Studies in Language. International Journal sponsored by the Foundation “Foundations of Language” - Volume 29, Issue 3, 2005

Contemporary approaches to Generative syntax lead to the expectation that WH in situ would be subject to few distributional restrictions; but a series of complex constraints apply to in-situ WH in subject position in Standard Indonesian. We argue that this distribution does not follow from principles of formal grammar, but rather from a constraint on the relationship between syntax and information structure. We then turn to Colloquial Jakarta Indonesian, a variety similar to Standard Indonesian with regard to grammatical restrictions on WH in situ, but lacking the constraint on the relationship between syntax and information structure found in Standard Indonesian. We contend that the seeming differences between the grammars of Standard Indonesian and Jakarta Indonesian do not reflect differences in grammar in the narrow sense but rather in how the dialects relate to formal grammar and pragmatics.

Recent work on the relationship between lexicalization and grammaticalization appears to yield a consensus compatible with traditional notions of gestalt linguistics. A survey of the literature indicates the complexity of this supposed dichotomy, and provides a starting point from which we may evaluate a diagrammatic approach to the problem. Examples focus on the German derivational suffix -heit, cognate with English -hood. Such suffixation has traditionally been difficult for scholars to assess in terms of lexicality and grammaticality. Taking the gestalt concepts of parts and wholes into account, the inherent flexibility and fluidity between lexicalization and grammaticalization becomes apparent.

Evidentiality is prone to diffusion and has been identified as a diagnostic feature of linguistic areas such as the Vaupés region of the Brazilian Amazon (e.g., Aikhenvald and Dixon 1998). This paper examines the processes by which a complex evidentiality system can develop in a particular language, catalyzed by language contact but fed by language-internal resources. The discussion considers data from Hup, a Vaupés language of the Vaupés-Japurá (Makú) family, and demonstrates that Hup has developed an evidentiality system parallel to those found in the two other unrelated language families of the region. Finally, a reconstruction of an evidentiality distinction for the Vaupés-Japurá family challenges Aikhenvald and Dixon’s (1998) claim that evidentiality had two independent points of innovation in northern Amazonia.