A change-of-state token and aspects of its sequential placement

Structures of Social Action Studies in Conversation Analysis 13. A change-of-state token and aspects of its sequential placement JOHN HERITAGE Edited by J. Maxwell Atkinson and John Heritage University of Wanvick 1. Introduction This chapter reports some preliminary findings on the work accomplished by the particle "oh" in natural conversation. Evidence from the placement of the particle in a range of conversational sequences shows that the particle is used to propose that its producer has undergone some kind of change in his or her locally current state of knowledge, information, orientation or awareness. Such a proposal is, in various sequence specific ways, informative for other participants and is implemented in, or accomplice to, the achievement of a variety of interactional tasks. 1 A preliminary sense of the way "oh" can be used to propose some kind of change of state is readily available from such fragments as the following. (1) 1l11 riflll tJ/tltll Unlursity of Cambrid111 ltlprlftt rmJ srlf 11/1 malfffllr of hooks WQ.f ftDrtltd by Htrtry V11ll11 1.5J4. Tlr11 U,inrsity .\a.r prlntNI (11ft/ publislt11d rontlftllt:I141Y -+ N: [GJ:.FNl ((three people are walking together: someone passes than wearing a photograph teeshirt)) Oh that teeshirt reminded rre (S'IDRY) (2) [Goodwin: G91: 250] A: Yeah I useta- This _girlfr- er _{eff's: gi.:rlfriend, the one he's gettin' married to, (0.9) s brother.= =he use' to uh, slntt U84. Cambridge University Press Cambridge London New York New Rochelle Melbourne Sydney Editions de Ia Maison des Sciences de !'Homme Paris Earlier versions of this chapter were presented at the Conference on the Possibilities and Limitations of Pragmatics, Urbino, Italy, July 1979, and at the British Sociological Association Sociology of Language Study Group Meetings, University of York, December 1980. I would like to thank the members of the Oxford-Warwick seminar in conversation analysis for their many informal observations and comments. Max Atkinson, Graham Button, Paul Drew, Bob Dunstan, and Tony Wootton have read earlier drafts and supplied valuable comments and examples. Finally, I am grateful to Gail Jefferson for her detailed comments and suggestions about data extracts and for her generous intellectual support over the past several years. 299 300 ｾａＺ＠ A: J. Heritage A change-of-state token ..... (( 13 lines of data orrdtted. During this ..... period the setting is disrupted by the ..... leaving of some of the participants)) What was I gonna say.= =Oh:: anyway.=She use'ta, (0.4) come 2_ver ...... . (3) -t In (1) a noticing is proposed with "oh." The noticing is subsequently described and used to furnish the basis for a storytelling. 2 In (2) a storytelling is temporarily abandoned in the face of the setting's disruption. The resumption of the story is achieved by a preliminary display that a search is being made for the next item of the narrative ("What was I gonna say.="). This is followed by "Oh:: anyway.=" and what is then subsequently recognizable as a resumption of the narrative thread. Here then the "oh::" displays a successful outcome of the search previously displayed as being in progress. The "oh's" produced in these fragments thus provide a fugitive commentary on the speaker's state of mind. Produced within ratified states of talk and as component elements of larger turns at talk, they are nonetheless fully fledged response cries: "signs meant to be taken to index directly the state of the transmitter" (Coffman 1981:116), through which evidence of an alignment taken to events is displayed, "the display taking the condensed, truncated form of a ... non-lexicalised expression" (ibid:100). In these cases, the speaker volunteers his production of "oh" and thereby injects an extraconversational contingency, adumbrated by the particle and subsequently elaborated upon, into the talk. However, the particle is also produced as a response to a variety of conversational actions, and it is these other occasions of its production that will be the central focus of this chapter. 3 The claim that "oh" makes a generic change-of-state proposal will be demonstrated by considering its placement in two major types of conversational environment - "informings" (Sections 2-4) and "repair" (Sections 5-7) - where the particle is regularly produced in response to prior turns at talk. In these sections, the sense of the particle's generic change-of-state proposal is particularized by reference to (1) the conversational sequences in which it occurs, (2) its precise placement within such sequences, and (3) the additional turn components that it commonly prefaces. In the final sections of the chapter a variety of aspects of the particle's sequential role are considered. 2. Informings A major conversational environment in which "oh" regularly occurs is in response to informings. Thus in (3) I's report of the arrival of some furniture is, initially at least, responded to with "oh." 4 [Rab:B:l:l:l2:1] J: Ye:h. 'h ｾＺｭ＠ (0.2) I've jis' rung tih teh- eh tell you (0.3) uh the things 'av arrived from Barker'n S!_one'ouf :se, Oh:::::. J: Oh c'n I c'm rou_:_nd, hh I: 301 ( .) Similarly in (4) J's informing "I rang y'earlier b'tchu w'r ou:t" gets an "oh" response as does I's subsequent accounting "I musta been at Dez's mu:m's." (4) J: .... I: .... J: [Rab: II: 11 =Hell2_ there I rang y'earlier b'tchu w'r ou:t, toh: I musta been at Dez's mu:m's= ="laGh : : . h= Moreover in (5) a more extended informing is similarly responded to with "oh." (5) [Trip to Syracuse:l] E: C: Goo:: d [hhhen heh · hhh I was urn: ( 0 . 3 ) I 1..en 1 u(.) I spoke t' the gir- I spoke to Caryn. (0.2) · hh andum i 1 w' z really bad because she decided of all weekends for this-one to go away E: Wha? (0.3) She decided to go away this 1..eekend. = =Yeah 'hhh (.)So that (.) y'know I really don' have a place ti' stay 'hO:: :h. (0.2) · hh So you're not gonna go up this weekend? (0.6) C: E: C: ... E: E: In each of these three cases, we can minimally characterize the data by suggesting that "oh" is used to mark the receipt of the informing delivered in the preceding turn or turns. Moreover we can additionally notice that these "oh" receipts (1) occur in response to complete chunks of information and (2) are produced at points at which the informings are possibly complete. In this context, it is striking that, in (5), Cs "She decided to go away this weekend" is "continuation" receipted with "yeah." This continuation also receipts a complete chunk of information A change-of-state token 302 J. Heritage but, in eliciting further talk from C, it seems designed to propose that C's informing is not yet possibly complete. By contrast, E's subsequent "oh" receipt does not prompt further talk from C, who thereby treats it as proposing that his informing is now complete at this stage and, relatedly, that E has now grasped what he has left unstated. 5 Free-standing "oh" receipts of prior informings, as illustrated in (3)(5) above, are comparatively rare in the conversational data to hand. More commonly, the turn initiated with an "oh" receipt contains additional components that achieve other tasks made relevant by the sequence in progress. For example, it is common for recipients to attend to, and deal with, informings as tellings of good or bad news. Recipients do so by assessing the news delivered, and, in such contexts, "oh" receipts are commonly combined with assessment components to give an oh-plus-assessment turn structure. In such turns, the "oh" invariably occurs in the turn-initial position. Fragments (6) and (7) illustrate this format. Such assessments commonly occur at the end of an informing and are regularly terminal or topic-curtailing in character (Jefferson 1981b). By contrast, where an informing is produced as a "hearably incomplete" news announcement, "oh" may co-occur with additional turn components that in various ways request or invite the informant to continue:6 (9) [JG:3C:5] R: I fergot t'tell y'the two best things that happen'tuh me t'day. Oh super,=What were they .., C: (10) A: -t B: [TG: 16] ... Well lately in the morning Rosemary's been picking ｾｵｰＮ＠ -Yihknow so I (haven' been) even takin' a train in [(the morning) hhOh that's grea:t. (11) (7) [Rah:1:1l J: I w'z j'st eh ringing up inn a rrDn-ent, (.) ｴＧｾ｡ｹ＠ I'll be comin' down - Ohgh goo_:_d, And in fragment (8) the format is deployed twice in succession: (8) [JG:3C:5 simplified] R: I fergot t' tell y' the two best things that happen'tuh me t'day. R: I gotta B plus on my math test, C: R: On yer final? Un huh? .,. C: R: C: R: _, C: Oh that's wonderful And I got athletic award. REALLY? Uh huh.=From sports club. Oh that's terrific Roger [HG:II:2] ((rea visit to a dermatologist)) My f:face hurts,= = 0 W't:=-o N: H: . ( ) Oh what'd'e do tih you. _, H: .... 1: [Rah:B:1DJ(12):2] I rsaw ｾ｡ｮｩ･＠ this morn!ng= Yes =in in: uh Marks'n sp encers 1Oh you 1 did didju yes, J: 1: J: ｾｉＺ＠ (6) 303 (12) [Rah:I:S] V: Oh I met Jani :e, eh:: :m yesterday an' she'd had a-f.s?_: rm from the Age Concern-about that jo:b.h= .... J: =Db she has? - - - In each of these cases, a chunk of information is "oh"-receipted by the recipient who subsequently proceeds to promote further continuation of the informing by the production of a question as in (9) and (11), or a "newsmark" (cf. Jefferson 1981a and note 13 to this chapter) as in (10) and (12). In order to develop an initial sense of the Work accomplished with the production of "oh" in these sequences, it can be noted that conversationalists exhibit a pervasive orientation to the tellability of information . A major aspect of this orientation involves avoiding telling recipients what they already know. Thus in (13), A and D have "good news" to tell to Band C, which is announced by A with: "Hey we got good news." J. 304 Heritage A change-of-state token When B then requests a telling of the news, C proposes (in overlap with B's request) to know it already. ( 13) [KC: 4] A: Hey we got good news . B: [What's the good ne ws, C: I ( A: D: .) k n o : w. 1 [Oh ya do::? Ya heard it? Rather than proceeding to tell the news, as requested by B, A and D both address C's proposal that he knows it already. 7 Here then the telling of the "good news" is deferred in the face of a claim by one of the recipients to have prior access to it. Although interactants may have a variety of resources with which they can infer, a priori, whether a candidate recipient is informed or uninformed about a potential "tellable," it is nonetheless the case that, with respect to the specifics of an informing, the informed or uninformed status of recipients is commonly the object of active negotiation and determination throughout the course of the informing itself. Negotiations over the informed/uninformed status of recipients have been shown powerfully to structure the design of storytelling, joke telling, and announcement sequences. 8 Through these negotiations, the parties to the talk establish local identities of informed teller and uninformed recipient with respect to the matter at hand, and these identities are commonly sustained through to the termination of the informing sequence. In this context, a particle that proposes that its producer has undergone a change of state may be nicely responsive to prior turns at talk that are produced as informings. With the act of informing, tellers propose to be knowledgeable about some matter concerning which, they also propose, recipients are ignorant. Correspondingly, in proposing a change of state with the production of "oh," recipients can confirm that, although they were previously uninformed on the matter at hand, they are now informed. With the use of "oh," recipients thus confirm the presupposition, relevance, and upshot of the prior act of informing as an action that has involved the transmission of information from an informed to an uninformed party. 