Deconstruction and the Philosophers

(The Very Idea)

2
ha ha (hah hah), int. repr. laughter.
ha-ha (hah'hah), n. Sunk fence bounding park or garden. [F]
h ~ m
1
, v.i. & t. (-mm-). Make continuous murmuring sound, as
of bee, spinning top, etc.; make low inarticulate vocal sound,
esp. (usu. - & haw or ha) of hesitation; sing with closed lips;
(colloq.) be in state of activity, as make things-; (sl.) smell un-
pleasantly; (v.t.) utter, sing, with closed lips. [ME, imit.; so
MHG hummen]
hum
2
, n. Humming sound esp. of hesitation (usu.-sand ha's),
applause, surprise, etc.; (sl.) bad smell. [imit.]
hum
3
(hem), int. expr. hesitation, dissent, etc.
hum
4
, n. (sl.). Sham, hoax. [=HUMBUG]
We knew it wouldn't last for ever. We 'literary critks' or even 'theorists'
had to move quickly while the going was good, make the most of our chan-
ces, rush on excitedly, trying not to take too much notice of the slow, heavy,
inexorable tread of the law somewhere behind. The philosophers were back
there somewhere, tortoise to our hare. In 1986 their books came out.
1
Not, of course, that we'd worried about the philosophers who seemed to
be too simply on our side: we knew clearly enough that Rorty, for example,
had got it all wrong, precisely because he didn't want Derrida to be doing
philosophy at all, but just to be telling stories? Others too had a silly idea
of 'the literary' as liberating and liberated, or else tried to pull Derrida into
quarrels about other things, calling him antifoundationalist or fallibilist,
sometimes even sceptic or relativist? That never bothered us much-we
knew that was wrong. Our problem was never with 'philosophers' who
74 Geoffrey Bennington .
wanted to throw over philosophy for literature, nor with those who wanted
to translate Derrida into analytic idioms (though we much preferred the lat-
ter). But we knew that we would get into trouble one day soon because of
a paradox which became immediately obvious to anyone 'teaching' decon-
struction (my first joke), and which Culler had formulated imperturbably
in On Decoftstruction: 'Derrida 's own discussions of literary works draw
attention to important problems, but they are not deconstructions as we
have been using the and a deconstructive literary criticism will be pri-
marily influenced by his readings of philosophical works'.
4
We had to get
the 'philosophical' readings straight (who, in any case, could teach from
something like Signsponge?), without being philosophers. It was all very
well of Culler to sidestep this in an evasive footnote earlier in his book
5
-.
we had to do better than that. And turning to one of the sources of informa-
tion indicated in that footnote (assuming we hadn't been secretly weeping
over the text in question for some time already), we found stern warnings
of things to come. Gasche's much-quoted 'Deconstruction as Criticism'
told us what we feared: that we were philosophically naive or at least 'un-
trained '
6
(we had been doing our best to catch up with some reading on the
side, be fair). This was chastening in the extreme. Not, of course, that we
had ever really been guilty of the confusions Gasche so severely denounced,
though we all knew someone who But as reproaches went, there was
a disturbing difference between these and the sort we had got used to from
the moralists such as Said or Jameson or Eagleton,
7
who simply kept get-
ting it wrong about reference or history or the political or the real (best not
even mention the Lacanians),
8
and who could be easily enough refuted,
however volubly they repeated their charges. Putting them right kept us
going happily enough (though tended to give some people the unfortunate
idea that something they kept calling a 'debate' was going on, or, even
worse, that talk of 'agendas' was in order). This was different, and some-
thing of a threat: which of us was going to deny the force of at least most
of Gasche 's demonstration? Far from being someone we could laugh at and
put right, Gasche clearly knew better. And there must be others too who
knew their three H's
9
better than we did and would be following soon
enough. Maybe we'd do better to stop talking so loud about 'Western Meta-
physics' and get back to poems-but we'd learned enough meantime to see
that that sort of opposition wouldn't hold, and that any such move, espe-
cially if seen as a turning from theory to practice (as in Norris 's New Ac-
cents book, which thus makes its first mistake in its subtitle, but also as in
Geoffrey Bennington 75
Ulmer, say)
10
was going badly wrong somewhere-to say nothing of the
fact that if we went back to poems we might have to teach with Signsponge
after all.
And then we felt fairly secure after all in the sense that one of the effects
(as we say) of Derrida's work was to put in question (as some others say)
the generic distinction of philosophy and literature-so we could assume
that no-one could simply take Derrida away from us (sympathy we felt for
that participant at the 1980 Cerisy conference, admitting in the final ses-
sion that he had feared to come, lest his 'petit Derrida interieur' fet taken
away from him; relief when he claimed this hadn't happened).
1
We also
had a good hunch that Derrida made notions such as authority and 'mas-
tery' difficult, so perhaps need not be unduly alarmed by Gasche's open-
ing remark in his book that 'to judge Derrida's writings as literary-to
exclude them from the sphere of "serious", that is, philosophical discus-
sion, or to recuperate them for literary criticism-is a feeble attempt to mas-
ter his work, one that cannot do justice to the complexity of the Derridean
enterprise' (1). And when Irene HaiVey, in the 'Open Letter to Literary
Critics' which opens her book (and we indeed spent some time wondering
whether there was some play on the 'open' here, for how could such a 'let-
ter', printed as here, be other than 'open' in the normal sense?) 'insists' 'that
justice has not yet been done to the seriousness of Derrida's project' (x),
we nodded at the repetition of the question of justice and noted the absence
of scare-quotes around the notion of the serious and wondered what to make
of that.
