in the broadest or narrowest sense of the term. For Chomsky, in fact, linguistics
can at best only be a provisional science; and, at worst, not a science at all –
or, rather, not a specific science. At worst, the I-language is an object for
scientific psychology, which will itself one day be reduced to biology. At best,
it is currently the object of the science of language, pending the day when
the advances in biology will render superfluous indirect description of the
language faculty via grammatical structures which, whatever level they are
envisaged at, can only be surface phenomena, effects of the material constitution
of the mind/brain.
This position has consequences for explaining linguistic phenomena. As
we have seen, the I-language excludes many of the phenomena that constitute
language for common sense. This is regrettable for Marxists, who are precisely
interested in the excluded phenomena, but it is perfectly licit: the linguistic
science constructed by Saussure or Milner proceeds no differently. But we
have also seen that Chomsky’s position does not explain the phenomena that
it seeks to explain. It will no doubt be objected here that the encyclopaedia
article whose analysis of reciprocal pronouns I have criticised, being already
rather old (it dates from 1987), cannot take account of the most recent state
of the doctrine – the minimalist programme. However, in addition to the fact
that the theory is revised from top to bottom every five years, and nevertheless
claims at each step to state the truth of the non-temporal faculty of language –
prompting me to take it with a pinch of salt – the evolution of the theory is
proceeding towards an ever greater abstraction, rendering its capacity to
explain grammatical constructions (which linguistic science must explain)
highly uncertain.
To summarise. It seems to me that the reason for these uncertainties about
the science of language is to be found in a decision that is not scientific, but
philosophical, and which guides and constrains Chomsky’s scientific practice:
the philosophical thesis of the innate character of the biological structure of
language. This thesis aims to account for an undeniable fact: chimpanzees do
not speak; language is specific to homo sapiens. It is defended with an argument
that is tirelessly repeated, but highly contestable, from the poverty of stimuli –
the under-determination of linguistic competence by experience. This argument
takes the form of a claim of impossibility: the linguistic clock is too complex,
and so on. But there are other explanations: capacities for learning peculiar to
the human species, but which do not concern the details of grammatical
Critique of Linguistics • 43