This course is designed to walk students, beginning with conceptual basics, through the myriad of complex issues which surround the relationship between the two distinct approaches to `universalism’ (typological generalization and formal model construction) and the task of syntactic reconstruction. There is considerable debate in the literature as to the possibility of actually reconstructing the syntax of a protolanguage, with a general split between nay-sayers (usually ‘formalists’, though I myself hate using labels) and advocates (usually ‘functionalists, though, ditto) regarding the process.
We will begin with a consideration of the relationship between typology, formal model construction and reconstruction methodology in a somewhat less controversial (though still subject to much debate) domain: that of phonological reconstruction, thus exploring the debate between typologists and formalists in a domain within which there is no serious dissent as to the practability of reconstruction.
We next turn to a survey of typological approaches to syntactic structure, including the wealth of new tools (e.g., the WALS database) now available to assist scholars in establishing an empirical foundation for their investigation. The general theoretical issue of the ‘grounding’ of typological generalizations will be raised at this juncture as well (since this forms part of the basis for the conflict between ‘functionalists’ and ‘formalists’ in reconstruction).
Next, we turn to the often very different kinds of generalizations that ‘formalist’ models seek to account for, the types of evidence which are offered up for such generalizations, and the ‘grounding’ (in this case, in UG) of the accounts offered.
We turn finally to the question of how these two types of approach play out for the issue of syntactic reconstruction. What are the ‘units of analysis’ in the two domains which COULD (in principle) be reconstructed? What would the successful reconstruction of such units tell us about the ‘syntax’ (in the descriptive sense) of the proto-language and what would it leave unclear?
In conclusion, we move to the practical consideration of three specific ‘test cases’:
(1) embedded clause stuctures in Proto-Indo-European,
(2) the ergative vs. accusative reconstruction of Proto-Polynesian and
(3) Wackernagel’s Law and the ‘left periphery’ (i.e., syntax-discourse interface) in Proto-Indo-European. We will conclude with some general lessons, open avenues for future research, etc.