e.g. masochistic, Schaden-freude, and Weltschmerz.

CONCLUSIONSEmotions function as two-fold communications that enable a repertoire ofbehaviours to be produced with a minimal load on the information-processing system within an organism and on the communicative systembetween organisms (see Oatley & Johnson-Laird, 1987). Human beingscan experience basic emotions for no apparent reason, but they can alsoexperience emotions that have an object, a cause, or a goal, and complexemotions that depend on high-level cognitive evaluations. All these typesof emotion depend on a small set of emotional modes.

We have approached the everyday language of emotions armed with thistheory, which rests on empirical evidence from outside the linguisticdomain, and we have shown how the different components of the theoryare reflected in the words that are used to describe emotional experiences.This language and its underlying conceptual apparatus is intimately relatedto the real nature of emotions, and the meanings of emotional terms areneither arbitrary nor unanalysable but do indeed relate to experience. Thefolk psychology embedded within the language is essentially correct thoughradically incomplete and seldom articulated. The semantic field is based onthe five emotional modes, and words that refer solely to them have nointernal semantic structure-the modes are primitive and unanalysablestates, at least from the standpoint of normal mental processing. Otherwords do refer, as we expected, to emotional experiences that combine a


26/43

106 JOHNSON-LAIRD AND OATLEYbasic mode with a knowledge-often partial,. and perhaps often erroneous-of the cognitive evaluation that led to the mode or that concerns theobject of the emotion. Likewise, there are terms that denote complexemotions that d epend on cognitive evaluations concerning the m odel of theself.What obscures the relatively simple structure of the semantic field is thediversity of terms that contain an emotional component. Likewise, thedivergent analyses of emotional terminology to be found in the literatureare a consequen ce, not of the absence of underlying order , but of the use ofdifferent m ethodologies lacking any common theory of emotions. Previousstudies have also erred by including components that are not truly emo-tions, such as characteristics of behaviour like cruelty, aggression, andvehem ence (see e.g. F rijda, 1970; Plutchik, 1962; Schlosberg, 1954).Although some of the details of our account may have to be revised, wehave corrobo rated o ur three major predictions. (1) Emotional terms relateto an organised semantic field, and are not an incoherent assemblage ofterms. (2) Their m eanings depend o n the five basic emotional m odes. (3)They divide up into coherent categories containing words denoting basicemotions, emotional relations, caused emo tions, causes of em otions, emo-tional goals, and complex emotions.

Manuscript received 29 September 1987Revised manuscript received 20 October 1988

REFERENCESArmstrong, S.L., Gleitman, L. R., & Gleitman, H. (1983). What some concepts might notBendix, E. H. (1966). Compo nential analysis of general voc abulary: The semantic structureChomsky, N . (1965). Aspecu of the theory of syntax. Cambridge, Mass.: MIT Press.Clore, G. L., Ortony, A., & Foss, M. A. (1987). The psychological foundations of theDavitz, J. R. (1969). The language of emotions. New York: Academic Press.Davitz, J . R. (1970). A dictionary and grammar of emotion. In M. L. Arnold, (Ed.),de Rivera, J . (1977). A srrucrural rheory ofrhe emotions. New York: International Univer-Fehr, B. & Russell, J . A. (1984). Concept of emotion viewed from a prototype perspective.Fodor. J. D. (1977). Semantics: Theories of meaning in generative grammar. New York:Frijda, N. H. (1970). Emotion and recognition of emotion. In M. L.Arnold, (Ed.), FeelingsFrijda, N. H . (1986). The emotions. Cambridge University Press.Frijda, N . H . (1987). Commenton Oatley and Johnson-Laird's "Towards a cognitive theory

be. Cognition, 13,263-308.of a ser of verbs in English, Hindi and Japanese.The Hague: Mouton.

affective lexicon. Journal of Personality and Social Psychology, 53, 751-766.

Feelings and emotions: The Loyola Symposium. New York:Academic Press.sities Press.Journal of Experimental Psychology: General. 113, 464-486.Crowell.and emorions: The Loyola Symposium. New York: Academic Press.

of the emotions". Cognition and Emotion. 1 . 51-59.