9 "Oh" is thus a means by which recipients can align themselves to, and confirm, a prior turn's proposal to have been informative. Furthermore, by the addition of specific types of 305 turn components, such as assessments or requests for further information, recipients can proceed to treat the local trajectory of the informing as complete (with assessments) or incomplete (with requests for further information). Moreover, it can be further noted that "oh" is a strong indication that its producer has been informed as a result of a prior turn's talk. Specifically "oh" is scarcely ever (see note 9) associated with further turn components that assert prior knowledge of "oh"-receipted information. By contrast, receipt objects such as "yes" and "mm hm" avoid or defer treating prior talk as informative. Thus "yes" is regularly, and in contrast to "oh," associated with additional turn components that assert prior knowledge of just delivered information. Thus in (14) and (15), "yes" -prefaced turns involve just such assertions. (14) [PD:250] H: Listen, Bud's alright. .., J: Yeah, I know, I just talked to 'm. (15) [Frankel:TC:l:l:l9] S: In any eve: :nt? hhhhh That's not all that's ne:w. G: W't-e:lse. S: · t · hhhl1h W' l ｗｾ､｡Ｇ＠ n I have been rilly having problems. G: M-hm, S: ..,.hh An' yesterday I talk' tih her. · hhhh A: n' (0.3) apparently her mother is terminal. -. G: "tch Yeh but we knew that befo:re. In (16) prior knowledge of the high cost of a train journey is asserted at a slight distance from a "yeh" receipt of a turn that was intendedly informative. (16) [WPC:1:W.J(1):18) M: Well u- she's goin' by: trai:n.h All the wa:y. • hh An I If'S Seventy lli ; nepOU; nds by trai ;ll. h . hh . hh= .... J: Yeh M. =[Isn't that aw:ful:. "h .... J: [Well I know it is.= A change-of-state token 306 J. Heritage M: e-eYe:s (5) ｾ＠ J: =[!e inquir:ed abou:t [th'trai:-n, Y e : :s::. 1 M: ｾＱｨｫｮｯｷ＠ before we went bx_: coach. E: Additionally receipt objects such as "yes" and "mm hm" are regularly used as continuers in extended tellings. In these tellings the production of "oh" or "oh" -prefaced turns is commonly reserved for significant story elements. Thus in (17) an initially bland description of a mother's state of health is continuation-receipted (arrow 1), whereas the subsequent detailing of the mother's progress gets a strong news receipt ("oh" + "newsmark") from M (arrow 2). . (17) 0 0 'O.v is yih !!Pther by: J: We:ll she's a:, h bit better:, (.) Ｑｾ＠ M: ｾＮｴｮ｛Ｚ＠ J: Ｑｾ＠ 2-, : ,- Un;: - - ｴｨＧｷｾＺｹＮ＠ -- - eh- She crure: do:wn on: Satidee:eveni[ng 2c M: ｾ＠ ｾｍＺ＠ toh: did [she : : ,hl J: for the fir :s'ti:me. Ye:s. J: Ye[s. ( )- I Qfn't !mow whether she crure I: ..... 5-. M: 0 h t This sequence is of additional interest is that M's "f Oh: did she::,h" overlaps a further piece of information that J appends to her prior turn, namely, that Saturday was the first evening the mother had risen from her sickbed (arrow 3). This information, however, initially receives a simple "continuation" response from M (arrow 4). In this context, J begins her next turn with a resumptive "yes," and midway through its production, M, hearing the "yes" as a possible recompletion of the prior turn's talk as "the news," revises her receipt of the prior to a stretched and rise-fall intoned "Oh f ::." (arrow 5). Here then is delicate, but dramatic, evidence for the use of "oh" to respond to prior talk as significant and, by picking it out for such treatment, to mark it as information to be "foregrounded" from surrounding talk. A return to (5) shows that E's contihuation receipt to Cs "She decided to go away this weekend" treats it as a "background" to an informing that is to come. "hO:: :h. (0.2) E: "hh So you're not gonna go up this weekend? Thus, E's later "oh" receipt to Cs subsequent detailing "foregrounds" the latter as informative, and thereby treats it as the burden of his informing in a way that a further continuation receipt would not. Subsequent to such a receipt, which is strongly hearable as proposing a realization, C may thus wait (arrow 1) to see what E has made of his informing (arrow 2). Here the type of receipt object, its positioning in the sequence, and the way in which it is intoned all contribute toward this outcome. In sum, it is proposed that "oh" specifically functions as an information receipt that is regularly used as a means of proposing that the talk to which it responds is, or has been, informative to the recipient. Such a proposal is not accomplished by objects such as "yes" or "mm hm," which avoid or defer treating prior talk as informative. Where tellings are chunked into segments, "oh" may co-occur with additional turn ｣ｯｾｰ＿ｮ･ｴＬ＠ ＢＧＺｾｩ｣ｨ＠ in ｩｮｶｾｴｧ＠ or curtailing further talk to an informing exhrbrt the oh -producers orientation to, and preparedness to collaborate in, the production of an informing as an event having recognizable s.tages of development. Finally, "oh" may be used by recipients to highlight or "foreground" particular elements of an informing. [WPC:l:MJ(l):2] M: "hhhh (.) [Trip to Syracuse:!] C: She decided to go away this weekend.= E: =Yeah C: ·hhh (.) So that (.) y'know I really don' have a place ti'stay (0.3) J: 307 3. Question-elicited informings Just as "oh" receipts regularly occur in the environment of informings that are, in various ways, initiated by the informant, so they also regularly occur in response to informings that are elicited by questions. In each of the following cases, a simple sequence of (question)- (answer)("oh" receipt) (arrows [1]-[3]) is deployed. (18) [Rah: I: 8] V: And she's got the application forms.= SO when is O er _!nterview dId she sa r..:X 2-. V: She Ｑｾ＠ J: =O::>h: : A change-of-state token J. Heritage 308 didn't u- Well u-she's gotta send their Sh[e d'sn't know w hen=O h : 3 ... J: : Oh it's just th' ( 19) (24) !back 1 : V: =the [interview i : s [yet. J: ｦｾＺｲｭ＠ fo: [Rah:C:1:JS(15):1] 1-'> J: Whadiyih doing eating yer breakfast or 1nn,= Ｒｾ＠ [Carrpbell: 4: 1 J sareth [ ing Yes, 3+ J: Ohgh(h)h hhuh huh huh 'huh [.hhh S: J: OU I re enti,Oying .!_ife aren I tchU, S: r Well lis: ten, (.) tiz you vicar ye:t, 1 -+A: 309 ｴｩ､ｪｾ＠ phone yer (25) [Rah:B:1:1DJ(12):4] (0.3) 2 ... B: No I ain't (A): ＨｯｾＩ＠ _._ ｾａＺ＠ 3 (20) 1-. J: Okay then I w' z askin' and she says you're working to!IDrrow aS well, 2-+ I: Yes I'm s'"Pose to be tarorrow yes, 3...,. J: O['g:::, I: Yeh, ("hhh)Ch:. (WPC:1:MJ(1):1] 1-+ J: ｾ･ｮ＠ d'z ｾｵｳＧｮ＠ g[o back.= ｾｮ･｣ｳ｡ｲｩｬｹ＠ M: ·hhhh J: - ( ) 2-> M: -[u-She: goes back on Satida:y= ｾ＠ J: =Q_[g:. M: A:n:' ｾｴ･ｶＧｮ＠ w'z here (.) all las'week (21) [Frankel:'IC:1:1:13-14] S: 'hh When d' ju get out.= 1-+ S: =Christmas week or-the week before Christmas (0.3) - 2 ... G: Uh: :m !,wo or three days before Ch ｲｾｩｳｴｭ｡Ｌ '0 3 ... S: h : , Ｑ＠ 'hh= S: =Tha rt'See, G: ( 22) [Rah : A: 1 : IMJ ( 2) : 2 ] 1... I : Ah thi- et-y- I : think there w' z only about three things ordered was it or four. 2-> J: eh-u-Four I think there w'z two: for Ken'n two for I:an. ｾｉＺ＠ Oh:. J: B1 t I(c) I don't know what quite. (23) Ｑｾ＠ IRah :B: 2: JV(14): 1] J: Oh:::. Have !,hey'ave yih visitiz g[one then !Jley've ｾ＠ 2.... V: Yes, J....J:Ch:ah V: [A:: n:' they've ｟ｧｾＺｭ･＠ to .... J :ne. Examining (18)-(25) shows that the production of an "oh" receipt is not associated with the degree to which an answer is unexpected. The cases are arrayed in an order that roughly approximates the degree to which the expectations of the questioner - exhibited in the design of the question - are met. Thus in (18) the answer undercuts one of the presuppositions of the question to which it responds (i.e., that an application for a job that might lead to an interview has, in fact, been lodged). This answer may be treated as "least expected" by the questioner who subsequently goes on to produce a display of understanding (see Sections 6 and 7) that is hearably corrective of her prior misapprehension. In (19) and (20), open requests for information are responded to with the information requested, while in (21) and (22) information requests that propose alternative possibilities are responded to with a selection of one of the alternatives. In (23) and (24), questionformed likely inferences receive confirmation, while in (25) a sourcecited report receives confirmation.1° In all cases, however, the responses are "oh" -receipted by the questioner at, or near, the first point at which the responses are possibly complete and the question is, at least in a minimal sense, answered. In these simple (question) - (answer) - ("oh" -receipt) sequences, which are massively recurrent in ordinary conversation, "oh" again functions as an information receipt by proposing a change of state of knowledge or information. Moreover, in proposing a change of state, the "oh" receipt is once more nicely fitted to the Q-A sequence in which it participates. For the producer of a question proposes, with the production of question, to assume the status of presently uninformed about its substance and thereby proposes as well that the respondent, in a A change-of-state token 310 J. Heritage answering the question, assume the status of informed with respect to the matter at hand. Given this organization, the questioner may be committed by the provision of an answer to have undergone a change in his or her state of information and may be required to propose just that. 11 Here then the production of "oh" confirms an answer as an action that has involved the transmission of information from an informed to an uninformed party .12 Although it is but one of a variety of resources for such proposals, 13 the particle is an economical resource for their accomplishment. Moreover it is one that, as already noted in the context of simple informings, is readily combined with additional turn components - such as assessments or requests for further informationthat accomplish associated sequential tasks. These considerations further suggest that the production of an "oh" receipt (or some equivalent) may be avoided by questioners so as to ·propose that they have not been informed. Thus in (26) three Q-A "oh" sequences (arrowed a, b, and c respectively) run to completion. Subsequently (arrowed d), a fourth such sequence, is apparently initiated with N's question-intoned utterance "Nice Jewish bo:y?." (26) [HG:II:25] = · hhh Dz he 1 av 1 iz o.vn aE_a: rt ｛ｾｨﾷ＠ a EN: H: N: =Ch:' (1.0) Hem didju gi t 1 iz nunber, b nt?] Yea:h,= tNH: (.) I(h) (.) c(h)alled infermation 1 n San Fr 1 ncissc(h) [uh! N: Oh::::. ( .) N: Very cleve:r, hh= H: =Thank you[: I- "hh- "hhhhhhhh hh= c EH:N: =Uh: : Freedia: nd. ｾｎＺ＠ =Nice Jewish bo:y? (26). (26) ·hh [hh (.) - H: O:f cou:rse,= N: = v [cou_:_rse,] hh-hh-hh hnh "hhhhh= H: N: ］ｾｩ｣･Ｍｊｗｳｨ＠ boy who doesn 1 like tih write letters? Detail] [HG: II: 25 N: =· hhh Dz he H: Oht(:r) Freedlind.= H: d "O:f cou:rse," which treats the inference as self-evident rather than merely likely. In turn, this confirmation is receipted by N with a repetition of the confirmation (" 'v cou:rse"), which preserves this treatment and asserts it on her own behalf. In effect, the recipient withholds a change-of-state proposal and thus retrospectively proposes that her previous, question-intoned inference is to be heard as having been a comment on something self-evident rather than an inference concerning something still in doubt. It was noted earlier in the context of informing sequences that the status of informed/uninformed may be the object of detailed negotiations over the course of a telling sequence. Similar issues can now be seen to arise in the context of questions and their answers where, for example, the avoidance of now-informed receipts can be used, as in (26), to implicate that an answer does not inform, or to confirm or revise the status of a prior question-formed utterance as one that did not request information.14 Finally, it may be noted that an "oh" receipt, in occurring at a point at which an answer is possibly complete, may be used to propose that its producer is prepared to treat the answer as, in fact, complete. In this context, it is relevant to observe that, whereas "oh" is routinely used to receipt information, its sequential role is essentially backward looking. Specifically, the particle does not invite or request further information. Since, in many Q-A sequences, questioners cite information gaps, which they request to be filled, an answerer/"oh" recipient may treat an "oh" receipt as proposing that the questioner's information gap has been made up and that the answer is, from the questioner's point of view, sufficient. Under these circumstances, the answerer may withhold the detailing of further tellable materials until invited to tell more, as in W1 ts 1 iz last nama, N: 311 1 av 1 iz own apa-rt rrn· nt? -. [ "] "hhhh Yea:h,= N: =Oh: (I-:-o) ｾ＠ N: !!ow didju gi t 1 iz mnrber, 1 This latter utterance appears little different from other likely inferences (e.g., [24]) whose confirmations are routinely "oh"-receipted. In this case, the respondent (H) confirms the inference with an utterance Alternatively where, as in (20) (see also [22] and [23]), the answerer does continue with further detailing after an "oh" receipt, (20) (WPC:l:MJ(l):l] J: When d 1 z Sus'n g[o back.