This art icle was downloaded by: [ Universit y of York] On: 28 May 2015, At : 15: 42 Publisher: Rout ledge I nform a Lt d Regist ered in England and Wales Regist ered Num ber: 1072954 Regist ered office: Mort im er House, 37- 41 Mort im er St reet , London W1T 3JH, UK Research on Language and Social Interaction Publicat ion det ails, including inst ruct ions f or aut hors and subscript ion inf ormat ion: ht t p: / / www. t andf online. com/ loi/ hrls20 Or-Prefacing in the Organization of SelfInitiated Repair a Gene H. Lerner & Celia Kit zinger a b Depart ment of Sociology, Universit y of Calif ornia, Sant a Barbara b Depart ment of Sociology, Universit y of York, Unit ed Kingdom Published online: 25 Feb 2015. Click for updates To cite this article: Gene H. Lerner & Celia Kit zinger (2015) Or-Pref acing in t he Organizat ion of Self -Init iat ed Repair, Research on Language and Social Int eract ion, 48: 1, 58-78, DOI: 10. 1080/ 08351813. 2015. 993844 To link to this article: ht t p: / / dx. doi. org/ 10. 1080/ 08351813. 2015. 993844 PLEASE SCROLL DOWN FOR ARTI CLE Taylor & Francis m akes every effort t o ensure t he accuracy of all t he inform at ion ( t he “ Cont ent ” ) cont ained in t he publicat ions on our plat form . However, Taylor & Francis, our agent s, and our licensors m ake no represent at ions or warrant ies what soever as t o t he accuracy, com plet eness, or suit abilit y for any purpose of t he Cont ent . Any opinions and views expressed in t his publicat ion are t he opinions and views of t he aut hors, and are not t he views of or endorsed by Taylor & Francis. The accuracy of t he Cont ent should not be relied upon and should be independent ly verified wit h prim ary sources of inform at ion. Taylor and Francis shall not be liable for any losses, act ions, claim s, proceedings, dem ands, cost s, expenses, dam ages, and ot her liabilit ies what soever or howsoever caused arising direct ly or indirect ly in connect ion wit h, in relat ion t o or arising out of t he use of t he Cont ent . This art icle m ay be used for research, t eaching, and privat e st udy purposes. Any subst ant ial or syst em at ic reproduct ion, redist ribut ion, reselling, loan, sub- licensing, syst em at ic supply, or dist ribut ion in any form t o anyone is expressly forbidden. Term s & Downloaded by [University of York] at 15:42 28 May 2015 Condit ions of access and use can be found at ht t p: / / www.t andfonline.com / page/ t erm sand- condit ions RESEARCH ON LANGUAGE AND SOCIAL INTERACTION, 48(1), 58–78, 2015 Copyright © Taylor & Francis Group, LLC ISSN: 0835-1813 print / 1532-7973 online DOI: 10.1080/08351813.2015.993844 Or-Prefacing in the Organization of Self-Initiated Repair Gene H. Lerner Department of Sociology University of California, Santa Barbara Downloaded by [University of York] at 15:42 28 May 2015 Celia Kitzinger Department of Sociology University of York, United Kingdom This report identifies a distinct, and distinctly positioned, element of the repair segment—the repair preface—and focuses on or-prefacing to introduce the practice of repair prefacing and to develop an analysis of one preface type. Although or-prefaced repairs do substitute one formulation for another, the or-preface shows that the trouble source formulation is not being discarded altogether, thereby mitigating the reparative character of the repair operation. We also examine expanded or-prefaced repair segments for what they reveal about the part or-prefacing plays in repair. Additionally, and as part of our explication of repair prefacing, we show how some same-TCU repairs (with and without or-prefaces) can be mounted without progressivity-disrupting hitches or alerts. Data are in American and British English. In reporting on their discovery of “an organization of repair in conversation,” Schegloff, Jefferson, and Sacks (1977, p. 376) note that “the basic format for same-turn repair is . . . self-initiation with a non-lexical initiator followed by candidate repair.” Extracts 1 and 2 show this “basic format”: In each the speaker initiates repair nonlexically by cutting off production of the trouble source word (“larger” in Extract 1 and “cop” in Extract 2) and then follows the initiation with the repair solution (the replacement words “smaller” in Extract 1 and “officer” in Extract 2).1 (1) [COH 2001.T1.1] A: . . . And it turned out we could have done mcht t about a size la:r- smaller. (2) [from Jefferson, 1974] Bassett: When thuh ku- officer came up . . . A version of this report was presented to the International Communication Association, Chicago, 2008. Correspondence should be sent to Gene H. Lerner, Department of Sociology, University of California, Santa Barbara, CA 93106-9430. E-mail: lerner@soc.ucsb.edu 1 Names and other identifiers have been anonymized with the exception of data cited from published sources or in the public domain. OR-PREFACING IN SELF-INITIATED REPAIR 59 This “basic format” for same-TCU repair (i.e., repair initiated in the same turn-constructional unit as the trouble source) can also include turn-constructional framing—as in Extracts 3 and 4 where speakers replace “we” with “I” repeating “c’d” as a preframe to the repair solution in (3) and repeating “were going” (with appropriate grammatical substitution) as a postframe to the repair solution in (4). (3) [Schegloff et al., 1977: 366] Naomi: But c’d we- c’d I stay u:p? (4) [Holt: X(C)1:1:3] Downloaded by [University of York] at 15:42 28 May 2015 L: .hhhh We were going t- <I wz going t’ring yesterday Some cases of self-initiated repair in the same TCU depart from this “basic format” in that an additional item—additional, that is, to the repair solution, nonlexical hitches and any pre- or postframe—is introduced into the repair segment. This added element—which we are calling the repair preface—is a distinct and distinctly positioned constituent of the repair segment and can take a variety of forms. The full range of forms repair prefacing can take (including or, well, no, actually, sorry, and I mean, among others) is presented in a companion report (Lerner & Kitzinger, 2010). In this report we concentrate on or-prefacing as a way to introduce the practice of repair prefacing and to develop an analysis of its features for a single preface type. Four cases of same-TCU or-prefaced repair are displayed as Extracts 5–8. (5) [999 UFO/Wales] Caller: I just need to tell y- or inform you: (.) that across the mountain there’s a very bright stationary object. (6) [DB01] Meg: I don’t know wh- or I didn’t know whether she might prefe:r (0.2) to speak to somebody who knows what she’s talking about . . . (7) [HB58] Stacey: An:d s:- six years on (.) or five ‘n a half years o:n hhhh uhm I’m now having my second chi:ld . . . (8) [PP5] Helen: He thought they were going to come out and tell- tell us: tha:t (.) or tell him that ‘e- ‘e could only save one of us As in Extracts 1–4, in each of Extracts 5–8 the speaker is partway through a TCU when repair is initiated nonlexically (e.g., with a cutoff in Extracts 5 and 6; a hesitation in Extracts 7 and 8) and this is followed by a repair solution that offers a different formulation of a trouble source term: “tell” is repaired to “inform” (5); “don’t” to “didn’t” (6); “six” to “five and a half” (7); and “us” to “him” (8). As in Extracts 3 and 4, the speakers in Extracts 5–8 use turn-constructional framing: the “y-” sound returns as the beginning of “you” in Extract 5, postframing the repair solution; “years on” postframes the repair solution in Extract 7; and Extracts 6 and 8 have both Downloaded by [University of York] at 15:42 28 May 2015 60 LERNER AND KITZINGER pre- and postframing (“I” and “know wh-,” Extract 6; “tell” and “that,” Extract 8). Crucially, however, Extracts 5–8 differ from Extracts 1–4 in that what first follows repair initiation is not the repair solution, nor its turn-constructional framing, but a newly introduced item: or. Because, as Extracts 1–4 demonstrate, speakers can (and recurrently do) go directly from repair initiation to repair solution (e.g., Extract 5 might have run off as: “I just need to tell y- inform you . . .”) the question arises for analysts of talk-in-interaction, as for recipients of the talk, what are speakers doing when they add or to an otherwise well-formed repair segment? Previous research on or has largely focused on its use without regard to its specific repair context. Linguists have suggested two uses of or: (a) its common use as a coordinating conjunction to introduce the second of two components where the second is cast as an alternative to the prior component, and (b) in “apposition” (Meyer, 1992), to cast the second as a reformulation of the prior component. An example of “the ordinary nonappositional or” (“You can use either butter or oil in this recipe,” Blakemore, 2007, p. 312) is contrasted with its appositional use in the following two examples: “That glissando, or slide, must be played with a flourish” (Blakemore, 2007, p. 333); “Cuts, or in more technical terms, lacerations, can be treated in two ways” (Blakemore, 2007, p. 324). Blakemore (2007) argues that appositions introduced by or are metalinguistic in character and are used “to communicate that there are alternative means of communicating ONE AND THE SAME CONCEPT” (Blakemore, 2007, p. 327, emphasis in original). Previous conversation analytic work has shown how or as a coordinating conjunction heralding an upcoming alternative can be deployed in other-initiated repair, e.g., in designing an “alternative question” (one inviting as an answer one of the two alternatives presented in the question) targeting a problematic hearing: for example, “He jumped outta the convertible goin sixty miles an hour,” followed by the repair initiation “Sixteen or sixty” (Koshik, 2005, p. 197). In addition, Lerner (2004, p. 179) shows that a stand-alone or can be used to initiate repair by prompting the prior speaker to expand the scope of an offer. In the research reported here we show that or-prefacing of repair solutions—and repair prefacing more generally—is a distinctive practice in the organization of self-initiated repair. This report draws on our analysis of a collection of about 250 instances of self-initiated or-prefaced repair—compiled from field recordings and broadcast media in American and British English2 — to explore how or-prefacing operates and what it can accomplish. We show how or-prefacing ties a yet-to-be-voiced repair solution to its associated trouble source in a particular way: It casts a relationship between a repair solution and its trouble source that exploits the routine use of or as an indicator of an upcoming alternative to something just said. All of the extracts shown so far display repairs that are initiated in the same TCU as the trouble source, but or can also be deployed in repairs that are initiated at subsequent positions for self-initiation of repair. Many cases in our collection are launched in the transition space, i.e., after the speaker has come to a possible completion of the TCU containing the trouble source but before they continue with further talk (and before—or sometimes simultaneously with—talk from another speaker). Instances of transition space or-prefaced repair are shown in Extracts 9 and 10.3 2 Although we restrict our investigation to or-prefacing in English, the practice of or-prefacing of self-initiated repairs can be seen to be employed across a range of diverse languages: e.g., in Norwegian (Sikveland & Ogden, 2012, p. 182), Finnish (Laakso & Sorjonen, 2010, p. 1164), and Japanese (Rosenthal, 2008, p. 229). 3 Repairs initiated in the transition space—whether or not or-prefaced—can be one of two distinct types. Both Extract 9 and Extract 10 can be described as incidentally transition space repairs in that each targets a trouble source that is the OR-PREFACING IN SELF-INITIATED REPAIR 61 (9) [SBL 1.1.12] B: I don’t think she’s asking so awfully much, she started out asking thirty thousan:. Or twenty nine thousand five hundred, but I think she’s gone down some. Downloaded by [University of York] at 15:42 28 May 2015 (10) [BCC 462] Clt: Mdw: Clt: Mdw: Clt: Mdw: .hhh Ho:w many weeks¿ Thirty nine=but she’s due on Boxing Da:y.=[ Or on]= [mm hm]: Christmas E:ve so: So why are they inducing. Basically ‘cause they didn’t want her to have her ba(h)by on Chri(h)stmas Da(h)y . . . Moreover, as is the case for other self-initiated repairs, or-prefaced repairs can also be launched after possible completion of the speaker’s subsequent TCU (Extract 11), or in third turn, after an intervening turn by another speaker (Extract 12). (11) [BBC4 Today 12 March 2011] Pre: War: Pre: As far as: uh Tokyo is concerned uh Mister Warren what- what’s happening there this [morn]ing. How is it.[ [Or this after[noon I]= [( )] [.hhh [( [Well u:h]) =think it is. (12) [BCC_257] Julie: Clt: Julie: She put on four ounces so she was allowed ↑ho:me! Uh- put on four ounces whe:n. Or ninety grams. Which I think is about four ounces. Additionally, we have found that or-prefacing can be used in association with a range of first-order repair operations (Schegloff, 2013, p. 64). Those we have seen in our collection so far are replacing (Extract 12), inserting (Extract 13), deleting (Extract 14), and searching (Extract 15). terminal item of the just-completed TCU, and so the repair is initiated just after the trouble source, as are many same-TCU repairs. It is only incidental to the position of the trouble source that these are initiated in the transition space. However, some repairs initiated in the transition space are not there incidentally but are initiated at some distance from the trouble source, as can be seen in the following extract. [Auto Discussion] 01 CUR: =en ’e siz yah, I gotta get rid’v it though. 02 (0.5) 03 CUR: I said why dihyou have tih get rid’v it. ’n ’e sid 04 well I’m afraid my wi:fe will get it. <er my ex wife. 62 LERNER AND KITZINGER (13) [BBC R4 World at One 20.11.08] Kearney: And presumably: the problem of piracy can’t be resolved .hhh unti:l (.) stability or more stability is brought to Soma:lia. (14) [Clacia] C: Marty Hill was saying that ’e had a really good articleor an article in the paper the other night. Downloaded by [University of York] at 15:42 28 May 2015 (15) [Kate: Birth B6 Audio 3:13] . . . I think they’ve developed that as a bit of a:: .hh stopgap but .hhh ◦ o:r not a stop gap in a negative way but you kno:w uhm .hhh an’ so I know that they’re quite into it over the:re ON THE REALIZATION OF REPAIR: FORMING A DISJUNCTION WITH THE TCU-SO-FAR It is important to note that some or-prefaced repairs are not introduced by a hitch in the talk. In this section we describe practices through which same-TCU repairs that are not heralded by any sort of hitch in the talk are nonetheless recognizable as repair. The repair segment begins when there is a “possible disjunction with the immediately preceding talk” (Schegloff, 2000, p. 207) and ends when the speaker resumes the talk that had been suspended for the purposes of repair. The initiation component of the repair segment is whatever is used “to put its recipient on alert that what follows may not be more of whatever unit has been transpiring, but that the progression of the talk may be being interrupted for repair” (Schegloff, 1992, p. 1315). In each of the same-TCU or-prefaced repairs shown thus far (Extracts 5–8) there was an initial hitch (e.g., a word cutoff or hesitation) that disrupted the pace of the ongoing talk, thereby alerting recipients to the possibility of repair. It is on the inspection of what follows on from such alerts that participants can determine if the TCU-in-progress has just continued on (as in Extract 6 where there is a sound stretch on “prefer” and then a pause, but subsequently the speaker continues on without repairing the talk) or whether some type of repair operation on a trouble source is now underway. Thus such hitches are not sufficient to indicate that repair is underway; but as we will now show, they are also not necessary to show that repair is underway. This observation expands our understanding of how repairs can be shown to be underway. This holds for both prefaced and unprefaced repairs alike. Some same-TCU repairs (including some or-prefaced repairs) lack any such hitch in the talk. Because this type of repair occurs in our data, but has not been the focus of previous research, we present some cases of it here and explain how they are, nonetheless, recognizable as repairs. We begin with a case that does not include or-prefacing. In Extract 16 there is no preliminary alert either on, or subsequent to, the trouble source term “said,” following which “told” is produced as a repair solution. OR-PREFACING IN SELF-INITIATED REPAIR 63 (16) [MDE:MTRAC:60-1:5] Downloaded by [University of York] at 15:42 28 May 2015 A: Did you hear what he said told Sarah .hhh that