480
PARTISAN REVIEW
and poems into elements which can be categorized is the result of this
desire for data; he wants units which can seem to be compassable in
isolation. Tellingly enough, he speaks of Barthes's distinction between
the
texte de plaisir
and the
texte de jouissance
as a distinction involving
either historical periods or the dominance of certain kinds of elements
in the novels. He does so by ignoring the power of the binary logic
which he recognizes as pervasive in structuralism, a logic which
enables one novel to be both kinds of text at different moments because
the context of the reader's choice changes (so that
Anna Karenina
might produce pleasure for a reader who had just finished
Don
Quixote,
and joy if the reader had just read a detective story). Thus,
when Culler is continually asserting that various examples of struc–
turalist practice vitiate the theories which they would exemplify, he is
both denying the inevitable difference between theory and practice and
is also restricting the range of possible meanings of those theories and
practices. As a critique of Derrida, Culler offers the observation that "to
replace a metaphysic of presence by a metaphysic of absence, to invert
the relation between speech and writing so that writing engulfs speech,
is to lose the distinction which translates a fact of our culture. " This is
a characteristic gesture on Culler's part: to introduce with the aid of
Wittgenstein a notion of the constraints upon our language (and also
to imply that Derrida believes that his work can reverse "reality" by
fiat, as if the binarism of Derrida's thought would not itself automati–
cally reintroduce that "fact of our culture"). But Cull er, in his concern
with constraint, tends too often to leave aside or slight another central
feature of Wittgenstein 's language games-the range of possibilities for
linguistic change. For him, the proper "role of linguistics is to
emphasize that one must construct a model to explain how sequences
have form and meaning for experienced readers, that one must start by
isolating a set of facts to be explained, and that hypotheses must be
tested by their ability to account for these effects." Literary meaning,
like meaning in general, is ru le-governed, but one of the rules is that
the irrelevant may become relevant, the nonfunctional functional to
such a degree that "isolating a set of facts
to
be explained" becomes an
illimitable exercise, like the structuralist project for which Culler
wants to establish boundaries. The most valuable insight of the book is
the prominence which structuralism and Culler ascribe to reading, but
Culler's desire to check the possibilities of infinite regress in structural–
ist thought finds its justification in an empiricism which demands
evidence when we are not certain that we understand even the criteria
for recognizing evidence about reading.
FRANCES FERGUSON
