emotion and language

7/27/2019 Emotion and Language

http://slidepdf.com/reader/full/emotion-and-language 1/6

EmotionndLanguage

EMOTIONNDLANGUAGE

A vastdomain ofresearch on emotion and languagecuts actoss

many disciplines,methodologies,and theoretical frameworks.

To render his topic coherentand manageable,we focus on the

curtent resurSence f researchon emotionalwords Emotional

words (e.g.,floLuer, shit) contrast with connotatively neutral

words (e.g., toaster,being) and include subcategories uch as

taboo words (insults, scatological eferences,and swea ng or

curse words), threatening words [e.8.,negativevalence wordsreferring to menacing situations such as murder and abuse),

and some EMoTroN woRDs (e.g., terror, dislust). In a con

tinuum of vocal emotional expression anging from nonverbal

[e.g.,screams) o abstractverbal (e.g., igurative anguage;see

IDIOMS,RONY,METAPHOR,ERBAL UMOR), . B. AY 2003)

argues, aboo words constitute he strongest orm of emotional

language:Taboo words are more arousing han figurative an

guageandyield reliable and robust emotional effectsmore ofien

than do threateningwords.

We review research on emotional words from hisiorical,

methodoiogical,and theoreticalperspectives.

HistoricalerspectivesHistoical perspectivesllustrate he multiple domainsand meth-

odologies of research on emotional words. In the mid 1800s,

neuropsychologicalcase studies of HughliDgsIackson [1958)

and others helped shape cu[ent ideas concerning automat]c

or uncontrollableproduction ofemotional words (see,e g.,Van

Lancker 1987).Carl Jung's [1910) work with emotional words

in free association asksalso shaped p(ocedures for diagnos'

ing clinical disorderssuch as schizophrenia seealso PsYcHo-

ANALYSTS ND LANGUAGE). rom 1950 o 1975,experimental

psychologistsused classicalconditioning concepts to analyze

the learning of emotionalwords (e.g.,Staats 968) and adopted

perceptual defenseparadi8ms to determine whether ego pro_

tectiveprocesses hield threatening stimuli (taboowords) from

awareness e.g.,Dixon 197r). Hou'ever, both lines of research

were argelyabandoned:perceptualdefense ecause fmethod-

ological laws and the Iearningofemotionalwords becaus€ om

putermetaphorsdominatedthestudyofanguageandcognit ion

and downplayed emotion during the Period of 1975 to 1990

(lay 20031.

MethodologicalerspectivesRATING TUDIES. ating studies prcvide a method for deter-

mining the emotional qualities of words. A classicexample s

the semantic differential (Osgood,Suci,and Tanenbaum 1957),

where atingsolwords on bipolar connotativescales eflect hree

underlying dimensions: eualuation (the valence compoDent,

e-g-,negative posit\ve); activity (e.g., fast slow); and potency

[e.g.,strong weak). L. H. wurm and D. A. Vakoch(1996)argued

thatevolutionaryconsidemtionsndrelationsbetweenprocess

ingtime data and the evaluation,activity,and potencyGtingsfor

words ndicale an affectjve exrcon for avoiding hreatsJhat dif

fers rom the general cicon (for obtainingvaluableresources).

Other rating studies involving thc affective lexicon include

Bellezza, Greenwald, and Banaji (1986), Bradley and t,ang

(1999),and tay (1992).Unrepresentedn current lating studies

278

aregender,age,psychological istory personality actors, ocial

context, political and religious affiliation, and cullural facto$

(seecuLTUREAND LANGUAGE), ll of which powerfully nflu-

encepeople'sperceptionofemotion linked words(Jay2000).

SELF-REP0RTNDFIELD TUDIES.ield studiesoftaboo word use

indicate that emotional langua8e s learned early and persists

well into old age lay 2000).Self-report tudiessuggesthat pun-

ishment for cursing ails to alter the actllal likelihood of swear-

ing but nevertheless ervesa function because he samepeqple

admit that theywould punish their own children for cursing Jay,

King, and Duncan2006).

NEUR0PSYCH0t0GICALTUDIES.europsycholo8ical tudies

have focusedon lwo primary dimensionsofemotions: arousal

(excitement)and valence [positive-negative).A primary neu

ropsychologicalmeasureof arousaland unconscious uto-

nomic act iv i ty s the skin conductance esponse SCR; ee ,

e.g.,LaBarand Phelps1998). or emotionalwordspresented

to bilinguals, the SCRdecreasesas a function of the order in

which a language is learned (Harris, Aycicegi, and Gleason

2003).he SCRalsovarieswith the estimated motional orce

of aversive words (Dewaele 2004) and occurs even whenpresentation imes are too briefforword identification (Silvert

et al.2004).

AMYGDALA ctivity also ndexesarousal:Threateningwords

t gger increased amygdalaractivation (Isenberget al. t999),

and amygdalardamage mpairs recognition of arousal but not

valence characteristicsof emotional words (Adolphs, Russell,

and Tranel 1999; ee also Lewisct al. 2007 or the role of other

subcortical slructures n arousal). Somecortical and subcorti_

cal areas espondonly to valence, ome espond only to arousal,

and some respond to an interaction of valence and arousal,

particularly when valence is negative (Lewis et al in press)

Finally, some cortical areasrespond to valence pcr se, while

othersrespond selectively o eilhcrpositiveor negativevalence

(Maddock,Garrett,and Buonocore2003).