they were considered the cutest couple in the [cla:ss.] Although there is no preliminary alert of any sort here, the repair is made apparent (and the trouble source is indicated) by the adjacent placement of a word that would not ordinarily follow on from “said.” So, here the pace of the talk continues, with a next word smoothly following “said,” but it is not a syntactically (i.e., projectably) suitable next word that is placed adjacent to it (see Lerner, 1996, pp. 256–258)—and this replacement repair is then confirmed by the subsequent word that recognizably resumes the TCU-in-progress. It is the status of “told” as a co-class member or synonym for “said” and the contrastive stress placed on it that reveals this adjacently placed word as a replacement for “said,” rather than as the next word for the TCU-so-far. Or-prefaced repairs in the same TCU can likewise be initiated without any preliminary production hitch. In Extract 17 or occurs as a next word with no preliminary hitch in the talk. (17) [ATC 2-25-05] Hut: ‘Bout a year ago uh I created or we: created .hhh a new (.) position of National Intelligence Officer fo:r Transnational Threats. As in (16), here too the pace of the talk continues, one word after another. However, following the or, a co-class formulation of self-reference (the more inclusive formulation, “we”) is produced with contrastive stress and postframed by “created,” displaying it as a replacement for “I,” rather than as the next word for the TCU-so-far. In addition to some same-TCU repairs, repair initiation without such hitch-generated alerts is frequently the case for transition space repairs. This is so because they are launched at or just after possible completion of a TCU, and so intra-TCU hitches are not an available resource.4 In Extract 18 the repair solution is produced without any prior indication that repair may be underway. (18) [SBL: 3:1:2 (from Schegloff et al., 1977)] B: . . . then more people will show up. Cuz they won’t feel obligated tuh sell. tuh buy. In addition, or-prefacing (as in Extracts 9 and 10) as well as other preface types are common in transition space repair. As Schegloff (personal communication) notes, many transition space repairs are begun with such lexical items, including those we have identified as repair prefaces (or, well, no, actually, etc.). Even when there is no hitch in the progressive realization of their talk, speakers can show that they are repairing what they have said by fashioning the repair solution with pre- and/or 4 Latching and speeding up of the pace of the talk do occur here. However, these are not repair-specific practices; they can be associated with any continuation of talk into the transition space—because it is a transition-relevance place. Downloaded by [University of York] at 15:42 28 May 2015 64 LERNER AND KITZINGER postframing or by making repair evident by using such tying devices as co-class and contrastclass membership or through a pro-form to full form substitution. Thus, “disjunction with the immediately preceding talk” is realized in and through the repair solution and any accompanying turn-constructional framing. That there is (self-initated) repair emerges from the repair segment rather than from an alert to the possibility of repair beforehand. Thus far we have introduced or-prefacing as a distinctive practice employed in self-initiated repair of talk-in-interaction; we have shown where or is positioned in the repair segment, and we have shown that or-prefacing is used across a range of positions for self-initiation of repair and in association with a range of repair operations—but we have not yet described what it adds to the actions those repair operations accomplish. In the rest of this article we will first describe the use of or in nonrepair contexts, then locate or-prefaces in relation to other types of repair prefacing before focusing on the use of or-prefacing in and as action. Finally we examine expanded orprefaced repair segments, ending with those devoted to indicating that “this is repair.” OR IN AND AS ACTION The use of or in self-initiated repair—in whatever position within the repair initiation opportunity space and irrespective of the first-order operation of the repair—is not far removed from its use elsewhere in talk-in-interaction. As linguists have noted, or can be understood either as introducing the second of two distinct alternatives5 or as introducing an alternative formulation of the first (i.e., in linguistic terms, either as a simple coordinator or as a coordinative apposition, Quirk, Greenbaum, Leech, & Svartvik, 1985, pp. 1301–1302). It is sometimes difficult, in practice, to distinguish between these two uses—and it can be the action context within which or is deployed that informs which of these two uses is relevant on any particular occasion. Thus, in Extract 19 a helpline call-taker is informing Fran (who doesn’t know the technical term for what went wrong with her labor) about what she may find written in her medical records; she gives two possible formulations of the same condition—a classic appositional construction (see Meyer, 1992, p. 44, ex. 128a). However, in terms of the action the speaker is engaged in here, these two formulations are also alternatives: One or the other may be on her medical records. (19) [BCC 508] Clt: Fra: Clt: Fra: And you may fi:nd on you::r (0.2) m::: labour report >you know< on- on- on the no:tes:: Yes that it says “deep transverse arrest” or “D-T-A”. Okay. I’m gonna check that. On the other hand, what might appear to be two distinct alternative referents may actually be alternative ways of referring to the same thing, as in Extract 20, where Lesley informs her husband about the reduced cost of their son’s driving lessons, if a family car is used. 5 Casting both components as “distinctive” alternatives does not imply that both are always understood as viable alternatives—e.g., if one alternative is seen as “not serious.” OR-PREFACING IN SELF-INITIATED REPAIR 65 (20) Holt SO88(II).2.3. Downloaded by [University of York] at 15:42 28 May 2015 01 Lesley: And he would (.) give him if he used his own ca:r 02 or Katherine’s .hh[h h h h (0.4) he would only charge= 03 Skip: [Yes, 04 =seven fif:ty for a lesson. The recipient’s understanding that “or Katherine’s” (line 2) is a repair, substituting the formulation “Katherine’s [car]” for “his own car”—and that or is used here in an “appositional” rather than “alternative” sense—depends upon his knowledge that their son does not own a car. In action terms, the trouble source (“his own car,” line 1) is a formulation selected to account for how the price of a driving lesson might be reduced (by using a car other than the driving instructor’s), and the repair (line 2) is a more precise formulation that correctly attributes ownership of the car to one member of the family rather than the other (although which member of the family actually owns the car is irrelevant to the cost of the driving lessons, which is why it was disattended in the trouble source formulation). In Extracts 19 and 20 the two items connected by or in a sense “belong together” as different ways to refer to the same thing (a condition [19], or a car [20]). However, “belonging together” is not necessarily found in the items themselves; it can be bestowed on them precisely because they are conjoined with or—which casts them as alternatives. This type of colligation (Jefferson, 1986) puts two or more things together as a way to propose that they go together in some fashion. Yet Jefferson (1986) also shows that when or is used as part of constructing the alternative to some formulation proposed by another speaker, it can implement a move away from that formulation without explicitly rejecting it. In Extract 21, Ben is advising his son Bill about how to construct a lamp and recommends the use of “thin pipe,” which he subsequently formulates (after Bill hesitates) as “quarter inch.” (21) [Sch:I:14-15 in Jefferson, 1986] 01 02 03 04 05 Ben: Bill: Ben: Bill: Ben: You ordinarily use thin pi:pe, Yeah that’s what I would want ve:ry thin. Like uh:: Quarter inch. Quarter inch, or ha- or half in[ch. [No half is too big. Bill’s response (“Quarter inch, or ha- or half inch,” line 4) is a repeat of his father’s proposal followed by an or that introduces an alternative width of pipe. Jefferson suggests that Bill may have formulated his turn in this way (casting the widths as alternative using