Linguistic Society of America

Response Cries Author(s): Erving Goffman Source: Language, Vol. 54, No. 4 (Dec., 1978), pp. 787-815 Published by: Linguistic Society of America Stable URL: http://www.jstor.org/stable/413235 . Accessed: 29/06/2011 10:17
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=lsa. . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Linguistic Society of America is collaborating with JSTOR to digitize, preserve and extend access to Language.

http://www.jstor.org

RESPONSE CRIES ERVING GOFFMAN Un vers tyof Pennsy van a Universityof Pennsylvania
U erancesarehousedno n paragraphs bu n urnsa a k occas ons mp y nga Utterancesare housed not in paragraphs,but in turnsat talk-occasions implyinga a erna ono akers hemse vesare temporarytaking of the floor, as well as an alternationof takers.Turnsthemselvesare emporary ak ng o he oor we nked n runsmarked na ura ycoup ed n o wo par y n erchangesIn erchanges naturallycoupled into two-partyinterchanges.Interchangesare linked in runs marked o off by some sort of topicality. One or more of these topical runs make up the body of sor o op ca y he o hese op ca o a conversation. This interactionistview assumes that every utterance is a statement conversa on Th s n erac on s v ew ha u erance s s a emen es ab sh ng he nex speaker s establishingthe next speaker'swords as a reply, or a reply to what the prior speaker rep y rep y o wha he pr or has just established, or a mixture of both. Utterances, then, do not stand by themno s and m x ureo bo h U erances hen hem us es ab shed heard bu cons ruc edand se ves ndeed hey o en selves-indeed, they often make no sense when so heard-but are constructed and timed to support the close social collaboration of speech turn-taking.In nature the med o suppor he c ose soc a co abora on o urn ak ng na ure he s on y ound n verba n erp ay be ng n egra ydes gned or spoken word is only found in verbal interplay, being integrally designed for such co ec vehab a s However h s collective habitats. However,this paperconsiderssome roguishutterancesthat appear cons derssome rogu shu erances ha to violate this interdependence,entering the stream of behavior at peculiar and uno v o a e h s n erdependence en er ng he s ream o behav ora pecu ar un na ura p aces produc ngcommun ca vee ec sbu naturalplaces, producingcommunicativeeffects but no dialog. The paper begins with paperbeg ns w h d a og a special class of spoken sentencesand ends with a special class of vocalizations-the sen encesand w h spec a c ass o voca za ons he spec a c ass o first failing to qualify as communication, the second failing not to.* rs a ng o qua y commun ca on he a ng no o *

1 1. To be all alone, to be a SOLITARYthe sense of being out of sight and sound a a one in n be ng s ght of everyone, is not to be alone in another way-namely as a SINGLE, party of one, a one n SINGLEa one everyone s way-name y a person not in a WITH, person unaccompanied 'socially' by others in some public n WITHa n pub c unaccompan ed soc a y undertak ng ( tse f undertaking (itself often crowded), such as sidewalk traffic, shopping in stores, and crowded) s dewa k traff c shopp ng n stores restaurantd n ng 1 restaurant dining.1 A ow ng Allowing the locution 'in our society'-and, incidentally, the use of WEas a ocut on n WEas soc ety -and nc denta y means of referring to individuals without specifying gender-it can be said that referr ng nd v dua s w thout spec fy ng gender- t sa d when we members are solitary, or at least assume we are, we can have occasion to so tary east are occas on make passing comments aloud. We kibitz our own undertakings, rehearse or relive pass ng a oud k b tz undertak ngs rehearseor re ve a run-in with someone, speak to ourselves judgmentally about our own doings run- n w th someone ourse ves judgmenta y do ngs (offer ng (offering words of encouragement or blame in an editorial voice that seems to be b ame n ed tor a vo ce that of an overseer, rather than ourselves), and verbally mark junctures in our overseer ourse ves) n verba y phys ca do ngs Speak ng aud b y physical doings. Speaking audibly, we address ourselves as the sole intended recipiourse ves so e ntended rec p ents of our own remarks. Or, speaking in our own name, we address a remark to remarks Or speak ng n name remarkto someone who isn't present to receive it. This is self-communication, specifically sn t rece ve t Th s s se f-commun cat on spec f ca y
SELF TALKA SELF-TALK. hough a conversation-like Although conversa on ke exchange of speaker-hearer o speaker hearer ro es roles may

somet mes occur th s sometimes occur, this seems unusual: either we address an absent other, or we unusua : e ther other address ourselves in the name of some standard-bearingvoice. Self-talk of one type ourse ves n standard-bear ngvo ce Se f-ta k seems rarely answered by self-talk of the other type. I might add that the voice or rare y se f-ta k type m ght vo ce name in which we address a remark to ourselves can be just what we might properly n wh ch addressa remarkto ourse vescan bejust m ght proper y
* I have incorporated,without specificacknowledgment,a large numberof suggestions,both ncorpora ed w hou spec cacknow edgmen argenumbero sugges ons bo h genera generaland particular,providedby John Carey,Lee Ann Draud, John Fought, Rochel Gelman, par cu ar prov dedby Carey Draud Fough Roche Ge man A en Gr mshaw Ga Je erson W am Labov G an Sanko Joe Sherzer Allen Grimshaw, Gail Jefferson, William Labov, Gillian Sankoff, Joel Sherzer, and W. John W Sm h Smith. I am grateful to this community of help; with it I have been able to progressfrom theft gra e u o h s commun yo he p w h ab e o rom he to pillage. Comments on broadcasters'talk are based on a study in progress. o p age Commen son broadcas ers a k s udy n progress
1 This easy contrast conceals some complications. A WITH-a party of more than one-can be Th s con ras concea s comp ca ons WITH a par y o han one can so ary oo solitary too, as when a lone couple picnics on a deserted beach. Strictly speaking, then, a SINGLE one coup e p cn cs deser ed beach S r c y speak ng hen S NGLE is a party of one present among other parties, whereas a solitary individual is a party of one s par y o presen o her par es so ary nd v dua s par y o

w h with no other parties present. o her par espresen 787

788

LANGUAGE, VOLUME 54, NUMBER 4 (1978)

use in addressing a remark to someone else, or what another might properly use in talking to us. It is not the perspective and standards that are peculiar, or the words and phrases through which they are realized, but only the fact that there are more roles than persons. To talk to oneself is to generate a full complement of two communication roles, speaker and hearer, without a full complement of roleperformers; and which of the two roles-speaker or hearer-is the one without its own real performer is not the primary issue. Self-talk could, of course, be characterized as a form of egocentricity-developmentally appropriate in childhood years, and re-appearing later only 'in certain men and women of a puerile disposition' (Piaget 1956:40). Common sense, after all, recommends that the purpose of speech is to convey thoughts to others; and a selftalker necessarily conveys them to someone who already knows them. To interrogate, inform, beseech, persuade, threaten, or command oneself is to push against oneself, or at best to get to where one already is, in either case with small chance of achieving movement. To say something to someone who can't hear it seems equally footless. Or worse, self-talk might appear to be a kind of perversion, a form of linguistic self-abuse. Solitary individuals who can be happily immersed in talking to themselves need not seek out the company of their fellows-a convenience that works to the general detriment of social life. Such home consumption in regard to the other kind of intercourse qualifies either as incest or masturbation. A more serious argument would be that self-talk is merely an out-loud version of reverie, the latter being the original form. Such a view, however, misses the sense in which daydreamingis differentfrom silent, fugue-like, well-reasoneddiscussion with oneself-let alone the point (on which Piaget 1962:7 and Vygotsky 1962:19-20 seem to agree) that the out-loud version of reverie and of constructive thought may precede the silent versions developmentally. It misses, too, the fact that both the autistic and constructive forms of 'inner speech' are considerably removed from facially animated talk in which the speaker overtly gives the appearance of being actively engrossed in a spirited exchange with invisible others, his eyes and lips alive with the proceedings. In any case, in our society at least, self-talk is not dignified as constituting an official claim upon its sender-recipient-which is true, incidentally, also of fantasy, 'wool gathering', and the like. There are no circumstances in which we can say I'm sorry, I can't come right now; I'm busy talking to myself. And anyway, hearers ordinarily do not REPLY to our self-talk, any more than to the words spoken by an actor on the stage, although they may REACTto both. Were a hearer to say, What?, that would stand as a rebuke to conduct, not a request for a rerun, much as when a teacher uses that response to squelch by-plays occurring at the back of the room; or, with a different intonation, it could mean that the self-talk had been misheard as the ordinary kind, a possibility which could itself induce a reply, such as Sorry, I was only talking to myself. Indeed, our society places a taboo on self-talk. Thus it is mainly through selfobservation and hearsay that one can find out that a considerable amount of this sort of thing goes on. Admittedly, the matter has a Lewis Carroll touch: the offense seems to be created by the very person who catches the offender since, it is the

RESPONSE CRIES

789

t nto wh ch w tness ng witnessing of the deed which transforms it into an improper one. (Solitary selfmproper one (So tary se fta kers talkers may occasionally find themselves terminating a spate of self-talk with a se f-ta k w th occas ona y f nd themse ves term nat ng se f-d rected self-directed reproach; but in doing so, they seem to be catching THEMSELVESn do ng so catch ng THEMSELVES somet mes emp oy ng se f-ta k sometimes employing self-talk to do so.) In point of fact, the misdoing is not so fact po nt so ) m sdo ng s
n pub c o much tied up with doing it in public as CONTINUING to do it in public. We are all, n pub c ed w h do ng a

it seems, allowed to be caught stopping talking to ourselves on one occasion or t seems a owed ourse ves occas on stopp ng ta k ng another another. the r m ts Str ct y speak ng Expected y Expectedly, there are questions of frames and their limits. Strictly speaking, quest ons m rror mach ne rehears ng p ay pray ng a oud d ctat ng etter dictating a letter to a machine, rehearsing a play to a mirror, and praying aloud at f our bedside are not examples of self-talk; but if others unexpectedly enter the scene beds de unexpected y examp es se f-ta k; of such solitary labor, we still feel a little uneasy and look for another type of work. tt e ook work st fee so tary abor s vu nerab e hav ng rout nes n wh ch S m ar y Similarly, there are comedy routines in which the butt is made vulnerable by having s h dden to sustain a full-blown discussion with someone who is hidden from general view. susta n fu -b own d scuss on w th genera v ew wh ch And there are well-known comic gestures by which someone caught talking to we -known com c ta k ng de ct nto yawn h mse f himself attempts to transform the delict into a yawn, or into the just-acceptable nto just-acceptab e beh nd voca zat ons wh st ng humm ng vocalizations of whistling, humming, or singing.2 But behind these risible issues of r s b e ssues s ng ng 2 fa s adu t concea h s se f-ta k frame is the serious fact that an adult who fails to attempt to conceal his self-talk, s ser ous or at least to desist quickly upon the appearance of another person, is in trouble. east des st qu ck y person s n troub e attr bute fa ure n Under the term verbal hallucination, we attribute failure in decorum here to verba ha uc nat on
'mental illness'.3 men a ness 3

se f-addressed we nto adu t fe G ven Given the solitary's recourse to self-addressed remarks well into adult life, and so tary s ta k s obv ous y g ven given that such talk is obviously not merely a transitional feature of primary mere y trans t ona pr mary s soc a zat on ( f ndeed socialization (if, indeed, a natural phase of childhood development), one is ennatura ch dhood deve opment) sh ft nteract ona approach Se f-ta k couraged to shift from a developmental to an interactional approach. Self-talk, deve opmenta when performed in its apparently permissible habitat-the self-talker all alone-is n ts apparent y perm ss b e hab tat-the se f-ta kera a one- s of being a mimicry of something that has its initial and natural provenance ts n t a natura by way be ng m m cry someth ng in speech between persons; this in turn implies a social encounter, and the arrangen th s n soc a encounter mp es ment of participantsthrough which encounters are sustained. (Such transplantation, part c pantsthroughwh ch encountersare susta ned transp antat on restr cted dev ant act v ty; wr ter t note s certa n y note, is certainly not restricted to deviant activity; thus a writer does it when he in the body of his own single sentence, an entire paragraph from a cited n hs ent re paragraphfrom c ted quotes quotes, s ng e sentence text-thereby pseudomorphically depositing in one form something that in nature pseudomorph ca y depos t ng n someth ng n to another.) be ongs belongs another ) W th se f-ta k then With self-talk, then, one might want to say that a sort of impersonation is occurm ght mpersonat on s r ng; ring; after all, we can best compliment or upbraid ourselves in the name of someone a comp ment upbra dourse ves n other than the self to whom the comments are directed. But what is intended here se f d rected s ntended is not so much the mere citation or recording of what a monitoring voice might say, s c tat on record ngof mon tor ng vo ce m ght say or what we would say to another if given a chance, but a stage-acted version of such wou d f g ven chance stage-actedvers on
2 Nor should the opposite framingissue be neglected. A man talking to himself at a bar may shou d he oppos e ram ng ssue neg ec ed a k ng o h mse a he bar ender o h nk h m drunk no pecu ar cause the bartenderto think him drunk, not peculiar; if he wants to continue drinking,he may wan s o con nue dr nk ng su ermore hardsh p rom he rs mpu a on han he second In suffermore hardshipfrom the first imputationthan the second. (In an instancereportedto me, ns ancerepor ed o me a bar-roomself-talker,misframedas having had too much, temporarilysolved this threat to his bar roomse a ker m s ramedas hav nghad oo much emporar yso ved h s hrea o h s dr nk ngr gh s re rea ng o he avern s e ephone boo h o drinkingrights by retreatingto the tavern'stelephone booth to do his self-talking.) h s se a k ng 3 I leave open the question of whetherthe individualwho engages in verbalhallucinationdoes eaveopen he ques on o whe her he nd v dua n verba ha uc na ondoes n so in order to create a disturbingimpression,or does so in ignoranceof the effects, or indiffero crea ea d s urb ng mpress on n gnoranceo he e ec s nd er o hem ence to them, or in spite of concern about them. I leave open too the question of whether, in n sp e o abou hem eave oo he ques on o whe her n treating unabashedself-talk as a naturalindex of alientation, we have (in our society) any good rea ngunabashedse a kas na ura ndexo a en a on n soc e y or nduc on grounds for our induction.