Relativcactivityin dre LEFTHEMISPHERELH) versusRIGHT

HEMIsPTTERERH) also indexes emoiional processing,albeil

less consistentlyacross studies, and the nature and scope of

emotionJinkcd processing n the RH is an onSoing ssue (see

Borod,Bloom, and Ha)wood I998).RI_I rain damage s associ

ated with enotional blunting (Gainotti 1972)and difficulties n

identiryingemotional \^,ordsor the emotion they represent, n

matchingwords and emotions, n interpreting emotional con_

tent, in describillg emotional autobio8raphical information,

in self expressionwith emotional words (Borod, Bloom, and

Ha).t'ood 1998),and in comprehendingand expressing umor

(Blake2003).The coRpus cALLosuNr hat links the RH and LH

also plays a role in comprehending emotion linked prosody,

humor,and igurativeusagesBrown t a1.2005;aul t a1.2003)

The FRoNTALLoBE seems o regulateor inhibit socially nap-

prop ate uses of ernotionalwords, lvith links between frontal

lobe damage and verbal aggression, uch as excessive ufsing

[e.9.,Grafman t al. 1996.).

CLINICALND NDIVIDUALIFFERENCETUDIES"l ienl-pat ient

intemctions ocuson emotions,and an inability to express ne s

i"c,^ : TLvLLary"<l-72 gChF J.o. t . t ,

('n ubio(ge ubio(q€ Ercqchloe/)a

{5)etties,*( , ?ah''h & -{ c.L.r

lLtt

cuPlk oga..



Emotion nd Language

emotions in words may reflect a serious psychiatric problem

known as alexith)'rnia.Alexith)'rnic ndividuals have few words

for describing their feelings and communicatin8 emotional

distress,are unable to identiq/ and descibe subjective states,

and have difficulty interacting with others, ncluding therapists

(Taylor,Bagby,and Parker 1997).

Clinicalstudieshave developed trategies orfacilitating her-

apeuticcommunication and emotional interactions n Seneral,for e$mple, use of metaphor (seeStine 2005).Clinical studies

havealsodevelopednew ways ofusing emotion-linked words o

diagnosepsychopathology.An example s the emotionalStloop

taskwhereclients name the font color ofwords while attempting

to ignore their meaning: Longercolor naming times for specific

word classese.9.,aeb, spidcr)arc associated ith clinical prob-

temssuch asphobias(e.9.,arachnophobia),anxietyand depressivedisorders,alexjthFnia, eating disorders,drug abuse,and a

rangeof other psychopathologies seeWilliams, Mathews, and

Macleod 1996 or a review).

EXPERIMENTALTUDIES.ecentexperimentalstudieshavemade

extensive xplorationsofthe effectsof emotional words on co8-

n;tive processes uch as memory and attention. For example,

in a variant of the emotional Stroop task known as the taboo

Stroop MacKayet al.2004),peoplenamethe font color of taboo

and neutralwords (equated or length, amiliarity, and category

coherence)while ignoring the meaning olthe words and their

screen ocation,Theythen receivea surprisememorytest for the

words, he font color ofparticular words, or the screen ocationofparticular words, and the results ndicate better memory for

taboo han neutral words and better memory for the font colors

and screen ocations of taboo than of neutral words [see,e,g.,

MacKay et al. 2004;MacKay and Aimetzanov 2005J. heseand

other results suggesl hat taboo words facilitate recall of con

textual d€tails n the sameway as do "flashbulb memories" for

traumatic events such as the September l, 2001 - tragedies,

after which people vividly recall contextual details associated

with the emotionlinked event, for example, how and when

rhey irst earned l theerent,where heyuere.what he ywere

doing,andwho elsewas present(seeMacKay and Ahmetzanov

2005).

Other results indicate that taboo words impair immediate

recallof pdor andsubsequentneutral words n rapidlypresented

mixed ists containing taboo and neutral words (e.g.,MacKat

Hadley,and Schwartz2005),without impairing recall of neigh-

boringwords in pure, all taboo lists(Hadleyand MacKay2006).

However, exical decision imes [the time to identifJra stimulus

asa word) do not differ for tabooversusneutral words (MacKay

et al. 2004).We discuss heoreticalperspectives n this pattern

ofresultsne)d.

TheoreticalerspectivesCurrent esearchon emotional words illustratesa gamut ofthe-

oreticalperspectiveshat differ in their scope and goais and inthe narure and specificity of the predictions they make. Iay's

(2000)neuro psychosocial heory of cursing summa zes'likeli-hood estimatesofvarious forms of cursing,based on neurological (e.g.,conscious state, brain damage),psychological e.g.,personality,age,history), and socialcontext(e.g.,culture, class)

factor's.American males are more likely to curse han females

both as children and as adults, although women also learn a

range of taboo words, whether they use them or not. Similarly,

Americanswith high sexualarLxiety ut no religious rainirg are

less ikely to use sex-linkedcurcewords than profanity or blas-

phemy,especially n conversations ith same-sex thers see ay

1992,2000,2003J.

W. Bucci 's (1997) mult iple code theory (MCT) of emo-

tional informationprocessing inks Freudian and coNNE c-

TIoNrsT concepts ia the conceptofreferentialact iv i ty RA).