or) to avoid proposing, “Not quarter inch, half inch,” and thereby explicitly disagreeing with or correcting Ben. Bill is: . . . making EQUIVALENT ALTERNATIVES out of what, for him, is “TOO THIN AND JUST ABOUT RIGHT.” That is, just as Ben proposes about a half inch that it’s “. . . too big,” Bill could have, but is avoiding saying “No, quarter is too thin.” (Jefferson, 1986, emphasis in original) So, in an environment in which another party has offered a candidate proposal, a speaker can first accept that candidate by repeating it but then move away from it by proposing another Downloaded by [University of York] at 15:42 28 May 2015 66 LERNER AND KITZINGER candidate as just another alternative. Although colligation—as a format—seems to indicate they are “equivalent alternatives” (to use Jefferson’s phrase), the very fact of adding the alternative weakens the acceptance of the other party’s proposal but does so without explicitly rejecting it (also see Lerner, 1994).6 In summary, when or is used in nonrepair contexts it can herald an alternative within the speech of a single speaker—casting what will come next as the alternative to what came before the or. And as Jefferson (1986) has shown (for a particular action environment), or can also be used to move away from an initial formulation without overtly rejecting it. As we will show, or-prefacing of repair solutions is convergent with both of these features in that it casts the repair solution (which comes after the or) as the alternative to a trouble source (that came before the or) while moving away from (but not rejecting) it. Casting repair solution and trouble source as alternatives downplays the difference between trouble source and repair solution, while still establishing the latter as the favored alternative. When or projects a next component, it can be understood as prefacing that component; or-prefacing establishes a relationship between the projected next component of the talk and a just-previous component of the TCU-so-far. As such or not only projects something next but also prepares the way for locating the trouble source in the TCU-so-far. A component of the TCU-so-far can come to be seen as the trouble source (once the alternative to it is produced) because or distinguishes a yet-to-be-specified element of the TCU-so-far as the component to which the projected next component will be the alternative. In short, the orpreface formally points forward to an alternative and in so doing also points backwards to what it will be an alternative to—and as such the or-preface casts repair solution and trouble source as alternatives. Furthermore, in comparison to its use elsewhere in talk-in-interaction, an or-preface—when employed as part of self-initiation of repair—forms an asymmetrical relationship between repair solution and its trouble source: It indicates that the repair solution should be understood as the favored alternative. This result rests on (a) the practice of using or to conjoin two constituents as alternatives to each other, and (b) its deployment within a repair segment, where the second alternative will therefore be understood as a repair solution that is thereby the favored alternative to the original constituent, now understood as its trouble source. In or-prefaced repair the or works, as it does in other-than-repair environments, to conjoin two components as alternatives to each other within the speech of a single speaker. In other environments these components may be treated either as equally valued or as differently valued. But, because the second component in a repair segment is cast as a repair solution, it is thereby systematically constituted as the favored alternative. And, as with the use of or in colligation (Jefferson, 1986), or-prefacing of repair diminishes the extent to which the repair rejects the trouble source formulation as wrong because the or-preface links the repair solution back to its trouble source as “belonging together”—and thus casts it as more of an alternative to, than as a replacement for, the trouble source. 6 Other conversation analytic work has focused on or—and its equivalents in other languages—in turn-final position (e.g., “are you a bit scared of it or?”) where the “alternative” projected by or is not actually produced. Rather or is used to relax the preference structure of the turn toward a nonaligning response (Blöndal, 2008; Drake, 2013; Lindström, 1999). OR-PREFACING IN SELF-INITIATED REPAIR 67 OR-PREFACING AMONG THE REPAIR PREFACE TYPES Different repair preface types can cast different kinds of relationships between repair solutions and their trouble sources. An examination of Extracts 22 and 23 allows for an informative comparison because the repairing action transforms the trouble source (a time formulation) in the same way: The trouble source indexes a time late on one day, and the repair solution indexes a time early the next day, but in the first case the repair solution is or-prefaced, whereas in the second case it is well-prefaced. Downloaded by [University of York] at 15:42 28 May 2015 (22) [BCC 367] 01 02 03 04 05 06 07 08 Sam: Clt: Sam: Clt: Sam: On the Thursday I discharged myse:[lf. ] [Mm] hm. I had o:ne good night’s slee(h)p! O:h! Thank heavens! Mm, Right. (0.2) Sam: Frida:y ni:ght (0.2) ◦ >or actually Saturday morning<◦ (0.2) I went into labour. (23) [Holt_SO88(II).2.7] 01 02 03 04 05 06 07 08 09 10 Les: Cla: Les: Cla: Les: Cla: Les: He then turned up very: very la:te >well very early in the morning really<= Yes= =after two: [.hhh] and it had- transpired= [yes ] th’t he’d been at the Waggon ‘n Horse:[s ] [Y]es .hhhh and he’d had ↑fa::r too much to drink .hhh and when he was in this happy state. . . . We note first, that the use of or-prefacing does not reflect a natural or determinate relationship between trouble source and repair solution—since here the relationship is identical, but a different preface is employed in each case. These speakers are designing their turns so as to instruct recipients about how to understand that relationship and are doing so with reference to the action in which they are otherwise engaged. In Extract 22 the caller, talking to a counselor on a birth crisis line, is about to describe her labor. Things had begun badly with premature rupture of her membranes, for which she was hospitalized for three nights, during which she was unable to get any sleep due to the noise on the ward and was dismayed at the thought of starting her labor in a state of exhaustion. She had discharged herself, against medical advice, on a Thursday, and—as she says—“had one good nights sleep” (line 3) (i.e., Thursday night), but only one because she went into labor the following night (Friday). The repair manages the precision of it having been after midnight on the Friday, hence properly referred to as “Saturday morning.” Or-prefacing here (as in the other cases we have shown) mitigates the reparative character of the repair operation—in terms of her need for sleep, it was Friday night that was disrupted, even if it was by then technically Saturday morning, Downloaded by [University of York] at 15:42 28 May 2015 68 LERNER AND KITZINGER so the trouble source formulation is not discarded altogether, and thus the more precise repair solution formulation can be understood as simply a technically better one. In Extract 23 the speaker is telling a story (she describes herself as “scandal-mongering”) about the outrageous late-night drunken behavior of a third party. The repair (from “very very late,” to “very early in the morning”) contributes to the depiction of the person’s bad behavior (drilling holes in a wall with a power drill while “in his cups”), in that it took place after midnight. Rather than use or-prefacing to treat the time-formulation repair as near equivalent to the trouble source (as in Extract 21), she well-prefaces it (and later respecifies it as “after two,” line 4). We suggest (Lerner & Kitzinger, 2010) that well-prefacing a repair solution treats the trouble source as having been problematic in some way insofar as it makes explicit that the speaker is reconsidering the initial formulation as not quite right and indicates that the repair solution will not be as straightforward as that (cf. Schegloff & Lerner’s, 2009, description of well as a responseturn preface that alerts recipients to an upcoming less-than-straightforward second pair-part). The well-preface projects the abandonment of the