equilibrium, of which one can construct the science. The remainder, the
variations, are assigned not to langue, but to parole. When one writes, one
is well aware that by its very nature language is far from being in a state
of equilibrium, but is in permanent disequilibrium. There is no difference
of level between langueand parole. Langueis composed of all sorts of
heterogeneous currents, which are in disequilibrium with one another.^2Deleuze, then, defines style by two characteristics: style makes the language
stammer, subjects it to an original syntactical treatment; and, by contorting
syntax, takes all language to its limits, towards the border separating it from
music.
These statements possess the advantages and disadvantages of exaggeration
and hence injustice. They have the advantage of suggesting the framework of
a philosophy of language that goes beyond the ordinary – that is, the dominant
philosophy that informs and deforms most versions of linguistics. It can be
expressed in typically Deleuzian fashion with the aid of a correlation:
linguistics/literature; system/variation; homogeneity/heterogeneity;
equilibrium/disequilibrium; langue/style; science/philosophy or art. We can
see the extent of the shift. What is being challenged is the Saussurian concept
of langue, in its opposition to parolein the technical sense – in other words,
the foundation of all linguistic science. And it will have been noted that Deleuze
refuses to construct an abstract object by reduction, and that he speaks in
terms of ‘language [langage]’.^3
This position also has the disadvantage of being unjust. It is unjust in two
ways. First of all, with respect to the science of language in general, whose
undeniable advances it ignores (and which authorise it to assume the name
of science: it will be recalled that, even if its star has dimmed, linguistics
during the structuralist moment was regarded as a model of scientificity for
all the other human sciences). All language teachers know full well that the
grammatical description of the language they teach has made gigantic strides
since Saussure. It also unjust in that it generalises and speaks of linguistics
in the singular, whereas it should be referred to in the plural, so numerous
and divergent are the theories. Deleuze was well aware of this since, under
16 • Chapter Two
(^2) This is a rough transcription of what Deleuze says: there exists no printed edition
of these interviews. 3
For a more detailed analysis, readers are referred to Lecercle 2002, Chapter 2.