RA is an index of the ability to link primary [e.9., emotional,

unconscious) nd secondary e.9.,verbal,conscious) evels

ofplocessingwithin a connectionis t etll 'ork.Applied n th e

domain of clinical psychology, MCT has provided explana-

tions lor negative psychologicalstates,such as repression, nterms of the nature or quality of connections between these

fundamentally inguist icversusemotional evelsof piocess-

ing. Under MCT, people with high versus owRA differ in their

abilityto express nd describe heir emotlons, n the structure

and organization oftheir narratives,and in their therapeutic

successales,

Resource theories of emotion and attention (e.g., Wells

and Matthews 1994)peihaps provide the bioadest conceptu-

alization of emotion and cogiitive processes.Under resource

theories, hreatening stimuli attract limited-capacity cognitive

resources, hereby reducing iesourcesavailable or processing

and responding o other stimuli, for example, ont color n clini-

cal, emotional, and taboo Stroop asks.This hlpothesis readilydescribesphenomena such as the taboo Stroop effect(longer

times for naming the font color oftaboo than of neutral words)

but cannot desc be other phenomena, for example, superior

memory for the font color and screen ocation of taboo than of

neutral words (seeMacKay et aI.2004).

Two exceptions o the descriptiveor post hoc approach hat

characterizes esource heories are noteworthy, One i s arousal

theory (e.9.,LeDou-x1996)as applied to emotional words (e.9.,

Kensingerand Corkin 2003).Under arousal theory, low-level

sensory aspects of emotional stimuli, such as taboo words,

directly engagean emotional reactionsystem in the amygdala)

independently of other stimulus factors, such as context and

presentation ate. The emotional reaction system hen t ggers

enhanced skin conductanceand facilitatesmemory consolida-

tion for the emotional stimuli and their contextofoccufience [in

th eETPPoCAMPUS).

Whatmakesarousal heoryattractives ts genemlityand test-

ability.Forexample,arousal heory explains lashbulb memories

under the hJpothesis hat arousaltends o induce storageofper-

ceptual images hat include both the emotional stimulus and

its contexl ofoccurence. However,arousal heory as applied o

emotional words has not fared well in recent tests:Contrary o

arcusal heory, fpresented in mixed taboo neutral ists at rela

tively slow ates e.9.,2,000ms/word) orifpresented inpure (all-

taboo orall-neutral) listsatrapidrates(e.g., 00ms/word), taboo

words are no better recalled than neutral words equated for

familiarity, ength,and category oherence Hadleyand MacKay

2006).Also contrary o arousal heory, recent data ndicate that

taboo words do not trigger imagelike memories (MacKay and

Ahmetzanov2005).

279



Emotion and Language

The second notable exception to the summary description

approach is node structure binding theory, or binding theory

for short (e.g.,Hadley, and MacKay 2006). Under binding the-

ory, emotion-linked stimuli, such as taboo words, engage the

emotional reaction system, which delays activation of binding

mechanisms (located in the hippocarrpus) for linking concur

rent neutral stimuli to their context of occurrence As a result,

(less mportant) neutral stimulionly form links to their context of

occurrence after links to context for (more important) emotion-

linked stimuli have been formed.

These binding rheory assumptions have generated counter-

intuitive predictions that subsequent experimental tests have

verified. For example, unlike other theories, binding theory

correctly predicted impaired recall ofneutral neiShbors before

and afterataboo word ifand only if mixed (taboo neutral)word

lists are presented raptdly (Hadley and MacKay 2006). Binding

theory also correctly predicted no difference in recall oi taboo

versusneutralwords n pure (all-tabooor all-neutral) istspre-

sented rapidly or slowly (Hadley and MacKay 2006). Unlikc

other theodes, binding theory also correctly pledicted no dif

ference in lexical dccision times (the time to identitT a stimulus

as a word) for taboo versusneutralwords (MacKay et aI..2004).

ConclusionBoth historical and contempomry researcll on emotional

words reflects a wide variety of theoretical and methodological

approaches in fields ranging ftom neutoscience to psycholin-

guistics to co8nitive and clinical psychology. Fulther research is

required to piece togetherthese multiple domains and to develop

a general understanding ofemotional words and their relation to

other cognitive processes. However, emotional words curlently

seem poised to resume their central position in the language sci-

ences and related disciplines.

Kristin L. Ianschewitz and Donald G MacKay

WORKS ITED NOSUGGESTIONSORFURTHER EADING

Adolphs, R., J. A. Russell, and D. Tranel. 1999 "A role for the human

amygdala in recognizing emotional aro\rsz.1." sychological Science

r0:167-71.

B€llezza,F. S., A. G. Grecnwald, and M R. Banaji 1986 "Words high and

low in pleasantness as rated by male and femalc college students."

Behavior Research Methods, Instrumenls, and Computers 18:299 t-103

Blake, M. I-.2003. "Affective language and hunor appreciation after

right hemisphere brain danage." Seminars in Speechancl I'anguage

24i107 19 .

Borod, I. C-, R. L. Bloon, and C. S. HalM'ood 1998. "Verbal aspcctsin emotional communication." In RiSht llentplErc Language

Comprehensiotl: Perspecti)reson Cagnitiue Neuroscience' ed. M.

