The appeal of the PDC program

In 2006, I gave a series of lectures in Paris arguing for a set of rather general strategies governing language production. I was inspired to do this largely by the work of John A. Hawkins (especially his 2004 book), which argues that many morphosyntactic properties of languages can be explained in terms of how they facilitate processing. Hawkins's thesis is that language users favor structures that are easy to parse; that many languages grammaticalize these parsing preferences over time; and that this can explain facts about linguistic typology and language change. This idea is deeply insightful and has been influential among typologists, syntacticians, and psycholinguists. But I have always felt uncomfortable about a number of the particular processing strategies Hawkins proposed, because they struck me as too focused on the problem of parsing—that is, on comprehension. The form that an utterance takes is determined by the speaker, not the listener. A theory that relies too heavily on comprehension-based considerations to explain properties of languages must assume that speakers design their utterances primarily to accommodate their audience's needs, rather than their own. Given the inherent difficulty of the task of articulating thoughts as fast as people do in ordinary speech, more production-oriented explanations in the Hawkinsian style would seem more convincing.