27/43

THE LANGUAGE OF EMOTIONS 107James, W. (1890). The principles of psycho logy. New York, Holt.Johnson-Laird, P. N. (1983). Mental models: Towards a cognitive science of language,Johnson-Laird, P. N. & Oatley. K. (1988). Are there only two primitive emotions? A replyKatz, J. M. (1980). Discrepancy, arousal and labelling: Towards a psychosocial theory ofKeil, F. C . (1979). Semantic and conceptual development: An ontological perspective.Cam-Mandler , G . (1962). Emotion. In R. W . Brown et al. (Eds), New directions in psychology.Mandler , G . (1984). Mind and body: Psychology of emotion and stress. New York: Norton.Mees, U. (1985). What do we mean when we speak of feelings? On he psychological texture

of words denoting emotions. Sprache & Kognition, I , 2-20.Meyer-Gross, W . , Slater, E ., & Roth, M. (1960). Clinicalpsychiatry (2nd . edn). London:Balliere, T indall & Cassell.Mdler, G. A . & Johnson-Laird, P. N . (1976). Language andp erception. Cambridge Univer-sity Press.Oatley, K. & Johnson-Laird, P. N. (1987). Towards a cognitive theory of the emotions.Cognition and Em otion, I , 29-50.Ortony, A., Clore, G. L . , & Collins, A. (1988). The Cognitive structure ofemotions. Ne wYork: Cambridge University Press.Ortony, A., Clore, G ., & Foss, M. A. (1987). The referential structure of the affectivelexicon. Cognitive Science, 11, 361-384.Plutchik, R. (1962). The emotions: Facts, theories and P new model. New York: RandomHouse .Rosch, E. 1973). On he internal structure of perceptual and semantic categories. In T. M .Moore (Ed.) , Cognitive development and the acquisition of language. New York: Aca-demic Press.

inference and conrciourncss. Cambridge University Press.to Frijda. Cognition and Emotion, 2, 89-93.emotion. Sociological Inquiry, 50 , 147-156.bridge, Mass.: Harvard University Press.New York: Holt.

Schlosberg, H. (1954). Th ree dimensions of emotions. Psychological Review, 61, 81-88.Shaver, P., Schwartz, J., OConnor, C., Kinon, D., Marsh, C., & Fischer, S . (in press).Emotions and emotion knowledge: A prototype approach. Journal of Personaliry andSocial Psyc hology.Tiller, D. K. (1988). Structure n the affective lexicon. D.Phil. Thesis. University of Oxford.Wierzbicka, A. (1972). Semantic primitives. Frankfurt: Athenaum Verlag.Wierzbicka, A. (1987). Protorypessave:On he uses and abusesof the concept prototypein current linguistics, philosophy, and psychology. Unpublished manuscript. Departmentof Linguistics. Australian National University, Canberra.

Wittgenstein, L. (1953). Philosophical investigations.Oxford: Blackwell

APPENDIX 1A Corpus of 59 0 Emotional Words and theirAnalyses in Terms of the Five Families ofEmotional Modes

The aim of this list is to establish that any word denoting an emotion can be analysedsemantically as based on one of five basic families of underlying emotions, which forconvenience we label as: happiness, sadness, anger, fear, and disgust.


28/43

108 JOHNSON-LAIRD AND OATLEYThe corpus was compiled Erom three main sources: those collected by Fehr and Russell(1984); Clorc et al. (1987);and Tiller (1988). We considered all the words that Clore et al. or

their subjects judged to contain an emotional component, and all the words in the other twocorpora. Of the resulting words 76 do not, in our opinion, denote emotions, and so we havelisted them separately in Appendix 2: None of them implicates any emotion outside of the fivemodes. The remaining 327 words are presented here together with a further 263 words notincluded in any of these corpora.In general, we have used the morphologically simplest term, e.g. happy rather thanhappiness, but, where different forms of the same root differ in meaning, we havesometimes used the more complex, e.g. exhaltation is a state of happiness whereas toexalt is to praise in order to cause happiness. Where a word has more than one meaning, wehave proposed only an analysisof its emotional meaning(s), e.g. worship can refer either toa feeling or, more often perhaps, to the forms and rituals associated with that feeling.Likewise, we have not indicated specifically that a word can refer to a personality trait if it canalso be used to refer to an emotional state or mood.Our semantic theory, which can be found in the body of this paper, distinguishes sevenmain types of emotion words, which can denote generic emotions, basic emotions, emotional

