Methodological individualism in linguistic matters has at least one advantage:
it readily explains the construction of the meaning of the utterance, which
for it is a function of the speaker ’s meaning. All things being equal, I say
what I say because I mean to say it; I put my linguistic resources to work in
the service of my intended meanings. Making language a collective and social
phenomenon before it is an individual one renders the issue less easy to
resolve. So how is the social construction of the meaning of the utterance to
be conceived? At the end of the previous section, I accounted for historical
change in language in terms of metaphorical drift and suggested that the
systematic construction of the meaning of the utterance was effected along a
Markov chain – a chain that is interrupted, or rather completed, at the moment
of the retrospective provision of meaning, from the nodal point. A Markov
chain is a mathematical entity which takes the form of a series of finished
states such that the choice of the first is wholly free, that of the second
determined by the first, and so on, with the choice of the final state being
constrained to the maximum. Applied to language, this means that the
speaker ’s freedom of expression diminishes as the sentence advances, from
left to right, along the line of language. But a Markov chain does not only
unfold according to the internal constraints of language, be they syntactical
(a transitive verb is followed by a direct object noun), syntactico-semantic (in
French a feminine article precedes a feminine noun), or semantic (which
explains why Chomsky’s famous sentence, ‘green ideas sleep furiously’, is
grammatical but meaningless). It is also constrained by the socio-collective
constraints of public meaning (and meaning is always public). To take the
standard example in English, take the beginning of the sentence ‘pride comes
before a.. .’. The appearance of the final word, which is going to complete
the sentence, obeys syntactical constraints (it must be a noun; any one will
do), and semantic constraints (here not any noun will do: it must be compatible
with the preposition ‘before’). But it is above all constrained by the
encyclopaedic knowledge that readers have of the proverb, which leads them
to expect as a matter of course the word ‘fall’ at the end of the sentence. For
in English proverbial pride always comes before a fall, just as in French the
Tarpeian Rock follows the Capitol (the height of glory is close to the nadir
of disaster). The speaker is therefore interpellated to her place, summoned
by her language to be a speaker of the proverb. Naturally, however, nothing
Propositions (1) • 173