472
PARTISAN REVIEW
(the change dated variously at 1848, Flaubert, Laurreamont , Mallarme)? Or
are these textual functions , so that all texts are both "writerly" and " read–
erly " ? (Or perhaps all texts are "writerly" but some are- or pretend to be?–
"readerly" too-or instead?) Or are these ways of approaching texts different
ways of reading? Are the two mutually exclusive? Interdependent? Dialecti–
cally connected? (Barthes would presumably dislike ' 'dialectic ," hinting as it
does at a Hegelian synthesizing totality .)
It
is as though from a debate in
several languages someone transcribed only snippets of each speaker filtered
through the neutralizing medium of a simultaneous translation. Again , the
fault here is not borrowing, not " eclecticism."
It
is premature synthesis, the
too hasty move from a still-developing speculation to an applied technology.
These objections may seem harsh.
It
is only fair to concede that to deal
successfully with the difficult issues now facing critical theory , a critic will
need exceptional philosophical talents along with literary experience . And
Barthes 's eminence- founded on a dozen books, half of them available in
English-is secure enough to bear a rigorous testing of the limits of his posi–
tions . More important , these objections are small change. The real question is
not whether Barthes does his work well or not, but whether the whole enter–
prise makes sense . I think it does not , and I think the trouble lies in that key
idea which meant so much to Auerbach and Lukacs and is so despicable
to
Barthes : " representation."
The attack on "representation" is quite widespread in contemporary
French thought . Ultimately , the critique is political, and presumably
"Marxist" (though very strangely so) : Derrida's analysis of Rousseau in
D e fa
grammatofogie,
for instance , subtly corrodes the theoretical bases of " repre–
sentational democracy ." Yet within literary theory , this attack is at its heart
entirely consonant with a formalist problematic. Barthes distinguishes a
"pure" or "idyllic" communication-that of mathematics, logic, the
formalized sciences-from "impure" communication , the "noisiness" of
writing, either limited , as in the "readerly ," or "massive , unnameable ," as
in the " writerly ." This antithesis is an old friend: practical vs . poetic lan–
guage (Russian formalism) ; referential vs . emotive statement (I.A . Richards) ;
scientific-conceptual vs. dramatic-metaphoric language (Cleanth Brooks) ;
logical vs . counterlogical style (W.K. Wimsatt). For Barthes, narrative - and
" writing " more generally-has references, but no referents : there is no
"reality" to which the text can innocently and transparently correspond .
Similarly , critical' 'writing" cannot give innocent access to a preexisting text.
Criticism either seeks conceptual power over texts-a sterile , pseudoscientific
classification of' 'structures" ; or it seeks the narcissistic pleasure of rewriting
the text in a "structuration, " which accepts the perverse complexities of
" reading" and" interpretation." That is , criticism must choose between a