= M: "hhhh 312 J. A change-of-state token Heritage J: -r< > M: - u-8he goes back on Satida:y= J: =O[h:. - A:n:' ｾｴ･ｶＧｮ＠ -+ M: - w'z here (.) all las'week ..... this detailing is volunteered, that is, produced at the teller's initiative. The issues raised by these observations will be considered in more detail in sections 8-10. In sum, "oh"-carried change-of-state proposals are commonly used to receipt answers-to-questions as informative, while withholdings or substitutions of "oh" receipts may be used to imply either that an answer was not, or not yet, informative or, alternatively, that a prior questionformed utterance did not request information. Produced as a free-standing object, the placement of "oh" at, or near, a first point at which an answer is possibly complete may be used to propose the "oh" producer's preparedness to treat the prior answer as complete for all practical purposes and may result in a curtailment in the production of further tellables. 4. Counterinformings Finally, in the context of informings, we briefly consider the placement of "oh" in sequences where a first statement is met by a second "counterinforming" statement that is contrastive with the first. In this environment, the "oh" is regularly produced as part of the turn with which one of the speakers revises a previous position. Once again, the use of "oh" as a change-of-state proposal that is responsive to the informative character of a prior turn at talk is strongly implicated. A simple case is shown below. (27) B: [Goodwin: Family Dinner: 13] It looks like beef'n bean curd. ( 1.0) -+ J: Well I wan' lots of beef. D: I think it's pork. B: Oh. Pork. D: Mn hm Here B identifies a package of Chinese food as "beef'n bean curd." Subsequently another speaker (D) counteridentifies it as "pork." B subsequently accepts D's identification by repeating "pork," marking a change of state of information with "oh." In this case, the production of "oh" is closely associated with the acceptance of the counterinforming as a correction. 313 In the following case, J's realization/recollection-intoned "oh" adumbrates a revision of her prior assertion, which is accomplished by an initial display of "thinking back" ("Well i W'z it j la:s' night.") and a subsequent revision of her position. ( 28) [Rah: I : 1] V: Where didje get to la:s' ｮｩＭｾｧｨｴＬ＠ (1.0) J: ｾＺｳｴＭＡ＠ (0.4) dit (0.2) I di'n't go anytwhere? ------ V: W'l Andrew r:ang t'see if you were there, (0.7) 0 .. J: ·hh 0hE...:_::. 0 hh Well i W'z it fla:s' night. (.) . J: tYes it w- Tha:t's right i' was la:s' nah.hh !io I 'd taken I :an: ..... -:-:-(continues) A more elaborate case is (29). Here I' s initial announcement is met, after a considerable gap, with a possible predisagreement object (see Pomerantz, Chapter 4 herein, and note 13 to this chapter) from J " j Janie has." (line 4) and a subsequent counterassertion "No she hasn't" (line 6). (29) [Rah:B:1:1DJ(12):2) 1 I: Ye- 'h Well she's gone to mn eh: eh: Chester:. (0.9) 2 3 I: Ja[nie:, 4 J: - tJanie has_ 5 I: tEy?6 J: No she hasn' t (0.8) 7 8 I: Yes. She's go: :ne, (5.7) -9 10 I: She went just before dinner. (0.2) 11 12... J: ｏｨｴｾＭ Oh fr (thought ), 13 She w'z in such a 1rush, After a further extended gap, I reasserts her own prior informing (line 8) and, after yet another gap, elaborates it with further detailing (line 10). It is only after this subsequent detailing that J announces the revision of her prior position with a stretched rise-fall intoned "oh" (line 12) and the beginning of a description of her prior misinformation which, in turn, is overlapped with further detailing from I (line 13). The use of "oh" to project acceptance of another speaker's position in the above sequences can be highlighted by examining comparable sequences in which "oh" is absent. Thus, in (30), D's proposal that Rice is A change-of-state token J. Heritage 314 in Louisiana is met with the intendedly corrective counterinforming from M and F that it is in Texas. In her subsequent turn, D revises her position by accepting that Rice is in Texas. (30) [Post Party:I:14) D: Rice? M: is in Louisiana.o Tex as 1Texa[s. Texas. Rice. F: M: -+ D: A: 0 No(. 0 Yeh that's (right). o ｈ･ｾｴｯＺｮ＠ Here it is noticeable that D manages to revise her position while avoiding the production of an "oh" -carried ｣ｨ｡ｮｧ･Ｍｯｦｾｳｴ＠ proposal. ｉｾｳｴ･｡､Ｌ＠ by repeating "Texas. Rice.," she produces a dtsplay of ｣ｯｮｳｵｾｴｭｧ＠ ｾ･ｲ＠ own knowledge of the location and only then produces a confmnahon 0 ( " 0 Yeh that's [right] . " ) which accepts the position asserted by M and F. In effect D manages to revise her prior assertion so as to show that, although the revision is an interactional consequence of ｴｾ･＠ inter:entions from M and F, it rests not on a simple acceptance of thetr assertions but, in part at least, on a consultation of her own independent ｫｮｯｾｬﾭ edge of the location of Rice. Thus whereas in (27) B ｡｣･ｾｴｳ＠ a _countermforming as a correction with her "oh" -plus-repeat .recetpt, t.n (30) ｾｨ･＠ repeat-plus-acceptance is managed so as to avotd, wtth the wtthholdmg of "oh," an acceptance that treats the counterinforming as an authoritative transmission of information (i.e., a correction) from M and F to D.. Relatedly, in (31), C's "yes" receipt (line 8) to a hearably complete and intendedly corrective counterinforming'from the OJ (in line.s 3-4 a.n.d 67) is treated as insufficient to propose a revision of her pnor posthon. (31) [JH:FN] ((Fran a radio phone-in caii>etition titled "Beat the Jock." Carla's question to ''beat the jock" was: "Name the second group to enter the British 'Top 20' at No.1") 1 2 3 4 5 6 7 8 DJ: You'd better tell me then. "Going Underground." Nineteen eighty. That's why I asked you i f you thought Slade were the first. C: Yes. DJ: 'Cbs the Beatles \vere first, Slade were second and the Jam were third. .., C: Yes. 9 10 315 DJ: No. The Jam were the third group to go straight in at nl.liTber one. Yeah? 11 ... C: Oh:. 12 DJ: See people forget that the Beatles were first ..... Here C's "yes" receipt to the DJ's itemized counterinforming is met (lines 9-10) with "no," a further assertion of the central component of the prior counterinforming (concerning "The Jam") and a tag-positioned request for acceptance/confirmation. Only after C's "oh" receipt of this turn does the OJ treat her position as having been adequately revised by observing that "people" (including C by implication) "forget that the Beatles were first." In sum, in sequences in which contrastive proposals concerning a state of affairs are being made, an "oh" -carried change-of-state proposal may be used by one of the parties to propose a revision of his or her position that overtly responds to the other's talk as corrective. The data suggest that, while a free-standing "oh" may be sufficient to propose acceptance of a counterinforming (as in [31]), it is more normally accompanied by turn components (usually repetitions) that explicitly accept the counterinformation. Similarly, although it is possible to accomplish acceptance of a counterinforming without the production of "oh" (as in [30]), such cases are rare and instance the accomplishment of rather special interactional work. Finally, as (31) illustrates, "yes" is insufficient to propose a revision of position. 1 s Overall, in each of the informing sequence types considered so far, "oh" is used to propose a change of state of information. In each case, the sequential role of the particle is, at the minimum, one of accomplishing a retrospective reconfirmation of both the prior and the current knowledge states of the participants. The conclusions of sections 2-4 can be drawn together and summarized as follows: Through the use of the particle, informed, counterinformed, or questioning parties can assert that, whereas they were previously ignorant, misinformed or uninformed, they are now informed. Correspondingly, the informing, counterinforming, or answering party is reconfirmed as having been the informative, knowledgeable, or authoritative party in the exchange. By means of the particle, the alignment of the speakers in their sequencespecific roles is confirmed and validated. C: The Jam. DJ: Uh no Carla. 5. Other-initiated repair A second major sequential environment in which "oh" is regularly used as a receipt object is that of other-initiated repair. 16 In each of the following fragments, a second speaker initiates repair on a prior speaker's turn J. 316 Heritage A change-of-state token A: Who? and, the repair having been performed by the first speaker, the second speaker receipts the repair with "oh." ( 32) A: Well who'r you workin' for. B: 'hhh Well I'm working through the Amfat Corporation. 1-> A: The who? 2-o B: Amfah Corpora[tion. T's a holding company. 3-> A: 14 8: Percy. . ( ) 8: That young fella thet uh- his daughter was !lllrdered, Ｒｾ＠ [ C & D : 9] 317 3--i B: (1.0) (And)= YEA:: :h. A: roo A: YEAH. 8: They, said sunp' n about his goin tuhgether uh-on th'ticket so, Oh A: Yeah Ｒｾ＠ (33) [TG:3] B: Where didju play ba:sk[et baw. A: (The) gy:m B: In the gym? (hh) A: [Yea:h. Like grou(h)p therapy. Yuh know [half the groulp thet ｾ＠ had la:s' term ... Ｑｾ＠ 3-. B: 0 h : : (34) . [HG:II:4) N: But he goes, (.)he:- he gpes ｾｩｨ＠ 'av a ｾｩｬｹ＠ mild case he gpes, (.) 1-t H: Of wha[:t. Yih 1 shN: ( .) 