May 30, 2004

Strict transitivity

It sounds like Geoff's recent posts on transitive verbs have inspired a flood of activity in which people are seeking real linguistic examples related to a syntactic phenomenon, and that's all to the good. However, a quick perusal of the discussion suggests that people are taking an overly simplistic approach to the phenomenon in question.

There is a whole body of literature, dating back at least to Fillmore's work on definite and indefinite null complements, demonstrating that it is dangerous to think of the presence -- and particularly the absence -- of direct objects as a single phenomenon. For a start, "Habitual" or "characteristic" uses are well known to permit even the most classicially "rigidly transitive" verbs to appear without an object ("Pussycats eat, but tigers devour"); in addition, a verb's degree of flexibility with respect to omitting objects is influenced by aspectual considerations and the strength with which it selects the semantic category of its argument. Discourse factors can play a role as well.

All this is to say that "transitivity", in the sense of a dictionary's v.i. versus v.t., or in the sense of strict subcategorization frames, is not all it's cracked up to be, and it hasn't been for quite some time, at least to lexical semanticists who work at the interface with syntax. I haven't done a careful analysis, but the counterexamples being sent to Geoff seem like they fall into categories we already knew about.

I would say more, but theoretically this is a holiday weekend and I'd rather spend my time playing than discussing.

Posted by Philip Resnik at May 30, 2004 12:42 PM