How much syntactic reconstruction is possible?

This paper explores ways to synthesize methods from generative linguistics and historical linguistics to develop explanatory criteria that need to be satisfied by different attempts to carry out syntactic reconstruction. It addresses various questions such as (i) the need to define exactly what it means to reconstruct a language; (ii) characterizing the formal entities that count as the basic elements in the analysis of the empirical evidence for syntactic reconstruction, and whether the units of analysis and the elements that need to be reconstructed are formal entities of the same kind; (iii) whether it is possible to establish general principles for reconstructing syntax; (iv) to which extent the methodology adopted for the reconstruction of other properties of a linguistic system can be applied successfully to the reconstruction of syntax; and (v) identifying methodological criteria to assess the success of a syntactic reconstruction, and devise empirical tests for the reconstruction model.