2+ N: A:cne-e,= ｾ＠ H: =Oh[:, Jlhhh] (hhh) N: seh ｾｯｵ＠ shouldn' even ｾｲｹ＠ abou: t it . In each of these cases, the producer of the repair initiation (arrow 1) proposes to have some difficulty with the prior turn's talk and specifically locates that difficulty through the repair initiation itself. In each case the producer of the prior turn remedies the difficulty (arrow 2) by repetition (32), elaboration (33), and specification (34) and in each case, the producer of the repair initiation receipts the repair with "oh," thereby proposing a change of state of information and, by implication, a resolution of the trouble previously indicated. These cases contrast with the following in which the one (A) who initiates repair with "Who?" fails to respond to B's first repair attempt. (35) [NB:II:1:10) B: If Percy goes with- Nixon I'd sure like that. Here it may be noticed that A's repair initiation ("Who?"), although locating a trouble source in a person reference, does not locate the trouble specifically (as between Percy and Nixon). Moreover the repair initiation does not discriminate the type of trouble being proposed as either a hearing problem or a recognition problem. In producing her repair, B first addresses the trouble as a hearing problem located to the person reference stressed in her prior turn by repeating "Percy" (arrow 1). Having got no immediate receipt, B then attempts to remedy a hypothesized recognition problem by elaborating additional particulars of the referenced person (arrow 2). A further period of one second elapses during which the initiator of the repair produces no receipt, after which B begins a ｾｨｩｲ､＠ attempt at repair (arrow 3) which is overlapped by an "oh" -initiated receipt that proposes recognition of the referenced person ("Percy"). A similar case is the following. Here A proposes, with the use of a questioning repeat ("Pixy dust?!") to be having difficulty in understanding R's prior turn (lines 1-2). In line 7, R proposes a remedy by elaborating an origin for the "pixy dust," namely, "the big boom." (36) 1 [GTS:2:2:19] R: But the air's gotta came in dere an' the air is 2 3 4 5 6 7 8 sorta infiltrated with little uh piXY dust. ( 1.0) K: A: K: R: D::>esn' bother ID2 any. Pi XY dust?! -[y-ain't gonna live in it. Y' know from the big boom? 9 D: 10 A: K: D: A: Ra[dioPiXY dust, hehl1h Radioactivity I think is what he means, (hh)OH. Okay. 11 12 13 (2.0) -t J. 318 Heritage A change-of-state token The absence of any receipt for this repair attempt from A engenders a further repair attempt ("Radio-") from a further speaker (D) which is cut off as A initiates repair on the trouble source for a second time by rerepeating "Pixy dust." Finally D produces a remedy proposed as on behalf of the producer of the trouble source ("Radioactivity I think is what he means") which is "oh"-receipted by AY Once again, in proposing a change of state of knowledge or information, the "oh" receipt is well fitted to the sequence of repair initiation repair- "oh" receipt in which it participates. For the initiator of a repair proposes, with the production of a repair initiation, to be undergoing some difficulty with the prior turn and thereby proposes that the respondent, in producing the repair, will resolve this difficulty. Given this organization, the initiator of a repair may be committed by the provision of a repair to have undergone a change in his or her state of information and may be required to propose just that. The particle "oh" is a major resource for the achievement of this proposal which, in turn, permits a mutually ratified exit from repair sequences. Although such exits may be achieved by other means, 18 "oh"-accomplished exits from repair sequences are a common form of exit in both simple repair sequences (32)(34) and their extended counterparts (35)-(36). 6. Understanding checks A closely related environment in which "oh" is used as a sequence exit device is instanced in (37). (37) tion in the design of other-initiated repairs sketched above. Whereas the latter form of repair initiation proposes, and commonly locates, a trouble with a previous turn's talk for which a remedy is solicited, the understanding check identifies a trouble with a previous turn's talk by proposing a solution to that trouble. The understanding-check sequence, however, is not properly complete at this stage. For in proposing a candidate understanding of what an earlier speaker had intended, the producer of an understanding check thereby invites that speaker to confirm (or disconfirm) the adequacy of that proposal. The locus of the completed repair, therefore, is to be found in the responsive confirmation/disconfirmation of the understanding check (arrow 3). This responsive confirmation/ disconfirmation is, once more, routinely receipted with an "oh" that reconfirms the previous understanding check as a candidate one. Thus a basic format for other-initiated repair sequences is: A: Repairable B: Repair initiation A: Repair B: "Oh" receipt 1. 2. 3. 4. Similarly, a basic format for repair sequences involving understanding checks is strikingly similar: 1. A: Repairable 2. B: Understanding check ((repair initiation)) 3. A: Confirmation/disconfirmation ((repair)) 4. B: "Oh" receipt The format is clearly evidenced in (37) and in (38)-(40). [NB:III:1:2] (( Re an invitation for F' s daughter to visit)) F: When didju want'er tih care do::w[n. 'hhh Oh any titre 1-+ S: between: now en nex 1 Saturday 1 hh Ｒｾ＠ F: A wee:k fran:: (0.3) this coming Saturdee. 3-» S: Yeah. ( .) ｾ＠ (38) Ｑｾ＠ [Rah: II: 7] J: Ierek 1 s ho:me? (0.5i 2-+ I: Yo:ur Ie[rek. ｾ＠ J: Ye:s rn rn 4-,I; [Oh:. F: 'hhhh Oh:: :. (39) In (37), F proposes a trouble (arrow 2) concerning S's prior time reference. Rather than initiating a repair with, for example, "Which Saturday," F proposes a remedy for the trouble by producing an understanding check which, in this case, takes the form of a best guess about the specific "Saturday" in question. This sequence involves a simple varia- 319 [SF:2:5:sirrplified] B: So::: \\12 thought thet yihknow= =if you wanna care on over early. C 1 rnon over. Ｒｾ＠ M: 'hhhh- 'hhhh:::::: Ah::: hhh fer dinner yih rrean? hh 34 B: No not fer dinner. h= 1.., 4-. M: =Oh- 320 J. Heritage ( 40 ) A change-of-state token [NB: II : 2 : 17 J N: I just uh, forward'iz mail, stick it in 'n envelope, (0.4) send it all on up to 'im en, 'hhhh 2- E: You know where 'e is then. Ｑ ｾ＠ (0.8) 3-o N: I have never had Ｔｾ＠ E: Cll::. ｾ＠ of it retmned Emm, It is noteworthy that this format is preserved throughout a wide variety of cases, varying from those, e.g., (37) and (38), in which the understanding check is transparently clarificatory and is confirmed, through (39) in which the understanding check is self-interested and disconfirmed, to cases such as (40) in which the understanding check, in topicalizing a presupposition of the prior turn, can be heard to be investigaｴｩｶｾ＠ in character. In both of the forms of other-initiated repair considered, "oh" is used as a repair receipt. In each of them, as in the Q-A sequences treated in Section 3, an information gap or difficulty is proposed and its subsequent remedy is receipted with an "oh"-accomplished change-of-state proposal. Once again, the alignment of the speakers in their sequencespecific roles is reconfirmed by this means. The work of such alignment, moreover, is not always simply formal in character. For example, to return to (40), E's investigative understanding check is deployed as a means of inquiring into the present relationship between N and the "he" of the fragment (N's ex-husband). N's reply, "I have never had any of it returned Emma," manages both to depict the absence of communication between them (and hence the state of the relationship) and to implicate an unwillingness to develop the topic further. In this context, E's "oh" receipt preserves the prototypical understanding-check sequence and hence avoids any overt or official treatment of N's unwillingness to elaborate. Through the understanding check an opportunity to develop the talk in a particular direction is offered (by E) and declined (by N) without the offer or its declination ever reaching the official surface of the talk. 7. Displays of understanding In Section 6, it was suggested that the provision of a sequence-terminating "oh" by the producer of an understanding check confirms the latter as a candidate display of understanding. That this is so is, we suggest, a result of the production of the change-of-state proposal after a responsive confirmation/disconfirmation of the check. However, there are oc- 321 casions in talk where recipients may wish to show that prior talk has been adequately descriptive and/or that they have competently understood its import. Such recipients require resources with which they can display confidence or certainty in their displayed understandings of what another had intended. One such resource simply involves the repositioning of "oh" from a fourth-turn receipt position to preface the turn within which the understanding is displayed. This use of "oh" is examined by reference to (41)(45). (41) G: [Goodwin: G84:M:3] He wz o:n the opposite side a'the driver (.) - ｲｩｾｧｨｴ＿＠ 0 With iz:: o (0.4) M: No he w- (.) E?-he wz on the sa- :me side ez the drive r ... G: foh on nuh ba:ck seat?= M: =Yeah i[n ､Ｇ｢｡ｾ｣ｫ＠ s[eat G: Wu:l C: [ hmn, hm-m-hrn-m-hm G: Example (41) is taken from an extended stretch of talk in which G is being told about an automobile escapade. It turns out that an understanding of the physical location of the major protagonist (the "he" of the fragment, who is not present in the conversation) within the automobile is critical for an appreciation of the story's details. G's first attempt to locate the protagonist's position within the car consists of an understanding check explicitly designed for confirmation: "He wz o:n the opposite side a'the driver ri:ght?." In disconfirming this attempt, M refers only to which side of the car the protagonist was located and does not explicitly state whether the latter was in the front or the back seat. He does, however, provide G with the resources with which to infer this ｾ｡ｳｴ＠ coordinate of the protagonist's position. G may thus, using the mformation provided, conclusively infer that "the back seat" is the location in question. G is enabled, in short, to work it out for himself. In this context, it can be seen that G prefaces his revised referencing of the location with "oh." Here he draws upon the "oh" -carried change-ofstate proposal to assert that then and there is the point at which he has determined the location, a point which is prior to any possible confirmaｾｩｯｮＮ＠ In proposing this as the point of realization, G thereby proposes its mdependence of subsequent confirmation and, hence, his confidence in his displayed grasp of the state of affairs. 322 J. A change-of-state token Heritage Other materials provide support for the proposal that "oh" -carried recognition/realization claims in turn-initial position are associated with subsequent referencings that are, in various ways, confidently produced. Thus in (42) a firm assertion of recognition is "oh"-prefaced and followed by a word search that is finally successful. (42) (4.0) A: What's that gal's name? Just went back to Michigan. ( 1. 0) A: Hilda, LnTI 1 .... B: Oh I know who you mean, ( 1. 0) 2.... · B: Grady-Grady. A: Yeah. Hilda Grady. In (43), the first speaker makes three attempts to secure evidence of the recipient's recognition of the location of "Pilgrim Lake." Following the third attempt, the recipient comes up with a firm re-referencing of the lake by reference to "Bakersfield" and thereby proposes independent recognition of the lake in question. Here again a turn-initial "oh" marks the initial realizing moment of the recognition that was solicited by the prior speaker. (43) [Northridge 2:3] D: Like yih know wherah:: P.!_lgrim ｾ･＠ is i(t)sthat's on the other si:de ｵＧｴｨｾ･ｶｩｮＮ］＠ =Yih know this side of the Gr_!9Jevine. _.. p: Oh the's jus' up to Bakersfield. A return to (41) further shows that M's confirmation of G's understanding check is not "oh"-receipted byG. A similar pattern can be observed in (44) and (45). ( 44) 14 B: Oh (.)from where she lives now. A: Yeah. ＲｾｂＺ＠ [JH::FN] A: She's rroving house soon. B: Where to. A: Just round the corner actually. B: Fran you? A: No. That won't be too difficult. A: No. ( 45) [DA:2 :2] F: How long yih gunna be he:re,= B: ｾﾷｨ＠ Uh:t's (.) not too lo:ng. Uh· · j_ustn'til: uh: : uh think Monday:- [8:1:1:12:23) A: Uh, she asked me to stop by, she brought a chest of drawers fran LnTI 323 (1.0) - 14 F: Til, oh yih mean like a week tomorrow. ( .) B: Yah. 2-+ (0.3) B: Mm:hm,= Ｓｾ＠ F: Now-you _!old me you .... In (44), as in (41), disconfirmation of a first understanding check (B's "From you?") is followed by a second attempt that is "realizing" "oh"prefaced as a confident inference. Confirmation of this second attempt is not subsequently "oh"-receipted (arrow 2). In (45), F revises a simple understanding check initiated with "Til" in favor of a "realizing" "oh"prefaced display of confident inference. Once again, the subsequent confirmation is not "oh"-receipted by F (arrows 2 and 3). Comparing (45) with the very similar instance in (37), we see a simple movement of the "oh" -carried change-of-state proposal from a fourth turn-receipt position (in [37]) to a turn-initial position to the understanding check (in [45]). This comparison yields the conclusion that the "oh" functions in (45) as a realization claim which, in occurring prior to the subsequent turn components conveying the substance of the understanding achieved, proposes confidence in the adequacy of the understanding ｾｵ｢ｳ･ｱｮｴｬｹ＠ displayed. Moreover, since a change of state is proposed' to have occuqed then and there, no further similar proposal subsequent to confirmation is required. By comparison, in (37) the producer of the understanding check that was not "oh"-prefaced thereby proposes the displayed understanding as a candidate understanding that requires subsequent confirmation before the process of realization is accountably complete. In this case the completion of the realization process is proposed with a fourth turn-receipting "oh." By means of these two alternative placements of "oh," therefore, a turn that proposes a confident display of understanding may be systematically discriminated from one embodying a less certain understanding check. 