Do the brain circuits involved in syntactic processing form a " module " in the sense of Fodor (1983) ? They undoubtedly meet several modularity criteria. Not only is their neural architecture fixed and reproducible from one individual to another , but their development is rapid and specific to the human species. Recent research suggests that, following in Steven Pinker's footsteps, we can speak of a " language instinct " or, with Chomsky, of a language acquisition device that enables very young children to make any language in the world their own.
From birth, in fact, babies prefer to listen to their mother tongue rather than a foreign language. At two months of age, brain imaging shows that listening to the mother tongue already activates a specific network in the left hemisphere, similar to that of adults. Mother tongue phonology stabilizes at around six months for vowels and twelve months for consonants. As early as six months, children can understand very frequent words (bottle, foot, eat, etc.). Towards the end of the first year of life, children begin to identify and use grammatical words. They identify them on the basis of their high frequency, length and presence at prosodic boundaries. At 11 months, they can already distinguish the most frequent determiners (le, la, des...) and pronouns (il, elle, ma, sa...), and use them to identify and segment nouns and verbs. In babies, comprehension long precedes language production. However, as early as 20 months, when children produce their first syntagms of two or three words, they respect the grammar. It is remarkable, for example, that the child places the negation correctly, i.e. before a verb in the infinitive form (" pas manger "), but after a conjugated verb (" pas mange "). Children's grammar is not identical to that of adults, but linguist Luigi Rizzi hypothesizes that children's productions correspond to a truncated version of the same arborescent structures.