790

LANGUAGE, VOLUME 54, NUMBER 4 (1978)

s nto a delivery, albeit only vaguely a version of its reception. What is set into the ongoing ongo ng de very a be t on y vague y vers on ts recept on who e nteract ona text is not merely words, but their animator also-indeed, the whole interactional s the r an mator a so- ndeed mere y words th s n wh ch br ef y sp t spoken m ght arrangement in which such words might get spoken. To this end we briefly split characterwho ta ks characterto ourse ves n two project ng ourselves in two, projecting the characterwho talks and the characterto whom such n words could be appropriately directed. Or we summon up the presence of others in cou d appropr ate y d rected them Se f-ta k then nvo ves order to say something to them. Self-talk, then, involves the lifting of a form of ft ng someth ng ts emp oyment n spec a way ts natura p ace interaction out of its natural place, and its employment in a special way. nteract on cons derat on Se f-ta k descr bed n th s Self-talk described in this way recommends consideration of the soliloquy, long a so oquy ong feature of Western drama, although not currently fashionable.4 An actor comes drama a though current y fash onab e 4 ength d vu g ng h s stage center and harangues himself, sometimes at enormous length, divulging his h mse f somet mes w th we -projectedaud b ty Th s behav or inner thoughts on a pertinent matter with well-projected audibility. This behavior, nner pert nent ru e aga nst pub c of course, is not really an exception to the application of the rule against public app cat on course s except on rea y se f s se f-ta k self-talk. The soliloquizer is really talking to self when no one is around; we so oqu zer s rea y ta k ng members of the audience are supernatural, out-of-frame eavesdroppers. Were a aud ence supernatura eavesdroppers wou d aud b y characterfrom characterfrom the dramatizedworld to approach, our speakerwould audibly (to us) dramat zedwor d approach evennow come se f-d rect warn ng: self-direct a warning: But soft, I see that Jeffrey even now doth come. To the appearsoft cont nue ance of innocentbusiness then-and would stop soliloquizing. Were he to continue nnocentbus nessthen-and wou d so oqu z ng nstructed h m fa not ce to self-talk, it would be because the script has instructed him to fail to notice the t wou d se f-ta k scr pt a seen approach ng f gure approaching figure whom all the rest of us have seen. conversat on if talking to oneself in private involves a mock-up of conversation and a f ta k ng onese f n pr vate nvo ves Now Now, ts comp ementar ty recast ng recasting of its complementarity, then the production of this recasting on the stage, product on th s recast ng stage in the bloated format of a soliloquy, obviously involves a further insetting, and a n b oated so oquy obv ous y nvo ves nsett ng transformed cou d transformat on transformation of what has already been transformed. The same could be said, a ready sa d incidentally, about a printed advertisementfeaturing realistically-posedlive models nc denta y pr nted advert sementfeatur ngrea st ca y-posed ve mode s n broken- neba oons nto we -art cu ated nner whose sentiments are cast into well-articulatedinner speech in broken-line balloons sent ments above their heads-providing a text that the other figures in the pictured world the r heads-prov d ng f gures n p ctured wor d can t see can't see, but we real people can, as distinguished from the continuous-line balloon cont nuous- ne ba oon rea peop e can d st ngu shedfrom another for containing words that one figure openly states to another. conta n ng f gure open y commun cat on Behav or Here be eve s cruc a Here, I believe, is a crucial feature of human communication. Behavior and r tua zed- n appearance are ritualized-in the ethological sense-through such ethologicallyetho og ca etho og ca ydef nedprocesses exaggerat on stereotyp ng standard zat onof ntens ty oosendefined processes as exaggeration, stereotyping, standardizationof intensity, loosencontextua requ rements etc ing of contextual requirements etc. In the case under question, however, these ng quest on however transformat ons transformations occur to a form of interaction, a communication arrangement, a nteract on commun cat on arrangement standardset part c panta gnments be eve standard set of participant alignments. I believe that any analysis of self-talk (or for ana ys s se f-ta k th s that matter, any other form of communication) that does not attend to this nonmatter commun cat on) transformat on s un ke y linguistic sense of embedding and transformation is unlikely to be satisfactory. ngu st c embedd ng sat sfactory se f-ta k prov de 2 2. These parables about self-talk provide entrance to a mundane text. First, text F rst parab es w th n wh ch def n t ons: definitions: By a SOCIALSITUATION, I mean any physical area within which two or SOCIALSITUATION phys ca f nd themse ves n v sua aura more persons find themselves in visual and aural range of one another. The term another bod es restr ct on s GATHERING can be used to refer to the bodies that are present. No restriction is present n implied about the relationship of those in the situation: they may all be involved in mp ed re at onsh p s tuat on: a nvo ved n the same conversational encounter, in the sense of being ratified participants of the conversat ona encounter n be ng rat f edpart c pantsof
4 It is never necessaryin novels and comics, where the author can open up a character'shead I s nevernecessary n nove s he au horcan charac er shead com cs n o he deas con a ns echno og ca y s so the readercan peer into the ideas it contains; technologically,it is no longer necessaryin the he readercan ongernecessary n he o commerc a make be eve mov es compe ng competing modes of commercial make-believe-movies and television. In these latter, a voicee ev s on hese a er vo ce over effect allows us to enter into the inner thoughts of a charactershown silently musing. e ec a ows o en er n o he nner hough s o charac ershown s en y mus ng

RESPONSE CRIES

791

same state of talk; or some may be in an encounter while others are not, or are in n wh e n ta k; not a different one; or no talk may be occurring. Some, all, or none of those present d fferent ta k occurr ng Some a def nab e n soc a part c pat on e n WITH may be definable as together in terms of social participation, i.e. in a WITH. k nd comm tted n soc a s tuat ons A though a most Although almost every kind of mayhem can be committed in social situations, one class of breaches bears specifically on social situations as such, i.e. on the social c ass breachesbears spec f ca y soc a s tuat ons such e soc a organ zat on organization common to face-to-face gatherings of all kinds. In a word, although word a though gather ngs a k nds
de c s SITUATIONAL soc a SITUATED on y many delicts are SITUATED, only some are SITUATIONAL. We owe, to any social owe o

s tuat on n wh ch situation in which we find ourselves, evidence that we are reasonably alive to what f nd ourse ves ev dence reasonab y a ve is already in it-and furthermore to what may arise, whether on schedule or s a ready n t-and schedu e ar se mmed ate act on s requ red us w unexpected y unexpectedly. If need for immediate action is required of us, we will be ready-if ready- f not mobilized, then mobilizable. A sort of communication tonus is implied. If mob zab e commun cat on s mp ed mob zed n shou d addressed by anyone in the situation, we should not have far to go to respond, s tuat on respond if not to reply. All in all, a certain respect and regard is to be shown to the situation f certa n s s tuat on rep y A n a at large. These demonstrations confirm that we are able and willing to enter into demonstrat ons conf rm ab e nto arge w ng the perspective of the others present, if no more than is required to collaborate in s requ red co aborate n present f perspect ve the intricacies of talk and pedestriantraffic.In our society, then, it is generally taboo ntr cac esof ta k pedestr antraff c soc ety then t s genera y in public to be drunk, to belch or pass wind perceptibly, to daydream or doze, or n pub c w nd percept b y be ch drunk doze to be in disarray with respect to clothing and cosmetics-and all these for the same n d sarray w th cosmet cs-and a c oth ng convent ona reperto re reason reason. These acts comprise our conventional repertoire, our prescribed stock of compr se prescr bed 'symptoms', for demonstrating a lack of respectful aliveness in and to the situation; symptoms s tuat on; demonstrat ng ack respectfu a veness n the r nh b t on s their inhibition is our way of 'doing' presence, and thereby self-respect. And the do ng presence se f-respect w th aud b e nd cators nvo ved we demonstrat on demonstration can be made with sound; audible indicators are involved as well as v sua ones visual ones. It is plain, then, that self-talk, in a central sense, is situational in character, not s p a n then se f-ta k n centra sense s s tuat ona n character mere y s tuated; ts occurrencestr kes d rect y merely situated; its occurrence strikes directly at our sense of the orientation of the or entat on speaker to the situation as a whole. Self-talk is taken to involve the talker in a s tuat on who e Se f-ta k s nvo ve ta ker n s tuat ona y nappropr ate way D fferent y put situationally inappropriate way. Differently put, our self-talk-like other 'mental se f-ta k- ke menta symptoms - s symptoms'-is a threat to intersubjectivity: it warns others that they might be ntersubject v ty: t m ght wrong in assuming a jointly-maintained base of ready mutual intelligibility among n assum ng jo nt y-ma nta ned mutua nte g b ty a all persons present. Understandably, self-talk is less an offense in private than in present Understandab y se f-ta k s ess n pr vate n pub c; public; after all, the sort of self-mobilization and readiness it is taken to disprove is a se f-mob zat on read ness t s d sprove s not much required when one is all alone. requ red s a a one Th s genera argumentmakes This general argument makes sense out of a considerable number of minor details. cons derab enumberof m nor deta s In a waiting room or on public transport, where it is evident that little personal wa t ng pub c transport t s ev dent tt e persona attent on pedestr antraff c s requ red(and attention to pedestrian trafficis required (and therefore less than a usual amount of ess usua a veness aliveness to the surround), reading is allowed in our society, along with such selfsurround) read ng s a owed n soc ety a ong w th se fw thdrawa withdrawal to a printed world as this makes possible. (Observe that reading itself pr nted wor d th s poss b e read ng tse f is institutionalized as something that can be set aside in a moment, should a reason s nst tut ona zed someth ng as de n moment shou d present itself-something that can be picked up and put down without ceremony. tse f-someth ng p cked w thout ceremony Th s def n t on This definition does not hold for all our pleasures.) However, chuckling aloud to ho d a p easures ) However chuck ng a oud ourse ves n ourselves in response to what we are reading is suspect; this can imply that we are read ng s th s mp y too freely immersed in the printed scene to retain dissociated concern for the scene free y mmersed n pr nted reta n d ssoc ated in which our reading occurs. Interestingly, if we mouth the read words to ourselves, n wh ch read ng occurs Interest ng y f ourse ves mak ng making the mouthings audible, we will be taken to be unschooled, not unhingedmouth ngs aud b e w unschoo ed unh ngedun ess unless, of course, our general appearance implies a high educational status and course genera mp es h gh educat ona therefore no 'natural' reason for uncontained reading. (This is not to deny that natura unconta ned read ng (Th s s

792

LANGUAGE, VOLUME 54, NUMBER 4 (1978)

s nvested n some mumbled reading gives the impression that too much effort is invested in the mumb ed read ng g ves mpress on sheer task of reading to allow a seemly reserve for the situation at large.) s tuat on read ng a ow seem y arge ) In public, we are allowed to become rather deeply involved in talk with others, a owed pub c deep y nvo ved n ta k w th others b ock traff cor ntrude ead prov d ng th s providing this does not lead us to block traffic or intrude on the sound preserve of ab e ta k w th others; presumably our capacity to share talk with one other implies we are able presumab y capac ty mp es to share it with those who see us talking. So, too, we can conduct a conversation t w th conversat on ta k ng So too a oud aloud over an un-boothed street-phone while either turning our back to the flow of wh e e ther turn ng f ow w pedestr an traff c watch ng t n pedestrian traffic or watching it in an abstracted way, and the words will not be way thought improper. Even though our co-participant is not visually present, a natural mproper v sua y present natura co-part c pant s one can be taken to exist, and an accounting is available as to where, cognitively ex st where cogn t ve y account ng s ava ab e gone Moreover th s s fam ar p ace wh ch speak ng speaking, we have gone. Moreover, this is a familiar place to which others can see themse ves trave ng themselves traveling, and from which we can be duly recalled, should events wh ch du y reca ed shou d warrant 5 warrant.5 Observe also that we can, with some impunity, address words in public to a peta so can w th mpun ty addresswords n pub c an ma affect ve e ement presumab y presumably on the grounds that the animal can appreciate the affective element of apprec ate the talk, if nothing else. In any case, although on these occasions a full-fledged occas ons fu -f edged ta k f noth ng e se case a though words t s c ear rec p ent sn t recipient isn't present to reply to our words, it is clear that no imagined person or rep y mag ned a en alien agency has captured our attention. On the other hand, to be seen walking attent on hand wa k ng down the street alone while SILENTLY cu at ng a conversation with an absent a one wh e gest gesticulating conversat on w th other is as much a breach as talking aloud to ourselves; it is taken to be evidence s ev dence ta k ng a oud ourse ves; t s of alienation, just as much as its audible counterpart. ts aud b e counterpart a enat on em t (somet mes F na y Finally, there are the words we emit (sometimes very loudly) to summon another oud y) into talk. Although such speaking begins by being outside of talk with actual others, nto ta k A though speak ngbeg ns be ng outs de ta k w th actua others its intended recipient is likely quickly to confirm-by ritualized orientation, if not ts ntended rec p ent s ke y qu ck y conf rm-by r tua zed or entat on f verba rep y-the ex stence by a verbal reply-the existence of the required environment, doing so before our requ redenv ronment do ng s comp eted 6 s open y utterance is completed.6 A summons that is openly snubbed or apparently unapparent y can leave us feeling that we have been caught engaging in eave detected however detected, however, fee ng engag ng n someth ng ke ta k ng ourse ves something like talking to ourselves, and moreover very noticeably.7 not ceab y 7
5 I once saw an adolescentblack girl make her male companion collapse in laughter,on a busy ado escen b ackg r makeherma ecompan onco apse n augh er rom h m o down owns ree downtown street, by moving away from him to a litter can in which she had spied a plastic toy ercan n wh chshe mov ng sp ed p as c oy o mou h wh e e ng he rema n n he can phone Ho d ng he phone. Holding the phone up to her mouth and ear while letting the cord remain in the cano v ew he pass ngparade n d ssoc a edmanner as and then, half-turningas if to view the passingparadein a dissociatedmanner(as one does when hen ha urn ng anchored to an open telephone kiosk)-she projected a loud and lively conversation into the o e ephone k osk she pro ec eda oud ve y conversa on n o he ac pu s on pub c n ra herdeep way s r k nga s accommo mou hp ece mouthpiece. Such an act 'puts on' public order in a ratherdeep way, strikingat its accommoda ve c ose read ngs ones dative close readings-ones we all ordinarilysupport without much awareness. a ord nar ysuppor w hou awareness 6 A sma ch d pe pet or a small child can be repeatedlysummoned with a loud cry when it is not in sight, repea ed ysummonedw h oud s no n s gh w h with some disturbanceto persons in range; but a 'mental' condition is not ordinarilyimputed. n range bu d s urbance o men a cond on s no ord nar y mpu ed Typ ca y s unders ood ha he Typically it is understood that the words are merely a signal (a toy whistle would do) to come mere ya s gna a oy wh s ewou d do o o n o v ew o rece vea message bu no o home home, or to come into view to receivea message-but not to come into protractedconversation n o pro rac edconversa on from whereverthe signal is heard. rom he s gna s heard 7 Such an occurrenceis but one instance of the deplorableclass of occasions when we throw s bu ns anceo he dep orab ec ass o occas ons hrow ourse ves u ace n o ourselves full-face into an encounter where none can be developed-as when we respond to a encoun erwhere deve oped as o summons that was meant for someone behind us, or warmly greet a total strangermistakenly ha mean or beh ndus warm ygree o a s rangerm s aken y taken to be someone we know well, or (as has already been mentioned) mistakenly reply to aken o we as a ready men oned m s aken yrep y o someone s se a k someone's self-talk.The standardstatementby which the individualwhom we have improperly s andards a emen wh ch he nd v dua mproper y en ang edse s r gh e g Sorry I m a ra dyou ve entangled sets us right-e.g. Sorry,I'm afraidyou've ...-itself has an uneasy existence. Such a se uneasyex s ence remark is fully housed within a conversationalexchange that was never properlyestablished, s u y w h n conversa ona ha proper yes ab shed and its purpose is to deny a relationshipthat is itself requiredfor the remarkto be made. s s o re a onsh p ha s se requ red or he o made