324 J. A change-of-state token Heritage 8. Aspects of the placement of "oh" in informing sequences The aim of this chapter thus far has been to demonstrate that the production of "oh" generically proposes that its producer has undergone some kind of change of state. In previous sections, it has been argued that this generic proposal is parVcularized by reference to the sequence types in which "oh" occurs and by the details of its placement in such sequences. Finally, some attempt has been made to characterize the formal or official sequence-specific tasks accomplished by the production of the particle. These tasks, however, are far from being the only ones that the production of "oh" may be used to accomplish; indeed they constitute the absolute minimum that may be claimed about the uses of the particle and its placement. In the remainder of this chapter, an attempt will be made to develop a broader appreciation of the particle and its uses by considering its placement in the context of a wider set of sequential relevancies than those treated so far. We begin by reemphasizing that, while the particle may propose a change of state that is appropriately responsive to a prior turn's informing or repair, its sequential role is essentially backward looking. Specifically, although the production of a free-standing "oh" is commonly used to establish or confirm current speaker alignments, the particle does not, of itself, request, invite, or promote any continuation of an informing. Thus in (38), the "oh" receipt of the repair on an initial news announcement ("Derek's ho:me?"), which also receipts the announcement itself, is not treated by the announcer (J) as requesting further elaboration. ( 38) [Rah: II: 7( extended)) (0.5) I: Yo:ur De[rek. ｙｾＺｳ＠ elapses before N initiates further on-topic talk with "How didju git 'iz number." (26) [HG:II:25 (detail)] N: ="hhh D'z he 'av 'iz own ｡ｅＺｲｴ｛ｾｨﾷｮ＿ｬ＠ Yea:h,= H: N: =Oh:' -+ cr:-o) N: !!ow didju git 'iz ｭｾ･ｲＬ＠ In both cases, the informative party withholds further on-topic talk after an "oh" receipt until receiving a request to do so. While one factor contributing to these post-"oh" hitches may be the informative party's wish to avoid proceeding unilaterally with further talk, additional considerations are undoubtedly at work. As noted in Section 2, free-standing "oh" receipts to informings are rare in the data to hand. Instead, the particle most regularly occurs in conjunction with additional turn components such as assessments or requests for further information. Moreover, in a range of instances (see, for example, [3], [18], [21], [24]), the production of "oh" is followed at a slight distance by further talk from the "oh" producer. Thus an informant/"oh" recipient may withhold further talk on the assumption that the "oh" already produced is prefatory to further turn components. And indeed such additional components are forthcoming, at a slight distance, in (38) and (26) above and (26 [detail]) below - (see also [5]): (26) [HG:II:25 (detail)) N: How didju git 'iz nllllber, (.) - J: Derek's ho:rre? J: 325 H: I(h) (.) c(h)alled infermation'n San Fr'ncissc(h) mfm I: Oh:. 1... (.) 2.... I: An'- is he a' ri:ght?= ｾ＠ J: =Oh he's fi:ne ..... Instead J withholds continuation or elaboration of the initial news announcement (arrow 1) until specifically invited to do so (arrow 2), whereupon she responds promptly with a latched utterance (arrow 3). Similarly, in (26), after N's "oh" receipt of N's affirmative answer to her prior question "D'z he 'av 'iz own apa:rtmint?," a one-second pause uh. -+N: [Oh::::. (-.)- -+ N · Very s:_leve_:_r, hh= Here it is the "oh" producer who, as in the previous two examples, resumes the talk with an assessment - another turn component that commonly co-occurs with "oh." In this context, it may further be noted that, whereas "oh" may propose a change of state in response to an informing, it is entirely opaque as to the quality or character of the change of state proposedly undergone by its producer. Thus an infor- 326 J. A change-of-state token Heritage mant/"oh" recipient may withhold from further talk with a view to permitting/inviting the "oh" producer to elaborate what lay behind the production of the particle. . . An elaborate version of this post-"oh" withholding IS the followmg: (Rah:B:2:JV(14):1] J: Oh: : : . Have they' av yih visitiz g r0ne then, ] V: IIJbey've go :ne. Yes, - - J: Oh [:ah . ... V: - A: :n:' they've _&one to ..... [Rah: II :2] ( 46) ( (Re a previously announced change of arrangerrents for coffee)) J: C's uh: I: [there's no ｢｡Ｚｾｮｴ ｯｮＬ＠ I:s Veras Ｑ (.) I: Pardon? J: There's no ba:dminton:, tarorrow so[: ｾＺｲＬ｝＠ 0 h : . I: . 1... ( ) 2-+ J: Yeh. ( .) 3-. 4-. J: S o I thought well] lah[e e- Y e h . I: J: It' 11 be-an opportunity for liE: to do it. Here J abandons a projected extension to her repaired accounting ('There's no ba:dminton, tomorrow so: wuh:r,") in the face .of ｉＧｾ＠ ､ＺＬｳ｡ｾ＠ pointed-sounding "oh" receipt. Here it ｡ｰ･ｲｾ＠ that J, heanng Is "oh" receipt, abandons her accounting so as to permit I to elaborate her, oh with some comment or query (arrow 1). In the absence of any move by I, J recompletes her prior turn (arrow 2) and thus creates a further opportunity for I to produce some talk (arrow 3) before proceeding with a continuation of her account (arrow 4). A final type of evidence that informants/"oh" recipients treat the production of "oh" as projecting further turn components arises from sequences such as (20) and (23). (20) (23) 327 [W:PC:l:(MJ)l:l] J: When d'z Sus'n g[o back.= M: "hhhh J: ( ) M: =ru-She: goes back on Satida:y= J: =Q[Q_:. ...,. M: A:n:' Stev' n w' z here (.) all las'week In each of these cases, the production of a receipting "oh" is intersected with a stretched version of "and" with which the prior speaker displays continued turn occupancy and a commitment to extend the prior question-initiated informing with further talk.I 9 In these cases, the informant/"oh" recipient's production of overlapping talk appears designed to stifle, or otherwise sequentially delete, the production of additional turn components projected by the production of "oh." In considering the tasks accomplished by this overlap competitive talk, it will be recalled that possible additional turn components projected by the production of "oh" include additional inquiries and assessments. Since either item may disrupt trajectory of talk intended by the informant - inquiries by redirecting the talk along a different track, assessments by being topic-curtailing - neither may be desired by an informant who wishes to elaborate on prior talk and to control its direction. Thus the "oh" -intersecting elaborations instanced in (20) and (23) appear designed to forestall the possible production of such additional turn components in the service of retaining control over the future development of topical talk. With these considerations to hand, we can now proceed to consider recipient conduct in two systematically organized sequence types used to develop new topics in conversation. We will find that a free-standing "oh" is an unsatisfactory receipt item, though for different reasons, in both sequence types. 9. Recipient conduct in new topic beginnings Button and Casey (forthcoming) have described two distinctively organized procedures- news announcements and itemized news inquiries through which speakers can initiate talk that involves an abrupt shift from an immediately preceding topic to a new one. In news announcements, intending informants initiate a new topic by partially describing, or headlining, events in which they, or known-torecipient third parties have been involved . J. 328 A change-of-state token Heritage (10) require a strong display of recipient commitment to the proposed topic before continuing. By contrast, a simple "oh" receipt may be insufficient to promote such topical development, as in (38). (Rah:B:1DJ(12):2] I: Yes !:,e's he[re' J: mHm. . ( ) I ｛ｾ＠ ｾ｡ｮｩ･＠ this l!X)rnJ:.ng= Yes .... J: =in in: uh Marks' n Sp [encers] 2 .... I: Qh you did ､ｩ｛ｪｾ＠ J· 1 [ ｉｾ＠ J: Mn. .Y[: s,)= . . hh J: Oh (we 11 ) _!et ' s hope sanething cares o: f i ( t Ye s [Derek's ho:rre? (0.5) I: Yo:ur De[rek. J: ｙｾＺｳ＠ m[m 0!:,:, I: J: Ｑｾ＠ o 'hhhhhh (.) 'lba.t 's a'right, 0 (.) N: u-h Ctl::, ｾｉＺ＠ (.) :3--. N: j<n:J'e dis ( .) prac'ly killed !1\Y dunb fa:ce,= (12) [Rah:I:8] J: ... 'cuz she said she wouldn' be going if Janie w' z going t' that keep fit thing. V: ullight yeh 'hh Oh I ｾｴ＠ ｊ｡ｮｩｾ･Ｌ＠ eh:: :m ｬＧ｟･ｳｴｲ､ｾ＠ an' she' d had a fo: rm fran the Age Concern about 1 that jo:b. h= 2 ... J: =Oh she has? 3 .... V: -So: ehshe w'z §_ending the f_Q:rm back ..... t In each of these cases the initial news announcement (arrow 1) is disjoined from previous talk and is hearably incomplete in intimating that there is more to be told than has emerged ｴｨｾｳ＠ far. Similarly, in each of the cases the recipient orients to this hearaqie incompleteness by creating a fu;ther opportunity for an elaboratibn of the news (arrow 2), whereupon the news announcer engages in such elaboration (arrow 3). The recipients of news announcements, in each case, progress the projected sequence with the use of a receipt of an "oh" plus inquiry (in [11]) or of "oh" plus newsmark (as in [10] or [12]). These receipts are well fitted to the news announcements to which they respond. For a news announcer unilaterally proposes a new topic of conversation and may (.) 3-t (.)- H: Bu:t 1-+ N: ］｛ｾ＠ f :_!:.ace hurts,= 0 H: = W't- 0 (.) 2-t H: Oh what'd'e do tih you. Yes:. I: J: ｾＺ｛Ｇｨ＠ (HG: II: 2] H: 'hhhh I c'n live without 'er, [Rah: II: 7 (extended)] ( 38) 3..., J: =She w' z buyin' a whole load of stuff ( 11) 329 I: An'- is he a' ri:ght?= J: =Oh he's fi:ne ..... Here, as we have already noted, I's "oh" receipt of J's (repaired) news announcement does not progress the topic. In this context, "oh" is systematically weaker than an "oh" plus inquiry or "oh" plus newsmark receipt in that (1) it fails to invite the informant/news announcer to tell more and (2) in projecting additional turn components, it may invite the announcer to await them by withholding from further talk. Given these features, in the environment of news announcements as unilateral new topic proposals, a mere "oh" receipt is systematically insufficient to promote further talk from the news announcer/"oh" recipient. In the context of itemized news inquiries, by contrast, the reverse is the case. Here it is the intending recipient who nominates a possibly newsworthy event by inquiring into a coparticipant-related event as in (17) and (47). (17) ｾ＠ [WPC:1:(MJ)1:2] M: 'hhhh (.) ( 4 7) Ｐ ｾＺ＠ 0 : 'Ow is yih ｾｴｨ･ｲ＠ by: ｴｨＧｷｾＺｹＮ＠ [Rah: II : 5] -+ J: !hen are you gettin' yer: tdining room suite. Such inquiries are regularly understood, not as requests for information to be answered in abbreviated form, but as news inquiries - requests to 330 J. A change-of-state token Heritage be brought up to date on current recipient circumstances or troubles to which the inquirer displays partial access (Button and Casey forthcoming). Recipients display this understanding by giving elaborated, but hearably incomplete, responses (arrow 2) and thus establish themselves as the intending tellers of further information topicalized by the prior inquiry (ibid.). In turn, inquirers regularly promote such further telling by means of the production of a continuation (arrow 3). (17) [WPC:l:(1U")1:2] 1..., M: 'hhhh (.) 