Is our mental grammar just a set of constructions?: Commentary on Evans 2014

According to this book, ‘each of us carries around in our heads a “mental grammar” far more impressive than any written grammar’(6); this grammar is made up of ‘abstract rules’that can generate a ‘seemingly infinite number of novel sentences’ (12). Since this mental grammar has properties, we could decide to be interested in them and seek to make them formally explicit. In that case, we could decide to investigate my mental grammar or yours, my grammar as an infant or as it is now, or even a more complex and abstract construct such as the ‘grammar of English’ or of French. Given that the above statements are made in a general fashion, however, that is, with respect to all (neurotypical) humans, we could also be interested in mental grammar as such, which exists in all human heads. We would then make assertions about grammar in general: this would be ‘universal’ grammar in a technical sense. A rational discussion can now arise over the properties ascribed to this mental grammar, and how, in general, it is acquired, for the goals of this research program are in fact shared with the generative tradition attacked in this book (though referring to it throughout as the ‘language-as-instinct crowd’ and presenting it as a bunch of lunatics living in a Hegelian dream world clearly will not help either the field or rational discussion).

Regarding how the grammar is acquired, Chomsky in particular has tirelessly argued for a half-century that the so-called ‘innateness hypothesis’ with respect to language acquisition is a trivial claim rather than a profoundly controversial one. All parties agree that acquiring a language is not like learning to play the piano. The former is universal (present in about 7 billion people), not only as such but also in terms of its development along biologically timed milestones. Hence it is not a particular achievement for which special effort or talents are required. Its acquisition is actually very hard to prevent—so hard that trying to do so would amount to an atrocity. It would be nice if the same were true of piano playing. So in whatever sense it would be true that language is not ‘something that emerges automatically, and effortlessly’ (3), it would also be false. The same applies to Evans’s fervent rejection of the claim that language is an ‘instinct’ (99–100). In whatever sense this claim would be false, it would also be true, for the reason just given: language is an instinct in the same sense in which piano playing or cooking is not. Does E really think that Pinker’s use of the term ‘instinct’ in informal expositions implies that he thinks there are no differences in how an infant acquires language and a spider spins a web (100–101)?

In the same sense in which language has features of an instinct, it is also uncontroversially not simply acquired from use, since cats, chimpanzees raised in a human home, or canaries do not acquire language from the use of it to which they are constantly exposed. Up to at least 25% of children on the autism spectrum (Tager-Flusberg et al. 2005), as well as children with nonsyndromic language delay due to specific genomic changes (Speevak & Farrell 2011), never acquire functional language either—it would be cynical to suggest they should pick it up from its use. Home-signing children do acquire language, but again not from its use, since they have not experienced any. Deaf language-less adults with otherwise normal cognitive abilities do not acquire language from its use either: even signed language can be a complete mystery to them, despite their normal desire and intent to communicate and to socially engage (Schaller 2012)—to make them even aware of the concept of language can require extraordinary efforts. Language is not experientially there in this sense, for those who do not already know or expect it, as very young infants instinctively do when distinguishing speech from nonspeech sounds long before even knowing the first words (Vouloumanos et al. 2014).

The moral...