RESPONSE CRIES

793

t s se f-ta k s s tuat ona mpropr ety s To say that self-talk is a situational impropriety is not to say that it is a aud b e decorum more e more, i.e., than any other audible breach of decorum, se f-ta k s such as an uncovered, audible yawn. Desisting from self-talk is not something we uncovered aud b e yawn Des st ng someth ng n the r owe our fellow conversationalists as such; i.e., it is not owed to them in their fe ow conversat ona sts e t s on y them C ear y capac ty co-part c pants n spec f c encounter capacity as co-participants in a specific encounter, and thus only to them. Clearly avo dance it is owed to all those in sight and sound of us, precisely as we owe them avoidance t s n s ght a us prec se y nd v dua of the other kinds of improper sounds. The individual who begins to talk to himself k nds beg ns ta k h mse f mproper sounds wh e n conversat ona encounterw while in a conversational encounter will cause the other participantsin the encounter part c pants n n n to think him odd; but for the same reason and in the same way, those not in the th nk h m way conversat ona t w a so th nk h m odd w th n encounter but within range of it will also think him odd. Here the conversational outs de ts soc a s tuat on s L ke sna c rc e s circle is not the relevant unit; the social situation is. Like a snail caught outside its re evant un t; outs de conversat ons outs de rat f ed she shell, words are here caught outside of conversations, outside of ratified states of th s s on y ta k; talk; one is saved from the linguistic horror of this fact only because the words ngu st c ta k s themse ves themselves ought not to have been spoken. In fact, here talk is no more conversafact spoken d fferent ref ects adverse y t ona tional than a belch; it merely lasts longer, and reflects adversely on a different part be ch; t mere y asts onger of personality. persona ty de ct-no delict-no CONVERSATIONAL
he ay ormu a on ONESELFIN PUBLIC Bu o So a rule: No TALKING TO ONESELFIN PUBLIC. But of course the lay formulation ru e

t mere y te s of a rule never gets to the bone; it merely tells us where to start digging. In linguistic ru e d gg ng ngu st c phras ng No ta k ng onese f n pub c s prescr pt veru e commun cat on phrasing, 'No talking to oneself in public' is a prescriptiverule of communication. ess neat The descriptive rule-the practice-is likely to be less neat, and certainly less descr pt ve ru e-the pract ce- s ke y certa n y ess ava ab e a ow ng ( f available, allowing (if not encouraging) variously-groundedexceptions. The frameencourag ng) var ous y-groundedexcept ons s nvo ved s work of normative understandings that is involved is not recorded, or cited, or normat ve understand ngs recorded c ted nformants ava ab e n summaryform available in summary form from informants. It must be pieced out by the studentp eced in part by uncovering, collecting, collating, and interpreting all possible exceptions n uncover ng co ect ng co at ng nterpret nga poss b e except ons to the stated rule. ru e others 3 3. An unaccompanied man-a single-is walking down the street past others. unaccompan ed s ng e- s wa k ng v ews h m w th ev dence H s genera His general dress and manner have provided anyone who views him with evidence prov ded of his sobriety, innocent intent, suitable aliveness to the situation, and general social h s sobr ety nnocent ntent su tab e a venessto s tuat on genera soc a stumb es str kes obtrud ng p ece competency H s eft competency. His left foot strikes an obtruding piece of pavement and he stumbles. cont nues on ess eff c ent y He instantly catches himself, rights himself more or less efficiently,and continues on. nstant ycatchesh mse f r ghtsh mse f Theretofore h s Theretofore, his competence at walking had been taken for granted by those who wa k ng grantedby sawh m conf rm ngthe rassessmentof h m n th s connect on H s tr pp ngsudden y saw him, confirmingtheir assessmentof him in this connection. His trippingsuddenly we casts these imputations into doubt. Therefore, before he continues, he may well mputat ons nto doubt Therefore cont nues n w th aws mechan cs act ons engage in some actions that have nothing to do with the laws of mechanics. The noth ng hs remed a remedial work he performs is likely to be aimed at correctingthe threat to his repus ke y a med correct ngthe f tat on tation, as well as his posture. He can pause for a moment to examine the walk, as if we h s posture exam ne wa k intellectually concerned (as competent persons with their wits about them would be) nte ectua yconcerned w th the r w ts wou d to discover what in the world could possibly have caused him to falter-the implid scover n wor d cou d poss b y h m fa ter-the mp cat on be ng cation being that anyone else would certainly have stumbled, too. Or he can address e se wou d certa n yhave stumb ed too a wry little smile to himself, to show that he himself takes the whole incident as a tt e sm e h mse f h mse f who e nc dent joke-someth ng qu teuncharacter st c someth ngthat joke-something quite uncharacteristic,something that can hardlytouch the security hard ytouch secur ty he feels in his own manifest competency, and therefore warranting no serious fee s n h s man fest competency warrant ng ser ous account account. Or he can 'overplay' his lurch, comically extending the disequilibrium, overp ay h s urch com ca y extend ng d sequ br um thereby concealing the actual deviation from normal ambulatory orientation with concea ng actua dev at on norma ambu atory or entat on w th c own ng movements mp y ng clowning movements, implying a persona obviously not his serious one. obv ous y h s ser ous one

794

LANGUAGE, VOLUME 54, NUMBER 4 (1978)

nward state externa zesa In brief, our subject externalizes a presumed inward state, and acts so as to make br ef te s tt e d scern b ethe spec a c rcumstanceswh ch presumab yproduced t discernible the special circumstanceswhich presumablyproduced it. He tells a little a n s tuat on rendersh mse f story to the situation. He rendershimself easy to assess by all those in the gathering, gather ng s the r assessment even as he guides what is to be their assessment. He presents an act specialized in a spec a zed n gu des convent ona conventional way for providing information-a DISPLAY-a communication in the prov d ng nformat on-a DISPLAY a commun cat on n behav or s an ma - ke ngu st c sense etho og ca ethological, not the linguistic, sense. The behavior here is very animal-like, except s an ma obv ous b othat what the human animal seems to respond to is not so much an obvious bioreputat on wh ch t wou d ord nar ytry ma nta n logical threat as a threat to the reputation which it would ordinarilytry to maintain og ca s t soc a competence nd v dua n in matters of social competence. Nor is it hard to catch the individual in a very n somet mes ook-the hasty surrept t ous standard look-the hasty, surreptitious survey sometimes made right after comr ght s w tness ng m tt ng f eet ng d scred tab edeed mitting a fleeting discreditable deed. The purpose is to see whether witnessing has remed a act on s th s tse f s occurred and remedial action is therefore necessary; and this assessment itself is done quickly enough so that a remedy, if necessary, can be provided with the same remedy f necessary prov ded w th qu ck y s tw necessary d spatch dispatch as when there is no doubt from the start that it will be necessary. instead of (or as a supplementto) engaging in a choreographedaccountnstead However However, supp ementto) engag ng n s v sua y ava ab e wonderment ing that is visually available, our subject may utter a cry of wonderment, such as ng wor d!Aga n the world!Again, he rendersreadily accessible to witnesses what he chooses n What in read y access b eto w tnesseswhat h s nward state d rects attent on to assign to his inward state, and directs attention to what produced it; but this th s t; ass gn d sp ay s arge yaud tory Moreover f non-voca gestures n conjunct on t me time the display is largely auditory. Moreover, if non-vocal gestures, in conjunction aud b e scene can t conven ent y prov de requ red nformaw th with the visible and audible scene, can't conveniently provide the required informavsbe nd cated a ternat ve nd v dua se f-ta k w sudden y t on tion, then self-talk will be the indicated alternative. An individual who suddenly on y gr mace c utch h s f s n hs stops in his tracks need only grimace and clutch at his heart if there is an open stopp ng h s remember ngthat manho e h s manhole at his feet; but the same stopping, consequent on his rememberingthat he somep ace e se s ke y was supposed to be someplace else, is more likely to be accounted for by words. words matter se f-remarksw (Presumab y (Presumably, the more obscure the matter, the more extended the self-remarkswill ess ke y nd v dua w them ) have to be-and perhaps the less likely the individual will be to offer them.) s ngu st cs n I am arguing here that what is part of the subject matter of linguistics, in some argu ng exam nat on re at on soc a s tuat ons arge requ re sense sense, can require the examination of our relation to social situations at large, not conversat ons Apparent y verba zat ons qu te outs de re at on mere y merely our relation to conversations. Apparently, verbalizations quite outside of p ay ro e bt non-voca conversat ons conversations can play much the same role as a choreographed bit of non-vocal ke s tuat ona propr etyand mpropr ety behav or Together behavior. Together, they are like other situational acts of propriety and impropriety ent re surround access b e n des gned in that they are accessible to the entire surround, and in a sense designed for it. n t speech un ke c oth ng cosmet cs ke c oth ng They are more like clothing than speech. But unlike clothing or cosmetics, these n pantom me-are nterpreted bear ng d sp ays-whether voca displays-whether vocal or in pantomime-are to be interpreted as bearing on a w th m ted n t me certa n y pass ng event passing event, one with a limited course in time. (What we wear can certainly be att tude soc a occas on hand hard y nd cat on taken as an indication of our attitude to the social occasion at hand, but hardly to occas on ) Necessar y f unant c patedpass ng occurr ngdur ng spec f c specific events occurring during the occasion.) Necessarily, if unanticipated passing ntroducedjust addressed markermust emp oyed events are to be addressed,a markermust be employed that can be introducedjust at occurs w thdrawnwhen ended the moment the event occurs, and withdrawnwhen concern for the event has ended. s proh b t on aga nst pub c se f-ta k 4 4. I have argued that there is a prohibition against public self-talk, and that d sp ay a so th s ru e soc a breach ngs breachings of this rule have a display character-yet also that there are social se f-ta k Indeed wh ch eads cou d s tuat ons n wh ch situations in which one could expect self-talk. Indeed, the very force which leads us ndu ge n to refrain from self-talk in almost all situations might itself cause us to indulge in refra nfrom se f-ta k n a most a s tuat ons m ght tse f th s ght cons der n greaterdeta dur ngcerta nexcept ona ones during certain exceptional ones. In this light, consider now in greater detail se f-ta k self-talk se f-ta k s frequent y found a few environments in which exposed self-talk is frequently found. env ronments n wh ch

RESPONSE CRIES

795

When we are 'informed' of the death of a loved one (only by accident are we nformed oved acc dent (on y n pass ng) br ef 'told', since this verb implies that the news might be conveyed in passing), a brief to d s nce th s m ght mp es nto wou d certa n y am ss n f ood ng flooding out into tears would certainly not be amiss in our society. As one might soc ety m ght expect t s expect, it is just then that public self-talk is also sanctioned. Thus Sudnow (1967:141) pub c se f-ta k s a so sanct oned n hosp ta s: descr bes describes the giving of bad news in hospitals: g v ng
'While no sympathy gesturesare made, neither does the doctor withdrawfrom the scene Wh e made ne herdoes he doc or w hdraw rom he sympa hyges ures he e egramde veryboy doc or a oge herby eav ng he room as or examp e altogetherby leaving the room, as, for example, does the telegramdeliveryboy. The doctor ha con a ned con ro s progress is concerned that the scene be contained and that he have some control over its progress, s ha he he rs genu ne that it not, for example, follow him out into the hall. In nearly all cases the first genuine ha no or examp e o ow h m ou n o he ha near y a he re a ve Dur ng he per od o cry ng here s interchangeof remarkswas initiated by the relative.During the period of crying, if thereis n erchangeo remarkswas n a ed can be eve " "I s us no a r" any re a ves requen y " a k" Examp es are any, relatives frequently "talk". Examples are: "I can't believe it", "It's just not fair",
"Goddamn" "Goddamn", "No "Not John ... no ..." These remarks are not responded to as they are not no o no " hey

addressed to anyone. Frequently, they are punctuated by crying. The physician remains o anyone Frequen y hey punc ua edby cry ng phys c anrema ns s en silent.'

The common-sense explanation here is that such informings strike at our self so s se f nform ngs str ke exp anat on s reasonab e-an excusab e that self-involvement immediately thereafter is reasonable-an excusable se f- nvo vement mmed ate y v o ent y violently e se n imposition of our own concerns upon everyone else in the gathering. Whatever the mpos t on gather ng Whateverthe estab sh c ass of a -too-human cr ses case convent on case, convention seems to establish a class of'all-too-human' crises, to be treated as something which anyone not directly involved should still appreciate; they give d rect y nvo ved shou d st apprec ate; someth ng wh ch g ve us victims the passing right to be momentary centers of sympathetic attention, v ct ms pass ng r ght sympathet c attent on and provide a legitimate place for anything 'uncontrolled' we do during the prov de eg t mate p ace anyth ng uncontro ed dur ng occas on Indeed occasion. Indeed, our utter self-containment during such moments might create se f-conta nment dur ng m ght uneas ness n uneasiness in others concerning our psychological habitat, causing them to wonder concern ng psycho og ca hab tat caus ng how responsive we might be to ordinary situated concerns directly involving respons ve m ght ord nary s tuated d rect y nvo v ng them them. a env ronmentswh ch favorse f-ta kare convent ona yunderstoodto so Not all environments which favor self-talkare conventionally understood to do so. For example, a podium speaker who suddenly finds that he has a page or line examp e pod um sudden y f nds ne h s text m ss ng missing from his text, or a faulty microphone, will sometimes elect to switch from fau ty m crophone w somet mes e ect sw tch aud ence ta k ng h mse f address nga fu ta k ng talking to the audience to talking to himself, addressing a full sentence of bewilderbew derhs ment chagr n ment, chagrin, or anger for his own ears and (apparently) his own benefit, albeit (apparent y) h s benef t a be t n room ha f-aud b y half-audibly to the room. Even in broadcast talk, when speakers lose their places, ta k ose the r p aces f nd themse ves w th ncoherent m sp ace the r scr pts misplace their scripts, or find themselves with incoherent texts or improperly mproper y n th s way seem ng funct on ng equ pment functioning equipment, they may radically break frame in this way, seeming sudrad ca y the r the r ob gat ons susta n ro e den y denly to turn their backs on their obligations to sustain the role of speaker-to-anaud ence audience. It is highly unprofessional, of course, to engage in sotto-voce, self-directed s h gh y unprofess ona n sotto-voce se f-d rected course remarksunder remarks underjust those microphonic conditions which ensure their audibility; but m crophon c cond t ons wh ch ensurethe r aud b ty; broadcasters may be more concerned at this point about showing that some part of th s po nt show ng them is shocked by the hitch, and in some way not responsible for it, than about s n h tch respons b e t so e sourceof mean ngfu ma nta n ngbroadcast ngdecorum A so be ng maintaining broadcastingdecorum. Also, being the sole source of meaningful events for their listeners, they may feel that the full text of their subjectiveresponse is better the r steners fee fu the r subject veresponse s than no text at all. Note that other social situations provide a speaker with an a soc a s tuat ons prov de w th aud ence audience that is captive and concerned, and thereby encourage self-talk: drivers s capt ve concerned se f-ta k: dr vers of buses, taxis, and private cars can shout unflatteringjudgments of other motorists buses tax s pr vatecars unf atter ngjudgments motor sts and pedestrianswhen they have passed out of range, and feel no compunction about pedestr answhen range fee compunct on the r passengers ta k ng a oud themse ves n talking aloud to themselves in the presence of their passengers. After all, there is a a s