0 Um::o 'Ow is yih (.) ＲｾｊＺ＠ - ｾｴｨ･ｲ＠ by: ｴｨＧｷｾＺｹＮ＠ We_:_l l _lihe' s a: , bit better_:_, 3-t M: Mn ::' J: [eli- She carne: do:wn on: Satidee:eveni[ng ..... 'f'Oh: M: (47) did (Rah:II:5] 1... J: When are you gettin' yer: tdining room suite. wee:k.h 2-+ I: Well not ye:t i- eh we ｣｡Ｚｬ･､ｾｴ＠ 3-. J: eYe[_:_s, 0 I: (8ut) Jilly_:_o ........ ( 48) 331 [Her: I: 11: 3] Ｑｾ＠ N: ="h have Are: you ex:pecting any (puppies)? 2.... I: f 'h _!ell I hope]so:_:_= 3....N: _ 0 0h. Howe xci ting.o 4-. I:-( u h:: n) --fu h:: lm: d-Tessa w'z mated ..... In this case of overlap competition, the intending informant (I) overtly contests both the "oh" receipt and the subsequent object it projects- an early and topic-curtailing assessment. Whereas N and I begin their overlapping turns simultaneously, I's first "uhm" is stretched across N's "oh" and the initiation of her assessment so as to claim continued turn occupancy, and this claim is reinforced by the initiation of a second stretched "uhm" in overlap with the final syllable of N's assessment. This overt contest with both the "oh" receipt and the intervention it projects contrasts with the immediately subsequent development of the talk, in (49), in which only the post-"oh" continuation undergoes competition. (49) [Her:I:ll:3] I: d-Tessa w'z mated urn (0.3) oh about three weeks ago:. In these cases, by contrast with the news-announcement sequences, the elaboration of the news is invited by the production of a more-or-less passive continuation object, most commonly, "yes" and "mm hm." 20 Again such continuation receipts are well fitted to the sequences in which they participate. For the would-be recipient of the informing has already displayed substantial commitment to the to-be-developed topic by the initial production of an itemized news inquiry. Moreover, since the projected informant has already begun a to-be-completed informing in response to the prior inquiry, the alignment of both speakers to a trajectory of topical development is largely accomplished and requires only that the inquirer/projected recipient sustain the role of recipient. The production of a continuation is the standard means to this end. In this context, an "oh" receipt is doubly inappropriate. Firstly, since it regularly co-occurs with additional turn components, it may be heard to project early recipient intervention in the informing sequence. Thus in (48) an itemized news inquiry (arrow 1) gets an elaborated but hearably incomplete response (arrow 2). The subsequent ( "oh" -plus-assessment) response (arrow 3) is overlapped by the projected informant with two stretched "uh::m"s, after which the informant proceeds in the clear with further detailing (arrow 4). h N: 2-t I: hhoh:. c ) 1And (.) Kizzy w'z mated about tv,.u weeks ago[:. 3-. N: I: Oh my ｾｯ､ｮ･ｳ＠ you do as [ :k for .:!:_] t eh-h e h Here N's appropriate "oh" receipt of the detailing of the puppy situation is permitted in the clear by I (arrow 1) whereas N's attempted continuation of the turn is cut off by I's overlapping continuation (arrow 2) of the detailing. Thus it is only after both dogs' matings have been detailed that N is permitted an ("oh" -plus-assessment) utterance in the clear (arrow 3). Example (48), in which both the "oh" and the utterance it projects undergo competition, suggests that the "oh" is being competed with in an attempt to forestall any subsequent turn components an attempt that is unsuccessful in this case and results in further overlap competitive activity. In short, the "oh" is being competed with for what it projects. This then is a first sense in which an "oh" receipt to a hearably incomplete response to an itemized news inquiry may be inappropriate. However, second, an "oh" receipt may be undesired and competed with for what it proposes in its own right and regardless of what it J. 332 A change-of-state token Heritage projects. Section 3 noted that an "oh" receipt of a possibly complete answer to a question may treat the answer as in fact complete - a satisfactory filling of a gap in information. A return to a sequence such as (26) shows that the apparently unproblematic fact that the question cites an information gap that is subsequently filled is in fact the product of fine-grained sequential negotiation. [HG:II:25 (detail)] (26) NH:: ='hhh D'z he 'av 'iz own ＡＱﾣ｡Ｚｲｴ｛ｾｨﾷｮ＿｝＠ Yea:h,= N: =Oh:, N: (1:-0) didju gi t 'iz mmber, !!ow In this sequence, the "oh"-carried now-informed proposal instructs the informant that a gap in information has now been filled and that the informant may lawfully withhold from further talk. By contrast, in (50), a bearably complete answer to a question that could have been referring to a similar information gap is continuationreceipted, and the question is thereby retrospectively formulated by the questioner as a topic-generative itemized news inquiry for which further detailing by the informant is appropriate. (50) When are you proposing setting off then.= = · t · Wuh we're t=-we're leaving...,.Su: nday J: M: ｾ＠ J: M: (W:PC:l:MJ(l) :11] ｾＺｭｩｮ＠ [g Ye:s. And we're ca: lling in Binningham ..... In (26), an "oh" receipt is produced at the first point at which a bearably complete answer is produced and results in the curtailment of any further detailing in response to the question. In (50), by contrast, a continuation receipt is produced at the first point at which a bearably complete answer is produced and, subsequent to this receipt, the informant (M) engages in further elaborate detailing of the planned trip. Returning to (20) and (23), we can now suggest that more is being done than simply stifling the production of post-"oh" tum components. (20) [W:PC: 1: (MJ)1: 1] J: When d'z Sus'n g[o back.= M: 'hhhh J: -r< 333 ) M: - u-She goes back on Satida:y= J: =O[h:. M: - A:n:' Stev'n w'z here (.) all las'week (23) [Rah:B:2:JV(14):1] rane then' ] V: %ey've go :ne. Yes, J: Cb :ah V: - [A::n:' they've _&l:>ne to ..... J: Cb: : : • Have .!_hey' av yih visitiz g In each of these cases, the questioner uses a now informed "oh" receipt to treat the prior answer as bearably complete while the answerer, with the production of an "oh" -intersecting continuation, treats the prior question as an itemized news inquiry generative of further topical elaboration. In each case, an in-the-course-of-being-produced "oh" is overlapped, not merely to stifle the production of further post-"oh" turn components, but also to preempt and, as far as possible, to delete the sequence curtailing implications of the now informed proposal accomplished by the use of a free-standing "oh" receipt. 21 In these cases, then, "oh" is being competed with for what it proposes in its own right and, through these cases, it can be seen that a questioner's choice between an "oh" and a continuation receipt is specifically consequential in proposing both how he viewed the initial question and how he is prepared to treat the answer that responds to it. In sum, where new topics are being developed with the use of news announcements or itemized news inquiries, a free-standing "oh" receipt is a systematically inadequate response. In the context of news announcements, it is generally insufficient to promote continuation whereas, conversely, in the case of itemized news inquiries it may constitute a curtailing intervention into the informing sequence and, in turn, may be systematically competed with. In this latter context, a free-standing "oh" receipt may be produced by a questioner who began with a gap in information and who is unaware of, or unwilling to collaborate with, an answerer's desire to respond to the question in an elaborated or topicgenerative fashion. 10. The production of free-standing "oh" as withholding In the preceding sections of this chapter, it has been repeatedly noted that "oh" is regularly used as a turn component prefatory to additional turn components and that when it is produced in free-standing form it is 334 J. A change-of-state token Heritage regularly the "oh" producer who subsequently progresses the talk. These sequential projections are, as we have seen, understood and traded upon by "oh" recipients who display this understanding in withholding further talk, or producing competing talk, in the immediate environment of the particle's production. In a number of cases, however, "oh" producers may refrain from the production of further talk in ways that are specifically identifiable by recipients as involving withholding. Thus the following sequence develops from a standard prerequest/pre-invitation object: "What are you doing?" (51) [NB: IV: 9: l) 1 E: Oh: I'm jis' sittin' here with Phil'n Martha'n 2 haa: eh fixin I a drink they I re goin I OUt tih 3 dinner:-:- m 4 (.) 5 E: H e' s6 P: [Oh: : : : . Oh. 7 E: Why: whiddiyih 8 (1-:-o)9 P: hhuhh !ell'(h I wa:nt. ｾ｡ｮｴ･､＠ Here E's response (lines 1-3) details a current activity that potentially conflicts with what P might have in mind (either an extended telephone conversation or, as it turns out, an immediate visit). P's disappointedsounding stretched "oh" receipt suggests that E's response does indeed pose some difficulty for her plans. However, rather than going on to detail these plans, P merely appends a shorter and terminally intoned "oh," thereby exhibiting a reluctance to elaborate. It is thus left toE (line 7) to inquire into the plan foreshadowed with the "pre-" object and depicted as frustrated with the stretched "oh." By this procedure, P manages the sequence so as begin her description of her prior intention (line 9) at the request of her recipient rather than on her own initiative. A more extended instance of withholding through the production of a free-standing "oh" is the following. ( 19) [Canpbell 4: 1 (extended)] 1 A: 2 3 4 B: .,.5 (A): (0.4) -'- Well lis: ten, (.) tiz you tidju phone yer vicar ye:t, (0. 3) No I ain't. (-:·hhh)- 6 A: 7 8 (A): 9 A: 10 B: 335 Oh:. (0.3) "hhhhhAh .. --:-. .. .. -.. - [i .w'z gonna ｾ｡ｩｴ＠ Here B indicates (line 4) that he has not fulfilled a previous undertaking to phone his vicar but does not account for this failure. After this reply, A permits a (0.4) gap to elapse before "oh"-receipting it and a further (0.3) gap ensues after this receipt. Both gaps constitute opportunities for the provision of the absent account and, after the second post-"oh" gap (line 7), A's subsequent nonlexicalized utterance (line 9) exhibits a continuing reluctance to advance the sequence. B's subsequently initiated account (line 10) can thus be seen to be elicited by the series of post-"oh" withholdings by A, but without the account being requested or demanded as such. In these cases, then, "oh" producers successfully rely upon the fact that the production of the particle routinely projects further talk as a means of inducing coparticipants to volunteer sequentially relevant activities. By not producing, and hence overtly withholding "oh" -projected talk that is "due" next, a speaker may induce a coparticipant to initiate or accomplish sequentially relevant activities that the withholding speaker would rather not initiate or request. 