trouble source formulation even if, on the face of it, there does not seem to be that much difference between the two formulations of extreme lateness. Whereas or-prefacing mitigates the repair operation, well-prefacing draws attention to it as more than merely a pro forma change—in this case one that gives more prominence to an element of a story that punches up its newsworthiness.7 Finally, it should be noted that speakers can draw on other resources—besides repair prefacing—to cast the relationship between a repair solution and its trouble source in a particular way. For instance, in the following case the speaker replaces a trouble source with a near-equivalent or synonymous formulation (changing only the register) but casts the original formulation as preposterous. (24) [TG] 01 Bee: nYeeah, .hh This feller I have- (nn)/(iv-)“felluh”, this 02 ma:n. (0.2) t! .hhh He ha::(s)- uff-eh-who-who I have fer 03 Linguistics is really too much, Here, what might have been treated as a relatively minor repair from “fellow” (line 1) to “man” (line 2), where the referent is retained and the trouble source is not actually “wrong” in the way that, say, “lady” would have been, is instead produced in such a way as to instruct the recipient to hear it as a notable correction of a mistake. By repeating the trouble source formulation, the speaker frames it as the wrong thing to have said and highlights the fact that she is dismissing it in favor of a replacement. Or-prefacing, by contrast, instructs recipients to treat the repair solution as making a relatively minor adjustment to provide a favored formulation of an essentially correct trouble source (whether or not it actually is a minor adjustment). In a sense then, or-prefaced selfrepair is a practical way for speakers to move on to a second (favored) formulation without really rejecting the original formulation. The upshot of these observations across Extracts 22–24 is that different prefaces or other prerepair-solution practices can be employed to establish a relationship between trouble source and repair solution—and each preface type can do so in ways that are separate from the relationship that might otherwise obtain between them. 7 See Lerner and Kitzinger (2010) for a description of a range of individual preface types as well as their combination in the “double prefacing” of a repair solution. OR-PREFACING IN SELF-INITIATED REPAIR 69 OR-PREFACING IN AND AS ACTION In this section we describe how the practice of or-prefacing a repair solution is employed. As might be expected from the use of or in apposition, or-prefaced repairs can be coreferential— that is, the or prefaces a repair solution that can be understood as denoting the same thing as (or is synonymous with) the trouble source.8 We begin by examining several cases (Extracts 25–28) where this seems to be the case, and then we move on to show that or-prefacing is also used to cast less-than-coreferential items as alternatives and what doing so accomplishes in those cases. Downloaded by [University of York] at 15:42 28 May 2015 Coreferential Repair Solutions Extract 25 is taken from a radio news program dealing with the crisis at Fukushima following the tsunami and nuclear instability in Japan. The speaker repairs a time formulation so as to refer to the same time (effectively, “now”) in a different way—reformulating “this morning” (line 2) as “this afternoon” (line 5) in recognition of the 8-hour time difference between London (where the presenter is) and Tokyo (where his co-conversationalist is). (25) [BBC4 Today 12 March 2011] 01 02 03 04 05 06 07 08 Pre: As far as: uh Tokyo is concerned uh Mister Warren what- what’s happening there this [morn]ing. How is it.[ [Or this after[noon I]= War: [( )] [.hhh [( [Well u:h]) Pre: =think it is. War: uh uhm I’ve followed uh as as your listeners have thee udevelopments at Fukushima:. The situation is not clear at the moment. This is a recipient-oriented repair that shifts a time formulation from one that is fitted to the speaker’s circumstances—and to the overhearing British audience—to an alternative that is better fitted to his recipient in Tokyo to whom he has addressed his question. The opening question to David Warren, asking “what’s happening there this morning” is designed to elicit up-the-minute news about current events in Tokyo: The place is referred to as “there” (i.e., from the perspective of a speaker who is not in that place) and the time also is formulated from the perspective of a speaker in a different time zone where it is still before 9 a.m. (“this morning”). By repairing the time formulation, the speaker displays an orientation to his addressee’s perspective, whereas or-prefacing the repair solution links it back to the trouble source and in so doing shows that the original was not wrong (for Britain). In this context “morning” and “afternoon” are two ways of 8 Traditionally apposition has been considered a relation consisting of two units that are coreferential, thereby, as Meyer (1992, p. 3) points out, severely circumscribing the class and number of constructions that can be admitted as appositions. Meyer (p. 57) argued in favor of expanding the semantic relations holding between units in apposition in order to admit as appositions a variety of different constructions, including adjective phrases, verb phrases, and adverb phrases (see examples in Meyer, 1992, p. 66) such as the verb phrase in Extract 5. Downloaded by [University of York] at 15:42 28 May 2015 70 LERNER AND KITZINGER formulating the same (current) time, but “afternoon” is cast as the favored alternative, as he is addressing a reporter in Tokyo.9 “At the moment” (lines 7 and 8) is Warren’s elegant solution to the same problem: He uses a formulation that is not location dependent. Why design the repair to retain a connection to the trouble source (“this morning”)? That is, what might occasion the use of an or-preface in this circumstance? The repair involved only changing the formulation and not the reference. Or-prefacing is a way to indicate that both formulations are appropriate in their own way—with the repair solution simply being better suited for this recipient. Although the repair does substitute one formulation for another, the or-preface shows that the trouble source formulation is not being discarded altogether, thus mitigating the reparative character of the repair operation. The repair solution in Extract 25 refers to the same time as the trouble source but uses a different formulation. Likewise in Extract 26 the repair solution refers to the same weight as the trouble source using a different formulation: A baby’s weight is referred to first with an imperial measurement (“four ounces,” line 1) and then with a metric one (“ninety grams,” line 3). Here a grandmother is calling the helpline to announce a positive outcome for her daughter and the new baby since the previous call. (26) [BCC_257] 01 02 03 04 05 06 Can: Clt: Can: Clt: Can: Clt: The baby put on four ounces so she was allowed ↑ho:me! Uh- put on four ounces whe:n. Or [ninety grams]. Which I think is about four ounces [Since when. ] Uhm: well between bi:rth ‘n: (.) A:h that’s lovely:. In this case, the speaker makes explicit that she means to be referring to the same referent (“which I think is about . . .,” line 3), albeit with reduced certainty (“I think”) and an acknowledged degree of approximation (“about”). The production of an alternative formulation to “four ounces”—of sufficient importance to the speaker that it displaces the production of a response to the question she has been asked (at line 2)—displays an orientation to the fact that she is reporting the baby’s weight as provided by hospital authorities. Again given that she is merely converting units, the circumstances furnish an occasion for designing the repair in a way that retains a connection with the trouble source, such that the repair solution is only a more apt way of delivering the (same) news—and the or-preface explicitly casts a relationship between repair solution and trouble source that indicates this. In Extract 27 the speaker refers to the same collectivity in two different ways: first as “three teenage sons,” and then—after a display of doubt (“I think it is”)—the or-prefaced version, “three teenagers.” The repaired version avoids committing to the gender of the teenagers, which is treated (via “anyway”) as irrelevant to the reported request to pray for them. 9 