But we knew this was stalling: we cared about doing justice to this work,
and we took irdesperately '"seriously"'. We knew, again with Culler, that
any simple denial of mastery to Derrida's workwould be foolish,
12
and that
our own efforts to teach that work depended on our ability to master it in
some significant sense. Something similar had held for Gasche 's insistence
in 'Deconstruction as Criticism' on 'the necessity of restoring [deconstruc-
tion's] rigorous meaning against its defenders as well as against those who
argue against it' ( 182) (we wondered here about clever things happening in
the redoubled 'against' in this syntax): we knew immediately, too comfort-
ably, that there was a funny smell about this type of restoration
when we read 'Restitutions', and got stuck trying to translate its subtitle):
1
but we weren't so stupid as to think some such move wasn't sometimes
necessary-we ourselves had to invoke some such notion to avoid the 'lib-
eralism' of which we were so often accused, and which would be an accur-
76
Geoffrey Bennington
ate label if we were liberal to the point of not arguing that the accusers were
just wrong. However much we hummed and ha' d at a certain relentlessness
and brutality of the writing of these people so much less literary than us,
n' est-ce pas? (and although this is true of Harvey and Gasche, who ex-
pound Derrida and philosophy with undertaker's gravitas and many a styl-
istic solecism, to the point of prov_oking an irrepressible hilarity in the
reader, no such preliminary comfort is available in Llewelyn's short, com-
plex, funny and eccentric book), we knew that in the circumstances such
aestheticism on our part would be rather vulgar. There is no alternative but
to take these books philosophically: the tortoise is upon us .at last. In the
closing 'Chronology' section of Norris's recent Modern Masters volume
on Derrida, we are told that these three books (which exhaust 1986 in the
list) 'all of which address [Derrida's] philosophical concerns' (emphasis
Norris 's), 'mark a decided shift in the Anglo-:American response to his
work'. No escape (especially, perhaps, for Norris, hare to everyone's tor-
toise).14
Don't get us wrong. Of course we' re delighted with these excellent
books. After years of warning students off almost all secondary reading on
Derrida, it's a relief to have these books to turn to. We had in any case been
arguing for years for the 'philosophical' importance of Derrida 's work,
often enough with philosophers. This didn't necessarily make us happy
with literary critics' attempts to get philosophical with Derrida: we admit
to wincing a bit at Barbara Johnson 's presentation of deconstruction as 'a
form of critique' in her introduction to her translation of Dissemination,
15
and we always thought there was something a bit iffy about Culler's use of
Nietzsche on cause and effect in On Deconstruction.
16
So we're delighted
with these books for knowing their philosophy and teaching us some of it.
But secretly delighted too that, with the possible exception of Llewelyn,
these philosophical presentations, in their acute sense of our philosophical
naivety, end up displaying their own philosophical naivety-which con-
sists precisely in their being too philosophical (naivety itself: 'the type of
the comic that stands nearest to jokes').
7
Perhaps there is no simple way
to avoid this-and especially not by being less philosophical. The fact re-
mains that all three books_are reductive in specific and different ways: Har-
vey comes round to an anthropologisation ofDerrida which is based on odd
misreadings of parts of the Grammatology and returns us to the subject of
all things; Gasche wants to place Derrida in a History of Philosophy in
which he will not be contained (Llewelyn is canny enough to be suspicious
Geoffrey Bennington 77
of such a desire (p.82)); Llewelyn reduces to various fonns of 'semiology'
the objects of Derrida 's readings.
2. Addresses
All three books are unphilosophical enough (as books of philosophy usually
are) to show a concern for their address or destination. In the case of Ga-
sche and Harvey, this is very much detennined by the (supposedly) pre-
dominantly 'literary' reception of Derrida in the United States, by that
increasingly tiresome scapegoat, so-called 'so-called American decon-
struction'. Llewelyn is rather different again, setting up a complex situation
by pointing to measured similarities between Derrida and the later Wittgen-
stein, Derrida and Peirce, Derrida and Quine, with the claim that 'this will
assist those of us who come to Detrida with the philosophical ethos pre-
dominant in English--speaking countries to feel less like innocents abroad'
(xi): but that 'us' is treacherous (as always: the 'we' used in our own writ-
. ing is difficult enough-to avoid misplaced anxieties, we recommend that
it be read throughout as usedin its modestly so-called 'Royal' fonn), for
Llewelyn himself, who ought grammatically to be the only indubitable
member of the group to which it refers, clearly escapes that group, and can
tell the others that 'We have not seen it all before', and that recent anglo-
phone philosophy is in any case more ,continental' than it knows (Ibid.)(a
welcome claim, this: analytic philosophers habitually manage to grasp that
part of Derrida 's work that overlaps with theirs, dismiss the rest on various
grounds, and then express surprise that when you get down to Derrida 's
real 'views' or 'beliefs' [Llewelyn's book is not exempt from these rather
touching idioms], they turn out to be what you've always thought anyway).
Such a stress on address, destination, and the question of context which
canhot be separated from them is of course strictly Derridean, as all three
authors know: Llewelyn quotes to this effect from the opening of 'La dif-
ferance' on the question of opening (38), and Harvey and Gasche both quote
at least part of a famous passage from the 'Question of Method' section of
the Grammatology (though only Gasche explicitly thematises 'method' as
such (121-4)), and which I quote first in what Norris's chronology refers
to judiciously as Spivak's 'landmark' translation:
18
We must begin wherever we are and the thought of the trace, which
cannot not take the. scent into account, has already taught us that it
78 Geoffrey Bennington
was impossible to justify a point of departure Wherever
we are: in a text where we already believe ourselves to be.