796

LANGUAGE, VOLUME 54, NUMBER 4 (1978)

n wh ch n traff cv s b y sense in which a contretemps in trafficvisibly and identically impinges on everyone dent ca y mp nges in a vehicle simultaneously.8 n veh c e s mu taneous y 8 The fact that drivers may actually wait until the apparent target of their remarks dr vers apparenttarget the r actua y wa t unt cannot hear them points to another location for self-talk, which is also suggested by ocat on se f-ta k wh ch s a so suggestedby po nts
the lay term MUTTERING. he ay erm MUTTERING Frus ra ed Frustrated by someone's authority, we can mutter words someone s au hor y mu er

of complaint under our breath as the target turns away, out of apparent conversacomp a nt away t ona earshot tional earshot. (Here is a structuralequivalent of what children do when they stick s structura equ va ent st ck ch dren out their tongues, or put their thumbs to their noses, just as their admonisher turns the r tongues the r the r noses the r admon sherturns sub-voca zat ons res de n nterst ce ta k away ) away.) These sub-vocalizations reside in the very interstice between a state of talk and mere co-presence-more specifically, in the transition from the first to the trans t on f rst spec f ca y n second second. Here function seems plain: in muttering, we convey that although we are funct on p a n: n mutter ng a though ne estab shed now going along with the line established by the speaker (and authority), our spirit go ng a ong w th author ty) sp r t has not been won over, and compliance is not to be counted on. The display is on over comp ance s d sp ay s a med e ther th rd part es aimed either at third parties or at the authority itself, but in such a way that we can n author ty tse f sa d ntent author ty deny our intent, and the authority can feign not hearing what we have said about fe gn hear ng h m Aga n th s s him. Again, this is a form of communication that hardly fits the linguistic model of commun cat on hard y f ts ngu st c mode s speaker and addressed recipient; here we provide a reply to the speaker that is rec p ent; prov de rep y h m th rd part es ourse ves d sp aced displaced from him to third parties, or to ourselves. Instead of being the recipient be ng rec p ent of our reply, the initial speaker becomes merely the object or target of our response. nta mere y response rep y L ke tongue-st ck ng mutter ng s t me- m ted commun cat on enter ng Like tongue-sticking, muttering is a time-limited communication, entering as a 'last ast t word word', a post-terminal dollop to a just-terminated encounter; it thus escapes, for post-term na do op just-term nated escapes incidental reasons, the injunction against persisting in public self-talk. nc denta reasons njunct on aga nst pers st ng n pub c se f-ta k k nd Cons derat on se f-ta k n Consideration of self-talk in one kind of interstice recommends its consideration nterst cerecommends ts cons derat on in others. For example, if we are stopped for a moment's friendly chat just before n others moment s fr end y examp e f estab shment turn ng enter ng entering or leaving an establishment or turning down a street, we may provide a eav ng street prov de one-sentence description of the business we are about to turn to; this account serves bus ness th s descr pt on as a rationale for our withdrawing, and as evidence that there are other calls upon rat ona e ev dence ca s w thdraw ng s somet mes our time. Interestingly enough, this utterance is sometimes postponed until the t me time Interest ng y enough th s unt s moment when the encounter is just ending, in which case we may mumble the mumb e end ng n wh ch account half-aloud and somewhat to ourselves. Here again is self-talk that is located ha f-a oud ourse ves aga n s se f-ta kthat s ocated ta k trans t ona y transitionally between a state of talk and mere co-presence; again, self-communiaga n se f-commun cat on s se f-term nat ng a though th s t me cation is self-terminating, although this time because the communicator, not the commun cator t s nescapab y c ear hearer s mov ng away hearer, is moving away. Here it is inescapably clear that the self-talker is providing se f-ta ker s prov d ng verba nformat on verbal information to others present, though not using the standard arrangementstandardarrangementpresent us ng a ratified state of talk-for doing so. rat f ed ta k-for do ng so it must be allowed that when circumstances conspire to thrust us into a t a owed c rcumstancesconsp re F na y Finally, nto course of action whose appearancemight raise questions about our moral character act on appearancem ght ra se quest ons mora or self-respect, we often prefer to be seen as self-talkers. If we stoop to pick up a se f-ta kers se f-respect p ck co n coin on a busy street, we may well identify its denomination to ourselves aloud, we dent fy ts denom nat on street ourse ves a oud
O course here w Of course, there will be occasions of equivalent license for non-verbal signs, both vocal and occas ons o equ va en cense or non verba s gns bo h voca na o gr maces ges cu a ory gesticulatory. In trying on a shoe, we can engage in all manner of grimaces and obscure soundry ng shoe sound ings, for these signs provide running evidence of fit; and such information is the official, chief ngs or hese s gns prov de runn ng ev dence o n orma on s he o c a ch e concern (at that moment) of all parties to the transaction, including the shoe clerk. Similarly, a ha momen o a par es o he ransac on nc ud ng he c erk S m ar y a sportsman or athlete is free to perform an enormous flailing about when he flubs; apart from a h e e s ree o per orm spor sman a ng abou ubs apar rom o her other reasons for this license, he can be sure (if anyone can) that his circumstances are fully or h s cense can ha h s c rcums ances u y a ended attended and appreciated by everyone who is watching the action. After all, such clarity of s wa ch ng he ac on A er a apprec a ed c ar y o intent is what sports are all about. n en s wha spor s a abou
8

RESPONSE CRIES

797

ourse ves s mu taneous y express ng surpr se simultaneously expressing surprise, even though we ourselves no longer need the onger t m ght information: the street is to be framed as a place of passage, not-as it might be to nformat on: s p ace passage a child or a bum-a hunting ground for bits of refuse. If what we thought was a ch d b ts refuse hunt ng co n coin turns out to be a worthless slug, then we may feel urged to externalize, through fee worth ess s ug externa ze sound and pantomime, that we can laugh at the fools we have made of ourselves.9 foo s ourse ves 9 pantom me augh m staken door-hand e own d scover ng Try ng Trying the door-handle of a car we have mistaken for our own, and discovering our se f-d rectedremarkthat proper y carefu b urt m stake mistake, we are careful to blurt out a self-directedremark that properly frames our act for those who witness it, advertising inadequate attentiveness in order to deny w tness t advert s ng nadequate attent veness n that we are thieves. th eves W th With these suggestions of where self-talk is to be found, we can take a second se f-ta k s found suggest ons ch dren n t aren t look at the conventional argument that children engage in it because they aren't yet ook convent ona modest es soc a zed nto socialized into the modesties of self-containment, the proprieties of persondom. se f-conta nment propr et es persondom P aget s pos t on ong Vygotsky respond ng Vygotsky, responding to what he took to be Piaget's position, long ago provided a prov ded lead ([1934] 1962:16): ead
'In order to determine what causes egocentric talk, what circumstancesprovoke it, we In o de erm ne wha egocen r c a k wha c rcums ances he ser es P age d d bu organ zed he ch dren sac v es n organized the children'sactivities in much the same way Piaget did, but we added a series o draw ch d o rus ra onsand d cu es of frustrationsand difficulties.For instance, when a child was getting ready to draw, he ns ance ge ng o her wou d sudden y nd ha herewas paper would suddenlyfind that there was no paper, or no pencil of the color he needed. In other penc o he co or needed h m ace prob ems words words, by obstructinghis free activity we made him face problems. obs ruc ngh s ree ac v y a mos 'We found that in these difficult situations the coefficient of egocentric speech almost We ound ha n hese d cu s ua ons he coe c en o egocen r c doub ed n compar sonw hP age snorma gure or hesameage anda so n compar son doubled,in comparisonwith Piaget'snormalfigurefor the same age, and also in comparison ch d wou d ry o graspand o w h with our figurefor childrennot facing these problems.The child would try to grasp and to gure or ch drenno ac ng hese prob ems b ue penc remedy the situation in talking to himself: "Where's the pencil? I need a blue pencil. he s ua on n a k ng o h mse "Where s he penc ? we w h wa er w Never mind, I'll draw with the red one and wet it with water; it will become dark and look darkand ook m nd I draww h he like blue."'10 ke b ue " 10 The implication is that, for children, the contingencies are so great in undertaking mp ca on s ha or ch dren he con ngenc es grea n under ak ng

w ke hood any task, and the likelihood so strong that they will be entirely discounted as task ent re y d scounted w f fa reasonab y- ntent oned reasonably-intentioned persons if they fail (or indeed, that they will be seen as just ndeed
9 Picking money off the street is, of course, a complicated matter. Pennies and even nickels o he s ree s o course comp ca ed ma er Penn esand n cke s P ck ng we might well forego, if the doubt cast on our conduct is of more concern to us than the money. he doub cas conduc s o moreconcern o han he money m gh we orego n sma (We accept the same small sums in change when paying for something in a shop, but there a We accep he pay ng or some h ng n shop bu here ransac on s he o c a bus nessa hand Shou dano her n money transactionis the officialbusinessat hand.) Should anotherin our sight drop such a coin, s gh dropsucha co n we might well be inclined to retrieveand returnit: we are allowed a distractiveorientationto the nc ned o re r eveand re urn area oweda d s rac veor en a on o he m gh we h s s pa en y n he n eres so o hers I don re r eveour groundwe wa k on ground we walk on, so long as this is patentlyin the interests of others. (If we don't retrieveour ong own small coins, then we run the risk that others will do so for us, and the consequentnecessity sma co ns hen he r sk ha o hers w or us he consequen necess y o show nggra ude I he of showing gratitude.)If the sum is large enough to qualifyas beyond the rule of finders-keepers, s argeenough o qua yas heru eo nders keepers we might quickly glance around to see if we have been seen, carefullyrefrainingfrom saying or o m gh qu ck yg ance seen care u yre ra n ng rom say ngor ac wheneverwe ges ur ngany h nge se Cover a so maybe gesturinganythingelse. Covertalso may be our act wheneverwe spy a coin of any denomination co n o denom na on to see if any additional ones are not to be found. o add ona no o ound 10Piaget, as his reply (1962:3-4) to a reading of Vygotsky's MS suggests, apparentlymeant h s rep y 1962 3 4 o read ngo Vygo sky s sugges s apparen ymean P age o herbehav or ha d d no ake n o cons dera on he 'egocentricity' to referto speech (or any other behavior)that did not take into considerationthe egocen r c y o re er o or perspec veo he o her n perspective of the other in some way, and only incidentally (if at all) to speech not openly way on y nc den a y a a o no open y addressed to others, the latter being what Vygotsky described, and which I call 'self-talk'. o o hers he a er be ng wha Vygo sky descr bed wh ch ca se a k P age s concep o egocen r c yhas ed o ano her con us on Piaget's concept of egocentricity has led to another confusion, a failure to discriminatetwo a ure o d scr m na e wo ma ers ak ng he po n o v ew o he o her n matters: taking the point of view of the other in order to discover what his attitude and action o d scoverwha h s a udeand ac on w be will be, and acceptingfor oneself or identifyingwith the perspectiveof the other. The classic con accep ng or onese den y ngw h he perspec veo he o her c ass ccon opera on us ra eshow u y he rs orm o sympa hymaybe requ redand producedw hou operation illustrateshow fully the firstform of sympathymay be requiredand producedwithout leading to the second. ead ng o he second

798

LANGUAGE, VOLUME 54, NUMBER 4 (1978)

offersome vo c ng a ways foo ng fooling around anyway), that they are always preparedto offer some voicing of what anyway) areabout adu t attempt ng earnto they are about. An adult attemptingto learn to skate might be equally self-talkative.11 m ght equa y se f-ta kat ve 11 Some loose generalizations might be drawn from these descriptions of places for oose genera zat onsm ght descr pt ons p aces se f-ta k F rst self-talk. First, when we address a remark to ourselves in public, we are likely to be remarkto ourse ves n pub c ke y in sudden need of re-establishing ourselves in the eyes and ears of witnesses as n w tnesses re-estab sh ng ourse ves n tr f ed w th; honest honest, competent persons not to be trifled with; and an expression of chagrin, express on chagr n etc wou d wonderment wonderment, anger etc. would seem to help in this-at least establishing what our he p n th s-at east estab sh ng f n th s can t ourse ves are susta ned Second expectat ons expectations for ourselves are, even if in this case they can't be sustained. Second, one could argue that self-talk occurs right at the moment when the predicament of cou d se f-ta k r ght pred cament the speaker is evident to the whole gathering in a flash, or can be made so-assuring s ev dent who e gather ng n f ash so-assur ng that the utterance will come as an understandablereaction to an understood event; w understandab ereact on it will come from a mind that has not drifted from the situation, a mind readily tw m nd dr fted s tuat on m nd read y tracked tracked. The alien world reflected in hallucinatory talk is therefore specifically a en wor d ref ected n ha uc natory ta k s spec f ca y s avo ded; avoided; and so too is some of the impropriety of talking outside the precincts of a mpropr ety ta k ng outs de prec ncts rat f ed conversat on ratified conversation. Nor is 'understandable' here merely a matter of cognition. s understandab e mere y cogn t on To quickly appreciate another's circumstances (it seems) is to be able to place s ab e qu ck y apprec ate another s c rcumstances ( t p ace ourse ves n ourselves in them empathetically. Correspondingly, the best assurance another can assuranceanother empathet ca y Correspond ng y have that we will understand him is to offer himself to us in a version with which we n vers on w th wh ch w understandh m s h mse f can identify. Instead of thinking of self-talk as something blurted out under pressure, dent fy th nk ng se f-ta kas someth ng b urtedout pressure then t m ght betterbe then, it might better be thought of as a mode of response constantly readiedfor those responseconstant y read edfor t me c rcumstances n wh ch t s excusab e Indeed circumstances in which it is excusable. Indeed, the time and place when our private p ace pr vate react on s reaction is what strangerspresent NEED to know about is the occasion when self-talk s occas on se f-ta k strangerspresent is more than excusable.12 s excusab e 12 5 5. It was suggested above that, when an unaccompanied man stumbles, he may that unaccompan ed stumb es se f-ta k nstead hs s ent gesture However s present his case by means of self-talk instead of silent gesture. However, there is another route to the advertisementof self-respect. He can emit one or two words of advert sementof se f-respect em t He ! Sh t! exc amatory mprecat on exclamatory imprecation, such as Hell! or Shit! Observe that these ejaculatory ejacu atory nd v dua express ons expressions are nothing like the pointed shout of warning which one individual noth ng ke po nted warn ng wh ch ke open y d rected broadcastto a n m ght might utter to and for another-nor even like an openly directed broadcast to all in shr ek hear ng ke street-vendor scry hearing, like a street-vendor'scry or a shriek for help. Talk in the ordinary sense is he p Ta k n ord narysense s ssue mmed ate apparent y apparently not at issue. In no immediate way do such utterances belong to a be ong conversat ona conversational encounter-a ritually ratified state of talk embracing ratifiedparticir tua yrat f edstate ta k embrac ngrat f edpart c one F rst speaker s pants-or to a summoning to one. First speaker's utterance does not officially summon ng off c a y
11Cook-Gumperz & Corsaro (1976:29) offer a more compelling account: 'We have found ound Cook Gumperz 1976 29 o er compe ng accoun We that children consistently provide verbal descriptions of their behavior at various points in ha ch dren cons s en y prov de verba descr p ons o he r behav or a var ous po n s n o her n erac an s o wha s presen yoccurr ngas we spon aneous an asy n ha spontaneous fantasy in that it cues other interactantsto what is presentlyoccurringas well as prov des poss b es or p ugg ng n o provides possibilities for plugging into and expanding upon the emerging social event.' The expand ng he emerg ng soc a even au hors mp y ha authors imply that if a fantasy world is to be built up during JOINT play, then words alone are bu an asy wor d s o dur ng JO NT p ay hen a one he likely to be the resourcethat must be employed, and an open recourseto self-talk then becomes ke y o ha mus emp oyed o se a k hen an effective way to flesh out what is supposed to be unfolding for all the participantsin the e ec ve o esh ou wha s o un o d ng or a he par c pan s n he fantasy. an asy A purelycognitiveinterpretation certainaction-oriented,self-directedwords('non-nominal o of pure ycogn ve n erpre a on cer a nac on or en ed se d rec edwords non nom na a so express ons expressions') has also been recently recommendedby Gopnik (1977:15-20). recen yrecommendedby Gopn k 1977 15 20 12 Unders andab y s age so oqu es Understandably,stage soliloquies occur only when the character'spersonal feelings about on y he charac er spersona ee ngsabou h s c rcums ancesareexac ywha his circumstancesare exactly what we membersof the audiencemust be privy to, to be properly memberso he aud encemus pr vy o o proper y pos oned n he un o d ngdrama positioned in the unfolding drama.