relations, caused emotions, causatives, emotional goals, or complex emotions. Each entryconsists of four components:1. An emotion word.2. Its type:generic, basic, relation, etc.3. A paraphrase of its meaning in terms of the five basic emotion modes (happiness,sadness, anger, fear, disgust). The paraphrase either directly uses a basic emotion word orelse, for convenience, a word that is analysed into one of the five modes in its own entry. Thelatter is indicated by italicising the word, and the reader is referred to that words analysis inorder to find the underlying basic emotion. For example, conceit is paraphrased as pridethat the speaker regards as unmerited, and pride in turn is paraphrased as happiness with

self as a result of a high opinion of self in relation to others.4. A code indicating which of the three corpora, if any, the word is to be found in: F = Fehrand Russells corpus; C = Clore et als corpus; and T = Tillers corpus.Finally, we have not included information about parts of speech since it can be found in anygood dictionary. Sometimes a word receives two analyses, one appropriate to it s use as onepart of speech and another appropriate to its use as another part of speech.

Words Denoting EmotionsAbandonAbashedAbhorAbominateAdmireAdoreAffectAffectionAffinityAfraidAfraid ofAffront

Generic: uncontrolled emotionComplex: athumed.Relation: to hate.Relation: to hare intensely.Relation: to take ple&urc from anothers achievements or characteristics(or o think that one ought to). (F, C)Relation: to love. (C)Generic: emotion.Relation: liking or love. (F, C)Relation: mutual liking.Caused emotion: fear for a known reason. (F, C, T)Relation: fear in relation to someone or something. (F , C, T)Causative: to offend.


29/43

THE LANGUAGE OF EMOTIONS 109AggravateAggrieveAgitateAgonyAlarmAlienateAlienatedAlleviateAmazeAmbivalentAmorousAmuseAngerAngryAngry withAnguishAnimosityAnnoyAntagonismAntipathyAnxiousApathyAplombAppeaseAppreciateApprehensionApprobationApprove ofArdourAshamed

APPal

AssuageAssuredAstonishAt-easeAt-peaceAttached toAttractAvariceAversionAwe(-struck)Bad bloodBad-temperedBeguileBelongingBewilderBewitch

Causative: to ang er. (C)Causative: to anger. (C)Causative: to cause fear. (C, T)Caused emotion: intense pain. (C )Causative: to frighten. (C )Causative: to cause to cease to like. (T)Complex: mild anxiety or depression as a result of an evaluation of self asnot in emotional relation with others. (T)Causative: to reduce pa in .Causative: to surprise. (F, C )Generic: an uncertainty about which emotion one feels. (F)Emotional goal: desiring love. (T)Causative: to entertain, perhaps by way of humour. (F , C)Basic emo tion.Causative: to cause anger in someone. (F, C , T)Basic emotion. (F, C , T)Relation: to feel anger towards som eone or something. (F, C , T)Caused emotion: intense pain. (F , C)Relation: hatred for someone that may be expressed in anger. (F , C)Causative: to anger. (F, C , T)Relation: harred, or its expression.Relation: dislike.Basic: fearful, mood. (F, C , T)Basic: mild depression or lack of response. (C)Complex: self-confident.Causative: to horrify.Causative: to calm anger by sa h f i in g a demand.Relation: to enjoy. (F, C )Caused emotion: mild fear about possible future events. (F, C)Relation: approval.Relation: to admire or respect. (C )Relation: love for someone.Complex: self-disgurt as a result of evaluation of self in relation to own andothers standards. (C , T)Causative: to relieve.Complex: confident.Causative: to surprise. (C )Basic: relaxed. (C )Basic: peaceful. (C )Relation: liking or love.Causative: to cause to desire. (F, C)Emotional goal: intense greed for money.Relation: dislike. (C)Caused emotion: asronished admiration. (C , T)Caused emotion: angry for a known reason.Basic: irritable.Causative: to charm or entertain.Complex: evaluation that self is happy in relation to others and viceversa. (F)Causative: to cause mental confusion and perhaps a n r i e f y . ( C )Causative: to charm intensely.