19
(We have often wondered what landmark-readers have made of this ref-
erence to scent: it might be thought for example that this passage gives some
needed support to Ulmer's insistence on the importance of the sense of
smell in Derrida
20
-but in fact quotes it only in his preface, omitting
the 'not' after. 'cannot' (xiii), which would create some difficulties for his
thesis were it not, alas, only a misprint. The French text of this 'motto'
(Ulmer, Ibid.) runs as follows:
11 faut commencer quelque part ou nous sommes et la pensee de la
trace, qui ne peut pas ne pas tenir compte du flair, nous a deja en-
seigne qu 'il etait impossible de justifier absolument un point de de-
part. Quelque part ou nous sommes: en un texte deja ou nous croyons
etre. (233)
'Quelque part ou nous sommes' is literally 'somewhere where we are',
which is perhaps less definite and singularising that 'wherever we are'; 'en
un texte deja ou nous croyons etre' is a little strange, but the 'deja' definite-
ly does not qualify our belief: literally, 'in a text already where we believe
we are'; 'flair' in the context of 'trace' certainly refers to the sort of scent
that hounds might be on or might have lost (maybe this was always clear)-
but sim pi y 'flair' might have been a safer translation ('Selective instinct for
what is excellent, paying, etc.' (COD)). For more details, see Volume 1 of
our forthcoming Some Errors in the Published Translations of the Works
of Jacques Derrida (Volume 7 will be devoted to The Truth in Painting).)
Harvey quotes this passage at the beginning of a sub-section entitled
"'Le Fil conducteur"' (24-8), suggests that this notion of a 'fil conducteur'
'is always Derrida 's point of departure for his deconstuctive projects', and
yet manages to end the same paragraph with the statement that 'we must
begin at the beginning'. Lest this last injunction be thought as a joke in the
circumstances (a reading perhaps encouraged by the specification of that
beginning as 'the opening of deconstruction', which leaves the question,
precisely, open), the next paragraph states that, 'As deconstruction moves
towards the text of its choice, it approaches armed with certain goals or in-
tentions' (25): the notions here of 'moving towards', 'approach', 'choice',
'goals' and 'intentions' are all extremely problematical. This is not a pure-
Geoffrey Bennington 79
ly verbal question (nothing in Derrida is), and Harvey's later finding diffi-
culties with Derrida's ascription of 'intentions' to Nietzsche (198: 'he con-
siders Nietzsche to have "intended" (a strange word for the Derridaen [sic]
discourse) to speak of the differance . .. ')is in its way just as problematic.
Derrida does indeed constantly use the word 'intention' in his earlier work,
of himself and of the authors being read, but not in way that would lend it-
selfto Harvey's description, which is, as we shall have occasion to verify,
rather intentional and goal-oriented in general. (Llewelyn has a helpful
gloss of Derrida's remarks about intention in 'Limited Inc.' (72-3).) Fur-
ther, the notion of afil conducteur, the metaphorical possibilities of which
Harvey spends a long time glossing, is, despite Derrida 's extensive use of
the term, much less a Derridean than a Kantian notion, which is also at least
once the object of an explicit distancing criticism by Derrida. This makes
us suspect that Harvey is unwise to make of the fil conducteur such a fil
conducteur for her reading_2
1
Gasche, more cautiously and accurately (Gasche 's book is almost un-
failingly accurate and reliable), quotes the same passage (without the
troublesome bit about flair), quite rightly suggests that this means that de-
construction 'proceeds in a radically empiricist manner' (170; see too Ga-
sche 's comments on this on pp. 80 and 186, which show him quite able to
deal with the obvious objection based on a passage in De la grammatologie
just prior to the famous quotation about where to start, namely that decon-
struction only looks 'empiricist' from within the closure of metaphysics),
and shows that the lack of an absolute starting-point does not leave us in
the grip of the arbitrary and the subjective or willful, linking this statement
to Derrida 's invocation of 'a certain historical necessity' (Ibid.). He might
well have referred to an earlier moment in the same work 'Yhen Derrida,
discussing his use of the word 'trace', says that 'If words and concepts re-
ceive meaning only in linkings of differences, one can justify one's lan-
guage, and the choice of terms, only within a topic and a historical strategy'
(102 [70]); and of course we none of us need reminding that 'strategy' for
Derrida is to be thought of as a strategy 'without finality' _2
2
This almost, but not quite, pragmatic insistence by Derrida (compare
with Peirce quoted by Samuel Weber: 'There is but one state of mind from
which you can "set out"-namely, the very state of mind in which you ac-
tually find yourself at the time you do "set out"' ;
23
Derrida is not of course
tallting about his or anyone's state of mind here, though he is sometimes
moved to do so )
24
on the impossibility of any absolutely justified starting
80 Geoffrey Bennington .
point is, in its simplicity, the source of untold problems to his expositors.
Whence, in the end, our hilarity. For if Derrida has no absolutely justifiable
starting-point, nor do those expounding his thought: Llewelyn solves the
problem by putting fairly independent and cross-referenced chapters in a
more-or-less arbitrary order, and suggests the possibility of reading them
in an order different from that in which they are printed (xiii); Harvey and
Gasche are committed to a semblance of ordered philosophical exposition,
and although they attempt to recognise the essentialltmits of such an en-
terprise, they are in fact caught up in it like a subtle bird-lime.
The reasons for this are perhaps best explained in Glas, where Derrida,
having watched the departure from the critical forum of all types of critic,
describes in an 'example of the reading-advice I am constantly erasing' how
he 'decapitates' metalanguage, 'or rather plunge its head back into the text
, to extract it again regularly, for the interval of a breath taken ... ' (neither
. Harvey nor Gasche does anything with Glas: Llewelyn, who makes more
use in general of the later texts, quotes it with no great fuss, thus helping at
long last to give the lie to prevailing myths about Glas (which the recent
Glassary will only reinforce, in spite of its explicit protestations, and des-
pite Derrida 's claim that, 'contrary to the rum or and to what some would
like you to believe, in [this] book there is not one single pun'
25
-(cf. too
Norris 's silly description of it as 'Joycean', in a note which then suggests
the book has to do with Derrida's early idea [in the Origin of Geometry in-
troduction] that 'thinking must in some sense choose' between the quest
for univocity and that of (Joycean) equivocality?