Beeman and C. Chiarello,285 307.Malr'!'ah, Nl: Erlbaum-

Bradley, M. M., and P- I. Lang. )999. Afectiue Nornxs or English luor^

(ANEW): Instruction Manual and Afecnue Ratngs Technical Report

C-r, The Center orResearch n Psychophysiology,University of Florida

Drou,n, W. S., L. K. Paul, M- Slmington, and R. Dietrich. 2005

"Comprehension of humor in primary agenesis of the corpus callo-

sum." Neuropsychologid43: 906 16.

Bucci, w. 1997. Symptoms and svmbols:A mulliple code theoryofsonr

atization." Psyclroana\tir Inq i! 17:151-72.

280

De}aele, J. 2004. "The emotional forc€ ofswearwords and taboo words

in the speech of multilirl9u.als."Journal of MultilinguaL and CuLtural

Deuelopment 25: 2O4 22.

Dixon, N. F. lqll. Subliminal perception: The Nature of a Controue$y.

l.ondon: Mccraw Hill.

Gainotti, G. l9?2. "Emotional behavior and the hemispheric side of the

lesion. Cor'aex8: l-55.

Grafman, t-, K- Schwab, D. Warden, A Pridgen, Il. R Brown, and

A. M. Salazar.1996. "Frontal lobe iniuries, violence, and aggression:A

report ofthe Vietnam Head Injury Study." Nelr"o og146,1237-8-

Hadley, C.8., and D. G. MacKay 2006. "Does emotio n help or hinder

immediate memory? Arousal versus priority_binding mechanisms "

lournal of ExperimentaL Psycholog: Learnin& Memory, and Cognition

32:79 88.

Harris, C. L., A. Aycicegi, and l. B. Gleason.2003 "Taboo words and rep_

rimands elicitgreater autonomic reactifity in afirstlanguage than in a

second anguage."Applied PsychoLinguistics 4: 561 79-

lsenberg, N., D. Silbersweig,A- Engelien, S. Emmerich, K- Malavade, B.

Bcattie, A. C. Leon, and E- Sterr. 1999. Linguistic threat activates he

human amygdala." ProceedingsoI the National Ande'ny of Science

96: 104569.

lackson, H. 1958. SeLectedWritings of Hughlings Jactson Vol 2. New

York BasicBooks.

Iay, T. B. 1992.Curstng n,4mertcd.Philadelphia: Iohn Benjamins.

-.2ooo. Why We C&rse.Philadelphia: Iohn Benjamins

-.2003. psychoLag]of LatlSuage.Uppet Saddle Rjver, NI: Prentice

Hal l .

Iay, T. B-, K. King, and T. Duncan. 2006. "Memories of punishment lbr

cursing." S€Jc oles32: 123-33.

Jung, C. G. 1910. "The association method,." American lournal of

PsycholoEt 31 219-69.

Kensinger, E. A., and S. Corkin. 2003. "Memory enhancemcnt for emo-

tional words: Are emotional words more vividly remembered than

neutral words?" Memory and Cognition 3t: I169-8O

LaBar, K., and E. Phelps. 1998. "Arousal mediated memory consolida

tion: Role of the medial temporal lobe in humans " psychaloSical

Science : 4SO-3.LeDorx, J. 1996. The emotional brain: The mystetiaus underpin lings of

emotional IiJe.New vork Simon and Schuster'

Lewis, P. 4., H. D. Cdtchley, P. Rotslrt€in,and R. I. Dolan.2007. "Neural

corelates of processing valence aDd arousal in affective words.

Cerebral Cortex17.742 L

MacKay, D. G., and M. V. Ahmetzanov. 2005. "Emotion, memory, and

anention in the taboo Stroop paradiSm: An experimental analoS of

flashbulb memories." psychologicalScience6t 25 32

MacKay, D. G., C. B. Hadley, and I. H. Schwartz.2005. Relationsbetlveen

emotion, illusoryword perception, and orthographic repetitjon blind-

ness: Tests of binding theory." QuafterLy lournal of Expetimental

Psycholog! a: r5l4-33.

MacKay, D. G., N,I.Shafto, J.K. Taylor, D. Marian, L. Abrams, and i- Dyer

2004. "Relations behveen emotion, memory and attention: Evici€ncefrom taboo Stroop, lexical decision, and immediate memory tasks.'

Memory and Cognition32:414 88.

Maddock, R. ., A. S. Garrett, and M. H. Buonocore. 2003. "Posterior cin

gulale cortex activation by emotional words: fMRI evidence from a

valenc€ decision task." Human Brain Mapping la.30-41.

Osgood, C. 8., G. t. Suci, and P. H. Tannenbaum. 1957 'lhe Measurement

of Meaning. Urbana: University oflllinois Press.

Paul, L. K., D. Van Lancker-Sidtis, B. Schieffer, R Dietricb and W S.

llrou,n- 2003. "Communicative deficits in agenesisofthe corpus callo-

sum: Nonliteral language and affective Prosody." Brain and Language

85i31: i 24 .



Emotion, peech, nd Writing

Sih'ert , . , S. De\tanque, H. Bouwalerh,C. Verpoort ,and IL sequeira.

2004."Aulonomic responding o aversivewords rv i thoul conscious

valence discriminatior\." ltxenntional loumal aJ PsycltopltJsiolosJ]

53 :135 45.

Staats,A. W. 196A.Language, Leanli,?& dnd Cogntttoll. New York Holt,

Rinehart and Winston.