30/43

110 JOHNSON-LAIRD AND OATLEYBitchyBitterBleakBlessedBlissBlitheBlueBoldnessBoreBoredomBotherBraveBroken-heartedBrowned offBuck upBuoy upBurdenedCalm

Bug

CapriceCaptivateCareCare forCarefreeCarewornChafeChagrinChargeCharmCheer upCheeredCheerfulCheerlessCherishChillCholericClosenessComfortComfortableCompassionComplacentComposedCompunctionConceitConcern

Basic: im'mble or hureful. (C)Complex: suppressed anger as a result of evaluation that one has beenwronged. (F, C, T)Basic: depressed.Causative: causing depression.Caused emotion: happiness for a known reason.Basic: intense happiness. (F, T)Basic: cheerful.Basic: sad, mood. (C)Relation: couruge.Causative: to cause boredom. (F)Complex: mild deprusion as a result of feeling that one has no goals. (F )Causative: to upscr.Relation: having couruge.Caused emotion: heun-broken. (C)Caused emotion: angry or depressed for a known reason.Causative: to cheer up.Causative: to im'rurc.Causative: to increase confidence or hope.Caused emotion or mood: anxious or depressed for known reasons. (C)Generic: not in extreme state of emotion.Causative: to reduce intensity of emotion. (F. C, T )Emotional goal: sudden wish.Causative: to c h u m .Caused emotion: fear or sadness for a known reason. (F, T)Relation: to have affection for. (F , T)Basic: cheerful. (C)Basic: fearful or sad, mood.Causative: to irritute.Caused emotion: sadness or anger for a known reason.Caused emotion: excitement.Causative: to pleuse or to unrucr. (C)Causative: to cause happiness in someone previously sad.Caused emotion: to experience happiness, where previously sad, for aknown reason. (C)Basic, happy, mood. (F. C, T)Basic: sad, mood. (C)Relation: to have uffecrion for. [to look after]Causative: to frighten intensely.Basic: angry, mood.Complex: evaluation of oneself as feeling mutual happiness and empathyin relation to someone. (F)Causative: to reduce puin. (C)Complex: belonging. (C)Complex:piry. (F, C)Complex: happiness from evaluation of one's current state, and, fromspeaker's point of view, ignoring dangers or difficulties. (F)Basic: culm and unworried.Complex: guilt that inhibits action.Complex: pride that the speaker regards as unmerited.Caused emotion: fear for a known reason. (F, C)


31/43

THE LANGUAGE OF EMOTIONS 111Concern forConfidentConscience-strickenConsoleConsternationContemptContentContentmentContriteConvivialCovetCourageCowardiceCrabbyCrankyCraveCravenCrossCrotchetyCrushCuriosityDampenDanderDashDauntDefeatedDeflateDegradeDejectionDelectationDelightDemoraliseDepressDepressionDesireDesolateDespairDespiseDespondentDetermined

Complex: anxiety or sympathy for someo ne else. (F, C)Complex: a mild happiness as a result of evaluating that one can cope witha situation. (F )Complex: guilt. (T)Causative: to reduce someone's sorrow by expressing sympathy. (C )Caused emotion: anxiety for a known reason.Relation: hatred for a known reason. (F , C , T)Causative: to s a t i r f . (F. C)Caused emotion: happiness for a known reason, not desiring more. (F, C )Complex: to feel or to express regret about one's actions. (C)Caused emotion: happiness caused by the company of others.Emotional goal: to want something that belongs to someone else.Relation: control, or lack, of fear in relation to danger. (C) [a lack of anemotion]Rela tion: inability to control fear, or actions motivated by it, in relation todanger. (C)Basic: irritable. (C )Basic: irritable.Emotional goal: to want.Basic: intensely fearful.Caused emotion: angry for a known reason. (T)Basic: irritable.Relation: intense im mature desire or love.Emotional goal: desire to know.Causative: to reduce happiness or enthusiarm.Basic: anger.Complex: self-confident.Causative: to frighten.Complex: depression from an evaluation of oneself as unable to cope. (F)Causative: to cause to feel less happy or less confident. (C )Causative: to humiliate.Caused emotion: depression for a known reason (F, C , T)Caused emotion: pleasure.Caused emotion: happiness for a known reasonCausative: to cause to feel happy. (F, C)Causative: to cause to have less courage or enthusiasm. an d to feelapprehension. (T)Causative: to cause depression. (C , T)Basic: sadness and lack of responsiveness, or psychopathological stateincluding sadness (C, T)Emotional goal: to have a goal, which may be sexual, an d which if attainedcauses happiness. (F, C , T)Caused emotion: intense sadness for a known reason.Causative: to cause intense sadness.Complex: intense sadness and lack of hope as a result of inability toachieve goals. (F, C , T)Relation: to hare. (C )Basic: depressed. (C )Emotional goal: having a desire with no intention of allowing oneself to beprevented from achieving it . (C )


