408
PARTISAN REVIEW
speech in each and every case." Third: it excludes the primary
intention of language, "which is
to
say something about something."
In
moving from the merely virtual system which structuralists call
"language"
to
its actual and actualized form which nearly everyone
would be willing to call "speech" or "discourse," Ricoeur finds
support in Emile Benveniste's
Problems in General Linguistics.
Benveniste makes a distinction between semiotics and semantics: the
sign is the unit of language, and its study is semiotics: the sentence is
the unit of discourse, and its study is semantics.
In
The Conflict of
Interpretations
and more specifically in
The Rule of Metaphor
Ricoeur calls upon Benveniste's distinction so that he can say, for
instance, that Saussure's ana lysis of the signifier and the signified as
constituting the sign holds good, if at a ll , on ly in semiotics, or in the
semiotic order, not in the semantic order. Benveniste points out that
in semiotics what the sign signifies does not have
to
be defined: it is
necessary and sufficient for a sign to exist, that it be accepted. But in
the semantic order we are moved beyond the yes or no of semiotics
into the repetitions of discourse which, when recognized as repeti–
tions, are received as meaning. Ricoeur's argument is that the referen–
tial function and power, vested in the sentence, compensates for
another characteristic of language, which is the separation of signs
from things: it is by means of reference that language, or rather
discourse, returns to reality, which it tried
to
grasp and express.
In
an
essay called "What is a Text?" he considers two possibilities which we
confront as readers. The first is that "we can remain in the suspense
of the text and treat it as a worldl ess and authorless text, in which case
we explain it by means of its internal relations, its structure." The
second is that "we can remove the text's suspense, accomplish it in a
way similar to speech, returning it to living communication, in
which case we interpret it. " This means that reading consists in a
dialectical interplay of these two attitudes: we may call them struc–
turalism and hermeneutics. Ricoeur has a larger investment in the
second than in the first, but he is happy to recommend that each be
retained and exercised within the act of reading.
So what have we suggested, as a program of reading? Literary
criticism cou ld reasonably settle for a procedure involving semiotics,
hermeneutics, and poetics, in that order. Semiotics would provoke the
considerations described by structuralism; hermeneutics would be
concerned with problems of interpretation, semantic problems for the
most part. And poetics would be concerned with an ever larger, more
comprehensive account of human relations and possibilities, exempli-
