Abstract
In languages with differential object marking, some objects are treated differently from others in terms of either casemarking or agreement. Dalrymple and Nikolaeva (2011) propose that variations in object casemarking or object agreement can often be explained by reference to information structure, a level of sentence structure grammar where propositions, as conceptual states of affairs, are structured in accordance with the informational value of sentence elements and contextual factors. Dalrymple and Nikolaeva present a theory of DOM which emphasises the role of information structure in the marking patterns of objects: marked objects are topical, while unmarked objects are nontopics.
The content and representation of information structure and its relation to other grammatical levels has been a central focus of work in LFG for many years. LFG's modular architecture reflects the LFG view that language is structured into a number of nearly-independent grammatical modules; the projection architecture allows for different aspects of linguistic structure to be represented separately, and relations between levels to be clearly defined. Dalrymple and Nikolaeva (2011, chapter 4) present a new view of information structure which incorporates Mycock's (2009) insight that the traditional semantic structure of LFG, an important component of the "glue" approach to the syntax-semantics interface (Dalrymple 1999, 2001), plays a central role in representing information structure relations. Information structure partitions "glue"-style meaning contributions into the information-structure categories of TOPIC, FOCUS, BACKGROUND, and COMPLETIVE (Butt and King 1996, Choi 1999). This is accomplished by specification of the value of the semantic structure attribute DF for a meaning constructor as one of these categories; this specification can be made by any component of the grammar, including phrase structure configuration, prosody, and -- crucially for the analysis of differential object marking -- agreement and casemarking, which specifies agreeing and casemarked objects as TOPIC. The value of DF is then re-used as the information-structure attribute specifying the information-structure role of the meaning constructor. The result is an information-structure representation in which the value of the TOPIC attribute is the set of meaning constructors for the topical elements in the sentence, and similarly for the other information-structure roles.