32/43

I 12 JOHNSON-LAIRDAND OATLEYDetestDevoted toDisaffectedDisappointDisappointmentDisapprobationDisapprove ofDiscomfitDiscomfortDiscomposureDiscontentDiscourageDisdainDisenchantDisfavourDisgraceDisgruntledDisgustDisheartenDisillusionDislikeDismayDispiritedDispleaseDisquietDissatisfiedDistaste forDistraughtDistressDisturbDivertDoldrumsDolefulDolourDote onDownDowncastDownheartedDreadDrearyDudgeonDullEagerEaseEcstaticEdgy

Relation: to h t e . (C)Relation: to love. (F, , T)Complex: alienated as a result of dissatisfaction.Causative: to sadden someone by failing to do something that they wanted(or doing something that they did not want). (F, C, T)Emotional goal: sadness caused by failure to achieve goal. (F, C, T)Relation: disapproval of.Relation: not to approve of. (C)Causative: to discomforr.Causative: to cause sadness or embarrussment.Caused emotion: mild anxiety for a known reason.Emotional goal: mild frustration. (C)Causative: to cause to lose hope or courage. (C) [to try to persuade not todo something]Relation: to lack respect for.Causative: to cause to lose desire or happiness. (C)Relation: disapprove of.Complex: shame.Causative: to shame. (C)Caused emotion: irritation for a known reason.Basic emotion.Causative: to cause disgust in someone. (F , C, T)Causative: to &courage. (C)Causative: to dumpen by revealing the truth. (C)Relation: not to l ike, or to hare. (F, C)Causative: to discourage. (F, C)Caused emotion: depressed. (T)Causative: to anger. (C)Causative: to cause anxiety.Emotional goal: frusrrared. (C, T)Relation: dislike.Caused emotion: intense grief, or anxiety for a known reason. (T)Caused emotion: to feel sadness or fear for a known reason.Causative: to cause someone to feel sadness or fear. (F , C, T)Causative: to upset. (F , C)Causative: to pleuse by distracting from sources of sadness or anxiery.Basic: sad, mood.Basic: sad, mood.Basic: intense sadness.Relation: to love.Basic: sad. [also "down in the dumps"]Basic: sad.Basic: sad. (C)Relation: intense fear of someone or something (F , C)Causative: causing boredom or depression.Caused emotion: anger for a known reason.Causative: causing boredom.Emotional goal: strongly desiring to do something (C)Causative: to make less anxious.Basic: intensely happy. (F. C, T)Basic: anxious. (F)


33/43

THE LANGUAGE OF EMOTIONS 113ElationEmbarrassmeEmboldenEmotionalEmpathyEnamourEnchantEncourageEndearEngagingEnjoyEnjoymentEnlivenEnmity forEnnuiEnrageEnrapturedEntertainEnthrallEnthuseEnthusiasmEnticeEntranceEnvyEquanimityEsteemEstrangeEuphoricExaltationExasperateExciteExcitementExecrateExhilarateExuberantExultantFancyFascinateFavourFearFed upFeel forFeelingsFelicityCE 3/2+

