Some Other “Uh(m)”s¹

¹The material discussed here was previously presented as part of Plenary Lectures at the “brandial” conference at the University of Potsdam, Germany, September, 2006 and at the 28th annual meeting of the Department of Linguistics, Aristotle University of Thessaloniki, Greece, April 2007. Parts of the text to follow appear in Studies in Greek Linguistics 28 (Proceedings of the annual meeting of the Department of Linguistics, School of Philology, Faculty of Philosophy, Aristotle University of Thessaloniki, April 21–22, 2007); Thessaloniki: Institute of Modern Greek Studies (The Manolis Triantafyllidis Foundation), 2008; or as Schegloff (2009) in Turner and Fraser (2009). My thanks to those who gave me helpful feedback in these several venues; to the editor of this journal and three anonymous referees; to Nick Enfield and J. P. de Ruiter of the Max Planck Institute, Nijmegen (and now at the University of Bielefeld), in particular; and to Gene Lerner for careful and detailed input when he had much else to do. Where I entertain alternative possible treatments of the data in the text or (most often) in a footnote and attribute them to “one” or “some” (as in “one might think …”), I am indebted to one or more of these colleagues who alerted me to these proposed alternatives.

Abstract

Recent work on the occurrence of “uh” and “uhm” in ordinary talk-in-interaction is concerned almost exclusively with its relation to trouble in the speech production process. After touching briefly on this environment of occurrence, this conversation-analytic article focuses attention on several interactional environments in which “uh(m)” figures in other ways—most extensively on its use to indicate the “reason-for-the-interaction's-launching.” The underlying theme is that accounts for what gets done and gets understood in talk-in-interaction must take into account not only its composition, but also its position—not only with respect to the grammar of sentences, but also with respect to the organization of turns at talk, of action sequences encompassing multiple turns at talk, and of occasions of talk, all of which are demonstrably oriented to by speakers in their production of the talk and by recipients in their analyzing of the talk.

Notes

¹The material discussed here was previously presented as part of Plenary Lectures at the “brandial” conference at the University of Potsdam, Germany, September, 2006 and at the 28th annual meeting of the Department of Linguistics, Aristotle University of Thessaloniki, Greece, April 2007. Parts of the text to follow appear in Studies in Greek Linguistics 28 (Proceedings of the annual meeting of the Department of Linguistics, School of Philology, Faculty of Philosophy, Aristotle University of Thessaloniki, April 21–22, 2007); Thessaloniki: Institute of Modern Greek Studies (The Manolis Triantafyllidis Foundation), 2008; or as Schegloff (2009) in Turner and Fraser (2009). My thanks to those who gave me helpful feedback in these several venues; to the editor of this journal and three anonymous referees; to Nick Enfield and J. P. de Ruiter of the Max Planck Institute, Nijmegen (and now at the University of Bielefeld), in particular; and to Gene Lerner for careful and detailed input when he had much else to do. Where I entertain alternative possible treatments of the data in the text or (most often) in a footnote and attribute them to “one” or “some” (as in “one might think …”), I am indebted to one or more of these colleagues who alerted me to these proposed alternatives.

²In different languages and different variants of English, this object is often rendered differently in print—for example, in British English it may be rendered as “erm.” The data addressed in what follows are taken from a variety of versions of American English. As noted in the text, I have amalgamated the two forms into “uh(m)” for convenience—which is to say that I do not now know whether the difference between various realizations is relevant for parties to talk-in-interaction and, if it is, how it is consequential, although there are claims about this in the literature (notably Clark & Fox Tree, 2002). Here I examine aspects of use that seem to apply to both forms.

³On assembling a conversation-analytic corpus cf. Schegloff, 1996a, pp. 176–181; Schegloff, 1997, pp. 501–502.

⁴Digitized files for data extracts are available at my Web site: http://www.sscnet.ucla.edu/soc/faculty/schegloff/

⁵For a more extended discussion, see Schegloff (2009).

⁶The first four paragraphs of this section present a minimal description of sequence organization, presented in considerably greater detail in Schegloff (2007b, especially chap. 5, pp. 58–96.