Stine, J. I. 2005. "The use of metaphors in the service of the therapelrtic

alliance and therapeutic colnmunicaiion." lournal of tl)e Anericatl

Academy of psychoanalysisand Dynamic Psrrclliatry33:53t 45.

Taylor, G. J., R. M. Bagby, and I. D. A. Parker. 1997. Disotders of

AJIect ReguLation: Alexithymia in Medical and Psychiattic lllness.

Cambridge: Cambridge University Press.

\ran Lancker, D. 198?. "Nonpropositional speech: Neurolinguistic stud-

ies." ln p/ogressi,] .fie Psyclnlog/ of La,lguage, ed. A.W. Ellis, 49-11B.

London:Erlbaum.

Wells,4., and G. Nlatthews. 1994. Attetltion and Enntion: A C Linical

Perspectiue. ove, llK Lawrence Erlbaum.

Wjlliams, t- M. G., A. Mathervs, and C. Macl-€od. l996. "Ilc cmotional

Stroop task and psycho pathology." PsychologicalBulletin l2O: 3-24.

Wunr, L. FL, and D. ,q..Vakoch. 1996. "Dimensions of speech percep

tion: Semantic associations in the affective lexjcon." Cognition and

Emotiatr l l :4Og 23 .

EMOTION,PEECH,NDWRITING

In our everydaylife,we are requently exposed o expressions uch

as "a thousand words cannot expressa single emotion," "what I

feel s somcthing that is beyond words," and so on. This kincl of

utterance, eflecting the difficulty ofexpressing emotions, evokes

special nterest: Can speechand r/!'riting eally expressemotions?

In today's world, with the increasing awaleness of emotion

as part of the self and the importance of expressing emotion as

part of human communication, Ianguage becomes vital to the

understanding and analysis of emotions. Lexical choices reflect

how people experience the world around them and, thus, con

stitute mediators between individuals' emotions, which are

intenal aDd subjective, and external entities, such as society and

enlrironment,

Theorists of emotion stress that language is the most con

venient channel lor approaching research into emotions and

that EMortoN woRDS are the best way to reflect the emotional

experience.Psychologists and psychoanalysts (Freud and his

followers) recognize that in spite of the importance of nonver-

bal behavior, \,vordsare the natural way ofexte orizing the inner

emodonalworld (seepsycHoANALysIS

AND LANGUAGE).Theoristsof emotion (e.g.,Ortony, Clore, and Foss1987)also

stresshat anguageoffers he most convenient accessor research

ing emotions,and that emotion words are the bestway ofreflect

ing emotional expeliences. Linguists such as N. J.Enfield and A.

Wierzbicka(2002)went further, stating that it would be impossi-

Dle o examinepeople's emotions without putting languageat the

ceqtel,both as he object ofthe researchand as he research ool.

One reasonfor the complexity ofsuch studies, according to severalresearchers,s linguistic usage hat confusesemotion terms.Criticismofpsychological research nto emotions focusesmainlyon the fact hat most research n this field relies argely on linguistic labelsand not on direct measurement ofthe emotion itself. f

this is, n fact, the case, t is pafticularly imporiant to investigatethe anguageofemotions as a discrete ssue,with tools exterior tone emodon itself such as hose of linguistics.

In hcr cr i t ica lessay, . T. l rv ine (1990)wrote that many lin-

guists end ro get "cold leet" when it comes o conside ring orv

emotions are expressedverbally. Accordingly, if we use the ter

minology of Ferdinand de Saussure,we can say that emotion

is accepted as integral to the parole, which is linguistically lessmeaningful than langue - language n its broadest sense. hus,

examinationof emotion is pushed to the periphery of l inBuis

tics. INine also remalks that though there are languages with

pHoNoLoGrcAL and MORPHOLOGTCALnits that iDdicate

emotional states n speech, inguists ftequently tend to combine

such elements with general descriptions of grammar, rather than

emphasizing sLlchverbal expressionsof emotion. I_Iou'ever, he

notes two importaDt linguistic texts that also deal with elnotion

in ianguage, namelt Edward Sapir's exicon ot' emotions as mir-

roring culture and Roman Iakobson'swork relating to the emo

tive function of language.

C. Caffi and R. W. Ianney (1994) examined the rhetorical

strategies or expressing emotion by comparing psychologicalcategodes of emotions with linguistic categories. They define

emotional communication as directed strategies or imparting

emotional information in speech or writing, insisting that such

expressionsmust be analyzed inguistically, because anguage

spoken or \ Titten is the means for conveying emotiol. Their

model comprises linguistic markers, including specific emotion

words, obligatorywords, SYNTAxmarkers, and spoken anguage

mechanisms i.e. ,ToNESand rNToNATroN,prosody, engthof

SYLLABLES, tc.).

Calfi and fanney's \,Titings infer that there are significant

connections bet\,veen extual linguistic usage and emotions,

as evidenced also in dia es, letters, and other autobiographi-

cal $ritings. For instance, language r4/as sed to measure emo-

tion in a study by G. Collier, D. Kuiken, and M.E. Enzle (1982).

The researchersnoted that when describin8 negative emotions,

people use more complex constructions than in describing posi-

tive emotions, and this also applies to expressions of negative

as opposed to positive personal qualities. fhe positive is always

more clearly expressed.Assessingdescdptions oflevels of posi-

tivity or negativity of emotions or traits indicated that the feel

iDgsortraits described n more complex detail tended to be more

negative, that is, the more complex the descriptions, the more

likely theywere to be negative.

