Unfinished Turns in French Conversation: How Context Matters

Abstract

On occasion, speakers do not complete their turns in conversation. Such syntactically incomplete turns are not treated with repair or misunderstanding. The responses that they receive display a clear understanding of the actions that the unfinished turns embodied. In this article, using conversation analysis, I describe the systematic occurrence of unfinished turns in French conversation. I show that context is necessary to the understanding of this type of turn, and I describe the nature of that context. Data analysis reveals that unfinished turns are understandable primarily by reference to their sequential position. I conclude that unfinished turns are a locally managed resource fitted to the particulars of the talk in progress and built on the context that the sequences that house them have so far provided.

Notes

1 The data has been transcribed using, for the most part, the Jeffersonian transcription system used within Conversation Analysis. Where necessary, I have adapted it. Modified or new symbols are listed in Appendix A. The data are presented with two lines of translation, except for lines composed only of pauses, which have not been repeated. Translation symbols are listed in Appendix B. The first line provides a literal word-by-word translation. The use of transcription symbols is limited in this first line to pauses, latches, and cutoffs to assist the reader in locating words. All transcription symbols have been used in the second translation line, which provides a fluent English translation. Translation permitting, these symbols were applied to the equivalent sounds on which they occur in French so as to give the reader a sense of the places where specific features occur. Word order has been preserved in the first translation line. It has been adapted to provide a fluent English translation in the second line. The first translation line also provides additional grammatical information in parentheses such as tense and number marking. In line with conversation-analytic transcription, an attempt has been made to capture the talk in a way that reflects how it sounds. This accounts for the variant renderings and nonconventional spellings of the same word, for instance, non (no) and its variants nan and nah; bon (good) and its variants ban, ben, or bah (well) and its variants beh and b'h. Discourse markers such as bon have also been translated in ways that fit the interactional tasks they accomplish and the places where they occur. Thus, bon has been translated both as good and okay or hein as uh or mm. See Bruxelles and Traverso (2001) for a report of the diversity of tasks that ben may be used to accomplish. Finally, the type of items that are left unarticulated in unfinished turns is indicated in square brackets in the two translation lines.

2 For an overview of the opposing arguments concerning what context consists of, see the Billig and Schegloff (1999) debate relating to the conversation analysis and discourse analysis positions. Also see Mandelbaum (1990–1991), Wetherell (1998), and Blommaert (2001).

3 Unfinished turns are characterized by the presence of euh (uh) and sound stretches. There have been attempts in psycholinguistics to address disfluent speech such as uhm and uh, although it tends to be treated as irrelevant, hindering comprehension, or as associated with topic unfamiliarity (Arnold, Fagnano, & Tanenhaus, 2003; Merlo & Mansur, 2004; also see Bailey & Ferreira, 2003; Brennan & Schober, 2001; and Clark & Fox Tree, 2002, for treatments of disfluent speech). In relevance theory, Stainton (1994, 1997, 2000) has addressed nonsentential or subsentential speech, relying, however, on isolated instances of elliptical sentences rendered out of context.

4 Koshik (2002) identified DIUs as a pedagogical practice used by teachers in a classroom setting to elicit self-correction of written-language errors and to prompt the students to correct the trouble source just before which the teacher's unfinished turn stops. Lerner's (1991, 1996) anticipatory completions focused on the compound TCU, a turn format exhibiting a two-part structure in which the preliminary component foreshadows and projects a final component. Examples include if X, then Y. This turn format affords a grammatically and socially provided place for “conditional entry into the turn space of another speaker” (Lerner, 1996, p. 238). Entry into another's turn space is, however, conditional in that it is regularly confined, as is the case with word searches, to the provision of the item(s) continuing the turn-so-far (i.e., the syntactic completion of the turn-so-far). Anticipatory completions are one way of doing affiliation with the compound-TCU speaker or of displaying alignment in conversation. They are also found in pedagogical settings with the same effect (Koshik, 2002; Lerner, 1995). Despite being deployed to accomplish different actions, DIUs and anticipatory completions share one similarity: The recipient provides a syntactic continuation of the disrupted turn. This is a major difference with the phenomenon under investigation. Unfinished turns are not pursued with syntactically fitted continuations. Also see Schegloff (1996b) for analyses of nonsentential TCUs, and Diaz, Antaki, and Collins (1996) on collaborative statement formulations.

5 Although unfinished turns are not specifically French, some aspects of the design of the turns reported here rely on the distinctive grammatical resources that the French language affords. These grammatical aspects are reported in Chevalier and Clift (in press).

6 In this corpus, the responses that unfinished turns receive tend not to be anticipatory completions (Lerner, 1991, 1996).

7 Preference organization refers to a structural relationship between the parts of a sequence and to the alternative responses that the first part of a sequence may make relevant. Both preferred and dispreferred responses have their own structural features. Here, although the action performed in line 30 is not a self-deprecation, it has resonances with it as it, too, invites a preferred disagreement (see Pomerantz, 1984, and Schegloff, 2007, for a description of preference organization).