RESPONSE CRIES

799

s s estab sh s ot wh ch establish a slot which second speaker is under some obligation to fill: there is no ob gat on f : rat f ed ratified speaker and recipient (not even imaginary ones), but merely actor and mere y mag nary ones) rec p ent t s w tness witness. To be sure, an interjection is involved; but it is one that interruptsa course sure nterrupts nterject on s nvo ved; of physical action, not an utterance. utterance phys ca act on who e ourse ves wa k When unaccompan ed When, unaccompanied, we trip and curse ourselves (or the walk, or the whole tr p ourse ves Therefore w de wor d) wide world), we curse TO ourselves; we appear to address ourselves. Therefore, a ourse ves; nvo ved L ke k nd kind of self-remarking seems to be involved. Like the publicly tolerated self-talk pub c y to erated se f-ta k se f-remark ng n gather ng sty ed a ready cons dered mprecat ons already considered, imprecations seem to be styled to be overheard in a gathering. s n nd v dua s Indeed Indeed, the styling is specificin this regard: when no one is present in the individual's sty ng s spec f c n th s om tted ch dren surround surround, the expression is quite likely to be omitted. If women and children are express on s qu te ke y h s cr es accord ng y: ma e se f-commun cator s qu te ke y present present, your male self-communicator is quite likely to censor his cries accordingly: s ke y avo d stumb es n a man who utters Fuck! when he stumbles in a foundry is likely to avoid that t s nursery on y tr ps n part cu ar exp et ve f particular expletive if he trips in a day nursery. If it is apparent that only very fa ed do) wh spered c ose-by close-by persons can see what we have just done (or failed to do), then whispered w away exp et ves expletives are possible; if witnesses are far away, then shouted sounds will be poss b e; f w tnesses Sacks term) qu ck y requ red Rec p ent des gn s nvo ved required. 'Recipient design' is involved (to use Harvey Sacks' term), and so quickly s tuat on s be ng susta ned on-go ng mon tor ng app ed applied as to suggest that on-going monitoring of the situation is being sustained, ar ses wh ch requ res t course enab ngjust th s enabling just this adjustment when the moment arises which requires it. Of course, ourvoca zat on n convent ona in any case we will have taken the time to encode our vocalization in the conventional n w takenthe t me oca s nstanlexicon of our language (which is likely to be the local one)-a feat that is instanex con anguage (wh ch s ke y somet mes taneous y accomp shed taneously accomplished, even sometimes by bilinguals who, in addition, must b ngua s who n add t on genera y se ect the r mprecat ons generally select their imprecations from the language of their witnesses.13(This is anguage the r w tnesses 13(Th s s not to say that bilinguals won't use a harsh imprecation from one language in place b ngua s won t mprecat on anguage n p ace of a less harsh one drawn from the language in use; foreignness apparently serves ess anguage n fore gnness apparent y behav or as a mitigation of strength.) Significantly, we have here a form of behavior whose m t gat on strength ) S gn f cant y t s someth ng b urtedout someth ng very meaning is that it is something blurted out, something that has escaped control; mean ng s escapedcontro ; and so such behavior very often is and has; but this impulsive feature marks not the s th s mpu s ve behav or s soc a y processed convent ona zed limits to which the utterance is socially processed, but rather the conventionalized m ts wh ch sty ng wh ch t s ob ged adhere styling to which it is obliged to adhere. It is plain that singles use imprecations in a variety of circumstances. Racing c rcumstances Rac ng s pan s ng es mprecat ons n var ety t automat ca y c oses turnst e t unsuccessfu y unsuccessfully to enter a turnstile before it automatically closes, or a door before it is locked for the evening, may do it; coming up to what has just now become a s ocked even ng t; com ng w th curse (Others br ck wa brick wall, we may exhibit frustration and chagrin, often with a curse. (Others, exh b t frustrat on chagr n f nd hav ng formu ated poss b e read ngof having formulated a possible reading of the precipitous rush we have made, can find prec p tousrush made that our imprecations are a way of confirming their interpretation, putting a period mprecat ons conf rm ngthe r nterpretat on putt ng per od to the behavioral sentence we have played out, bringing the little vignette to a close, behav ora p ayed out br ng ngthe tt e v gnette c ose and converting us back to someone easily disattendable.) Precariously carrying too convert ng eas y d sattendab e ) Precar ous ycarry ngtoo fa many parcels, we may curse at the moment they fall. When the horse we have bet on parce s is nosed out at the finish line, we may damn our misfortune while tearing up our s f n sh ne m sfortune wh e tear ng t ckets; s nce tickets; since our cause for disappointment, anger, and chagrin is amply evident, or d sappo ntment anger chagr n s amp y ev dent at least easily surmisable, we have license to wail to the world. Walking along a cense wa east eas y surm sab e wor d Wa k ng a ong carr es record-break ngsnow n w ntry wintry street that carries a record-breakingsnow now turned to slush, we are in a s ush n t happens pos t on position to cry God! in open private response-but as it happens, we do so just at pr vate the point of passing another-the cause of our remark and the state of our mind m nd po nt pass ng
13 It would be interestingto know whether or not bilingual children who self-talk select the I wou d whe heror no b ngua ch drenwho se a k se ec he n eres ng o code likely to be employed by others in their presence. ke y o emp oyed o hers n he r presence

800

LANGUAGE, VOLUME 54, NUMBER 4 (1978)

being perfectly plain and understandable. It might be added that the particular imprecations I have used so far as illustrations seem in our society to be the special domain of males: females, traditionally at least, use softer expressions. As is now well known, this gender convention is not impervious to rapid politically-inspired change. Finally, I want to note that although imprecations and extended self-remarkscan be found in much the same slot, do much the same work, and indeed often appear together-raising the question as to why they should be described separatelyjudgment should be reserved concerning their equivalence. Other questions must be considered first. 6. The functioning of imprecations raises the question of an allied set of acts that
can be performed by singles: RESPONSECRIES, i.e. exclamatory interjections which

are not full-fledgedwords. Oops!is an example. These non-lexicalized, discreteinterjections-like certain unsegmented, tonal, prosodic features of speech-comport neatly with our doctrine of human nature. We see such 'expression' as a natural overflowing, a flooding up of previously contained feeling, a bursting of normal restraints, a case of being caught off-guard. That is what would be learned by asking the man in the street if he uses these forms-and, if so, what he means by them. I am assuming, of course, that this common-sense view of response cries should give way to the co-occurrence analysis that sociolinguists have brought to their problems. But although this naturalistic method is encouraged by sociolinguists, here the subject matter moves one away from their traditional concern. A response cry doesn't seem to be a statement in the linguistic sense (even a heavily elided one), purportedly doing its work through the concatenated semantic reference of words. A remark is not being addressed to another-not even, it seems, to oneself. So, on the face of it at least, even self-communication is not involved, but only a simpler sign process whereby emissions from a source inform us about the state of the source-a case of exuded expressions, not intentionally sent messages. One might better refer to a 'vocalizer' or 'sounder' than to a speaker. This, of course, is not to deny the capacity of a well-formed, conventionally-directed sentence to inform us about the state of the protagonist who serves as its subject, nor to deny that the speaker and protagonist can be the' same'-for indeed, through the use of 1stperson pronouns, they routinely are. But this latter arrangement brings us information through a message, not an expression. This route is fundamentally different from and less direct than the one apparently employed in response cries-even though, admittedly, such cries routinely come to be employed just in order to give a desired impression. Witnesses can seize the occasion of certain response cries to shake their heads in sympathy, cluck, and generally feel that the way has been made easy for them to initiate passing remarks, attesting to fellow-feeling; but they aren't obliged to do so. A response cry may be uttered in the hope that this half-license it gives to hearers to strike up a conversation will be exercised; but, of course, this stratagem for getting talk started could not work if an innocent reading were not the official one. Expectedly, the circumstances which allow us to utter a response cry are often just the ones that mitigate the impropriety of a different tack we could take, that of opening up an encounter by addressinga remarkto an unacquainted other; but that

RESPONSE CRIES

801

fact doesn't relieve one of the necessity to distinguish between this fully social sort of comment and the kind that is apparently not even directed to oneself. A response cry is (if anything is) a ritualized act in the ethological sense of that term. Unable to shape the world the way we want to, we displace our manipulation of it to the verbal channel, displaying evidence of our alignment to the on-going events; the display takes the condensed, truncated form of a discretely-articulated, non-lexicalized expression. Or, suddenly able to manage a tricky, threatening set of circumstances, we deflect into non-lexicalized sound a dramatization of our relief and self-congratulation in the achievement. 7. Consider now some standard cries:
7.1. THETRANSITION DISPLAY. Entering or leaving what can be taken as a state of

marked natural discomfort-wind, rain, heat, or cold-we seem to have the license (in our society) to externalize an expression of our inner state. Brr! is a standard term for wind and cold upon leaving such an atmosphere. (Other choices are less easily reproduced in print.) Ahh! and Phew! are heard when leaving a hot place for a cool one. Function is not clear. Perhaps the sounding gives us a moment to orient ourselves to the new climatic circumstancesand to fall into cadence with the others in the room; these are not ordinarily taxing matters, and thus do not ordinarily require a pause for their accomplishment. Perhaps the concentration, the 'holding ourselves in' sometimes employed in inclement places (as a sort of support for the body) gets released with a flourish when we escape from such environments. In any case, we can be presumed to be in a state of mind that those already safe might well appreciate (for after all, weather envelops everyone in the vicinity), and so selfexpression concerning our feelings does not take us to a place mysterious to our hearers. It appears that, unlike strong imprecations, transition displays in our society are not particularly sex-typed. 7.2. THESPILL CRY. Here the central examples Oops! and Whoops!are phonetict ally well-formed, although not in every sense words. They are as much (perhaps even more) the practice of females as males. Spill cries are emitted to accompany our having, for a moment, lost guiding control of some feature of the world around us, including ourselves. Thus a woman, rapidly walking to a museum exit, passes the door, catches her mistake, utters Oops!, and backtracks to the right place. A man, dropping a piece of meat through the grill to coals below, uttersOops!, and then spears the meat to safety with his grill fork. On the face of it, the sound advertises our loss of control, raising the question of why we should want to defame ourselves through this publicity. An obvious possibility is that the Oops! defines the event as a mere accident, shows we know it has happened, and hopefully insulates it from the rest of our behavior-indicating that failure of control was not generated by some obscure intent unfamiliarto humanity, or some general defect in competence. Behind this possibility is another: the expression is presumably used for MINOR failings of environmental control. So, in the face of a more serious failure, Oops! has the effect of downplaying import, and hence implication as evidence of our incompetence. (It follows that, to show we take a mishap VERY seriously, we might feel constrained to omit the cry.) Another reason

802

LANGUAGE,

VOLUME 54, NUMBER

4 (1978)

for (and function of) spill-crying is that, since a specific vocalization is involved, we necessarily demonstrate that at least our vocal channel is functioning-and behind this, at least some presence of mind. A part of us proves to be organized and standing watch over the part of us that apparently isn't watchful. Finally, and significantly, the sound can provide a warning to others present that a piece of the world has gotten loose, and that they might best be advised to take care. Indeed, close observation shows that the oo in Oops! may be nicely prolonged to cover the period of time during which that something is out of control. Note that, when we utter Oops! as we slip on the ice, we can be making a plea to the closest other for a steadying hand, and simultaneously warning others to watch c osest steady ng hand s mu taneous ywarn ng c rcumstances sure y voca zat ons n out; these circumstances surely open up our surround for vocalizations. When in fact there is no danger to oneself, we may respond to ANOTHER'S s ANOTHER S oss onese f momentary loss of contro w th Oops!a so-prov d ng h m warn ng control with an Oops!also-providing him a warning that he is in trouble, a readied s n troub e read ed framework within which he can define the mishap, and a collectively established w th n wh ch def ne m shap co ect ve y estab shed cadence for his anticipated response. That some sort of help for others is thus h s ant c pated response s he p intended seems to be borne out by the fact that men seem more likely to oops for ntended ke y another when that other is a child or a female, and thus definable as someone for s ch d def nab e fema e whom responsibility can be taken. Indeed, when a parent plucks up a toddler and taken Indeed todd er respons b ty p ucks t n rap d y sh fts t rapidly shifts it from one point to another, or 'playfully' swings or tosses it in the po nt another p ayfu y sw ngs ch d s ar air, the prime mover may utter an Oopsadaisy!-stretched out to cover the child's pr me Oopsada sy!-stretched per od ground essness counter-act ng ts fee ng be ng period of groundlessness, counter-acting its feeling of being out of control, and at contro the same time instructing the child in the terminology and role of spill cries. In any t me nstruct ngthe ch d n ro e sp cr es term no ogy surv va va ue case t s case, it is apparent that oopsing is an adaptive practice with some survival value. oops ng s adapt ve pract ce w th And the fact that individuals prove (when the occasion does arise) to have been nd v dua s occas on ar se) a a ong ready all along to oops for themselves, or for an appropriate other, suggests that themse ves appropr ate other when nothing eventful is occurring, persons in one another's presence are still n another s st noth ng eventfu s occurr ng nonethe ess track ng nonetheless tracking one another and acting so as to make themselves trackable. themse ves trackab e act ng
73 7.3. THE THREAT cr es sex STARTLE STARTLE, ab y Eek! or Yipe! These response cries are sexno notably Eek Y pe

east be eved) fem n ne use Surpr seand stated- n ay typed (or at least so believed) for feminine use. Surprise and fear are stated-in lay terms expressed terms, 'expressed'. But the surprise or fear are very much under control-indeed, surpr se contro - ndeed about wa k noth ng nothing to be really concerned about. A very high open stairwell, or a walk that rea y h gh sta rwe leads to a precipice, can routinely evoke yipes from us as we survey what might havne eads prec p ce rout ne y y pes surveywhat m ght been our doom, but from a position of support; we have had ample time to secure doom pos t on amp e t me ourse ves ourselves. A notion of what a fear response would be is used as a pattern for not on wou d s actua m m cry mimicry. A sort of overplaying occurs that covers any actual concern by extending, overp ay ngoccurs extend ng w th obv ous unser ousness with obvious unseriousness, the expressed form which this concern would take. We wou d take expressedform wh ch th s demonstrate that we are alive to the fearsome implications of the event, but not a ve mp cat ons event overwhe med overwhelmed by them-that we have seen the trouble and by implication will troub e mp cat on w n th s re eases assured y contro assuredly control for it, and are therefore in need of no warning; all of this releases t warn ng; a others from closely tracking us. The moment it takes to say the sound is one we can t s c ose y track ng us use to compose ourselves. In a very subtle way, then, a verbal 'expression' of our ourse ves subt e way then verba express on state is a means of rising above it-and a release of concern now no longer necess t-and re ease r s ng onger s rea y over sary com ng sary, coming after the emergency is really over. Here an argument made earlier about multiple transformations can be taken up. ear erabout mu t p e transformat onscan up nd v dua Prec p tous Precipitous drops are the sorts of things to which an individual can be very close th ngs wh ch c ose