⁷Not all reasons for the call are marked by “uh(m),” and it is not yet clear what differentiates those that are so marked from those that are not. The absence of “uh(m)” does not preclude a turn appropriately positioned and composed from being taken as the reason for the call; its presence in the positions described in the text makes it a strong candidate for being so understood by interlocutors.

⁸Some might propose that the “uh” is placed before the main business of the call, not before the reason for the call, which was absent. In conversation-analytic treatment of the overall structural organization of the unit, “a single conversation,” however, the reason for the call is what is presented by the initiator of the contact as its main business; the point in this section of the article is that “uh(m)” is one resource—indeed, is one practice—for doing this “presenting.”

⁹It is worth mentioning here convergent evidence from a different institutional setting, operating on a different aspect of talk-in-interaction, but also organized by reference to reason for the call. Couper-Kuhlen (2001) examined data from radio call-in shows—a special, institutional setting with its own mandates about what should and should not occur in the openings. She showed that high-pitch onset is a feature of talk being offered by callers as the reason for the call, with other modes of delivery being employed otherwise.

¹⁰Regarding the second of these two exemplars, one might ask whether this “um” (line 18) could be accounted for instead as a dispreferred first pair part (FPP)—that is, that Bonnie wants to display tentativeness in introducing the pre-invitation. There are several reasons for not adopting such a line, of which only two are mentioned here: First, ordinarily, invitations are not done as dispreferred FPPs; unlike requests (the closest FPP “relative”), for example, they are rarely reserved until late in the conversation (unless generated in the course of the conversation) or otherwise delayed. But, second, that does not preclude, in particular cases, an invitation being done “with reservations”—that being displayed by an “uh(m)” delay, as might be conjectured. One of the points of this article is to make available to research colleagues a range of practices of talking in which “uh(m)” can figure and which a recipient has to “solve” in grasping the interactional import of a turn, the response to which they will have to supply directly on that turn's completion.

¹¹The text that follows focuses on the trajectory that may come between the “uh(m)”-marked pre-expansion and what follows, eventuating in the key action (the base first pair part) that was pre-monitored by the “uhm.” But, how about turns that precede the “uhm”-marked pre-expansion? Might not turns preceding that also have been pre-expansions for the eventual reason-for-the-call “business?” Might one not, for example, treat line 09, “Yer not busy are yuh,” as a pre-expansion to the eventual reason-for-the-call turn?; and if so, why is it not marked by a reason-for-the-call-alerting “uh(m)”?

First of all, as was noted earlier (see footnote 7), not all reasons for the call are marked by “uh(m),” and it is not yet clear what differentiates those that are so marked from those that are not. Therefore, the absence of an “uh(m)” has, at present, no probative value; its absence is not its “missing-ness.”

Second of all, we investigators have access to the way the conversation unfolded; the parties to the interaction did not. We can entertain the possibility of line 09, “Yer not busy are yuh,” as a pre-expansion to the eventual reason-for-the-call turn. For the participants, the eventual reason-for-the-call turn had not yet occurred. Therefore we have to ask: as Alan says, “Yer not busy are yuh” after what has just preceded it, what might he analyzably be doing and how might Karen analyze what his turn-in-context might be doing? Although it turns out, with the wisdom of hindsight, to have had prospective relevance, its prima facie rationale on its occurrence is as a possible account of the long delay in the call target's coming to the phone (at line 5) and its bearing on how he (Alan) should proceed. Its negative construction is designed for a “no” response, but gets a dispreferred one instead—delayed by both a gap and a “well,” and that prompts his pre-reassurance of brevity at line 12 and his launch of the reason-for-the-call sequence, which is “uh(m)”-marked.

Finally, not only are they oriented to the talk forward in real time rather than in retrospect, but we have to figure that they are not attending to the talk specifically with reference to “is this marking the reason for the call,” as we investigators writing and reading this article are. They are oriented to “what is this spate of talk doing here?”—for all the orders of “here-ness.”