Secondary Topic as a Relation in Information Structure*

Although information structure (information packaging, functional sentence perspective, discourse pragmatics, or informatics) has been studied for many years and from many angles, it remains a subject of great debate. In recent years researchers have fallen into two principal schools of thought. Under the more semantically oriented approach, information structure (IS) categories, such as focus, are understood as quantificational elements which affect the propositional content of an utterance (Rooth 1992; É. Kiss 1995; Erteschik-Shir 1997, and numerous works cited therein). On the other hand, the “pragmatic” approach deals with how identical propositions receive a different formal expression in accordance with the speaker’s assumptions about the addressee’s state of mind. Understood in the latter way, IS serves to optimize the communication process, but is essentially independent of the truth-conditional aspect of meaning (Chafe 1976; Prince 1981; Reinhart 1982; Gundel 1988a; Vallduví 1992; Lambrecht 1994; McNally 1998, and references therein). This paper intends to provide further evidence for the contention that IS in this second sense is an integral part of grammar, working both in parallel and in competition with its other components (semantics and syntax). My analysis will raise some fundamental issues about the content of IS and its basic units. I will argue that its commonly assumed binominal partition into topic-focus (theme-rheme, topiccomment, focus-open proposition, or focus-presupposition) may sometimes be insufficient. In particular, I hope to show that the IS relation of secondary topic can help us to describe IS in a more systematic way. Secondary topic, although sometimes used in research on IS, has been subject to rather little attention (for certain notable exceptions see Section 2.2). I will suggest some independent criteria for identifying secondary topic based on its IS role and interaction with other IS relations. Furthermore, I will show how secondary topic correlates with syntax and semantics. It should be mentioned that my notion of secondary topic is functionally similar (though not completely identical, see 2.2) to Vallduví’s ‘tail’ (Vallduví 1992, and other works). However, as noticed by McNally (1998: 177), tail has never been identified with any positive overt marking in languages other than Catalan, on which Vallduví’s original proposal is based. My analysis provides further grammatical evidence for tail/secondary topic, and demonstrates that it receives a systematic syntactic expression in at least one additional language. I will mainly be referring to optional object agreement in Northern Ostyak. Section 2 outlines the terminology relevant for the paper and the previous research on secondary topic. Section 3 demonstrates the systematic IS difference in Ostyak between the objects that trigger agreement and the objects that do not. The latter bear the focus status and the former have some topical properties. In Section 4 I argue that objects that trigger agreement bear the IS role of secondary topic; secondary topic is defined as an element under the scope of the pragmatic presupposition such that the utterance is construed to be about the relation that holds between it and the primary topic. In Section 5 I discuss some further syntactic and semantic implications of this analysis. Section 6 draws general conclusion.