RESPONSE CRIES

803

without the slightest danger of falling over, or intent to do so. In these circumstances, it would seem that imagery of accident would come to the fore, or at least be very readily available. It is this easily-achieved mental set that the threat startle seems to participate in. Thus the uncompelling character of the actual circumstances can be nicely reflected in the light and almost relaxed character of the cry. One has, then, a warning-LIKE circumstances. Ritualization begins to give signal in dangerous-LIKE way to a copy of itself-a playful version of what is already a formalized version, a display that has been retransformed and reset, a second-order ritualization.
7.4. REVULSION SOUNDS,such as Eeuw!, are heard from a person who has by

necessity or inadvertence come in contact with something contaminating. Females in our society, being defined as more vulnerable in this way than males, might seem to have a special claim on the expression. Often, once we make the sound, we can be excused for a moment while decontamination is attempted. At other times, our voice performs what our physical behavior can't, as when our hands must keep busy cleaning a fish, leaving only the auditory and other unrequired channels to correct the picture-to show that indelicate, dirty work need not define the person who is besmeared by it. Observe again that there is an unserious note, a hint of 'hyperritualization': often the contamination that calls forth an Eeuw! is not REALLY believed to contaminate. Perhaps only germ contamination retains that literal power in our secular world. So again, a protective-like cry is uttered in response to a contaminating-like contact. 8. So far, response crying has been largely considered as something available to someone who is present to others, but not 'with' any of them. If one picks accompanied individuals, not singles, the behavior is still to be found; indeed, response crying is, if anything, encouraged. So also, response cries are commonly made by
persons in an OPENSTATE TALK,persons having the right but not the obligation OF

to address remarks to the other participants; this is a condition that commonly prevails among individuals jointly engaged in a common task (or even similarly engaged in like ones) when this work situates them in immediate reach of one another. Examples follow.
8.1. THE STRAINGRUNT. Lifting or pushing something heavy, or wielding a

sledgehammer with all our might, we emit a grunt attesting the presumed peak and consummation of our fully-extended exertion. The sound seems to serve as a warning that, at the moment, nothing else can claim our concern-and sometimes as a reminder that others should stand clear. No doubt the cry also serves as a means by which joint efforts can be temporallycoordinated, as is said to be true of work songs. Observe that these sounds are felt to be entirely unintended, even though the glottis must be partially closed off to produce them, and presumably could be fully opened or closed to avoid doing so. In any case, it could be argued that the expression of ultimate exertion these sounds provide may be essentially overstated. I might add that strain grunts are routinely guyed, employed in what is to be taken as an unserious way-often as a cover for a task that is reckoned as undemanding but may indeed require some exertion: another case of retransformation. Note too that strain grunts are employed during solitary doings that can

804

LANGUAGE,

VOLUME 54, NUMBER

4 (1978)

be construed as involving a peaking of effort. The rise and falling away of effort r se nvo v ng peak ng effort fa ng contoured in sound dramatizes our acts, filling out the setting with their execution. n dramat zesour acts f ng sett ng w th the r execut on I suppose the common example is the vocal accompaniment we sometimes provide voca accompan ment somet mes prov de examp e s ourse ves ourselves on passing a hard stool. stoo pass ng 8 2 THEPAINCRY 8.2. THEPAINCRY,Oww! or Ouch!l4 The functioning of this exclamation is Ouch! 4 th s exc amat on s funct on ng rather clear. Ensconced in a dentist's chair, we use a pain cry as a warning that the n dent st s cha r c ear pa n warn ng dr drill has begun to hurt. When a finger is firmly held by a nurse, we ouch when the hurt ouchwhen nurse f nger s f rm y he d need e prob ng needle probing for a sliver goes too deep. Plainly, the cry in these cases can serve as s ver n deep P a n y a self-regulated indicator of what is happening-providing a reading for the instis happen ng-prov d ng read ng nst se f-regu ated nd cator otherw se nformat on needed gator of the pain, who might not otherwise have access to the information needed. pa n m ght The meaning, then, may not be 'I have been hurt', but rather, 'You are just now I mean ng then hurt rather You me Th s mean ng nc denta y a so com ng coming to hurt me.' This meaning, incidentally, may also be true of the response that a dog or cat gives us when we have begun accidentally to step on its tail, ts ta g ves acc denta y seemsto ate a though although THAT often seems to come too late. In any case, these are good examples case cry examp es of how closely a vocalizer can collaborate with another person in the situation. co aborate w th n s tuat on c ose y voca zer 83 8.3. THE Th This subvoca track ng SEXUAL MOAN MOAN. s subvocal tracking of the course of sexually climactic sexua y c mact c exper ence d sp ay ava ab e experience, a display available to both sexes, is said to be increasingly fashionable sexes s sa d ncreas ng y fash onab e for females-among whom, of course, the sound tracing can be strategically course fema es-among whom trac ng strateg ca y de neate dea deve opment n emp oyed employed to delineate an ideal development in the marked absence of anything anyth ng like the reality. ke rea ty 84 8.4. FLOOR CUES CUES. worker in a typing pool makes a mistake on a clean copy and A n typ ng poo m stake c ean em ts emits an imprecation; this leads to, and apparently is designed to lead to, a colmprecat on; th s eads to apparent y s des gned ead to co league's query as to what went wrong. A fully-communicated statement of disgust eague s wrong fu y-commun cated d sgust and displease can then be introduced, but now ostensibly as a reply to a request for d sp ease ntroduced ostens b y rep y requestfor information. A husband reading the evening paper suddenly brays out a laugh or a nformat on read ng even ng sudden y augh GoodGod! therebycaus ng h s w fe or ent Good God!,thereby causing his wife to orient her listening, or even to ease the transisten ng trans t on nto ta k tion into talk by asking what it is. (A middle-class WIFE ght be less successful in ask ng m dd e-c ass t s m might ess successfu n hav ng having her floor cues picked up.) Wanting to avoid being thought self-centered, inf oor p cked up ) Want ng avo d be ng se f-centered ntrus ve garru ous trusive, garrulous, or whatever-and consequently feeling uneasy about making an consequent y fee ng mak ng open request for a hearing in the particularcircumstances-we act so as to encourage requestfor hear ng n part cu arc rcumstances-we our putative listeners to make the initial move, inviting us to let them in on what we putat ve steners n t a move nv t ng et n areexper enc ng Interest ng y a though n are experiencing.Interestingly,although in our society marriedcouples may routinely soc etymarr edcoup es mayrout ne y breach many of the standard situational proprieties when alone together-this s tuat ona propr et es a one together-th s mark ngthe gradua extens on symmetr ca r tua cense marking the gradual extension of symmetricalritual license between them-the rule ru e aga nstpers st ng n pub c se f-ta kmay reta ned w th against persistingin public self-talk may be retained, with the incidental consequence nc denta that the couple can continue to use response crying as a floor cue. coup e cont nue cry ng f oor cue 85 8.5. AUDIBLE LEE lower middle-class adolescent girl, sitting with four friends G GLEE. ower m dd e-c assado escent g r s tt ng w th A fr ends at a table in a crowded creperie, is brought her order, a large crepe covered with tab e n creper e s order arge w th ice cream and nuts. As the dish is set before her, she is transfixed for a moment, ce nuts d sh s her s transf xedfor moment
14 Solitarily experiencinga bout of intense pain, we sometimes follow its course with a So ar yexper enc nga bou o n ensepa n some mes o ow s w h halfha moaned ha grun edsound rac ng moaned, half-gruntedsound-tracing,as though casting the experiencein a sort of dialogic form houghcas ng he exper ence n sor o d a og c orm werea were a way to get throughthe moment and to maintainmorale.We sometimes also employ such o ge hrough he momen o ma n a nmora e some mesa so emp oysuch sound rac ngswhen w nessesare perce ved ypresen produc ng n hese c rcums ancesa rea sound-tracingswhen witnessesare perceivedlypresent, producing in these circumstancesa real scene s opper mp y ng ha scene-stopper-implying that our current, inner, acutely-painfulstate is the business everyone curren nner acu e y pa n u s a e s he bus ness shou d hang ng on should be hanging on.

RESPONSE CRIES

805

and wonder and pleasure escape with an Oooooo! In a casino, an elderly woman playing the slots alongside a friend hits a twenty-dollarpay-off, and above the sound of silver dropping in her tray peeps out a Wheee!Tarzan, besting a lion, roars out a Hollywood version of the human version of a lay version of a mammalian triumph call. 9. It is important, I believe, to examine the functioning of response cries when the crier is a ratifiedparticipant of on-going talk-being a participant of a conversational social encounter, as opposed to a task-structured one. While walking along talking to a friend, we can, tripping, unceremoniously interrupt our words to utter Oops!, even as the hand of our friend comes out to support us; as soon as this little flurry has passed, we revert back to our talk. All that this reveals, of course, is that when we are present to others as a fellow conversationalist, we are also present to them-as well as to all others in the situation-as fellow members of the gathering. The conversational role (short of what the phone allows) can never be the only accessible one in which we are active. So response cries can function in work encounters, and can obtrude into conversational ones. Now we move on to a closer issue. If these responses are to be seen as ritualized expressions-and some as standardized vocal comments on circumstances that are not, or are no longer, beyond our emotional and physical controlthen there is reason to expect that such cries will be used at still-further remove, namely in response to a VERBALLYPRESENTEDreview of something settled long ago, at a place quite removed. A broker tells a client over the phone that his stock has dropped; the client, well socialized in this sort of thing, says Yipe! orEek! (Jack Benny made a specialty of this response cry.) A plumber tells us what our bill will be, and we say Ouch! Indeed, response cries are often employed thrice-removed from the crisis to which they are supposed to be a blurted response: a friend tells us about something startling and costly that happened to him, and at the point of disclosure we utter a response cry-on his behalf, as it were, out of sympathetic identification and as a sign that we are fully following his exposition. In fact, we may offer a response cry when he recounts something that happened to someone ELSE. In these cases, we are certainly far removed from the exigent event being replayed, and just as far removed from its consequences, including any question of having to take immediate rescuing action. Interestingly, there are some cries which seem to occur more commonly in our response to another's fate as it is recounted to us (good or bad), than they do in our response to our own. Oh, wow! is an example. We can play all these response games because our choice of vocalization allows the recipient, or rather hearer, to treat the sound as something to which a specific spoken reply is not required. To the plumber, we are precisely NOT saying: 'Does the bill have to be that high?'-such a statement would require a reply, to the possible embarrassment of all. Having started with response cries in the street, our topic has moved into the shelter of conversations. But it should not be assumed from this that the behaviors in question-response cries-have somehow been transmuted into full-fledged creatures of discourse. That is not the way they function. These cries are conventionalized utterances which are specialized for an informative role; but in the

806

LANGUAGE, VOLUME 54, NUMBER 4 (1978)

statements Obv ous y nformat on linguistic and propositional sense, they are not statements. Obviously, information ngu st c propos t ona sense cr es n is provided when we utter response cries in the presence of others, whether or not s prov ded others s ta k t me we are in a state of talk at the time. That is about the only reason we utter them in n n on y the first place, and the reason they are worth studying. But to understandhow these understandhow f rstp ace study ng sounds function in social situations, particularly during talk, one must first underfunct on n soc a s tuat ons part cu ar ydur ng ta k f rst wh ch stand the source of the prototypes of which they are designed to be recognizable des gned recogn zab e vers ons versions. What comes to be made of a particular individual's show of 'natural natura part cu ar nd v dua s occas on s emot ona express on emotional expression' on any occasion is a considerably awesome thing, not cons derab y th ng ex stence natura emot ona express ons dependent on the existence anywhere of natural emotional expressions. But whatts ever is made of such an act by its maker and its witnesses is different from what is ts w tnesses s d fferentfrom s s made of openly-designed and openly-directed communication. open y-des gned open y-d rected commun cat on t 10 10. At the beginning of this paper it was argued that extended self-talk, if disse f-ta k f d sbeg nn ng th s t ref ects bad y reflects badly on the talker. Then it was observed that elements in the ta ker e ements n covered covered, s tuat on situation can considerably mitigate the improprietyof talking to ourselvespubliclycons derab ym t gate mpropr ety ta k ng ourse vespub c yand that, in any case, we are prepared to breach the injunction against public selfthat n case njunct on aga nst pub c se fta k when n effect susta n th s part cu arpropr ety wou d talk when, in effect, to sustain this particularpropriety would go even harder on our harderon Much the same position could be taken with respect to interjected cou d w th reputat ons reputations. pos t on nterjected h tch n cases imprecations. In both cases, one can point to some hitch in the well-managed flow mprecat ons po nt we -managed f ow of controlled events, and the quick application of an ostensibly self-directed procontro ed events qu ck app cat on ostens b y se f-d rected nouncement to establish evidence-a veneer-of control, poise, and competency. estab sh ev dence-a contro po se competency A though Although response cries do not, on the surface, involve words uttered even to cr es not surface nvo ve onese f-be ng oneself-being IN PROTOTYPEmerely a matter of non-symbolic emotional expresmere y non-symbo c emot ona s on-they apparent ycome funct on sion-they apparentlycome to function as means of strikinga self-defensibleposture str k nga se f-defens b eposture in the face of extraordinary events, much as does exposed self-talk. However, one n extraord naryevents se f-ta k However rout ne routine source of trouble in the management of the world is, interestingly enough, troub e n wor d s nterest ng yenough the management of talk itself. So again we have response cries, but this time ones ta k tse f aga n cr es th s t me that are constantly uttered. constant y uttered F rst First, there is the well-known filled pause (usually written ah or uh or urn) s we -known f ed (usua y wr tten emp oyed employed by speakers when they have lost their places, can't find a word, are ost the r p aces can t f nd word momentar y d stracted otherw sef nd momentarily distracted, or otherwise find they are departing from fluently-sustained depart ngfrom f uent y-susta ned speech speech. Response CRIESseems an awkwardterm for such unblurtedsubvocalizations; seemsan awkwardtermfor suchunb urtedsubvoca zat ons; but they do, I think, function like response cries, if only in that they facilitate do th nk funct on ke cr es f on y n fac tate track ng tracking. In effect, speakers make it evident that, although they do not now have the effect speakersmake t ev dent that a though word or phrase they want, they are giving their attention to the matter and have not phrasethey want g v ng the r attent on matterand cut themselves adrift from the effort at hand. A word search, invisible and inaudible effortat hand search nv s b eand naud b e themse ves adr ft in itself, is thus voluntarily accompanied by a sound shadow-a sound, incidentally, n tse f s vo untar yaccompan ed sound nc denta y that could easily be withheld merely by otherwise managing the larynx-all to the cou d eas y w thhe d mere y otherw se manag ng arynx-a end of assuring that something worse than a temporary loss of words has not assur ng someth ng oss happened happened, and incidentally holding the speaker's claim to the floor.15(Interestingly, nc denta yho d ng speaker sc a m f oor 15(Interest ng y in radio broadcasting, where visual facial signs of a word search can't be effective, n rad o broadcast ng v sua fac a s gns can t effect ve
15 A case can be made that, in some English-speaking ha n Eng sh speak ng circles, the familiar hesitation markers c rc es he am ar hes a on are systematically employed in slightly different ways. For example, uh might be heard when the sys ema ca y emp oyed n s gh y d eren ways examp e m gh he speaker had forgotten a proper name; oh might occur when he knew a series of facts, but was orgo en name m gh ser es o ac s bu trying to decide which of them could be appropriately cited or best described for the hearers. ry ng o dec de wh ch o hem cou d appropr a e y c ed bes descr bed or he hearers The unfilled or silent pause participates in this specialization-giving un ed s en par c pa es n h s spec a za on g v ng one reason, alas, to think reason a as o h nk o of it as a response cry, too. Here see the useful paper of James 1972. cry oo he use u o 1972