Basic: intense happiness. (F, C , T)Complex: mild fear or shame as a result of evaluating self in relation toothers. (F, C , T)Causative: to cause to feel courage.Generic: feeling or causing emotions. (T)Complex: sharing and understanding the same emotion as someone else asa result of imagining oneself in their situation. (F, C)Causative: to cause love or desire for.Causative: to charm intensely.Causative: to increase courage or hope. (C )Causative: to cause liking or love.Causative: causing pleasure or attraction.Relation: to take pleasure in an activity (F, C)Caused emotion: pleusure for a known reason, as a result of an activity.Causative: to make happier.Relation: feeling or expressing hatred. (F, C)Complex: boredom.Causative: to anger intensely.Caused emotion: intense happiness or attraction for a known reason. (T)Causative: to please someone by an activity.Causative: to arrracr or cause intense pleasure. (T)Causative: to cause to feel enthusiasm. (F, C , T)Emotional goal: strong desire to do things. (F, C , T)Causative: to amacr, or to get someone to desire to do something.Causative: to anracr or to make intensely happy.Complex: hatred of someone because one desires some of their propertiesor possessions. (F, C , T)Generic: calm in a difficult situation.Relation: to admire.Causative: to cause people to cease to like one another .Basic: intensely happy. (F, C , T)Caused emotion: intense happiness for a known reason.Causative: to anger or frustrate. (C )Causative: to cause an emotion or excitement. (F, C, T)Caused em otion: intense happiness in anticipation or experience of events.(F, , T)Relation: to feel or express hatred.Causative: to cause intense happiness. (F)Basic: happy, mood. (F, T)Caused emotion: feel or express intense happiness for a known reason. (T)Emotional goal: to desire.Causative: to arrracr.Complex: to prefer.Basic emo tion.Relation: to feel fear of someone or something. (F, C , T)Basic: sad, mood.Relation: anger in relation to object, or person. (C. T)Complex: to have symparhy for.Generic: emotions or bodily sensations. (F )Basic: happiness.

nt

(F. )


34/43


35/43

THE LANGUAGE OF EMOTIONS 115Heart-strickenHeatedHeavy-heartedHelplessnessHighHilarityHold dearHomesickHopelessnessHope

HorrifyHorrorHostileHubrisHuffHuffyHumble

HumiliateHumilityHurtHystericalIdoliseIgnominyIll-at-easeIlkhumouredIll-temperedIll-willImpassionedImpassiveImpatientImpulseIn loveIncenseInclinationInconsolableIndifferentIndignantInfatuateInflameInfuriateInjureInsecureInsouciance

Caused emotion: grief. (C)Caused emotion: angry for a known reason.Basic: sad. (C)Complex: depression from evaluation that one is unable to cope withevents. (F).Basic: happy, mood. (F, C)Caused emotion: mirth.Relation: to be anached to.Emotional goal: longing for home. (C, T)Complex: oprimirm in relation to ones goals. (F , C, T)Complex: sadness from evaluation that events in relation to ones goalswill not occur. (F, C, T)Causative: to cause horror. (C)Caused emotion: intense fear or disgust for a known reason. (C)Relation: feeling enmiry for. (F, C, T)Complex: pride which the speaker regards as unmerited.Caused emotion: brief anger for a known reason.Basic: irritable.Complex: lacking pride as a result of having a low opinion of oneself inrelation to others.Causative: to cause to feel humble. (C)Causative: to cause to feel shame. (C, T)Complex: lack of pride from a low opinion of oneself in relation toothers. (F).Causative: to cause pain. (F, C)Generic: intense and uncontrollable emotion. (T)Relation: to love as a result of evaluating others achievementsor charac-teristics.Complex: shame.Basic: anxious, mood. ( C )Basic: angry, mood.Basic: angry, mood.Relation: hatred.Caused emotion: feeling or expressing intense excitement, anger, orhatred.Generic: without emotion.Emotional goal: irrirable desire to do something. (C)Emotional goal: sudden wish.Relation: love. (C. T)Causative: to anger intensely. (C)Emotional goal: wish.Caused emotion: having intense sadness for a known reason and thatcannot be consoled. (T)Relation: not caring for.Caused emotion: angry for a known reason. (C, T)Causative: to attract intensely. (C, T )Causative: to cause intense anger, desire, or hatred.Causative: to make furious.Causative: to hurt.Basic: anxious, mood. (F, C, T)Basic: happy, mood.