It seems clear that “this morning” in the presenter’s question is referring to the speaker’s present time, rather than to the recipient’s past (i.e., Warren’s morning), given the design of the question with the present tense (the contracted “is” in line 2)—as opposed to “what happened there this morning.” OR-PREFACING IN SELF-INITIATED REPAIR 71 (27) [SBL 1.12.2] Downloaded by [University of York] at 15:42 28 May 2015 01 02 03 04 05 06 A: B: A: B: And so she called one day and was talking, and so she said, “Oh Bea”, she said, uh “Will you uhm in your prayers remember our son, and daughter in law, their three teenage Mm [hm, ] [sons,] I think it is, or three teenagers, anyway, Mm hm, The repair here is designed to manage the problem of uncertain knowledge about the intended referents by reducing the level of specificity of the reference (removing gender). There is a similar repair on person reference in Extract 32 where a person referred to initially as an “anaesthetist” (a type of doctor) is subsequently referred to in an or-prefaced repair solution as simply a “doctor”—which, like “teenagers” in place of “teenage sons” is likewise a broader and thus a more epistemically cautious formulation. In neither case, however, is the original formulation discarded altogether: The or-prefacing casts it as still possibly correct. (See Extract 33 for another instance of or-prefaced repair managing the issue of epistemic uncertainty.) The extracts shown so far have concerned coreferential reformulations of a time (25), an amount (26), and a collectivity (27). In this final case or prefaces a coreferential reformulation that concerns the ascription of an action. In Extract 28, from the beginning of a call to the emergency services, the caller delivers the reason for the call. In so doing, he first represents himself as “needing to tell” but then shortly thereafter cuts off his TCU and using an or-preface substitutes “inform” for “tell.” (28) [999_UFO/Wales] 01 PCT: P’lice what’s your emergency. 02 Clr: No::: it’s not really. I just need 03 to tell y- or inform yo:u (.) that 04 across the mountain there’s a 05 very bright stationary object. 06 PCT: Ri:ght¿ The action the caller is ascribing to himself (telling) is repaired (to informing) to better fit the occasion of a call to emergency services: He changes the action he is ascribing to himself to a less informal, more institutionally appropriate manner of speaking. This caller may be especially concerned to do so, given that he is reporting something that he himself acknowledges is not really an emergency (line 2) and as such is an institutionally inapposite reason for calling. (The “bright stationary object” about which he is “informing” the police turns out to be the moon.) Without or-prefacing (like the repair of “said” to “told” in Extract 16) this repair would have dispensed with the trouble source (“tell”) in replacing it with the repair solution (“inform”). The caller could, in effect, be heard as discarding his original choice of words. By or-prefacing the repair solution, he can indicate that the projected alternative (as a repair solution) should be understood as an alternative to what was said before the or. In this way, the caller intimates that the original formulation, although targeted as a source of trouble, is not being discarded 72 LERNER AND KITZINGER as altogether wrong. In effect, or-prefacing can be a method for mitigating repair and downplaying the difference between trouble source and repair solution otherwise spotlighted in same-TCU repair. And in this case the caller thereby downplays a difference that—in doing replacing—might otherwise foreground the institutional inappositeness of his reason for calling. (In a similar vein, or-prefacing a repair on a quantity can reveal the trouble source to have been an approximation and the repair solution to be just a more precise alternative as in Extracts 7 and 9.) Downloaded by [University of York] at 15:42 28 May 2015 Or -Prefaced Repair Solutions That Amend What Is Being Referred To In each of the following repair segments (Extracts 29–31), an or-preface casts a less-thancoreferential formulation as an alternative formulation. In Extract 29 (line 1), an or-prefaced repair that aggregates the speaker with others (Lerner & Kitzinger, 2007) substitutes an institutional self-reference (“we”) for a personal self-reference (“I”).10 This case comes from a radio report discussing the lack of long-term planning by the U.S. government concerning the threat of terrorism. The former chairman of the National Intelligence Council (an agency responsible for long-term intelligence planning) begins to say that he, personally as chairman, created a position (and office) to analyze such threats but then repairs this individual self-reference to one that encompasses an organizational structure and process. By doing so, he constitutes the decision (and credit for it) as an institutional process, rather than as accomplished exclusively by his individual action. (29) [ATC 2-25-05] 01 Hut: ‘Bout a year ago uh I created or we: created .hhh a 02 new (.) position of National Intelligence Officer 03 fo:r Transnational Threats. Mainly terrorism. .hh 04 Precisely because we felt there was a need to have 05 (.) a whole office devoted to: .h looking longer term 06 at the kind of terrorist .hh challenge we might be 07 facing five or even ten years: .hh down the roa:d .hh 08 and ho:w (.) as a policy matter we might better 09 prepa:re for that world. Here the or-preface casts the repair solution (“we”) as a better way to formulate who was responsible, but without drawing attention to his earlier self-reference by overtly abandoning it. “I” was not wrong, or even inappropriate; rather “we”—which as an aggregation situates the exchairman himself in his organization—shares the credit and awareness for the need for long-term intelligence in this area. This repair solution shows him as a team player rather than someone trying to put himself in the spotlight by taking individual credit for an action. And the or-preface shows that the original reference is not being discarded altogether and casts the replacement as just a better alternative. A bald replacement might make it appear as if he were withdrawing 10 Linguists have sometimes identified the “part/whole” relationship (here, between trouble source and repair solution) as the “least appositive” (Quirk et al., 1985, p. 1308) on the “semantic gradient of apposition” (Meyer, 1992, p. 90). OR-PREFACING IN SELF-INITIATED REPAIR 73 an ill-chosen reference (and all that that might imply) by drawing attention to the difference between trouble source and repair solution. Or-prefacing downplays the difference by fashioning a structural—i.e., grammatical—correspondence, even when trouble source and repair solution are not actually coreferential. As Extract 30 begins, the caller to a helpline is pinpointing where she is experiencing pain. She then goes on to explain with what authority she is making this claim. She invokes her profession as “a physio” (line 3) to warrant her entitlement to make such a claim and thereby to buttress the soundness of the claim itself. Having done so, she then repairs the formulation to reveal she is actually no longer a practicing professional. Downloaded by [University of York] at 15:42 28 May 2015 (30) [PP123] 01 02 03 04 05 Han: Mine (.) is my sacroiliac joint as we[ll (because)]= Clt: [ Y e : : s. ] Han: =I am: a physio myse:lf? Clt: Ri:ght. Han: ◦ >Or I was before I had my childre(h)n< [huh huh]< Here again, or-prefacing casts the repair solution as merely a temporal adjustment to the speaker’s initial claim—one that retains the professional standing that underwrites her ability to speak authoritatively about her own medical condition but removes the defeasible, merely temporal, aspect of the inference that she is still practicing her profession. Again, the or-preface shows that the trouble source formulation is not being discarded altogether, thereby mitigating the reparative character of the repair operation. Here or is used to preface a repair solution that adjusts a temporal formulation (from current to past) in the interests of precision. Unlike the temporal reformulation in Extract 25, where the trouble source and repair solution both reference the same time, here the repair solution amends the temporal referent. Nevertheless the claim made by or-prefacing that amendment is that the trouble source is not wholly wrong for the purposes of the action (warranting a claim) for which it was produced. A similar claim is made by or-prefacing a noncoreferential repair solution in Extract 31. Here the repair solution is an epistemically downgraded reformulation (“not that I’m aware of,” line 7) of her original reponse (“No,” line 4). In this case the or-preface“ is followed by “put it like this” (line 7), which explicitly casts the upcoming repair solution as another way of putting her original answer. So, here we have a case in which the speaker herself makes explict what orprefacing otherwise seems to accomplish on its own. (31) [PP5 Helen.23:00] 01 02 03 04 05 06 Clt: Hel: Hel: Clt: Hel: .hhh Do th- Do they have anything like a: a Bi:rth Afterthou:ghts: (.) service at- at [the hospit]al.