RESPONSE CRIES

807

the filling of pauses by a search sound or a prolongation of a vowel has much to recommend it: speakers are under obligation to confirm that nothing has gone wrong with the studio's equipment, as well as their own-the floor in this case being a radio station. If only inexperienced broadcastersfrequently employ filled pauses, it is because professionals can manage speech flow, especially reading aloud, without the hitches in encoding which, were they to occur, would equally give professionals reasons to ritualize evidence of what was occurring.) In addition to the filled-pause phenomenon, consider the very standard form of self-correctionwhich involves the breaking-offof a word or phrase that is apparently not the one we wanted, and our hammering home of a corrected version with increased loudness and tempo, as if to catch the error before it hit the ground and shattered the desired meaning. Here the effect is to show that we are very much alive to the way our words should have come out; we are somewhat shocked and surprised at our failure properly to encode an appropriate formulation the first time round, and the rapidity and force of the correct version seem to suggest how much on our toes we really are. We display our concern and the mobilization of our effort at the expense of smooth speech-production-electing to save a little of our reputation for presence of mind, over and against that for fluency. Again, as with filled pauses, one has what is ostensibly a bit of pure expression, i.e. a transmission providing direct evidence (not relayed through semantic reference) of the state of the transmitter,but now an expression that has been cut and polished into a standard shape to serve the reputational contingencies of its emitter. 11. Earlier it was suggested that imprecations were somewhat like truncated, self-addressed statements, but not wholly so. Later these lexicalized exclamations were shown to function not unlike response cries. Now we must try to decide where they belong. Suppose that someone brings you the news that he has failed in a task you have seriously set him. Your response to the news can be: I knew it! Did you have to? In the styling I have in mind, this turn at talk contains two moves and a change of 'footing': the first move (uttered half under the breath with the eyes turned upward) is a bit of self-talk, or something presented in that guise-the sort of open aside which adults are especially prone to employ in exasperated response to children, servants, foreigners, and other grades who easily qualify for moments of non-person treatment. The second move (Did you have to?) is conventionally directed communication. Observe that such a turn at talk will oblige its recipient to offer an apology or a counter-account, locking the participants into an interchange. But although the recipient of the initial two-move turn will be understood to have overheard the self-addressedsegment, he will have neither the right nor the obligation to reply to it specifically, at least in the sense that he does in regardto the conventionally communicated second portion. Now shift from extended self-talk to the truncated form-imprecation: Shit! Did you have to? Given the same histrionics, one again has a two-move turn, with a first move that can't be answered in a conventional way. If the respondent does address the remark to this blurted-out portion, it will be to the psychic state presumably indexed by it-much as when we comfort people who have burst into tears, or when

808

LANGUAGE, VOLUME 54, NUMBER 4 (1978)

we upbraid them for loss of self-control. Or the respondent may have to venture a frame ploy, attempting to counter a move by forcing its maker to change the interpretive conventions that apply to it-as in the snappy comeback Not here, injected immediately after the expletive. In all this, and in the fact that standard lexicalizations are employed, I knew it! and Shit! are similar. However, although I knew it! follows grammaticalconstraints for well-formed sentences, Shit! need not, even if one appeals to the context in order to see how it might be expanded into a statement. Shit! need no more elide a sentence than need a laugh, groan, sob, snicker, or giggle-all vocalizations that occur frequently, except in the utterances ordinarily presented for analysis by linguists. Nor, I think, does it help understanding very much to define Shit! as a well-formed sentence with NP! as its structure. Here, of course, imprecations are exactly like response cries. For it is the essence of response cries that they be presented as if mere expression-and not recipient-directed, proposition-like statements-were involved, at least on the face of it. Imprecations, then, might best be considered not as a form of self-talk at all, but rather as a type of response cry. Unlexicalized cries have come to be somewhat conventionalized, and imprecations have merely extended the tendency, further ritualizing ritualizations. Since religious life already sets aside a class of words to be treated with reserve and ranked with respect to severity, response crying has borrowed them-or so it would seem. Insofar as self-talk is structurally different from the normal kind, imprecatory utterances (like other response cries) are too, only more so. And because of this sharp, underlyingdifferencebetween conventionally directed statementsand imprecatory interjections, the two can be given radically different roles in the functioning of particular interaction systems; they serve close together, in complementary distribution and without confusion. Consider tennis: during the open state of talk sustained in such a game, a player who misses an 'easy' shot can response-cry an imprecation loudly enough for opponents and partner to hear. On the other hand, a player making a 'good' shot is not likely to be surprisedif an opponent offers a complimentary statement about him to him. (As these two forms of social control help frame his own play, so he will participate in the two forms that frame his opponents'.) But, of course, good taste forbids a player to address opponents in praise of his own efforts-just as they must allow him elbow room, and not reply directly to his cries of self-disgust. A player may, however, use directed, full-fledgedstatements to convey self-castigation and (when directed to his partner) apology. Response cries and directed statements here comprise a closely-working pair of practices, part of the ritual resources of a single interaction system. And their workings can be intermingled because of their structural difference,not in spite of it. Given this arrangement,it is understandable that a player will feel rather free to make a pass at ironically praising himself in statements made to opponents or partner, correctly sensing that his words could hardly be misframedas literal ones. (That he might employ this devicejust to induce others to communicate a mitigated view of his failure merely attests again to the various conveniences that can be made of forms of interaction.) Just as response cries can form a complementary resource with conventionally

RESPONSE CRIES

809

d rected statements directed statements, so they can with self-directed ones. For example, in casino w th se f-d rected ones examp e n cas no craps craps, a shooter has a right to preface a roll, especially a 'come out', with selfro espec a y come out w th se fr ght trad t ona k nd-d rected d ce encourag ng encouraging statements of a traditional kind-directed to the fates, the dice, or fates some other ethereal recipient; this grandstanding (as dignified gamblers call this etherea rec p ent; th s grandstand ng d gn f ed gamb ers ca th s se f-ta k) somet mes servesto br ng self-talk) sometimes serves to bring the other players into a cadence and peaking of p ayers nto peak ng attent on When short y attention. When, shortly, the shooter 'craps out', he is allowed a well-fleshed s a owed we -f eshed craps out d sso ut on tab e s coord nated nvo veimprecation, coincidental with the dissolution of the table's coordinated involvemprecat on co nc denta w th ment ment. So again we find complementarity and a division of labor, with self-talk f nd comp ementar ty d v s on abor w th se f-ta k aga n located where collective hope is to be built up, and imprecatory response cry where ocated co ect ve s bu t up mprecatoryresponse it is to be abandoned. t s abandoned 12 1 DISCUSSION 12.1. DISCUSSION. r tten versions of response cries seem to have a speechW Written vers ons cr es cr es-so that contam nat ng effect conso dat ng contaminating effect, consolidating and codifying actual response cries-so that, cod fy ng actua in many cases, reality begins to mimic artifice, as in Ugh!, Pant pant, Gulp, and n m m c art f ce n Ugh! cases rea ty beg ns pant Gu p Tsk tsk; this route to ritualization is presumably unavailable to non-human th s r tua zat on s presumab y unava ab e an ma s 16Th s animals.16This easy change is only to be expected: response cries themselves are by s on y cr es themse ves unser ous ess r tua zat ons a ready be ng way of being second-order ritualizations, already part of an unserious (or less than ser ous) doma n serious) domain. Here cartoons and comics are to be taken seriously. These printed pictures must com cs ser ous y pr ntedp ctures ent re scenar os sma present entire scenarios through a small number of 'panels' or frozen moments, pane s moments somet mes on y one sometimes only one. The cartoonist has great need, then, for expressions that will cartoon st w need then express ons c ear y clearly document the presumed inner state of his figures, and clearly display the nner h s f gures c ear y d sp ay po nt point of the action. Thus, if individuals in real life need response cries to clarify the act on Thus f nd v dua s n rea fe cr es c ar fy drama of their circumstances, cartoon figures need them even more. So we obtain the r c rcumstances f gures more obta n wr tten vers ons written versions of something that could be thought originally to have no set, someth ng cou d or g na y set wr ttenform Moreover written form. Moreover, cartoon figuresportrayedas alone must be portrayedacting f guresportrayedas a one portrayedact ng in such a way as to make their circumstancesand inner states available to the viewer n the r c rcumstancesand nner ava ab eto v ewer (much as real persons do when in the presence of others), and included in this rea n others) nc uded n th s s tuat ona - ke behav or situational-like behavior are response cries. (So also in the case of movies showing cr es a so n mov es show ng persons ostensibly alone.) In consequence, the practice of emitting response cries ostens b y a one ) consequence pract ce em tt ng cr es when alone is tacitly assumed to be normal, presumably with at least some cona one s tac t y norma presumab y w th east tam nat ng taminating effect upon actual behavior when alone. actua behav or a one 12 2 12.2. A point might be made about the utterances used in response cries. As po nt m ght n cr es suggested suggested, they seem to be drawn from two sources: taboo but full-fledged words fu -f edged ( nvo v ng b asphemy and- n Eng sh-Ang o-Saxon (involving blasphemy and-in English-Anglo-Saxon terms for body functions) funct ons) and from the broad class of non-word vocalizations ('vocal segregates', to employ c ass voca zat ons ( voca segregates emp oy Trager sterm Trager's term, 1958:1-12)-of which response cries are one, but only one, variety. wh ch cr es one on y one var ety There is a nice division of linguistic labor here. Full-fledged words that are s n ce d v s on ngu st c abor here Fu -f edged soc a y acceptab e socially acceptable are allocated to communication in the openly directed sense, a ocated commun cat on n open y d rected sense wh e while taboo words and non-words are specialized for the more ritualized kind of spec a zed r tua zed k nd
16 The carry-back from the written to the spoken form is especially marked in the carry back rom he wr en o he orm s espec a y n he matter of ma er o punc ua on marks or punctuation marks, for here writing has something that speaking hasn't. Commonly used wr ng some h ng ha speak ng hasn Common y lexicalizations are: underline, footnote, period, question mark, quotes, and parenthetically. Written ex ca za ons are under ne oo no e per od ques on mark quo es paren he ca y Wr en abbrev a ons such abbreviations (such as British p for pence) also enter the spoken domain. Moreover, there is a Br sh or pence a so en er he doma n Moreover here s carry back o he carry-back to the spoken form of the pictorial-orthographic form of the presumed approximated orm o he p c or a or hograph c orm o he approx ma ed sound e ec s o sound-effects of an action: Pow!, Barn! are examples. ac on Pow Barn examp es

810

LANGUAGE, VOLUME 54, NUMBER 4 (1978)

commun cat on communication. In brief, the character of the word bears the mark of the use that is characterof bearsthe s br ef dest ned destined for it; and we have a case of complementary distribution on a grand scale. t; sca e comp ementary d str but on Non-words as a class are not productive in the linguistic sense, their role as c ass product ve n ngu st c sense the r ro e evo ved them (Th s s interjections being one of the few that has evolved for them. (This is not to say that nterject ons be ng a particular vocal segregate can't have a very lively career, quickly spreading from can t part cu ar voca ve y career qu ck y spread ngfrom one segment of a language community to others; the response cry Wow!is a recent s anguage commun ty examp e ) example.) Many taboo words, however, are quite productive, especially in the qu te product ve espec a y n words however trad t on ma nta ned n certa n subcu tures tradition maintained in certain subcultures,where some of these words occur (if not (f funct on) n a most function) in almost every syntactic position.'7 Furthermore,curse words are drawn syntact c pos t on 7 Furthermore from familiar scales of such words, and choice will sharply reflect (in the sense of fam ar sca es cho ce w sharp y ref ect ( n words d sp ay negot ate etc ) display, negotiate etc.) the terms of the relationship between speaker and hearer; re at onsh p non-words don't function very effectively in this way. don t funct on effect ve y n th s way Note that non-words can't quite be called part of a language. For example, there can t qu te ca ed anguage examp e tends to be no canonical 'correct' spelling. When and where convention does begin canon ca correct spe ng convent on beg n to establish heavily a particular form and spelling, the term can continue to be estab sh heav y part cu ar cont nue spe ng ts users f wr tten vers on cont nue thought of as not a word by its users, as if any written version must continue to that a rough-and-readyeffort at transcriptionis at work. (I take it here that, t convey rough-and-readyeffort transcr pt on s work that in our society, a feature of what we think of as regular words is that we feel the n th nk s fee regu ar soc ety wr tten written form is as 'real' a version as the spoken.) Further, although we have s vers on rea spoken ) Further a though eff c ent efficient means of reporting another's use of an expletive (either literally or by report ng another s exp et ve (e ther tera y estab shed paraphrast c form) th s s established paraphrastic form), this is not the case with non-words. So, too, the w th non-words So too vo ced voiced and orthographic realizations of some of these constructions involve conorthograph c rea zat ons construct ons nvo ve sonant clusters that are phonotactically irregular; furthermore,their utterance can c usters phonotact ca y rregu ar;furthermore the r a ow allow the speaker to chase after the course of an action analogically with stretches, act on ana og ca y w th stretches g des turns glides, turns, and heights of pitch foreign to his ordinary speech. Yet the sound he ghts p tch fore gn h s ord nary speech that covers any particular non-word can stand by itself, is standardized within a part cu ar tse f s standard zed w th n g ven anguage commun ty given language community, and varies from one language community to another, var es anguage commun ty another in each case like full-fledged words.'8 And the non-words of a particular language n ke fu -f edgedwords 8 part cu ar anguage comp y w th comply with and introduce certain of the same phonotactic constraints as do ntroduce certa n phonotact c constra nts its regular words (Jefferson 1974:183-6). Interestingly, there is some evidence ts regu ar 1974:183-6) Interest ng y s ev dence that what one language community handles with a non-word, other language anguage commun ty hand es w th non-word anguage commun t es communities do too. too On the whole, then, non-word vocalizations might best be thought of as semiwho e then voca zat ons m ght sem words words. Observe that the characterization provided here (and by linguists) of these character zat onprov ded ngu sts) ha f-caste express ons half-caste expressions takes no note that some (such as Uh? and Shh!) are clearly Uh? c ear y part of directed speech, and are often interchangeable with a well-formed word d rected speech nterchangeab e w th we -formed (here What? and Hush!); but others (such as the uh as filled pause) belong to a f ed be ong rad ca y d fferent spec es radically different species of action-viz., putatively pure expression, response act on-v z putat ve y express on
17 Admittedly, even in these productive cases, taboo words are not entirely Adm ed y n hese produc ve cases aboo no en re y vulnerable to vu nerab e o syn ac c ana ys s Say ng ha he uck n sen ence ke Wha he uck syntactic analysis. Saying that the fuck in a sentence like What the fuck are you doing ? is adjecdo ng s ad ec tival in function, or that bloody in What are you bloody well doing? is an adverb, misses someva n unc on ha b oody n Wha b oody we do ng? s adverb m sses some thing of the point. Here specific syntactic location seems to be made into a convenience; h ng o he po n spec c syn ac c oca on o no conven ence somehow the intensifying word is meant to color uniformly the whole of the utterance in which he n ens y ng s mean o co or un orm y he who e o he u erance n wh ch it occurs (cf. Quang Phuc Dong 1971). 1971 c 18 Quine (1959:6) has an example: '"'Ouch" Qu ne 1959 6 examp e " Ouch" is not independent of social training. One need s no ndependen o soc a ra n ng on y o pr ck only to prick a foreigner to appreciate that it is an English word.' ore gner o apprec a e ha s Eng sh word

