Architectural constraints vs. efficiency constraints: Argument coding and verbal voice coding

In this talk, I discuss several universals of argument coding and verbal voice coding and argue that they can be explained on the basis of efficiency of coding, i.e. a tradeoff between speaker and hearer needs (a functional-adaptive explanation). These constructions have often been discussed in a generative context, where explanations have been expected to come from innate architectural and substantive biocognitive constraints (“UG”). I observe that the generative proposals have not converged, while the coding asymmetries that are the focus of my work have proved robust and ubiquitous in grammar.