Explanation and Constructions: Response to Adger

David Adger asks why it is preferable to appeal to processes or biases of categorization, social cognition, and statistical learning in order to account for our remarkable ability to learn and use language rather than positing innate stipulations that are specifically about syntax. The answer is four-fold. First, these other processes are independently needed and it is self-evident that explanations that account for a broader range of phenomena are to be preferred over those that account for a more narrow slice of data. Secondly, the ability to categorize exemplars, share information, cooperate, and anticipate upcoming events by making various types of predictions are all clearly directly advantageous to the individual or group and so may plausibly have evolved, as Adger assumes.1 On the other hand, while our linguistic ability taken as a whole is clearly evolutionarily advantageous, the type of syntactic stipulations that have traditionally been claimed to comprise UG are not; in fact, by hypothesis, the type of modular, autonomous syntactic generalizations that have been posited over the past several decades serve no functions. For example, Adger assumes that constraints on long-distance dependencies are part of UG, but surely it is neither life-threatening nor sexually unattractive to produce an ill-formed dependency. An alternative might be to suggest that the properties of UG comprise a set of spandrels; that is, UG may contain the byproducts of other features that do have evolutionary advantages, but such an account has not been made explicit. Constructionists often look for explanations in terms of functions of the relevant constructions involved. The discovery of functional motivations obviates the need to stipulate that the constraints must be given a priori (Beckner et al., 2009; Bybee,