M.H. ABRAMS
571
g iving reasons for the working of langu age, " the spade turns" before
we reach an ultima te reason . But Wittgenstein 's stance is tha t language
is "a practi ce" tha t occurs as part of a shared " form of life," and that
this p racti ce works; as he puts it, " thi s game is played." His
Investiga–
tions
are design ed to get u s to recognize when lan guage works, and
when it doesn 't-"when language is like an engine idlin g, not when it
is do ing work"
- to
get us to understand how the slippage occurred.
Derrida of course acknowledges tha t langu age works, or as he puts
it, tha t it "fun cti on s" -tha t we con stantl y perform what we take to be
successful speech acts and successful instances of oral communicati on ,
and tha t a written text is
lisi ble,
"legibl e," tha t is, strikes us as h aving
determin abl y specifi c meanings. But he accounts for this working as no
more th an " the
effects
of ideality, of sign ifi cation, of meaning and of
reference" - effects whi ch are engendered by the play of differences
within langu age itself; he then p roceeds
to
" deconstruct" these effects
by undertaking
to
show tha t, since they lack a ground in presence, their
specifi city of meaning is onl y a simula ti on. Derrida's procedure mi ght
be summari zed as foll ows. He agrees tha t language works, then asks,
"But i it poss ibl e that it reall y works?" He concludes tha t, lacking an
u ltimate ground, it is absolutely not possibl e that it works, hen ce tha t
its working is onl y a seeming -tha t, in short, though texts may be
legibl e, they are not intelligibl e, or determin a tely signifi cant.
Of each of th e tradition al terms and distinctions used to analyze
the working of language-terms such as "communi ca ti on," "context,"
" in tention," " meaning," and oppositions such as speech-writing,
literal-metaphori cal, nonfi cti onal-fi cti onal-Derrida requires not onl y
that they be grounded in absolute presence, but also tha t they be
certi fied by criteri a of wh at he call s " idea l purity" and " ultima te rigor"
if they are to be determina tely used and understood. For exampl e: in
order to communi cate "a determina te content, an identifi able mean–
ing," each of these words must sign ify a concept " that is unique,
uni voca l, ri gorou sly contro ll abl e," and its contextual conditi ons of use
mu st be "absolutely determinable" and "entirely certain "; whil e the
utterance of a determina te speech act must be ti ed
to
" the p ure
singul arity of th e evenl. " Of course such analytic words cannot meet
these criteria o f absolute fi xity, p urity, and singul arity, nor can any
words, for it is an essenti al conditi on of a language tha t a fi nite set of
words, manageabl e in accordance with a finite set of regul ariti es, be
capabl e of genera ting an unlimited variety of utterances adaptable to
an unlimited di versity of circumstances, p urposes, and applications.
Bu t Derrida 's all -or-non e p rincip le admits of no alterna ti ve: failing to