11. Conclusion Although it has been almost traditional to treat "oh" and related utterances (such as "yes," "uh huh," "mm hm," etc.) as an undifferentiated collection of "back channels" or "signals of continued attention," the observations presented in this chapter suggest that such treatments seriously underestimate the diversity and complexity of the tasks that these objects are used to accomplish. In both their variety and their placement in a range of sequence types, these objects are used to achieve a systematically differentiated range of objectives which, in turn, are specifically consequential for the onward development of the sequences in which they are employed. Within this collection, "oh" is unique in making a change-of-state proposal which is most commonly used to accept prior talk as informative. Such a proposal is, in certain of the sequence types discussed here, strongly required and regularly used. In others, the production of a free-standing "oh" may be disruptive of the development of talk, competed with, or produced in the service of special interactional objectives. All of these variations, however, testify to the 336 J. Notes Heritage deeply structured and conventionalized character of the particle's production and interpretation in ordinary talk. Unlike such objects as "yes" and "mm hm," "oh" in conversation is essentially backward looking and scarcely ever continuative. "Oh" appears to share this characteristic with other response cries discussed by Coffman (1981) and this characteristic is associated with the fact that, when it (and they) are uttered in ordinary talk, the utterer will commonly have more to say that is richer in content and more overtly directed to a recipient. In this regard, the routine use of "oh" in conversation can be viewed as instancing an exceptionally ritualized use of response crying which is nonetheless betrayed as such in the standard occurrence of additional conjoined or postpositioned turn components. "Oh" occurs most densely perhaps in the environment of questions and their answers. Consideration of its placement in these sequences may contribute toward a broader line of inquiry and such inquiry suggests that the action of questioning is not only or fully accomplished within the span of a single utterance. Rather, if the observations of this chapter are correct, the action of questioning is, even in the simplest cases, the reciprocal achievement of two turns in a sequence having, at the minimum, a prototypical Q-A- "Oh" structure. Moreover, questions and answers are themselves the media through which a variety of activities are transacted in conversation, and choices among the (thirdturn) receipt objects discussed in this chapter play a considerable role in the determination of what these activities have been, or will come to. Included within the scope of this determination will be whether a syntactically formed question was produced so as to accomplish questioning. Finally, questions and answers are also the means by which other, nonconversational or quasiconversational, activities are accomplished. Medical consultations, news interviews, courtroom examinations, and classroom interaction all fall into these latter categories. All are marked by the absence of "oh" as a routine third-turn receipt object and, in certain cases, of other routine receipt objects as well. These absences may represent specific "identifying details" (Garfinkel forthcoming) that, in combination with others, are characteristic of the management of particular kinds of institutionalized interaction, such as a news interview or cross-examination. As such, they would necessarily contribute to the maintenance and reproduction within the talk that it is some special institutionalized activity which is in progress and, of course, to the pervasive sense within such contexts that something other than conversation is in progress. 337 In the Philosophical Grammar (1974:67), Wittgenstein observes: If we were asked about it, we would probably say "Oh!" is a sigh; we say, for instance, "Oh, it is raining again already" and similar things. In that way we would have described the use of the word. But now what corresponds to the calculus, to the complicated game which we play with other words? In the use of the words "oh" or "hurrah" or "hm" there is nothing comparable. On the evidence in the present chapter, Wittgenstein's judgment would appear to be premature and indeed the uses of "oh" appear to be considerably more complicated than he suggests. For the particle participates in a wide variety of "language games": noticing; having one's attention drawn to something; remembering; being reminded, informed, or corrected; arriving at discoveries and realizations of various kinds, and many more. "Oh" is perhaps as deeply implicated in the behaviors of "coming to see something" as "Ouch" is in the domain of pain behaviors. In this way, then, the final word can perhaps remain with Wittgenstein. For it is the unreflecting, routine anchorage of "oh"s in transparent and unproblematic contexts that permits their confident interpretation in ever more complex and reticulated contexts of use. Notes 1. "Oh" is listed by the linguist Charles Fries among a collection of "signals of ｣ｯｮｾ･､＠ ｡ｴＬｾｮｩｯＢ＠ ｨｩ｣＠ｾ also includes "yes," "unh hunh," "yeah," "I 2. 3. 4. 5. see, good, etc. (Fnes 1952:49). Subsequently, "oh" is also listed as a member of a class of "noncommunicative" utterances including "wow " "zowie," "r;!-Y ｾｯＬＧ＠ and so ｯｾ＠ ｾｩ､ＮＩ＠ B.Y. contrast, Deborah James ｰｲｾﾭ ｾ＿ｳ･Ｌ｟ｴｨ｡＠ oh Ｌｾ＠ m_ ｾｯｴｨ＠ ｴｵｲｮＭｉｾ｟｡ｬ＠ ｰｯｳｩｴｾｮ＠ and in free-standing form ､･ｾｉＬｭｴ＠ semantic mformahon and are appropriate in cer( _oh 2 ) ｾｮｶ･ｹ＠ ｾ＿ｭＬｰ･ｃｩｦ｣＠ ｣ｯｮｴ･ｸｾ＠ (James 1972:163). The present chapter, in arguing that oh. ｭ｡ｾ･ｳ＠ a l?enenc c_hange-of-state proposal that is made relevant by, and particulanzed m, certam contexts, takes James's view of the matter. Gail Jefferson (1978a:221-2) has cited the use of "oh" as a "disjunct marker" and notes that its use is associated with a "display of sudden remembering." She ｨ｡ｾ＠ also noted its use as a token of special interest (Jefferson ＱＹＷＲＺｾＳＭＴＩＬ＠ Its role in conjunction with "newsmarks" (Jefferson 1981a), and, m the context of repair, its use as a token of "prior trouble now resolved" (personal communication). An intermediate case between a volunteered and a responsive production of :'oh".is perhaps that in which the recipient of a telephone call recognizes the Identity of the caller from a sample of the latter's voice. Such recognition is ｣ｯｾｭｮｬｹ＠ asse.rted with "oh." See Schegloff (1979a) for a range of instances. This sequence IS treated in more detail in Drew (Chapter 6 herein). See also the treatment of this ｳ･ｱｵｾ｣＠ in Drew (Chapter 6 herein); Schegloff (1982) and Jefferson (1981b) have discussed the placement of continuations in extended talk by a single speaker. J. 338 Notes Heritage 6. See Button and Casey (forthcoming) for a characterization of hearably incomplete news announcements. 7. See Terasaki (1976) for a range of similar instances and a characterization of their sequential organization. 8. Sacks (1974), Terasaki (1976), and Jefferson (1978a) discuss aspects of these negotiations with reference to joke-telling, announcements, and storytelling respectively. 9. An "oh"-carried proposal that its producer was previously uninformed can, of course, be modified by additional turn components. In the following instance, F qualifies her "oh" proposal with additional turn components that refer to her prior information. [NB: 1:2] 10. 11. 12. F: ((f)) Wul when didju guys ,IE:::. S: Ah: Saturday?hh F: ((f)) Oh: fer, crying out loud. I thought it wz !he end'v th'IID:nth you were go:: :i:n, ｾ＠ By these means, F asserts herself to have been previously misinformed rather than simply uninformed. Similarly, in each of the next two cases, rejections of invitations are accompanied by candidate known-to-recipient accounts. In the first, the recipient of the account is specifically invited to "remember" the circumstances (the "two other kids") which are invoked to reject the invitation. [NB:I:5] F: 'hhh Oh: care o:n. T: [ I could] n' j' s COllE dov.n= Hm:? F: =the : re, hn · t · hh I got two other kids. Reneni:Jer? -t T: Oh: :-that's ri[..:_ght, F: - eYe:: 1::: :ah:: Here the recipient's ("oh" + "that's right") receipt accepts the prior account by treating its informing as a "just-now-recollected-as-relevant" remembering of previously and independently known information. And, in the following, a similar format is deployed and emphasized with an additional postpositioned "I FER i GO:T. Completely." [Frankel:TC:l:l:l5-16] S: 'hhh So if you guys want a place tuh sta:y. (0.3) G: 't'hhh Oh well thank you but you we ha- yihknow Victor. S: tOll that's 'J'RI:GIT.= G: =:!_hat's why we were Ｌｾｅｩｮｧ＠ -. S: -t [(we) I FERtOO:T. ｾｬｹＮ＠ In various ways, then, recipients can qualify an "oh" -carried change-ofstate proposal so as to formulate it as proposing misinformation or recollection rather than simply involving a prior lack of information. Moreover, just as a subsequently produced "that's right" may qualify a turn-initial "oh," so also a turn-initial "oh" qualifies the sense of "that's right" as involving a 13. 339 "just now" recollection of something known but not previously taken into account as relevant, rather than the sense of "independent confirmation" that "that's right," unprefaced by "oh," would otherwise convey. For a characterization of how a reference to an expectably known-to-recipient event, by a party who asserts limited access to that event, solicits information from the recipient, see Pomerantz (1980a). Some discussion of source-cited reports as alternatively displaying limited access or sensitivity to the matters reported is presented in Pomerantz (1981, forthcoming). The data also evidence the prospective readiness of questioners to assume this now informed status. A substantial number of "oh" receipts occur early, that is, in "latched" or slightly overlapped positions relative to the answers they receipt. Such receipts are rarely delayed longer than a micropause. It may further be suggested that "oh" receipts, in proposing a questioner's now informed status, also implicate the questioner's acceptance of an answer as fact. In this context, informings of various kinds that are not "oh"-(or some equivalent) receipted (see note 13) are often subsequently contested. Moreover, in the environment of contested informings, those who seek to remain neutral may systematically avoid "oh" receipts. Thus Max Atkinson (1979b) has noted that arbitrators in British small claims courts, who question both plaintiff and defendent in the presence of the other, avoid the production of "oh" to receipt answers to their questions in favor of more neutral objects like "yes" and "certainly." In British news broadcasts, which are required by statute to exhibit balance and impartiality, interviewers entirely avoid such receipts. Here the avoidance of "oh" production serves both to sustain the interviewer's neutral posture and to maintain the interview as an event in which the "overhearing" audience, rather than the interviewer, is the target of the informing and in which the interviewer's role is restricted to eliciting such informings (see Heritage forthcoming). Finally, lay characterizations of talk also treat "oh" as routinely accepting what is asserted in the prior talk. Thus in Northanger Abbey (pp. 97-8 of the 1972 Penguin edition), Jane Austen depicts the following exchange: Soon after their reaching the bottom of the set, Catherine perceived herself to be earnestly regarded by a gentleman who stood among the lookers-on, immediately behind her partner ... Confused by his notice, and blushing from the fear of its being excited by something wrong in her appearance, she turned away her head. But while she did so, the gentleman retreated, and her partner coming nearer, said, "I see that you guess what I have just been asked. That gentleman knows your name, and you have a right to know his. It is General Tilney, my father." Catherine's answer was only "Oh!" -but it was an "Oh!" expressing everything needful; attention to his words and perfect reliance on their truth. In the· following report from The Times (of London), a bride tells of her reaction to the discovery that her serviceman husband will have to depart for a war zone immediately after the wedding reception. The bride's gloss is a similar one: "As we left the church I whispered: 'When have you got to go?' He simply replied 'Tonight.' I said 'Oh' and accepted it." In this connection, a strong distinction is to be maintained between freestanding "oh" that centrally functions as a backward-looking information receipt and a variety of assertions of ritualized disbelief, e.g., "yer kidding," "really?" "did you," etc., that treat a prior utterance as news for recipient. 