6
in fact Derrida 's point
is precisely that no such choice is possible insofar as each paradigm is in-
filtrated by the other-the same goes for the famous 'two interpretations
of interpretation' passage at the end of 'Structure, Sign and Play'
27
-but
then Norris seems to know this too insofar as he implicitly chides Derri-
da's American readers with assuming that Derrida has made the second,
'Joycean' choice: but it is Norris who starts off by describing Glas as uni-
vocally Joycean. The point about Glas, which is not essentially different
from any other text by Derrida, is that it presents a certain original and ex-
treme tension or torsion or stricture of the univocal and the equivocal, and
thereby is subject to a stricture in terms of its explicability. But this is true
of all Derrida 's texts, and of all texts: being able to understand this ques-
tion of differential stricture (worked out most fully in the' Speculer ... ' sec-
tion of The Post Card and the 'Restitutions' essay of The Truth in
Painting,l
8
as well as in Glas itself) and its relationship with a thought of
Geoffrey Bennington 81
structure seems to us one of the more promising ways of going about read-
ing Derrida-it is disappointing .and significant that none of these three
books ever thematises stricture at all: Harvey's section called 'The Struc-
ture of Differance' (203-44) begins by recognizing problems in using the
term structure, but persists (although the 'Economy' of the book's title
should already have disrupted structure); Gasche 's use of the term 'infra-
structure' only escapes the same problems on a very generous reading.)
3. Context
These questions of address and destination, of starting-points and strategy
of exposition arc clearly bound up with Derrida's thinking on context. Ac-
cording to the argument with Austin and then Searle, there are only ever
contexts (and this is the sense of the earlier 'dans un texte deja') but no con-
text is saturable or exhaustively determinable?
9
Llewelyn usefully con-
trasts the force of this with post-Fregean thinking in the analytical tradition
which, 'Having been persuaded that the name has meaning only in the con-
text of a sentence ... has gone on to teach that it has meaning only in the
context of an entire form of life' (112), only to hang on to the 'dream' of
being able to totalise and determine the form of life in question. Derrida 's
double argument about context and the necessary possibility of grafting
away from context into other contexts effectively undermines that dream.
It is perfectly reasonable to read Derrida 's entire output as working between
the attachment to and detachment from context or rather, for again this is a
differential stricture, as exploiting the economy of forces of detachment
and the forces of attachment, depropriation and reappropriation, letters
never quite arriving, arguments never quite understood. Harvey and Ga-
sche also explore this argument in some detail. Harvey's explication here
is particularly useful:
Where does a context begin or end? Derrida will say always-in con-
text. Quite simply, the notion of context as such cannot and indeed
does not [hum] exist. There is no such thing as a context-in-general,
by definition. Yet the term exists as such. So one must define con-
text as such always according to the context. (239)
Reapplying this type of argument to Derrida's own texts, however, Harvey
and Gasche are motivated, in the first instance at least, more by attachment
82
Geoffrey Bennington
than detachment. Gasche in particular is concerned that literary criticism
has culpably taken Derrida out of his proper philosophical context, into
which he must be returned if understanding is to be achieved. Late in his
book, for example, insisting against stupid literary people that the 'general
text' is not a particularly literary notion, he writes:
Needless to say, [if we had the time and energy, we would list all of
Gasche 's slightly weary comments of this type, which all imply to
his chosen addressee, 'if only you weren't so ignorant I wouldn't
have to go into this (again)'] if the notion of the general text is to
become an operative concept of literary criticism at all, its function
in Derrida's debate with Husserlian phenomenology and Heideg-
ger's philosophy cannot simply be overlooked. The context of this
. debate alone makes the general text a significant term, and deter-
mines its specific features and implications. No mere invocation or
magical con juration of this term can make up for the indispensable
reconstruction of its actual context. (293)
This insistence on context is a leitmotif of Gasche 's book, and the putting
into context is always presented as an antidote to epidemic and probably
contagious misconceptions (although we fear, with some indirect support
from Llewelyn (pp. 39 and 82), that this antidote will be a pharmakon).
This does seem, in certain contexts, to be entirely justified. For example,
Gasche thinks that a common misunderstanding of deconstruction (all our
friends think this) is that it consists in the neutralization of contradictory
concepts or 'textual strata' that the reading has unearthed: Gasche links this
to the idea of 'self-deconstuction' which is indeed wides8read (and in fact
this is one of the few serious blunders of Culler's book:
3
Llewelyn opens
his preface by apparently falling into the same trap by arguing that the texts
Derrida reads themselves' or 'de-construct themselves' (x),
but ends the same preface by saying 'let us be clear from the start [this being
precisely what was anything but clear at the start], when Derrida refers to
the deconstruction or dissemination of something, the "of' marks a geni-
tive which is both objective and subjective' (xiii: this could be linked to the
later discussion of 'the middle voice' (90-4)), and proceeds, devastatingly,
as follows:
Geoffrey BenningtOn 83
The suspicion that antagonistic positions or opposite concepts are
identical arises from a neglect of the historical and pragmatic aspects.
of the contexts in which they are expressed. [A certain literary-criti-
cal revenge might be exercised on that 'expressed', with some help
from 'Limited Inc. '
31
] Only through such a simplification or reduc-
tion of this context. .. is such a dramatization of antithetical positions
of criticism possible. By neglecting the pragmatic and historical con-
text of the utterance of what is dramatized in such a manner as to
cancel it out, the criticism in question reveals its origins in Roman-
tic (as well as, in a certain interpretation, Idealist) philosophy. (139:
those checking our references will see that we have sigificantly ex-
cised the word 'ideologies' from within this quotation, and cut it short
just before the word 'ideology'; there might be a point to be made
about context here.)
