Abstract (eng)
This dissertation covers two central areas for two fundamental objectives. The first is to provide a general discussion on aspects of the grammar of Kusaal, a Mabia (Gur) language spoken in Ghana, Burkina Faso and Togo. This is relevant since not much is known in the literature on this language, especially speaking of the Kusaal spoken in Ghana. The section on Kusaal grammar gives a general perspective on the phonology, morphology as well as the syntax of the language.
The second objective is to draw attention to some issues of current interest in both descriptive and theoretical linguistics. Thus, the second section is of more interest to linguists working on information structure as well as those using the Lexical Functional Grammar (LFG) framework (Bresnan et al 2016; Bresnan 2001; Kaplan and Bresnan 1982). Kusaal combines various strategies: morphological, syntactic and prosodic means, for packaging discourse related information. Whilst information focus is morphologically null, contrast and exhaustivity is marked using the particles ń, nɛ́ and kà.
Furthermore, the dissertation fills a vacuum in the literature on topic constructions in Kusaal and by extension some Mabia languages. Topic constituents in Kusaal and other Mabia languages are generally qualified by special topic phrases or particles. Familiarity topics are morphologically null in Kusaal whilst contrastive topics are marked using the phrase yáˈá àn ‘if be’. It is generally observed that Mabia languages have features that place them somewhere in between topic prominent languages and subject prominent languages unlike previous attempt to entirely classify all Niger-Congo languages as subject prominent (Li and Thompson 1976).
In addition to the above, the dissertation also draws attention to some issues in previous analysis of information structure within the Lexical Functional Grammar framework. It is observed that there are mismatches between the c-structure and the i-Structure leading to instances of ambiguities in the interpretations of mostly contrastive focus constructions as opposed to information focus constructions. The i-structure is argued to be inadequately resourced to capture the different subtypes of focus constructions in Kusaal. Using prominence to differentiate subtypes of focus, (Choi 1996), does not solve the problem since all focus types receive some degree of prominence in Kusaal.
In addressing the above problem, the dissertation provides alternative suggestions by building on the proposals of King (1996) and Choi (1996). It is suggested that an additional predicate attribute referred to as discourse type (DTYPE), with a value that subcategorizes subtypes of focus and topic notions be introduced in the i-structure. DTYPE will have attributes that provide finer grained details of the discourse subtype: contrastive focus, information focus, contrastive topic and familiarity topic. The value for DTYPE will conform with the discourse status of the constituent in question together with the corresponding particle if any or the feature specification of the said discourse status determined by the language in question. For instance a DTYPE can have the value {contrastive focus: nɛ́} for Kusaal and {contrastive focus: +NEW +PROM} for German.
The value of DTYPE may be morphologically, phonologically, or syntactically encoded in the particle used or the phonological features associated with the said notion. This will also be entirely language dependent since different languages have different discourse particles that may also be tied to specific discourse strategies. This approach is intended to make the i-structure a complete, a comprehensive and an independent projection capable of disseminating full discourse interpretation of constituents.
The suggested proposal when adopted has a cross linguistic tendency of eradicating ambiguities as well as mismatches in the interpretations of constructions relating to various aspects of information structure.