An earlier study by C. E. Osgood(1958)dealtwith the connec

rion betlveen EMorroN AND LANGUAGE,establishing the linkbetween the lingual characteristics of a text and the motivation

ievel of the author when writing it. The research studied suicide

l]otes, written under the influence ofvery strong emotion - "the

last letter" as compared with ordinary correspondence \\ith

f ' m i lw^r . lnep

Fr ian. lc

In her chapter "How and why is emotion communicated?"

S. Planalp (1999) writes that verbal expression of feelings is

endemic to the process of communication, even though peo-

ple do not always use words. They do not, as a rule, announce

that "I'm angry" or "l'rn feeling depressed at ihe moment," but

there are other verbal indications, like swearing or extravagant

oulbur5ts uchas l couldUllh iml

I. W. Pennebaker and M. E. Francis (1996) aialyzed per-

sonal texts describing thoughts about commencing higher stud-

ies by first-year students at college. Linguistic and cognitive

281



Emotion,Speech,and Writing

parameterswere classifiedaccording o specificverbal catego-r ies.Thi5 ncludedclassifyingmotionwordsusedbr rnesuU_iecrs in panicular, osit ive nd/or negarive \pressions),h i lethe cognitive parameters ncluded clarity, accessibiliw o thereader, nd schemaric rganizar ionl the texr. hg 66n.aa,i.rlbetlveen these linguistic and cognitive variables and mentalhealthwas then examined,aswere the academicachievementsofthe subjectsn their ongoingstudies.

A. Boalsand K. Klein (2005)examinedhow the words usedin a nanative can

conveystressor distess in regard o levelsofpain afiera negative motional venl.Theitsuby-ectser emorethan 200 studentswho had undelgone a romantic crisis or thebreakingup of a relationship. he studentswere asked o writeabout both the relationship and the effectsof the separation.The researchers ound that there was conspicuouslymore useof negativeemotion expressions, oth of ,,physical,,

words andof first personutterances,as comparedwith desc ptions of therelationshipbeforebreakingup. It was possible o pinpoint lin_guist icdiftelences etween ejecrion/repressionnd intensireinternalizationof an expe ence. he ,,rejecters,,

tended to usemore casual anguage,negativeemotion words, and the firstperson slngular,as well as pronouns when referringto others,but alsoused ewercognitivewords.Cognitiveexpressions

mplyactivelysearching or meaningand comprehensionofa stressfulevent and of depression, o that using them is characteristicofpeoplewho thoroughly,,work

tl.rrough,,such n event.Psychotherapy lso offerssources or researchingemotions

via anguage. he herapisfsdiagnoses nd f ierape-uric erh_od sare requently ased n patiens, anguage choice fwords,slips ofthe tongue, narratives n their stories, epetitionswhendescribing traumd, nd oLhcrmarlers[see, or example, ucci.200t .

D. D. Danner,D.A. Snowdon,andW. V. Friesen 2001)exam_ined autobiographiesof nuns as part of researchknown as the''Nun Study." They examinedthe link betweenwdtinss aboutpo5ir ive motional ncidenr<, xpressedn posit ive erms,an d

th e ife.panofthe f i i ters.Emo jonal ta i l i ty,measured ccord_lng to the useofemotion ll,ords,was found to havea sisnificantposir ive ffecl.

Useofnonautobiographical extsnot ptoducedunder abora_toryconditionsare derived rom Intemetchat, blogs,and e_mailsin var iousramework.Forexample, . A.Cohn,M. R.Mehl,an d.1.W' Pennebaker 2004)examinedchangesn the languageuseofAmerican citizensafter September l, Z00l..lhe researchersstudiedrandomly selectedblogsover a period of four months _two months pdor to and two months after the traumatic event.Linguistic markers to psychologicaldifferenceswere studied.Two week after he attack, he wdters had returned to baselinein regard o useof lexical expressions f emotion. pennebaker.

who collaboratedboth with Francis in researchingacademicacmevementaccording to linguistic parametersand in Cohn,swork on linguistic markers in blogs written after 9/ll, developed a tool for categorizingvadous O?es of text according osryle, a l led Le LntC {l inSuisr icnquiryan dwordcounr; ,wh-ichi< avai lable nl ine (pennebaker 0OZJ. ennebakermainrainsthat his computer program ,,collates,,

wolds from variouscategoriesand "translates,'hem, according o theh relativenumberin the text, o psychologicalmeanings.Main categoriesnclude,

28 2

for example,words relating to the self,,,social,,

words diiectedto others,words expressing ositive or negativeemotion, cog_nitive words, "long',words (more than sL( etters),and others.He believes hat constantuse of posralve xpressionsndicatesoptimism, whereasnegativeexpressions ndicate depression.Cognitive erms ("In my opinion,,, ,It

seems o me,,, ,Ithir*,,,

etc.) ndicate hatthew ter doesa ot ofconsideringand prepa_ration when writing and s thus more thoughtfuland self_aware.Constantuseoflong wordssuggestshat the \,\,riters alienated,

keepinghis/her distance.Another group of researcheN (pennebaker,Slatcher,and

Chung2005)atempred to learn from the speeches fU.S. presi_dentialcandidates bout heir personalities nd emotions.Itwasshownrhatin spireofadvice eceivedbypolificalcandidatesromtheir advisorsabout using words correctly (e.g.,using the frrstpersonplural insteadof singular), hey sometimesspeakmorefreeiy, revealing more about theh personalities.pennebakerand his colleagues mphasized he useof functionalwords thatindicate the ability to absorband organize houghtsand ideas.Using he program developed or this study, hey alsoexaminedpositive and negativeexpressions f emotion, cognitivewords,e\clusi\,es,ingular nd pluralexpressions,t|-.

