However, this is not exactly true. For Chomsky naturally constructs his I-
language on the basis of phenomena that interest me. So, at least in part, we
are seeking to explain the same phenomena, even if we are not seeking the
explanation in the same place or in the same theoretical language. My second
example will bring this out clearly. It comes at the end of the entry on
‘Language’ in The Oxford Companion to the Mind, when Chomsky moves from
the abstract description of his theory to its empirical justification by means
of an example.
The example Chomsky takes is reciprocal pronouns (‘each other’, ‘one
another’). The rule of grammar governing the use of such pronouns is that
they must have an antecedent in the plural. The problem is to find the
antecedent. In most instances, it is present in the same clause (Chomsky gives
his examples in English, which for him has no consequences because he is
describing an aspect of universal grammar):
(5) The men recognised each other. (Note a difference between English and
French here, since the latter does not normally employ an explicit reciprocal
pronoun in this sentence: being highly intelligent, the French speaker
immediately understands if I say to her ils se sont reconnusthat the men
recognised each other – not that they recognised themselves.)
But there are definitely complications, because the antecedent is not always
in the same clause. We must therefore account for the fact that sentences (6)
and (7) are grammatical, while (8) and (9) are not:
(6) The candidates wanted each other to win.
(7) The candidates believed each other to be dishonest.
(8) The candidates believed each other were dishonest.
(9) The candidates wanted me to vote for each other. (As is customary, the
asterisk signals an ungrammatical sentence.)
In (8) ‘each other’ is the subject of the substantival clause ‘(that) X were
dishonest’; and its antecedent is therefore not situated in the same clause, but
in the main clause. In (9) the antecedent of ‘each other’ is the subject of the
main clause. However, the explanation that I have just implicitly given (the
relationship of antecedent does not hold across the clause boundary) is
insufficient, for it should lead me to conclude that sentences (6) and (7) are
equally ungrammatical, since in both cases the reciprocal pronoun is in the
infinitive subordinate clause and the antecedent in the main clause. A complex
grammatical rule appears to be at work here. Native English speakers apply
Critique of Linguistics • 27