On the innate building blocks of language and scientific explanation

Although Haspelmath’s target article does not explicitly say it, the conception of languages (and of linguistics as a science) that the author presents is the same as the one found in Saussure’s Cours de linguistique generale (Saussure 1916): Languages are social constructs and linguistics is a social science. This view is not demonstrably wrong in and of itself, but I do believe that it is an incomplete and insufficient foundation on which to build a comprehensive science of language. By establishing a Saussurean notion of languages, Haspelmath effectively hints that the naturalistic and internalist conception of language developed by generative grammar (GG) does not represent progress in the evolution of our discipline. I will argue here that Haspelmath’s assessment of GG is inadequate, because it is based on a misconception of the scientific nature of GG and of the assumptions under which it operates. In short, Haspelmath’s argument is that for GG tomake sense, it should be true that “grammatical systems are constructed from a rich set of innate building blocks of universal grammar” (§ 3). Given that a rich innate grammatical blueprint is biologically implausible (and has even been challenged within Chomsky’s minimalist approach), Haspelmath’s conclusion is that GG cannot contribute to the study of human language and has largely failed, whereas the functionalist approach is more appropriate, in that it is based on the functional-adaptive explanation of the universals discovered through the comparison of languages described in their own terms. But note that Haspelmath assumes that GG and functional linguistics offer alternative or competing answers to the samequestions,