ABSTRACT
Discourse markers, as has been pointed out on numerous occasions, encompass a wide range of grammatical categories including adverbs of various sorts (time, manner, mode), interjections, exclamations, conjunctions, and idiomatized chunks such as “you know.” They have been opposed to nouns and verbs as
having, variously, a metalinguistic (Maschler 1994) or procedural function vs. a propositional, conceptual, or denotative function (Fraser 1990; Schourup 1999; see Wertheim 2003: 154-64 for recent summary). They generally are represented as adventitious upon propositional/conceptual structure, relating it to and integrating it in the textual and interpersonal context. This chapter is concerned with a subset of discourse markers which we identify as conversation markers, and in the rest of this chapter “marker” means “conversation marker.” In the speaker sequence A B, they have a typical function of modulating the flow of propositions to and from speaker B. That is, speaker B uses a marker to respond in ways which will be summarized in Section 3 to what speaker A has said. The basic schema can be represented as follows:
(Speaker) A: utterance . . . (Speaker) B: marker (+ utterance . . .)
More so than most discourse markers, therefore, they are intimately bound up with the turn structure (Schiffrin’s exchange structure) of a conversation. In the current corpus, to give an idea of the prevalence of these markers, of the 8,471 speaker turn changes in our corpus, 847 (1 in 10) are marked initially by one of the markers summarized here. While it is not necessary to enter into a detailed debate here, we prefer (echoing Levinson 1983) to consider discourse markers as having text-internal reference. They refer to elements of texts, the propositions embodied in grammatical units (words, phrases, clauses), however these should be characterized. They can be likened to pronouns: pronouns cross-reference entities, or less frequently propositions. Conversation markers can be thought of as referencing propositions in various ways. For example, aaha “yes” uttered by HL in (39) on p. 225 references the entire proposition expressed. The markers to and iyo “oh, right, gottcha” will be seen below (2.1, 2.5) to have a completive component which references the preceding prepositional content.2 Note that “text” in this sense refers to the form-function correlation in all its manifestations. While we represent our texts with a normalized transcription, there exists behind these a spoken text which ultimately rests at the base of our analysis.