Gaschc 's Chapter 7 opens with the need to contextualise in the sense of es-
tablishing a 'conceptual filiation' of' deconstruction' in terms of its 'antece-
dents' (' Abbau' and 'Destruktion ') in Husserl and Heidegger (of whom
Harvey says, absolutely mysteriously, 'the fact that they are historical
antecedents of Hegel is of little significance here' (107)), and in fact this
particular type of 'contextualisation' determines Gasche 's project as a
whole, which tries to situate Derrida in terms of (a particular reading of)
'the' philosophical tradition, and specifically in terms of a particularly
powerful modem inflection of that tradition in terms of reflection. 'Decon-
struction as Criticism' had already taken its stand against so-called decon-
structive literary criticism for confusing daconstruction and
reflection-this book wants to set up a tradition of philosophy as a philos-
. ophy of reflection and then show how Derrida exceeds reflection. Whence
the stress on context. The problem with this approach is its particular deter-
mination of context as history as 'filiation': Gasche's commitment to seri-
ousness, contextualizing and the philosophical tradition leads him to
provide, in his first section, an anything-but-Derridean history of the phil-
osophy of reflection (first was Descartes, then came a man called Kant,
then two you might not have heard of, called Fichte and Schelling, then a
really very important one, Hegel. .. [who does he take us for?], and event-
ually some modem German people who sound terrible)?
2
This is the con-
text. To set it up, Gasche is obliged to forget everything about reading he
(and some literary critics) might have learned from Derrida, so that Derri-
84
Geoffrey Bennington
da can appear in this dismal procession in his rightful place, and no sooner.
If Gasche anticipated on a Derridean reading of the history of philosophy
he would already be breaking the rules of contextualisation h ~ has set him-
self (but that such rules can always be broken is one of Derrida 's essential
points, which has inevitable retroactive effects on 'history' (via the famil-
iar' always already' operator) and should prevent Gasche from getting away
with it. Gasche must have some sense of this problem: on pp. 64-5 he ar-
gues that 'Total dialectical mediation as it characterizes absolute reflection
is, in principle, of unequaled superiority', and suggests rather warily that
its bad reputation is due to certain 'aberrational consequences to which it
has led '-it is not entirely clear how in his perspective he could ever ac-
count for the possibility of such aberrations).
This is not all. A necessary corollary of this particular reductive form of
contextualisation is that when we do finally arrive at Derrida, Gasche has
largely to shift abruptly from attachment to detachment, and de-contex-
tualise Derrida 's thought. The point is that not only in his explicit reflec-
tion on signatures, events, contexts and dares,
33
but in his own practice of
contextual stricture we have described with the help of Glas, Derrida
necessarily escapes the type of simple naming and dating that Gasche must
presuppose in his account. To the extent (and Llewelyn has a nice play on
this type of expression in Derrida (p. 45)) that Derrida decapitates metal-
anguage or stuffs its head into the flow of the object-text, to the extent that
the genitive is both subjective and objective (i.e. neither), then it is entire-
ly misleading to suggestthat 'Derrida' or 'Derrida's thought' happen in the
simply linear historical way that Gasche has to assume. This severely phil-
osophical presentation relies on a pre-philosophical (or maybe all too phil-
osophical) notion of history, which is quite as debilitating as the 'Romantic'
aberrations of the literary critics he is chiding with such authority. Further
still, if it is true that that aberration was linked to a certain ignorance of his-
torical context and produced a 'suprahistorical criticism that pretends to
speak from a position free of ideology' (there's one of the ones that got
away), then we must assume that Gasche has no such pretension and that
his view of history must come from somewhere historically specific-it's
just embarrassing that that view is pre-Derridean and in fact essentially
Hegelian. There is really no way that Gasche can understand in this per-
spective the fact that Derrida's descriptions of the supplementary structure
of history should be worked out in and through Rousseau, that real pre-
Kantian, post-Hegelian 'antecedent' of He gel. The same type of point could
Geoffrey Bennington 85
be made for practically any of the texts Derrida reads (this is the point of
the 'Plato's signature is not yet complete' remark quoted by Llewelyn (89) ),
which implies that the double movement of the argument about context (at-
tachment/detachment; stricture) always already fouls up Gasche's reduc-
tive determination of it.
Whence the effect of decontextualisation: Gasche is committed to a vi-
sion of philosophers 'doing' philosophy, thinking in terms of what they
have inherited from 'the tradition', never essentially engaged in 'actively'
reading the tradition, as is Derrida. It is a horrible philosophical travesty of
Derrida to present his work as essentially coming out in the form of ideas,
thoughts or even arguments at the end of the digestive tract of history. So
Gasche's extremely useful (the question would be whether the price paid
in naivety and hilarity balances that usefulness in the restricted economy
of academic exchange) exposition of Derrida's 'infrastructures' contains,
as the (luminous) explication of the re-mark gets under way, the following
aside, of great moment for us:
As in our previous analyses, I shall discuss this infrastructure in ab-
straction from the rich context in which it is produced within Derri-
da 's work. Consequently, I shall not take advantage of the examples
of the re-mark, such as the fan, the blank, or the fold in the work of
Mallarme, on which Derrida relies in formulating this general law. .
(218)
What's this? Abstraction? From context? Fan, blank and fold as examples?
What's going on here? This is just the mirror-reflection (ha ha) ofGasche 's
previous insistence on context. The point ofDerrida's work would be that,
as usual, there is no question here of a choice between that sort of context
and that sort o( abstraction, between doing philosophy and reading (phil-
osophical) texts, between Gasche and, say, Descombes.