From all thesediversestudies,we learn that language, hedominant aspectof iniercommunication between ndividuals,is the simplestmethod of revealinga differentsystem,one thathas ts own attdbutesand nfluencesall aspects fourlives, thatis, he emotionalsystem,One can, asa rule,consciously ontrolthe conrentofthe storyone ells,but it is more difficult to controlthe exactchoiceofeach word. When it comes o ,,what

to writeor say,"we are awareofwhatwe are doing,butthis is not alwaysthe casewith houlwedo t. Even he mostpracticedspeakerfindsit difficult to monitor all the \,vords e or she selects n order tocommunicate. hus, linguisticmarkersare, n fact, he buildingblocks hat must be usedas he foundationfor researching molion in language.

OsnatArgaman

WORrcCITEDNDSUGCESTIONSOR URTHEREADING

aoals,A., andK.Klein.2005.Worduse n emoxonal arrativesbourfailed omanticelationshipsndsubsequet ne nralheatt\.,,ournatoJLanquage ndSociat sychotagt24:252_68.

Bucci,W.2001.pathwaysf emotjonatommuni ation.,,sychoanalyticInquiry2r.r:40-70.

Caffi,C.,andR.W lanney.994..,Towa.dpl?gmaticsfemotive om_mul]icario ." ou na ofpra ma . s 22:325 7 .

Cohn,M. A.,M. R_Mehl,andJ.W' pennebaker.2004.,LinSuisticmark-

ers of psychologicalhangesunoundingSeptember 1, 200r.,,PsychologicalcienceS 68z 93.

Collier, ., D. Kuiken, ndM. E.Enzle.982.,,Ihe oleof grammatical

qualificatio n theexpressionndperceptionf emotion.,,aurnaloJPsycholinguisticesearchlt a3]l so.

Danner,D. D., D. A. Snolvdon, ' V. Friesen.00. .posiriveemotions

in early ife and ongevfty: indingsrom he Nun Srndy."ournat fpenanalityandSociat sychdlo&,/s0:804 13.

Enfield, .J., ndA_Wierzbicka.002.,Introduction:Thebodyndescrip,

tionofemotion." r"4gmaticsandCognitiono.t/2: 25.Irvine,.T.1990.Registe.ingffecrHeteroglossian the inguisticxpres

sionof emorion. In adng,ageand thepotiticsafEmotion,ed.C.A. LutzandL.Abu-Lughod,26-6 Cambridge:ambridgeniversityress_

)



I

Ernotion ords

Oftony, 4., c. 1..Clorc,and NI.A. Iioss.1987.,,.Iheef€rentjatstmcrln.eof

thc affecrivc exjcon." CognitiueScience t.34t_64.Osgood,C. L. 19s8. Some effecrsof morivatjon on sryle of encodin8.,,

In Style n t .anqlnge,e<1. . A. Sebeok,293 306. Cambridge,MA: MrT

Pennebaker .UI. 2007, .ftrelvorld ofwords.,' Avajlabteontine ar hftp://

\\'l!w.liwc.nct/

Pennebake.,J. W, and M. E. Fnncis. i996. ',Cognitive,emorjonal andlanguSe p.ocessesn di sclasrtre.'Cogtlitionand Emotion rclt601 26.

Pennebaker, . W., R. B. Statcher,and C. K. Cbung. 2005. ,,Linguistic

markerc ofpsychological srare hrough mcdia Interviews: ohn Kerryand John Ed\^,ardsn 2004,Al core in 2000.',4natrses J Social rssuesa,rclPublicPolicys: r97-zoa.

planalp, S. 1999.CoDurnnica ng Emotion: Social,Moral and CrtturalP/oc€.rses.ambridge:Camb.idgeUn versirypress.

EMOTIONORDS

What ountssan Emotion ord?Languages iffer in rhe sizeand rangeoftheir emotion vocabularies. here are, for example,more thaD500 words in Bnglish,750 n TaiwaneseChinesefRussell 1991),and 256 in Filipino(Church, Katigbak, and Reyes1996). In addirion, translationequivalentsoften cover overlappingbut not identical semantjcspace Wicrzbicka 1999).Cleafly, he investigarionof the emotion lexicon requires he carefuldelimitation of what countsasan emotionword.

Empirical approaches Io this question are driven byPRo oTypE rheory F-ehr nd Russell1984;Rosch 97B),accord

ing to which semantic categoriesare recognizednot by lists ofNECESSARYND SUFFTCIENTeaturesbut in terms of a gestaltor conligurationalwhole. 'Ihis approachsuggestshat emotionis a "fuzzy category,"and emotion words fit the category n aSradedmanner.