8 Prior turn does not necessarily refer to the immediately prior turn. Only systematic, detailed analysis will reveal the relationship that exists between turns. Further, subsequent position is not equivalent to second position. Although it is true that many unfinished turns occur in second position in a sequence (i.e., mostly as second-pair parts), they can also occur in other subsequent positions such that the subsequent position is to be understood as non-initial position in sequence.

9 An adjacency pair can be expanded in all its axes: before the first-pair part with a presequence, between the parts of the main sequence with an insertion sequence, and after the second-pair part with a post-expansion sequence (see Schegloff, 2007, for examples). I turn to presequences in a subsequent section of this article.

10 The first TCU of line 14 brings the problem of identification that has occupied most of the talk since the beginning of the call (Mireille mistaking Erica's mother for Erica) to an end, whereas the “oui” (“yes”) in the second TCU responds to the question Mireille asked at the beginning of the call, c'est toi Erica qui a app'llé? (is it you Erica who called?; data not shown).

11 As I discuss presequences in detail in this article, the reader is referred to the section “presequences as context.” Briefly, however, the status of lines 14 through 17 as a pre is related to Mme E mentioning something Mireille already knows and to the past tense on “dire” (to tell) with which the turn projects that it is contrastive with what is to come. See Schegloff (1980, 1988, 1990, 2007) as well as Terasaki (1976/2004), for presequences.

12 This is reminiscent of Drew's (1984) reportings in invitations in which speakers produce a reporting but not its upshot, which the recipient is left to find for himself or herself.

13 Vernacular characterizations of French culture regularly focus on issues of argumentativeness, rudeness, and bluntness. For example, one recent survey reported in the British broadsheet newspaper The Guardian regarding the way in which other European nations viewed the French, listed arrogance, rudeness, and disobedience as three of this nation's top five characteristics. Elsewhere, I show that (a) differences in the ways in which people mobilize and adapt interactional structures and resources in culturally sensitive ways may partly account for cultural perceptions and (b) that misunderstanding may emanate partly from differing interactional styles.

14 I can only register here that the organization of turn taking (Sacks et al., 1974), along with the sound stretch and the “euh,” contribute to this invitation.

The term fishing is used vernacularly here and is not meant to refer to the practice by the same name described by Pomerantz (1980).

16 Bergmann (1992) noted that “… the delicate and notorious character of an event is constituted by the very act of talking about it cautiously and discreetly” (p.154). The features of delicacy identified in English talk include markers of hesitation such as hum, uh, sound stretches, cutoffs, and pauses (Pomerantz, 1984; Schegloff, 1996b). It appears that they may also be used in French to similar ends. Also see Schegloff (1980) for pre-pres acting as pre-delicates and Silverman (1994) for practices involved in the production and management of delicate items by counsellors and their patients.

17 This is not to say that pre-sequences develop in a fixed manner but, rather, that participants orient to the possible normative development of such sequences and the moves they project. In this respect, pre-sequences are sequential routines that may be said to display some element of interactive planning. See Drew (1995), Streeck (1995), and Levinson (1995) who take up this theme in Goody's (1995) edited collection of essays. Also see Heritage (1990–1991).

18 For a contrastive account of the types of sequences at hand from the perspectives of both speech act theory (indirect speech acts) and CA (pre-sequences), see Schegloff (1988) who argued that in focusing on the relationship between the form of an utterance and its function (indirection), speech act theory misses the sequential context in which pre-sequences occur. Levinson (1983) also provides an analysis of indirect speech acts from a conversation-analytic perspective. Also see Cooren (2005) for an exposition of the ways in which speech act theory may contribute to the understanding of pre-sequences if one goes beyond the notion of indirection.

19 A particular type of pre-sequence is the pre-pre (Schegloff, 1980), a pre-sequence that allows some preliminary element pertinent to the projected sequence to be established, typically a pre-mention or a pre-condition, before the sequence projected by the pre-pre is articulated. It is preliminary to a preliminary, not to the main action itself, and commonly takes the form of can I ask you a question, can I make a suggestion, and so forth.

20 The structural features of the talk suggest that Karine may have a primary claim over the next turn. Karine has made the call. She has initiated the pre, with which she has projected an underlying action. She has also closed the sequence at line 67. The next turn is also a place where a main action may occur.

21 See Schegloff (1987b, 1996b) for examples of how reusing the same words can constitute redoing the same action.

22 See Jefferson (1984) for ways in which laughter may be used in talk about troubles, and by extension, in delicate talk.

23 In the case of pre-announcements/tellings, this can clearly be observed in the fact that the announcement itself is vulnerable to repair, if issued without a pre, as can be seen in the response provided to the announcement that was delivered neat in the first line of the following example: [TG, 18:34–19:28] Bee: Oh Sibbie's sistuh hadda ba:by bo:way. Ava: Who¿ Bee: Sibbie's sister. Ava: Oh really? Bee: Myeah, Ava: [º(That's nice.)/[º(Sibbie's sistuh.) See Schegloff (2007) about the pre-knowness and the “on-delivery recognizability” (p. 39) of tellings.

24 For a clear example of a practice used to accomplish one type of action, see Schegloff's (1996a) work on confirming allusions.