34
Gasche knows all this too. Arguing a little later that in a complex way
Derrida is in some sense continuing (as much as criticizing) Husserl 's pro-
ject of a 'pure logical grammar', Gasche interpolates the following remark-
able paragraph before proceeding with his exposition, apparently unaware
of the effects his last sentence should have on his entire project:
I should add a brief note on Derrida 's indebtedness to Husserlian
phenomenology. Not infrequently one hears the opinion that Husser-
86 Geoffrey Bennington
lian phenomenology ~ s a dead end and that, consequently, any at-
tempt to continue the questioning of that philosophy is doomed to
failure from the start ... [We indeed heard this opinion uttered just the
other day, and immediately turned to the opiner, saying:] As far as
Derrida is concerned, his relation to Husserl is at least threefold.
First, it is a relation, to use Granel's words, to "simply the greatest
philosopher who appeared since the Greeks". To put it differently, it
is a relati9n to the philosophical as a battle of gods and giants about
being (gigantomachia peri tes ousias), as Plato calls it in Sophist, as
well as to the philosophical in all its technical and thematic richness.
[Something a bit funny about this first reaction, which begins like a
prize-giving and goes on to sound like something out of Harold
Bloom.] Second, Derrida's relation to Husserlian thought is radically
critical of the metaphysical implications of the project of phenom-
enology itself, as has been amply documented. Third, it is a continu-
ation and radicalization of a number of motifs in Husserl 's own
works that are capable of unhinging the major metaphysical themes
at the center of his philosophy, such as the idea of a primordial axio-
matical grounding, the ideal of dedu,ctivity in general, the idea of evi-
dence, and the idea of the idea itself. Yet to contend, as I do here, that
Derrida continues Husserl (and this is true of his relation to Heideg-
ger as well) precisely on those issues that foreground the classical
ethico-theoretical decisions constitutive of philosophy as philosophy
is also to say that such a continuation is at the same time a decisive
break with the idea of tradition, continuity, Oedipality, and so on. In-
deed, the motifs in question are of such a nature that they themselves
are radically more fundamental than the possibility of continuity, and
since, moreover, they cannot be developed within the philosophical
discourse as SU;Ch, their continuation is possible only from a perspec-
tive that is marginal with respect to the history of philosophical de-
velopment. From this standpoint, the fact that Derrida may have
discovered these motifs in Husserl's works is, in a certain way, radi-
cally contingent. (246, our emphasis of last sentence)
One consquence of this is that Gasche's own placing ofDerrida in a tradi-
tion dominated by reflection is also contingent. Derrida himself has been
as careful as possible to determine his 'object' very broadly as 'the meta-
physics of presence' (of which the 'metaphysics of reflection' is a particu-
Geoffrey Bennington 87
lar, 'modern' inflection), aiming to secure the possibility of dealing with
'the greatest totality' .
35
Gasche implies that in some way Derrida is wrong
about this, that reflection is what he's really after: he quotes Dissemination
on this, and he might have added a moment in 'Pas' pointed out by Llewe-
lyn (74):
Giving no order, receiving no order from the law of laws, from the
order of language, giving none because it receives none, 'Come' ex-
changes nothing, it says nothing, shows, describes, defines, states
nothing, at the moment it is pronounced, nothing that is something
or somebody lending or communicating themselves. It doesn't even
call somone who would be there before the call. To say that it calls
the call, that it calls itself would be more accurate so long .as no specu-
lar reflexion is heard here ... For the same reasons, which I should
like to pronounce clearly, simply, in limpidity, 'Come', which is not
an order, is no more a prayer, a request, a desire. Although it makes
all these modes possible, it is not for all that anterior to them, up-
stream like a transcendental origin, pure and for itself like a primi-
tive word or an a priori. It is each time a singular event on condition
of a 'Come', each time unique but eternally repeated ...
36
4. Kant
We winced, we said, at Barbara Johnson 's description of deconstruction as
critique, if only because it made the students recalcitrant and less ready to
believe us that 'critique' was a word they used (sometimes, alack, as a verb,
which goes with debates and agendas) far too much out of excitement.
'Critique' is also notoriously slippery between a Kantian sense and a Mar-
xist inflexion of that sense. Here she is:
Deconstruction is a form of what has long been called a critique. A
critique of any theoretical system is not an examination of its flaws
or imperfections. It is not a set of criticisms designed to make the
system better. It is an analysis that focuses on the grounds of that sys-
tem's possibility. [Best now take a deep breath and read the next sen-
tence quickly, for despite its return to the question of beginnings, it
is really part of an introduction to Barthes 's Mythologies which has
strayed into the wrong text here.] The critique reads backwards from
88
Geoffrey Bennington
what seems natural, obvious, self-evident, or universal, in order to
show that these things have their history, their reasons for being the
way they are, their effects o ~ what follows from them[?!] and that
the starting point is not a (natural) given but a (cultural) construct,
usually blind to itself. For example, Copemicus can be said to have
written a critique of the Ptolemaic conception of the universe. (xv)
'Grounds' of possibility and the reference to the 'Copernican Revolution'
lead us rapidly enough to Kant. We think there has been a general sense
that Hegel is the one to get at, and Hegel is still big for Gasche, but oddly
enough it seems that it is Kant \Yho provides the language for discussing
Derrida. It is true that all three authors duly _explain that deconstruction in
some sense 'exceeds' the Hegelian dialectic (see Harvey 76-7; Llewelyn's
idea that 'On the scale of continuity-discontinuity Bataille's transgression
and Derrida 's displacement are somewhere between a Hegelian trans-
missive Aufhebung and a Bachelardian or Kuhnian intermissive break
(coupure) (12; cf 46 and further discussion on 84) is perhaps best not taken
too literally-talk of discontinuity is not much help in relation to Derrida
(cf. too Harvey p. 8, wrong, we think, about historicity); by far the most use-
ful discussion of deconstruction and dialectic is provided by Gasc:he 's ana-
lysis of the Platonic symploke, weaving together opposites on the basis of
a prior violent exclusion of non-oppositional differences. In principle at
least the difference of deconstruction from the dialectic is simply enough
fofllJ.ulated in terms of a resistance to the determination of difference as op-
position which gets the dialectic going?