RESPONSE CRIES

811

crying. (Imprecations and some other well-formed interjections provide an even more extreme case, for exactly the same such word may sometimes serve as an ostensibly undirected cry, but at other times be integrated directly into a recipientdirected sentence under a single intonation contour.) Here again, one can see a surface similarity covering a deep underlying difference,but not the kind ordinarily addressed by transformationalists. Apart from qualifying as semi-words, response cries can be identified in another way, namely as articulated free-standing examples of the large class of presumed 'natural expressions' or signs meant to be taken to index directly the state of the transmitter-some of which, like voice qualifiers, can paralinguisticallyride roughshod across natural syntactic units of speech. Although gender differences in the basic semantic features of speech seem not very marked in our society, response cries and other paralinguistic features of communication are. Indeed, speech AS A WHOLE might not be a useful base to employ in considering gender differences, since it cancels sharp contrasts revealable in special components of discourse. 12.3. Earlier, I suggested that a response cry can draw on the cooperation of listeners-requiring that they hear and understand the cry, but act as though it had not been uttered in their hearing. It is in this way that such a form of behavior, ostensibly not designed for directed linguistic communication at all, can be injected into public life-in certain cases, even into conversations and broadcasts. In brief, a form of response perceived as native to one set of circumstancesis set into another. In the case of blasphemous cries, what is inserted is already something that has been borrowed from another realm, semantic communication; so the behavior can be said to have been returned to its natural place, but now so much transformed as to be little like a native. This structural reflexivity is, I believe, a fundamental fact of our communicative life. What is ritualized here, in the last analysis, is not an expression, but a self-other alignment-an interactional arrangement. Nor, as earlier suggested, is that the bottom of embedding. For example, when a speaker finds he has skated rather close to the edge of discretion or tact, he may give belated recognition to where his words have gone-marking a halt by utteringa plaintive Oops!,meant to evoke the image of someone who has need of this particular response cry, the whole enactment having an unserious, openly theatrical character.Similarly,in the face of another's reminder that we have failed in fulfilling some obligation, we can utter Darn it in an openly false manner-as a taunting, even insolent, denial of the imprecation we might normally be expected to employ. In brief, what is placed into the directed discourse in such cases is not a response cry, but a mocked-up individual uttering a mocked-up response cry. (All of this is especially evident when the cry itself is a spoken version of the written version of the cry, as when a listenerresponds to the telling of another's near-disaster by ungulpingly uttering the word Gulp.) So, too, the filled pause uh, presumably a self-expression designed to allow hearers to track speaker's engagement in relevant (albeit silent) production work, can apparently be employed with malice aforethought to show that the word that does follow (and is ostensibly the one wanted all along) is to be heard as one which the speaker might not naturally use (Jefferson, 192-4). In this case a 'correction format' has been used as a

812

LANGUAGE, VOLUME 54, NUMBER 4 (1978)

convenience, its work set into an environment for which it was not originally designed. Similarly, on discovering that he has said April the 21st instead of May the 21st, an announcer may (as one type of remedial work) repeat the error immediately, this time with a quizzical, speaking-to-oneself tone of voice, as though this sort of error were enough of a rarityto cause him to break frame; but this response itself he may try to guy, satirizing self-talk (and self-talkers) even as he engages in it, the transformation confirmed by the little laugh he gives thereafter, to mark the end to error-making AND playful correction. The moral is that what is sometimes put into a sentence may first have to be analysed as something that could not occur naturally in such a setting, just as a solitary's self-comments may first have to be analysed as something found exclusively in social intercourse. And the transformationswhich these alien bits of speech undergo, when set into their new milieu, speak as much to the competence of ethologists as grammarians. A turn at talk that contains a directed statement ANDa segment of self-talk (or an imprecation or a non-lexicalized response cry) does not merely involve two different
moves, but MOVESOF TWO DIFFERENTORDERS.This is very clear, e.g., when someone

in or out of a conversation finds cause to blurt out Shit!-and then, in apparent embarrassment,quickly adds Excuse me, sometimes specificallydirectingthe apology to the person most likely to have been offended. Here, patently, the first move is an exposed response cry; the second is a directed message whose implied referent happens to be the first. The two moves fit together nicely-indeed, some speakers essay an imprecation knowing that they will have a directed apology to compensate for it; but this fit pertains to how the two moves function as an action-response pair, self-contained within a single turn at talk, and not to any ultimate commonality of form. So, too, when an announcer coughs ratherloudly, says Excuse me with greater urgency of tone than he likes, and then follows with a well-designed giggle: here we have a three-move sequence of sounded interference, directed statement, and response cry-the second move a comment on the first, and the third move a comment on the second move's comment. Any effort to analyse such strips of talk linguistically by trying to uncover a single deep structure that accounts for the surface sequence of words is destined to obscure the very archeological issues which the generative approach was designed to develop. A blender makes a mush of apples and oranges; a student shouldn't. A student shouldn't, even when there is no obvious segmentation to help with the sorting. For now it is to be admitted that through the WAY we say something that is part of our avowedly directed discourse we can speak-ostensibly at least-for our own benefit at the same time, displaying our self-directed (and/or non-directed) response to what is occurring. We thereby simultaneously cast an officially-intended recipient of our proposition-like avowals as an overhearerof our self-talk. The issue is not merely that of the difference between what is said and what is meant-i.e. the issue of implicature; rather, the issue is that one stream of information is conveyed as avowedly-intended verbal communication, while the other is simultaneously conveyed through a structural ruse, i.e. our allowing witnesses a glimpse into the dealings we are having with ourselves. It is in this way that one can account for the apparently anomalous character of imprecations of the Fuck you! form. It might

RESPONSE CRIES

813

appear as if one person were making a directed verbal avowal to another by means of an imperative statement with deleted subject. But in fact the format is restricted to a relatively small list of expletives, such as screw; and none qualify as ordinary verbs, being constrained in regard to embedded and conjoined forms in ways that standard verbs in the elided imperative form are not (cf. Quang Phuc Dong). Nor is this analysis of the unconversational aspects of certain conversational utterances meant to deny the traditional concept of transformation and embedding; rather, the power of the latter is displayed. Waiting with her husband and a friend for the casino cashier to count down her bucket of silver, a happy player says, And when I saw the third 7 come up and stop, I just let out 'Eeeee'. Here, through direct quotation, a speaker brings to well-circumscribedthree-persontalk what was, a few minutes before, the broadly accessible eruption of a single. This shows clearly that what starts out as a response cry (or, for that matter, as any sounded occurrence-human, animal, or inanimate) can be conversationally replayed-can be reset into ordinary discourse through the unlimited power of sound mimicry. The 13. CONCLUSION. public utterance of self-talk, imprecations, and response cries constitutes a special variety of impulsive, blurted actions-namely, vocalized ones. Our tacit theory of human nature recommends that these actions are 'purely expressive', 'primitive', or 'unsocialized', violating in some way the self-control and self-possession we are expected to maintain in the presence of others, providing witnesses with a momentary glimpse behind our masks. However, the point about these blurtings is not that they are particularly 'expressive'. Obviously, in this sense of that word, ordinary talk is necessarily expressive, too. Naked feelings can agitate a paragraph of discourse almost as well as they can a solitary imprecation. Indeed, it is impossible to utter a sentence without coloring the utterance with some kind of perceivable affect-even if (in special cases) only with the emotionally distinctive aura of affectlessness. Nor is the point about segmented blurtings that they are particularly unsocialized, for obviously they come to us as our language does, not from our own invention. Their point lies elsewhere. One must look to the light these ventings provide, not to the heat they dispel. In every society, one can contrast occasions and moments for silence with occasions and moments for talk. In our own, one can go on to say that by and large (and especially among the unacquainted) silence is the norm, and talk something for which warrant must be present. Silence, after all, is very often the deference we owe in a social situation to others present. In holding our tongue, we give evidence that such thought as we are giving to our own concerns is not presumed by us to be of any moment to the others present, and that the feelings which these concerns invoke in ourselves are owed no sympathy. Without such enjoined modesty, there could be no public life, but only a babble of childish adults pulling at one another's sleeves for attention. The mother to whom we would be saying Look, no handscould not look or reply, for she would be saying Look, no hands to someone else. Talk, however, presumes that our thoughts and concerns will have some relevance, interest, or weight for others; and in this we can hardly avoid presuming a little. Talk, of course, in binding others to us, can also do so for protracted periods of

814

LANGUAGE,VOLUME54, NUMBER4 (1978)

time. The compensation is that we can sharplyrestrictthis demand to a small portion of those present-indeed, often to only one. The fugitive communications I have been considering constitute a third possibility-minor, no doubt, but of some significanceif only because of what they tell us about silence and talk. Our blurtings make a claim of sorts upon the attention of everyone in the social situation-a claim that our inner concerns should be theirs, too; but unlike the claim made by talk, ours here is only for a limited period of attention. Simply put, this invitation into our interiors tends to be made only when it will be easy for other persons present to see where the voyage takes them. What is precipitous about these expressions, then, is not the way they are emitted, but the circumstances which render their occurrence acceptable. The invitation we are free to extend in these situations we would be insane to extend in others. Just as most public arrangements oblige and induce us to be silent, and many other arrangementsto talk, so a third set allows and obliges us momentarily to open up our thoughts and feelings and ourselves, through sound, to whoever is present. Response cries, then, do not mark a flooding of emotion outward, but a flooding of relevance in. There is linguistic point to the consideration of this genre of behavior. Response cries such asEek! might be seen as peripheralto the linguist's domain; but imprecations and self-talk are more germane, passing beyond semi-word vocal segregatesto the traditional materials of linguistic analysis. The point is that all three forms of this blurted vocalization-semi-word response cries, imprecations, and self-talkare creatures of social situations, not states of talk. A closed circle of ratified participants oriented to engaging exclusively with one another in avowedly-directed communications is not the base; a gathering, with its variously-oriented, ofttimes silent and unacquainted members, is. Further, all three varieties of ejaculatory expression are conventionalized as to form, occasion of occurrence, and social function. Finally, these utterances are too commonly met with in daily life, surely, to justify scholarly neglect. Once we recognize that there is a set of conventionalized expressions that must be referred to social situations, not conversations-i.e. once we appreciate that there are communications specifically designed for use outside states of talk-then it is but a step to seeing that ritualized versions of these expressions may themselves be embedded in the conventionally-directedtalk to be found in standardconversational encounters. Appreciating this, we can then go on to see that, even though these interjections come to be employed in conversational environments, they cannot be adequately analysed there without reference to their original functioning outside states of talk. It is recommended, then, that linguists broaden their net, to bring in uttering that is not talking, and to deal with social situations-not merely with jointly sustained talk. Incidentally, linguists might then be better able to countenance inroads that others can be expected to make into their conventional domain; for I believe that talk itself is intimately regulated and closely geared to its context through non-vocal gestures which are very differently distributed than the particular language and subcodes employed by any set of participants-although just where these boundaries of gesture-use ARE to be drawn remains an unstudied question.

RESPONSE CRIES REFERENCES

815

and WILLIAM 1976. Social-ecological constraints on CORSARO. COOK-GUMPERZ, JENNY, children's communicative strategies. Papers on language and context, by Jenny Cook-Gumperz & John Gumperz (Working paper 46). Berkeley: Language Behavior Research Laboratory, University of California. 1977. No, there, more, and allgone: Why the first words aren't about ALISON. GOPNIK, things. Nottingham Linguistic Circular 6.15-20. JAMES,DEBORAH.1972. Some aspects of the syntax and semantics of interjections. Papers from the 8th Regional Meeting, Chicago Linguistic Society, 162-72. JEFFERSON,GAIL. 1974. Error correction as an interactional resource. Language in Society 3.181-200. PIAGET,JEAN. 1956. Le langage et la pensee chez l'enfant. 4th ed. Neuchatel. [Trans. by Marjorie Gabain. New York: Meridian, 1957.] --. 1962. Comments on Vygotsky's critical remarks concerning The language and thought of the child, and Judgment and reasoning in the child. (Appendix to Vygotsky, 1962 edition.) Cambridge: MIT Press. PHUCDONG.1971. English sentences without overt grammatical subject. Studies QUANG out in left field, ed. by A. M. Zwicky et al., 3-9. Edmonton: Linguistic Research, Inc. QUINE,W. V. 1959. Word and object. New York: Wiley. SUDNOW, DAVID. 1967. Passing on: the social organization of dying. Englewood Cliffs, NJ: Prentice-Hall. TRAGER, GEORGE 1958. Paralanguage: a first approximation. Studies in Linguistics L. 13.1-12. LEV VYGOTSKY, SEMENOVICH. Thought and language. Trans. by Eugenia Hanfmann 1962. & Gertrude Vakar. New York: Wiley. [Received 3 October 1977.]