36/43

116 JOHNSON-LAIRD AND OATLEYInspireIntimacyIntimidateIrascibleIrateIrkImtableIrritateJealousyJitteryJocundJollyJovialJoylessJubilantKeenKeen onKickLanguorLe cherousLet downLibidinousLightenLight-heartedLike

JOY

LividLoatheLonelyLonging forLovableLoveLovesickLowLustMadMaddenMaliceMawkishMeeknessMelancholicMiserableMiffedMirth

M e V

Causative: to enthuse. (T)Complex: evaluation that self and other feel mutual empathy in relation toone another. (C, T )causative: to fnghren. (C, T)Basic: intensely angry, mood.Basic: angry. (C)Causative: to i r r ime. (C)Basic: mild anger, mood. (F, C, T)Causative: to cause mild anger. (F. C, T )Complex: h rre d for someone who is evaluated as supplanting oneself inrelation to an urtached person. (F, C, T )Basic: anxious. (C)Basic: happy, mood.Basic: happy, mood.Basic: happy, mood.Basic: intensely happy. (F, C, T)Basic: sad. F, C, T)Caused emotion: intense happiness for a known reason. (F, C)Emotional goal: strongly desiring to do things.Relation: to like.Caused emotion: excitemenr.Basic: relaxed mild happiness, mood.Emotional goal: feeling or expressing lust.Causative: to disappoint.Emotional goal: feeling or expressing lust.Causative: to make less sad or worried.Basic: happy, mood. (C)Relation: to feel happiness in relation to someone or something.(F,C)Caused emotion: intensely angry for a known reason. (C)Relation: to hate intensely. (C)Complex: sadness from evaluation of self as not in emotional relation withothers. (F, C, T)Emotional goal: feeling sad as a result of unfulfilled desire for someone orsomething. (F, C, T)Causative: causing love. (F , C, T)Relation: to experience intense happiness in relation to object, or person,who may also be object of sexual desire. (F, C, T)Emotional goal: state of longingfor anuched person,with possible adverseeffect on health. (C. T)Basic: sad, mood. (C)Emotional goal: intense desire for sex. (F , C, T)Caused emotion: angry for a known reason. (F , C)Causative: to anger. (F, C)Emotional goal: desire to harm someone. (F. C)Caused emotion: intensely sentimental.Relation: lack of anger or aggression in situations likely to cause them.Basic: sad, mood. (F, C)Basic: happy, mood. (C, T)Basic: sad, mood. (F, C, T)Caused emotion: mild anger for some reason.Caused emotion: happiness caused by humour.


37/43

THE LANGUAGE OF EMOTION S 117MissMollifyMoodyMopeMortlfyMournMournfulMoveNarkNauseaNauseateNeedNeedleNervousNettleNostalgiaNuisanceObnoxiousOdiumOffendOnedgeOppressOpprobriumOptimismOutrageOverconfidentOverjoyedOverwhelmPacifyPain

PanicPanickyPartial toPassionPassionatePatiencePatrioticPeacefulPeevePenitentPensivePerk upPerplex

Relation: to feel sadness as a result of separation from ottuched person orCausative: to make less angry.Generic: sad or irritable moods. (F)Basic: to be in a sad mood.Causative: to c a w intense s h a m . (C)Relation: to feel or to express grief. (F, C)Basic: sad, mood. (F, C)Causative: to caw to feel an emotion. (C)Causative: to irritute.Basic: disgust. (C)Causative: to disgust. (C)Emotional goal: to have a goal which if attained causes happiness (ormakes good deficiency). (F)Causative: to irritute.Basic: anxious. (F , C, T )Causative: to irritate.Complex: to feel mildly sad as a result of remembering one's happiness inpast situation. (C, T)Causative: c a w of im'rution.Causative; causing disgust or hatred.Relation: hancd .Causative: to anger or disgust. (C)Basic: anxious.Causative: to depress or worry.Complex: shame.Complex: happiness from positive evaluation of events in relation to one'sCausative: to offend intensely. (C, T)Complex: confident to a degree judged to be excessive by the speaker. (C)Caused emotion: intense happiness for a known reason. (C, T)Causative: to cause an intense and uncontrollable emotion. (C. T)Causative: to make peaceful.Caused emotion: sadness or fear for a known reason. [also bodily sen-sation]Causative: to cause sadness or fear. (F, C)Caused emotion: intense un

Language of Emotions

Documents

Transcript of Language of Emotions