Syntax, prosody and function of discontinuous noun phrases in Jaminjung

Like other Australian languages, Jaminjung, a Non-Pama-Nyungan language of Northern Australia, exhibits considerable freedom of word order. This extends to the possibility of distributing nominal constituents across a clause which would belong to a single noun phrase in the English translation equivalent, e.g. 'fierce you have dog' instead of 'you have a fierce dog' (the second order is also acceptable, and in fact more common, in Jaminjung).

Faced with the phenomenon of split or discontinuous NPs, which challenges theoretical assumptions about the hierarchical nature of syntactic structure, various suggestions have been made in the literature. One proposal has been that Australian languages are 'non-configurational', i.e. have no phrase structure at all, but for a number of reasons this has not been tenable in its strict form. Most current analyses of split NP structures - which are also found in European languages like German, although in more restricted functions, e.g. 'split topicalisation' - are based on syntactic movement. Current analyses of split NPs in Australian languages do not necessarily involve movement, but converge on the view that the elements of a split NP have distinct values in terms of information structure, e.g. topic vs. focus.

My hypothesis for Jaminjung is that there is more than one subtype of split NP, possibly to be distinguished by prosodic characteristics, which are associated with different pragmatic functions. Specifically, I have evidence that one of the subtypes is employed in a subtype of thetic or 'all-new' clauses which announce the presence of a new entity in the discourse - e.g. 'raincloud is coming big' (in Engl. 'there's a big raincloud coming'). In this type of clause, the elements of the split NP do not differ in their information structure status, which makes both a movement-based analysis and an analysis in terms of two distinct, co-referential phrases implausible. In this project I will explore the application of Construction Grammar, a model entirely based on surface structure, to describe the syntactic make-up and prosodic characteristics of each of the subtypes of split NPs identified, to associate them with a discourse-pragmatic function, and to distinguish them from superficially similar constructions such as secondary predicates and dislocated NPs. The data come from a corpus of naturalistic texts compiled during my own fieldwork over the last 14 years, as well as from a field trip conducted specifically in order to explore the structural possibilities of both contiguous and discontinuous noun phrases, to be undertaken during the Institutional Research Leave.

This is the first time an analysis of split NPs as one of the strategies of marking thetic clauses will be proposed in the literature, although examples corroborating this analysis can also be found in descriptions of other Australian languages. The results of the project, to be disseminated in conference papers and three articles submitted to refereed journals, will be of interest to typologists, Australianist linguists, syntacticians, and any linguist interested in the interface between syntax, prosody and information structure, and in the application of construction-based frameworks to language description. Taken together, the publications will constitute one of the few in-depth studies of split NPs in any Australian language, and contribute to the documentation and description of a highly endangered language.