340 J. Notes Heritage The latter systematically advance the sequences in which they participate by inviting prior speakers to, at minimum, reconfirm the substance of the prior turn's talk. Commonly, speakers in receipt of such objects reconfirm the prior and advance the informing as in (a) and (b). thus replacing the sequence-terminal 'assessment' component with pursuit of further talk vis-a-vis the news" (ibid.:63-4). These alternative possibilities are displayed in (e) and (f). (e) (NB:II:2:12] (a) 1-> A: An' Warden, had to physically rarove ' im from ' iz office, 'hhhh ｾ＠ E: Really? -. N: Yeh they'd had quite a scuffle, a:nd ..... N: (TCI(a):14:2] We're havin a h-buncha B: Ｒｾ＠ [NB:II:4:10 (r)] N: But uh I didn't get home til' · hhh two las' night I met avery,h very n:ni:ce gu:y. - -> D: Di(. )dju:::. .., N: - I: really did through the..:_se ｦｲｩｾＺｮ､ｳ＠ Jefferson (1981a:62-6) refers to these objects as "newsmarks," that is, objects that specifically treat a prior turn's talk as news for the recipient rather than merely informative. In this regard, all newsmarks project further talk by the news deliverer/newsmark recipient by reference to the news but, Jefferson reports, different newsmarks project different trajectories for such talk. Any newsmark may be prefixed by "oh" and, in many cases, the presence or absence of such a prefix plays a role in projecting different trajectories. Jefferson notes, for example, that "oh really?" regularly occurs in sequences that run as follows: (1) news announcement, (2) "oh really?" (3) reconfirmation, and (4) assessment (which is generally terminal or topiccurtailing). This sequence type is instanced in (c) and (d). (c) [NB: IV: 7: 5-6] M: How many cigarettes yih had. (0.8) 1-> E: NO:NE. 2-. M: Oh reallz? 3-+ E: No:. 4-. M: Very _good. (d) [NB:II:2:3] Hey that was the same spot we took off for Honolulu. (0.4) Where they put him on, (1.0) at that chartered place, 2-+ N: Oh real.!l? 3-+ E: y: :Yea: :h. 4-> N: Oh::?for heaven sakes. By contrast, Jefferson proposes, newsmarks formed as "partial repeats" and produced in a format "Oh" -plus-partial repeat regularly occur in sequences "within which further talk by reference to the 'news' is done" and, she continues: "That talk is either volunteered by the news-deliverer/ newsmark recipient in Slot (3), accompanying the 'confirmation' component ... or is solicited by the news recipient/newsmark deliverer in Slot (4), ｾｬ･＠ . over too(:. Oh ｾｲ･＠ yih? ( ) A: Yeh it sort'v ｾ＠ (f) (b) 341 (NB:I:1:17] 1-> E: They ｾ＠ 2... G: Oh do 3-+ E: Yeh I too much Guy, they? think so, 4-1- G: WhaCdo they cha:rge. In sum, while free-standing "oh" functions as an information receipt, "oh" -plus-newsmark regularly functions as a news receipt, with different newsmark types standardly projecting different sequential outcomes for the onward course of talk by reference to the news. Finally Jefferson notes that free-standing (i.e., not prefixed by "oh") partial repeat newsmarks regularly engender sequences such as the following: (g) (NB: IV:3: 1] (L: I'm gonna take them up to Anthony's and dye them because they dye uhb- uhb, the- perfect rna: tch. Ｒｾ＠ E: ｦｵｴｾＬ＠ ｾｌＺ＠ Yeah, Ｔｾ＠ E: Ah hah, S.... L: I mean sometimes you buy them at these places ..... Here the slot (4) "is occupied by an acknowledgment token, in contrast to() terminal assessments ... and () solicitations of further talk. .. And (such objects) tend to be followed by a 'voluntary' production of more talk by reference to the news, now in Slot (5)" (ibid., p. 65). It may be concluded from these observations that "oh" -plus-partial repeat more strongly projects recipient commitment to further talk by reference to the news than either a free-standing "oh" or a free-standing partial repeat. These considerations can be taken a further step by comparing two alternative syntactic designs of partial repeats - those that are syntactically formed as questions (e.g., "did you?") and those that are not (e.g., "you did?"). The latter form of free-standing partial repeat regularly engenders the kind of topic curtailment already observed in the case of "oh really?" Thus in (h) and (i), it can be observed that after such a newsmark, the news deliverer/newsmark recipient merely reconfirms (slot 3) the prior turn's talk and subsequently the news recipient/ newsmark deliverer produces a (slot 4) sequence curtailing assessment. (h) [TC II(a):14:15] C: Th'reason they're vacant is becuz they got'm 1 t all ｴｯｲｮｾﾷ＠ - (0.6) C: Re£lumbing the whole £lace. J. 342 Heritage 2-> E: You are? 3-lo C: Yeah. 4-> 1:: Wul goo_;_:d.h ( i) Ｒｾ＠ A: I sound ha:p[py? 3-> B: Ye:uh (0.3) 4-> A: No:, [Adeto: 2: 15-16] (n) 1-> J: Think ｾﾷｲ･＠ gonna get a ra.!_:se, first of next rronth. 2-> G: You are? ((pause)) 3-t S: We are, 4-> G: Congratulations. . In (j), after a. similar free-standing newsmark, the news recip､･ｨｶｾｲ＠ curtails elaboration of the prior turn's talk by reIssumg a pnor query m the slot 4 position. ｾ･ｮｴＺｷｳｭｲｫ＠ (j ) [NB: I : 1 : 5] G: Think he'd like 1r E: to go? [Played golf with 'irn yes:terday 4 at San Cletrente. 2.... G: Yuh did. 3-t E: Uh huh, 4 .. G: Think he'd like tub gu? . Additionally, however, this form of free-standing partial repeat may proJect a further sequence type that is unique to it: disagreement. Thus in each of the following cases, after a free-standing partial repeat newsmark, the news recipient/newsmark deliverer moves to contradict the prior (slot 1) assertion and its subsequent (slot 3) reconfirmation. (k) [Rah:B:1:1DJ(12):2] (0.9) 3.., I: tEy?4-+ J: No she hasn't? (1) 2-> B: I do? :>. A : treaiun. Ｔｾ＠ B: mNo? I'm no:t, In the following case, an initial assertion that gets a similar free-standing partial repeat is guardedly moderated in slot 3. (o) [Travel Agency:10:ST] 1-+ A: Derek we have no hea: :t. 2 ... D: Yih have no hea:t? 3-t A: We, can't feel any.= Subsequent disagreement by a newsmark producer is, in the data to hand, uniquely associated with free-standing partial repeats that are not syntactically produced as queries. Where such partial repeats are syntactically produced as queries, subsequent disagreement does not occur. Thus a freestanding partial repeat that is not syntactically produced as a query alternatively projects either (1) sequence curtailment (as in [h]-[j]), or (2) disagreement (as in [k]-[o]); see also Pomerantz (Chapter 4 herein). By contrast, cases in which a syntactically nonquery-formed partial repeat is prefixed by "oh" run similarly to "oh"-plus-query-formed-partial-repeats (see cases [e] and [f]); either slot (3) volunteered continuations develop (e.g., [p]) or such continuation is solicited by the news recipient!newsmark deliverer in slot 4 (as in [q]). (p) [Rah:1:8] V: Oh I ｾｴ＠ Jani_;_e, ･ｨＺｲｮｸ｟ｳｴｾｹ＠ an' she'd had a fo: nn fran the Age Conoern about that jo:b. h= 2-> J: =Oh she has?:>. V: sO: eh she w'z ｾｮ､ｩｧ＠ 1 E [Earthquake Broan:1] (q) That broom you lookin' for is on the slanding a'the stairs. (0.3) 2-+ J: It i:s? (0.2) 3-t T: Yea(h 4-> J: I don' t see any broom there, (rn) ｡ｾｹＮ＠ (0. 7) -- I: Ja[nie:, 2... J: tJanie has. [TG:3] 1.... A: ... You sound very far .I: Well she's gone torn: eh : eh : Chester: . { 343 Notes [TG:1] 1-> B: Why wh:at'sa matter with y-Yih sound HA:PPY hh A: [Nothing. ' Ｒｾ＠ [NB:IV:14:1] 1-> E: Well, we just got do:wn,hh L: Oh you di_:_[d? Yea:uh. 3-lo E: 4-> L: Oh ha.v co: liE' In sum, whereas "oh you did?" appears effectively equivalent in sequential terms to "oh did you?" a parallel equivalence does not hold between "you did?" and "did you?" Whereas "you did?" may project disagreement and, in projecting disagreement, may license "paranoid" responses as in (o), "did you?" does not project the possibility of upcoming disagreement. Two conclusions may be drawn from these observations. (1) Whereas a 344 J. Heritage free-standing "oh" rarely promotes the onward course of an informing sequence, it is instrumental in combination with most newsmarks (except "really?") in promoting substantial further talk to the news receipted with the "oh" -plus-newsmark combination. "Oh" thus generally strengthens a newsmark's proposal of commitment to the materials it receipts as a potential topic for further talk. (2) Whereas a free-standing syntactically nonquery-formed partial repeat may project disagreement, an "oh" prefix to this form of newsmark entirely eliminates this possibility and constitutes further indirect evidence for the possibility that "oh" functions as an information receipt that is used to accept the information receipted as fact (see also note 12). Finally, it may be noted that where a free-standing "oh" is itself queryintoned, it may function as a newsmark that promotes further talk to the news it marks. Such a use of "oh" is rare in the data to hand, and no attempt is made here to characterize its functioning. 14. In this context, answers to exam questions (Searle 1969:66-7) are never receipted with "oh," but with some version of confirmation/ disconfirmation (see McHoul1978, Mehan 1979, for a range of instances). By this means, among others, the pedagogical frame of classroom interaction is continuously sustained within the talk. 15. None of these conclusions should be taken as implying that, in turns responsive to counterinformings, an "oh" preface invariably projects acceptance of the counterinforming by the counterinformed party. In the following case, involving conflicting identifications of bird song, Ben's "Oh yeh?" challenges Bill's counterproposal that the birds are quail, and Ben subsequently follows it with a reassertion of his initial identification that they are "pigeons." [JS: II: 219-20] Lissena E:_geons . Ben: (0.7) Ellen: ｲｾｯＭ｣Ｚ＠ :coo::: Bill: 'Quail, I think. _. Ben: Oh yeh? (1.5) Ben: No that 's not quai 1 , that 's a pigeon, In the small number of cases to hand, "oh" -prefaced challenges to counterinformings are invariably question-intoned, b1ft no further observations as to their character can be offered at the present. 16. On the types of repair and their initiation, see Schegloff, Jefferson, and Sacks (1977). 17. In the extended repair sequences of (35) and (36), more than "oh" receipts are provided for the finally successful repair event, e.g., "oh yeah" and "oh okay." 18. The major alternative means of achieving exit from a repair sequence involves simple continuation of the sequence in progress prior to the repair. This is illustrated in the sequence below. ｛ｆｲ｡ｮｫ･ｬＺｔｃＱＲｾＳ＠ S: G: =·hhh Uh:m, ·tch"llltllh Who w'yih fta:lking to. (0.6) Jis' no:w? Notes S: 345 · hhhh No I called be-like between e leI ｶｾｮ＠ .. G: en' I: wasn'= =talking to a: nybody ..... Here G initiates repair on S's prior question. FollowingS's provision of the repair, G proceeds immediately to answer the now repaired question and hence accomplishes an exit to the repair sequence that she initiated. 19. For a discussion of overlap competition, see Jefferson and Schegloff (1975). 20. Gail Jefferson (1981b) has distinguished between "mm hm" as projecting "passive recipiency" and "yeah" as implying that its producer may shortly assume "active speakership." 21. In the case of (23) J's "Oh:ah" is sympathetically intoned and seems to treat the unexpectedly early departure of V's visitors as a source of disappointment to V. Here V's overlapped talk may be designed to stifle any further such expression and to delete such sympathy as is carried in the intonation of J's "oh." V's overlapped talk may, in sum, be designed to avoid or curtail any treatment of the visitors' departure as a source of disappointment for which sympathy might be appropriate. Later in the call, V again refers to the visitors' early departure but again discourages her coparticipant from treating the matter as a complaint or as a subject for sympathetic affiliation.