7
if we \\:'ere hoping to amuse here,
we would quote criticisms of Derrida's thinking on the grounds that it is
'undialectical', as if that automatically disqualified it). But it seems as
though the philosophical stakes of writing a philosophical book on Derri-
da might be much more that of showing why his work is not the Kantian-
type transcendental philosophy which the most sympathetic construal of
Johnson's description of critique would make it. The challenge to any phil-
osopher attempting to present Derrida 's work is that of explicating why the
'conditions of possibility' discovered by that work are always also simul-
taneously 'conditions of impossibility' (roughly: what makes it possible for
a letter to arrive at its destination necessarily includes the possibility that it
might go astray; this necessary possibility means that it never completely
arrives; or, what makes it possible for a perfonnative to be brought off 'hap-
.. pily' necessarily includes the possibility of recitation outside the 'correct'
Geoffrey Bennington 89
c o n t e x t ~ this necessary possibility means that it is never completely
happy),
8
and what effects this has on thinking. All three of our authors
duly spend time explicating conditions of (im)possibility, with varying de-
grees of clarity and accuracy (e.g. Gasche 174-5, 289, 308; Harvey 68, 202;
Llewelyn 80, 98).
This tendency to turn to Kant as a means of setting off what Derrida is
doing is most clearly exemplified in Harvey's book. She is able to state, for
example, that unlike Heidegger, Derrida 'does not rely on the "essential"
in his analysis, but on what Kant called "the conditions of the possibility
of things'" (p. 185: Gasche also identifies Heidegger's attachment to essen-
tiality as a point of difference with Derrida (119-20)). The 'Introduction'
to her book in fact opens with a recognition of apparent similarities between
Kant's 'program' and Derrida's, and the announcement of an attempt, des-
pite these appearances, 'to radically distinguish Derrida's project from
Kant's' (3; we have already noted difficulties around terms such as 'pro-
ject' in Harvey's desc1iption). (rhe gesture here is not so different from that
of Derrida himself at the beginning of 'Freud and the Scene of Writing',
locating and then dissipating merely apparent ways in which deconstruc-
tion might be thought of as a psychoanalysis-that other 'Copernican rev-
olution'-of Western Metaphysics.)
39
A first move here is to argue that
Derrida might be considered to be doing a 'critique of critique' (6), show-
ing the conditions of possibility of Kant's critique. But this can only be a
first move, insofar as.critique is always a digging for foundations, a search
for firm ground on which the edifice of metaphysics might subsequently
be (re )built, and the point is that deconstruction is not even a metacritique
in this sense.
40
And Harvey duly ends her Introduction with the question
of impossibility raised above: Kant is obliged to stop with a notion of 'con-
stitution' he cannot question further; Derrida goes on to show that anything
like a constitution is made possible and impossible by differance:
That which leads Kant to rely on the notion of constitution ( Beschaf-
fenheit) as such, which cannot be known further. .. is that which Der-
rida aims to reveal the conditions of the possibility of and in turn,
necessarily, the conditions of the-more rigorously speaking-im-
possibility of [it cannot of course be more rigorous to talk simply of
conditions of impossibility which on _their own wouldn't get very
far]. [Differance] is not a centre or a ground for Derrida but the un-
grounded ground of ground, or that which allows for the constitu-
90 Geoffrey Bennington
tion of the notion of ground itself and in turn for the notion of con-
stitution itself-Kant's ultimate ground' (20) (Cf. Gasche's more
precise though more cumbersome explanation ( 100, and 157 for Der-
rida's replacement of 'constitution' by 'inscription' after early ac-
ceptance of the language of constitution or production), and
Llewelyn's usefully perplexed account (63)).
Llewelyn ends his paragraph on this point by quoting (as he does more than
once) Derrida quoting Husserl saying that 'for this all names are lacking':
no-one here is going to pretend that this lack is a provisional misfortune to
be got over by a little more inventiveness, and this essential lack of a proper
name for the groundless ground of grounds is one reason for the inevita-
bility of approaching deconstruction via the analogy of critique, and for the
almost intolerable tension of deconstruction which prevents such an ap-
proach ever being 'the truth'.
Gasche especially struggles with this tension and, casting in his lot with
the dignity of the philosophical (which he clearly does not find funny at all,
unlike Derrida and the rest of us) puts his money on what he calls 'infra-
structures'. This term is supposed to do the work we are looking for:
An infrastructure is not what is called a ground i_n traditional philos-
ophical language. It is, on the contrary, a nonfundamental structure,
or an abyssal structure, to the extent that it is without a bottom.: .. In-
frastructures are not deeper, supraessential origins. If they are said
to ground origins, it must be added that they unground them at the
same time. Infrastructures are conditions as much of the impossi-
bility as of the possibility of origins and grounds. (155, 161)
Gasche justifies his privileging of this term from Derrida himself: some-
what to our surprise, because it is hardly a common word in Dcrrida's writ-
ing. Yet Gasche is prepared to claim that' the very concept of infrastructure,
as the formal rule that each time regulates differently the play of the con-
tradictions in question, is an intrinsic part of [Derrida's] original contribu-
tion to philosophy' (142). Here again, as with reflection, it would appear
that Derrida gets it slightly wrong about his own work, which would on this
account be more or less secretly ruled by a low-profile 'concept', produced
as a 'transformed concept of"infrastructure"', as Derrida says in Positions,
Geoffrey Bennington 91
in a transformation which Derrida 'has never explicitly outlined' (147), but
which Gasche is generous enough to outline for him and us

'Infrastructure' is, then, the 'concept' which will on Gasche 's account
go down in Noddy's History of Philosophy with the name 'Derrida' attached
to it. This is all rather strange: even on Gasche's own account, this cannot
be a concept at all. The 'cases' 'subsumed' under this 'concept' are called
by names such as 'Trace', 'Differance', 'Supplementarity' and 'Re-Mark'
(sec G asche 's wonderful! y clear expositions of these in his Chapter 9: we' re
all for clarity): these are not names of concepts, but of what Gasche calls
'philosophical quasiconcepts' (156; cp. 172 for an unexplained extra
hyphen). These quasi-concepts have no proper name, as Gasche knows very
well: they are not entities, but are preontological, prelogical, presemantic.