A number of taxonomieshave been proposed.G. L. Clore,A. Ortony, and M. A. Foss r9BZ)disringuishedeight categoriesin English: ) pure affectivesrates e.g.,happy),2) affective_behavioralstates e.g.,cheerful),3) affecrivecognitivestates e.g.,encouraged),4) cognitive srates (e.g., cerrain), 5) cognitive_behavioralstares e.g.,caurious),6) bodily stares e.g.sleepy),?) subjectiveevaluationsof character (e.g.,attractive),and B)object ive ondit ions[e.g.,abandoned). nalyses f protorrpi_

callty ratings of 585 candidaie emotion words confirmed theempirjcaldiscriminabiiity of the eight categories, nd words inthe first three ("affective")categorieshad the highest ypicalityraongs.

Phillip Shaverand colleagues 1987)usedclusteranalysisofprotot'?icaliryrarinbsofEnglishemotionwords o displayapfo-lotlpe hierarchy,with two superortlinatecategodes ncompassingpositiveversusnegative ermsand fivebasic-lerelerms: ove,joy, anger,sadness,and fear.The rest of the terms are srrror_drr?atesnder thesebasic terms (shaver et al. 1982;Storm andStorm 9B7).t isinterestingto otethatnegativeemotionwordsSeneralty utnumber positiveemotionwords,perhapsexplainedby the greatercognitiveprocessing equired by negativeeventsrn comparison

u. l thpo\irit .e venr. schraufandSanchez 004).rne Indone\ian mo{ion cxiconhas he samcoverall tructure(Shaver,Murdaya,and Fraley 2001),bur in the Chinese exicon,a "love" categorydoes not emerge separate rom happiness

relatedwords (Shaver',Wu, and Schwartz1992).Recentstudieson the Iralian lzammuner I99B)and rhe French(Niedentha]eral.2004J motion exiconssuggesthai prototl/pi al i y ratjngsaredrivenbyvaience, ntensity,duration, familiarity,ageofacquisitioD,and ftequencyin the corpus.

How reEmotion ords epresentedn heMind?PSYCHOLTNGUTSTICSistinguishes abstract and concretewordsas separate classes ofwoads, and receDt [,ork suggests that emotion words may form yet a thi.d class oIu,ords. In general, con-crete versus abstract words are easier to imagine, lnore qujcklyrecallcd, and more easily recognized.ln addition, concrete wordsare more easily associated with a context, perhaps because ofprior association with those contexts (Schwanenf]ugel, Akin, and

Luh 1992). When approached as a third class ofwords, emotionswords are rated as ess concrete and lower in context availabilitythan abstact and concrete words. Nevefiheless, they are ratedhigher in imageability than abstract words, perhaps becauseof some connection to scRIpTs or typical situations jn whichthey are experienced. Furthel, when participants give the firsthrord that comes to mind in response to concrete, ernotion, andabstract words, emotion words garner the highest nurnber ofdifferent associates (Altariba and Bauer 2004). If associates arestored together fas an associative model of memory suggests),then emotion wordswould seem to be liDked to a Iicher concep_tual base than either of the other th,o word ty?es. k is interest_ing that when Spanish English bilinguals perform rhese tasks inSpanish, ratings ofcontext availability

in Spanish are rtgfterthanin English. This raises rhe possibiliry rhar emotion words mighrbe eicoded in language specific \arays Alrarriba 2006).

- RobertW. Schrauf

WORI(S ITED NDSUGCESTTONSOR URTHEREADING

Altarfiba, t. 2006. Cognirivc app.oaches o ihe study ofemorion ladenand emotionwords n monotinguatand bilinguatm cmary.,,n BitingualMinds: Emotional hperience. Expression,and Reprcsentation, ed.A.Pavlenko,232 s6. Clevedon,UK: MultitinguatMatrers.

Altarriba, J.,and L. M. Bauer.2004_,Thedislincrivcness femotion con_

cepts:Acomparisonbetweenemotion, absrmct,and concreten ords.,,Alneican Journal oJpsycholog tt7:3a9 1IO.

Church,A. T., M. S.Katigbah t. A. S. Reyes.996.',Toward a ta{ononryoftrait adjectives n Irilipino: Comparingpersonatity exiconsacrosscui_tutes." European lournat of personntity ro: 3_24.

Clore,c. L.,A. Ortony,M. A. Foss. 9s7_ ,.nrepsychotogi€aloundationsofthe affective exicon." ournat of persanatiryand Sociatpsychotogy

53:75166 .

Fehr, B., and t. A. Russell.1984. .Conceptof emotion vier{ed from a

protoq?e perspective."/oumat oI L:tperimentalpsychotoyt:Generall l3:464-86.

Niedenthal,P. M., C. Au-,rierte, . Nugier, N. Dalle, p. Bonin, M. Fayol.2004. "A protoo?e analysjs of the French caregory ,emotion.,,,

Cognitionand Emotion r8.3:289,312.Rosch, E. 1978. "Principles of categorization., t| Cognitit)n and

Categorization,ed. E. Rosch and B. B. Lloyd, 27-48. Hiltsdale,Nl: Law.rcnce rlbaum.

Russ€ll,JamesA. 1991. ,Cultureand the caregorizationof enotioni.,,prchological BuUetin tro.3t 426 SO.

Schraul R. W., J. Sanctrez.2004. ,,Thepreponderance of negative

emotion words in the ernorion lexicon: A cross_generational nd

283

emotion and language

Documents