Discourse Markers as Turn-Transition Devices: Evidence From Speech and Instant Messaging

Abstract

In this article we investigate the relation between discourse markers and turn-transition strategies in face-to-face conversations and Instant Messaging (IM), that is, unplanned, real-time, text-based, computer-mediated communication. By means of a quantitative corpus study of utterances containing a discourse marker, we show that utterance-final discourse markers are used more often in IM than in face-to-face conversations. Moreover, utterance-final discourse markers are shown to occur more often at points of turn-transition compared with points of turn-maintenance in both types of conversation. From our results we conclude that the discourse markers in utterance-final position can function as a turn-transition mechanism, signaling that the turn is over and the floor is open to the hearer. We argue that this linguistic turn-taking strategy is essentially similar in face-to-face and IM communication. Our results add to the evidence that communication in IM is more like speech than like writing.

Acknowledgments

We thank Vera van 't Hoff and Charlotte Poulisse (both student assistants at the Max Planck Institute) for their help in the annotation of the data. We are furthermore indebted to Elizabeth Traugott (Stanford University) for comments on an earlier version of this article. Two anonymous reviewers and an associate editor of Discourse Processes provided us with many detailed comments and suggestions that substantially improved earlier versions for this paper. All remaining errors and inadequacies are, of course, ours.

Notes

¹ TCUs have been operationalized in this analysis as grammatically or pragmatically complete. To allow for the comparison with the IM data, intonational completeness has not been taken into account.

³ In spoken language; to our knowledge, such analyses have not been performed for CMC.

⁴ Degand and Fagard (2012) argue that the notion of intersubjectivity should be extended to account for so-called metatextual relations too (such as digressions, reformulations, or metadiscursive comments), because they ‘‘materialize the strategic interaction between speaker and hearer and reflect the active role of the speaker to orient and to guide the hearer in his interpretational tasks' (Carlier & De Mulder, 2010, p. 269; see also Breban, 2010).

⁵ To avoid terminological confusion, we use the term “utterance” to refer to a sentence consisting of at least one independent clause or clause fragment in the IM data and to a grammatically or pragmatically complete TCU in the FTF data.

⁶ For semantic studies of the DMs under investigation, see van Bergen et al. (2011) and Mortier and Degand (2009) on eigenlijk; van Bergen et al. (2010) and Degand (2009) on dan; Evers-Vermeul (2005), Pander Maat and Degand (2001), and Pander Maat and Sanders (2000) on dus; and Hogeweg et al. (2011) and Kirsner and van Heuven (1996) on toch.

⁷ The relative frequency of DMs is higher in FTF conversations than in IM. A comparable medium difference is reported in Condon and Čech (2007), who found that the discourse-structuring use of English ok is much less frequent in text-based CMC than in FTF communication. They hypothesize that when processing demands increase (e.g., typing vs. speaking), conversational partners prefer more explicit discourse management strategies.

⁸ From the 400 utterances extracted from the DIM corpus, 394 utterances came from dyadic conversations, whereas 6 utterances were produced in multiparty conversations. To make sure these few multiparty instances did not alter the attested pattern, we ran all statistical analyses on both the full and the reduced data set (the remaining 394 utterances). Because these yielded similar results, we only report the results on the full data set.

⁹ In Degand's (2014) classification, the utterance is operationalized in clausal terms, whereby a clause generally corresponds to a grammatically complete TCU in speech (Ford & Thompson, 1996; Selting, 2000).

¹⁰ The model was fit using the glmer function from the lmerTest package (Kuznetsova et al., 2014), building on the lme4 package (Bates et al., 2014) in R (version 3.1.1; R Core Team, 2014).

¹¹ Model quality: baseline model comparison: χ² (3) = 34.7, p <  .001; Concordance index C = .87, D_xy = .74.

Additional information

Funding

This work was sutpported by grant 12/17-044 from the Fédération Wallonie-Bruxelles (to L.D.) and by the Netherlands Organization for Scientific Research (Veni grant 275-89-022 [to G.v.B.]) and the Max Planck Society (to G.v.B.).