= [N o : : . ] =No. No because often ther- there is a (.) [uhm] [Or ] 74 Downloaded by [University of York] at 15:42 28 May 2015 07 08 09 LERNER AND KITZINGER Clt: >put it like this< not that I’m awa:re of. .hhh Yeah. Often there is:. Uhm but not very well adverti:sed. In response to the call-taker’s resistance to her previous answer (line 5), the caller now reveals her orientation to what she is doing as “saying it again in a different (now less certain) manner.” She reasserts the same negative response but now with a caveat that her knowledge might be limited. She is backing down from her previously strong and unelaborated claim (lines 3 and 4) but making her backdown less overt, both by or-prefacing it and by designing her turn to be heard (via the coordinating use of or) as continuing on from her “No” and thereby, in a sense sequentially deleting the call-taker pushback. You will have noticed that, in this case, something comes between the or-preface and the repair solution. In the next section we examine such expanded or-prefaced repairs. EXPANDED OR-PREFACED REPAIR SEGMENTS We find that speakers recurrently introduce one or more of three elements into or-prefaced repair segments. These are: (a) additional repair prefaces so as to produce a “double preface” (e.g., or actually, or rather); (b) indicators of epistemic uncertainty (e.g., or maybe, or I think); and (c) explicit commentary on the repair operation, the trouble source, and/or the repair solution. Double prefacing is described in Lerner and Kitzinger (2010). Here we consider indicators of uncertainty and explicit commentary for what they reveal about the part or-prefacing plays in repair.11 In Extract 32 the speaker’s repair effects a shift from a specific category (the membership category, “anaesthetist,” line 1) to the encompassing category, “doctor” (line 1, that covers a range of specialist categories). She then provides a commentary (line 2) that both accounts for the substitution of “doctor” (a lack of certainty) as well as reestablishes “anaesthetist” as (probably) correct—that is, reestablishes it with a reduced degree of certainty. (32) [BCC02] 01 May: .hhhh U:m An’ this anaesthetist: or doctor 02 >I sup-presume he’s an anaesthetist< uhm 03 has said .hh that it’s not necessarily the epidu:ral . . . Here the speaker reveals, after the or-prefaced repair solution, how to regard that solution: as a broader and thus a more epistemically cautious formulation—but as one that does not discard the original formulation, and in fact its continued viability is made explicit. Repairs that adjust the certainty of an assertion seem well-suited for or-prefacing where the solution to uncertainty is the broadening (rather than abandoning) of the original formulation. In Extract 33 the trouble source is “left” (line 7) (which entails a bequest) and the repair solution is “gave” (which encompasses both lifetime gifts as well as legacies). Between the trouble source 11 Note that these two elements are repeatedly found together as in Extract 32 and 33 but do occur independently. OR-PREFACING IN SELF-INITIATED REPAIR 75 and the repair solution, the speaker produces an account of what may be “technically” wrong with the trouble source term (i.e., that the person “is still alive perhaps”). Downloaded by [University of York] at 15:42 28 May 2015 (33) [MDC 60-7b] 01 02 03 04 05 06 07 08 L: Who is Hi:rsch. (.) M: Didi Hirsch is thee[: ] L: [I ] know the name is very fam[ilia:r (to me)] M: [Well she:’s ] a very wea:lthy: Jewish lady who le:ft: o:r is still alive perhaps but gave an enormous amount of money to establish a mental health clinic. Repair is initiated on the trouble source “left” (with a stretch), and its implication is excavated in the act of rebutting it (“is still alive”) but with epistemic uncertainty (“perhaps”) such that “left” can be heard as still possibly true. Substituting “gave” for “left” solves the problem of epistemic uncertainty by employing an encompassing term (as with “doctor” in Extract 32) that does not abandon the trouble source term altogether. Here the speaker produces the account for the repair after the or-preface and before the repair solution—and so makes explicit her reason for the repair beforehand. Finally, Extract 34 presents a failed attempt at searching for a word. Here the term (“stopgap,” line 6) is produced by a helpline call-taker as an initial candidate search solution shortly after the speaker begins the kind of hesitation that can here (at precompletion position in the TCU-so-far) indicate the beginning of a possible search for a precise term. (34) [BCC:6:3:13] 01 Bet: They’re big on doulas out in Canada a[ctually ] 02 Clt: [They a:]re 03 qui:te I think because they’ve been so:: .hh 04 low on having midwi(h)ves for quite a whi(h)le 05 so I think they’ve developed that as a bit of 06 a:: .hh stopgap but .hhh ◦ o:r not a stop gap 07 in a negative way but you kno:w uhm .hhh an’ 08 so I know that they’re quite into it over the:re Although “stopgap” is not discarded altogether (as it is apparently on the right path to an apt term and thus hints at what the speaker is aiming for), the call-taker explicitly disclaims its negative connotation (as it implies a comparatively lower opinion of doulas). This disclaimer further specifies what kind of term she is aiming to substitute. In this case, there is no satisfying solution to the search, yet it is not completely unsuccessful in that the original candidate together with its or-prefaced detoxification conveys what is being searched for and what it is not. So, here again, the first alternative is not entirely discarded, even as a better alternative is sought, but not found. 76 LERNER AND KITZINGER INDICATING “THIS IS REPAIR” Some or-prefaces are treated by their speakers as vulnerable to being heard as (merely) proposing a viable alternative when they are in fact prefacing a repair.12 In this final section we examine two cases in which a repair segment expansion establishes that or is prefacing a repair, when it might otherwise not be understood in that way even after the putative repair solution is produced. In Extract 35, Ron expands the repair segment after the or (line 2) in a way that makes clear he is not just offering a second possible city for the first of two yearly events. Downloaded by [University of York] at 15:42 28 May 2015 (35) [CTFL5] 01 Ron: So: .hh he had a thousand booth sho:w twice a yea:r. 02 One in Vega:s ‘r ‘scuse me one in San Diego: an’ 03 one in Atlanta:. And it evolved . . . In this position in Ron’s TCU-so-far or—on its occurrence—still projects a second alternative in a way that retains “Vegas” as a viable first alternative (e.g., “One in Vegas or [in San Diego and one in Atlanta]”), rather than revealing that repair is underway. And, in this case, what is then produced as a repair solution will not itself defeat that hearing. Thus, the component following the or may very well be heard as an alternative to what had been said before the or, rather than as the alternative saying of what had been said before the or—as a second city, rather than a repair solution. The speaker seems oriented to this possibility when he expands the preface by adding “excuse me.” This makes clear that “San Diego” is not a second viable location but a replacement for “Vegas.” This can also be seen in Extract 36, but whereas in Extract 35 the repair-indicating “excuse me” was delivered in advance of the repair solution, here (at line 8) the repair-indicating “rather” is added subsequent to the projected alternative. (36) [Upholstery Shop] 01 J: 02 03 04 C: I wouldn’ a:rgue a[bout it (because) I’m not (particularly) sure myse:lf. [I only think it’s got tuh do with ] si:ze. [Bet you both are the sa:me hassock otto]man. 12 Not every instance of or that occurs at the outset of a repair solution is a repair preface. Sometimes or begins the repair solution itself rather than prefaces it. This use of or—aimed at inserting an alternative—can be seen in the following transition space repair (line 7). [Debbie and Shelley] 01 S: . . . come out to his house and do:, .hh like 02 spend a whole day o:n putting everything together 03 cause we don’t get the shit done while we’re at work= 04 D: =[Mm hm] 05 S: =[and if ] we stay late we’re doin’ it, and he’s like 06 well: this Saturday’s like the only time we could 07 do it.