Constructions and Grammatical Explanation
David Adger
October 2012 

I raise a number of issues here for especially the usage based variants of Construction Grammar defended by Goldberg, Tomasello and others. I point out that proponents of such varieties of Construction Grammar assume massive amounts of innate constraints on both learning mechanisms and on cognitive representations, many of which are specific to those capacities. They just take language to be unlike other cognitive capacities in that it does not involve such constraints. There is no a priori argument for this position. I also show that no empirical argument can be made on the basis of claims that semantics attaches to structures rather than words, given that both Construction Grammar and Generative Grammar allow this. Further, the evidence we have about language conflicts with the idea that it is unlike other cognitive capacities in having no innate domain-specific constraints. The facts suggest the need for a theory of structure, and the usage-based varieties of construction grammar provide no such thing.

Format:	[ pdf ]
Reference:	lingbuzz/001633 (please use that when you cite this article)
Published in:	submitted to Mind and Language
keywords:	construction grammar, usage-based grammar, generative grammar, semantics, morphology, syntax
Downloaded:	2846 times

 

[ edit this article | back to article list ]