=Or Sunday, and I’m like we:ll .hhh I do’kno::w. Here “or Sunday” is not replacing “Saturday” but adding an alternative to it as a way to broaden the reference to encompass the entire weekend. OR-PREFACING IN SELF-INITIATED REPAIR Downloaded by [University of York] at 15:42 28 May 2015 05 06 J: 07 C: 08 J: 77 Cuz I figure it’s th’ same, ((Sneeze)) [(‘scuse me) [Sa:me thing. Two diff’rent na:mes. Or shape rather. In adding “rather” J treats his third-turn addition of “or shape” as vulnerable to being heard as the incremental addition of an alternative to “size” at the end of his previous TCU; adding on “rather” shows it to be a replacement for “size.” Here the postpositioned expansion might be thought of as somewhat “remedial,” whereas the prepositioned expansion in Extract (35) might be characterized as “preventive.” In summary, when the component projected to follow, or actually following, the or-preface can be understood as an alternative to what was said before it and not as the alternative saying of it, then an explicit repair-indicating component can be added to specify “this is repair.” In Extracts (35) and (36) this repair-indicating addition casts (or recasts) the trouble source as one of correcting a mistake in speaking. It is worth entertaining the possibility that such instances may actually be misapplications of the practice of or-prefacing, that have to be “walked back.” That is, the expansion is itself added as a (second-order) repair of the repair segment. It is not just that here or (and what follows on from it) can be understood as constituting a distinct alternative but that this is so because the trouble source is being discarded altogether in favor of the repair solution that replaces it. In other words, these may not have been suitable candidates for or-prefacing in the first place. CONCLUDING REMARKS By examining the use of or in the context of repair, we have been able to identify and describe a distinct, and distinctly positioned, element of the repair segment—the repair preface. Although or-prefaced repair (qua repair) does substitute one formulation for another, or-prefacing indicates that the trouble source formulation is not being discarded altogether, thereby mitigating the reparative character of the repair operation. In addition, we have shown how a repair segment can be initiated without the use of a hitch that alerts recipients to the possibility of repair ahead but can nevertheless come to be understood as repair. Or-prefaces may occur after such alerts, but they (along with other preface types) can also occur with no other disruption to the progressivity of the pace of the talk. A key contribution we have made here is to understanding the relationship between grammar and action. Our analysis explains how or can do what linguists working on “apposition” (e.g., Blakemore, 2007; Burton-Roberts, 1975; Meyer, 1992) have discerned it can do as a “marker of reformulation” (Blakemore, 2007, p. 132) and what doing so can accomplish irrespective of whether or not the alternatives are coreferential. We have done this by examining instances taken from field recordings of talk-in-interaction and then by focusing on the use of or in the organization of repair, showing that it operates as a repair preface that reaches back into the TCU to connect the trouble source and repair solution in a particular way. By casting its two alternatives as trouble source and repair solution, the second alternative is thereby established as not just a reformulation, but as the favored formulation. 78 LERNER AND KITZINGER Downloaded by [University of York] at 15:42 28 May 2015 REFERENCES Blakemore, D. (2007). “Or”-parentheticals, “that is”-parentheticals and the pragmatics of reformulation. Journal of Linguistics, 43(2), 311–339. doi:10.1017/S0022226707004598 Blöndal, Þ. (2008). Turn-final eda (‘or’) in spoken Icelandic. In J. Lindström (Ed.), Språk och interaction 1 [Language and interaction 1] (pp. 151–168). Helsinki, Finland: HELDA-Digital Repository of the University of Helsinki. Burton-Roberts, N. (1975). Nominal apposition. Foundations of Language, 13, 391–419. Drake, V. (2013). Turn-final or in English: A conversation analytic perspective (Unpublished doctoral dissertation), University of Wisconsin–Madison. Jefferson, G. (1974). Error correction as an interactional resource. Language in Society, 3(2), 181–199. Jefferson, G. (1986, March). Colligation as a device for minimizing repair or disagreement. Paper presented at the Talk and Social Structure conference, Santa Barbara, CA. Koshik, I. (2005). Alternative questions used in conversational repair. Discourse Studies, 7(2), 193–211. doi:10.1177/ 1461445605050366 Laakso, M., & Sorjonen, M.-L. (2010). Cut-off or particle—Devices for initiating self-repair in conversation. Journal of Pragmatics, 42, 1151–1172. doi:10.1016/j.pragma.2009.09.004 Lerner, G. H. (1994). Responsive list construction: A conversational resource for accomplishing multifaceted social action. Journal of Language and Social Psychology, 13(1), 20–33. doi:10.1177/0261927X94131002 Lerner, G. H. (1996). On the “semi-permeable” character of grammatical units in conversation: Conditional entry into the turn space of another speaker. In E. Ochs, E. A. Schegloff, & S. A. Thompson (Eds.), Interaction and grammar (pp. 238–276). Cambridge, England: Cambridge University Press. Lerner, G. H. (2004). On the place of linguistic resources in the organization of talk-in-interaction: Grammar as action in prompting a speaker to elaborate. Research on Language and Social Interaction, 37(2), 151–184. doi:10.1207/s15327973rlsi3702_3 Lerner, G. H., & Kitzinger, C. (2007). Extraction and aggregation in the repair of individual and collective self-reference. Discourse Studies, 9(4), 526–557. doi:10.1177/1461445607079165 Lerner, G. H., & Kitzinger, C. (2010, July). Repair prefacing: Preparing the way for same-turn self-repair. Paper presented to the International Conference for Conversation Analysis, Mannheim, Germany. Lindström, A. (1999). Language as social action: Grammar, prosody, and interaction in Swedish conversation (Doctoral dissertation). Uppsala, Sweden: Skrifter utgivna av Institutionen för nordiska språk vid Uppsala universitet 46. Meyer, C. C. (1992). Apposition in contemporary English. Cambridge, England: Cambridge University Press. Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1985). A comprehensive grammar of the English language. London, England: Longman. Rosenthal, B. M. (2008). A resource for repair in Japanese talk-in-interaction: The phrase tte yuu ka. Research on Language and Social Interaction, 41(2), 227–240. doi:10.1080/08351810802028787 Schegloff, E. A. (1992). Repair after next turn: The last structurally provided defense of intersubjectivity in conversation. American Journal of Sociology, 97, 1295–1345. doi:10.1086/229903 Schegloff, E. A. (2000). When “others” initiate repair. Applied Linguistics, 21(2), 205–243. doi:10.1093/applin/21.2.205 Schegloff, E. A. (2013). Ten operations in self-initiated, same-turn repair. In M. Hayashi, G. Raymond, & J. Sidnell (Eds.), Conversational repair and human understanding (pp. 41–70). Cambridge, England: Cambridge University Press. Schegloff, E. A., Jefferson, G., & Sacks, H. (1977). The preference for self-correction in the organization of repair in conversation. Language, 53, 361–382. doi:10.1353/lan.1977.0041 Schegloff, E. A., & Lerner, G. H. (2009). Beginning to respond: Well-prefaced responses to wh-questions. Research on Language and Social Interaction, 42(2), 91–115. doi:10.1080/08351810902864511 Sikveland, R., & Ogden, R. (2012). Holding gestures across turns: Moments to generate shared understanding. Gesture, 12(2), 166–199. doi:10.1075/gest.12.2.03sik