The name they are given (and here we return to the questions about begin-
nings and contexts) cannot name anything like essence, but is borrowed,
'strategically', from the discourse being read (this is the same structure as
the 'decapitation' quoted from Glas, where the name 'decapitation', con-
enough, is borrowed from the contextofGenet's texts for the oc-
casion).
2
We take it that this is familiar to all readers of Derrida, and that
there is no need to explain again why it is simply a mistake to assume that
this signals some unfortunate or culpable collusion with what we should all
be trying to denounce or overthrow. (Though, oddly enough, Gasche him-
self gives an incomplete account of the retention of the name 'writing':
Derrida explains that he keeps to the term because writing was always deter-
mined as the signifier of a signifier; that's what he argues all signifiers are,
and that's why he retains the term.
43
gets a bit tangled up here,
saying merely, 'Writing is thus a phantom name for the X, des-
pite the fact that it is very different from what has always been called writ-
ing. What makes it susceptible to being named X is that, within the discourse
of metaphysics, it marks the repression of X and therefore essentially com-
municates with it' (167; cf. the similarly truncated account on 275). This is
too simple an account, and would scarcely motivate the retention of the
word. At least Gasche avoids the error made by Christopher Norris on this
point: in both his books he clearly confuses the parole of the parole!langue
couple with that opposed to writing, and appeals to a vague analogy be-
tween the system of langue and the common conception of writing which
confuses everything in the very inaccurate account of the earlier book, and
clouds the much more secure exposition of the later.
44
(We are of course
aware from previous OLR experience that even to hint that Norris has got
92 Geoffrey Bennington
things wrong is an unpardonable crime of elitism committed against the
generous populism of the 'prestigious' New Accents series and the principle
that accessibility e x c ~ s e s inaccuracy.)
45
While on the languelparole ques-
tion, it is worth noting Llewelyn's questioning of the necessity of anything
as complicated as 'differance' to escape the notorious problem of priority
between langue and parole (why not just say they are equiprimordial?), and
the slightly unsatisfactory sidestepping of the problem posed by an appar-
ent non-sequitur in Derrida's argument-if non-sequitur there be at this
point, it is certainly vulnerable to logical objection, and it will not do to ap-
peal to Derrida's notion of a-thetic writing to suggest that such objections
are inappropriate or illegitimate. Differential stricture again: Derrida 's texts
are all written in tension with logic-they do follow recognizable argumen-
tative structures up to a point, and this is true of the argument about langue
and parole. Descombes also suggests that there is some confusion in Der-
rida 's account of this, but you'll have to wait for the next OLR to see what
we think of it: we can't get it all in here.)
But if the 'irreducibly plural' and only partially systematizable 'general
system' of these infrastructures is thus shot through with resistance to con-
ceptual determination (which implies that each Derridean reading is some-
thing of an event, rather than just another example of a general law), then
it seems difficult to regard the general term 'infrastructure' as a concept
too. Some of the problems of attempting to do so appear later in Gasche 's
book,_ around the explication of metaphoricity, and why there is nothing
specifically literary about Derrida 's exploration of it:
Metaphoricity, because of its structure and the problems it accounts
for, is thus not to be confused with its empirical (philosophic or lite-
rary) homologue. In Derrida's sense, metaphoricity is a structure of
referral that accounts for the possibility and impossibility of the phil-
osophical discourse, yet not insofar as this discourse may be con-
strued as literary (sensible, fictional, and so on) because of its
inevitable recourse to metaphor and poetic devices,
46
but insofar as
it is a general discourse on the universal. The literary dimension of
the philosophical text is by nature incapable of pointing to, let alone
accounting for, this constituting nonorigin of philosophy. Seen in this
perspective, "llletaphoricity is a transcendental concept of sorts. Al-
though it is likely that the term I propose will meet with a good bit
of disapproval, I shall call metaphoricity a qtiasitranscendental.
Geoffrey Bennington 93
With quasi- I wish to indicate that metaphoricity has a structure and
a function similar to transcendentals without actually being one.
[This last sentence is one of those that occasion our occasional hi-
larity: this description would give the quasi transcendental metapho-
ricitya metaphorical relationship with 'actual' transcendentals, even
though presumably it makes any such relationship possible: this
priority of the quasi consistently doubles up Gasche's exposition,
and its reader] (316)
Now in fact Gasche has used this notion of the quasi transcendental earlier,
with no such precautions, and it seems a perfectly good word to us (much
better than infrastructure, at any rate, which is the one he should be apolo-
gizing for): in this earlier occurrence, Gasche writes,' Arche-writing is only,
if one may say so, the quasitranscendental synthesis that accounts for the
necessary corruption of the idealities, or transcendentals of all sorts, by
what they are defined against, and at the very moment of their constitution'
(27 4; this is, mutatis mutandis, the structure which motivates Derrida 'slater
quarrel with Heidegger's assumption that the 'essence of technology is not
technological', and could have motivated a Derridean refutation of Al-
thusser's assumption that the concept of history is not historical).
47
This
quasitranscendental ensures that all transcendentals are 'only' quasitran-
scendentals, being originarily contaminated by what they transcend. This
is the decapitation again, the radical empiricism, the impossibility of ab-
straction from context, of having a true starting-point, of determining Der-
rida 's terms as concepts. Just as the philosophical urge of Gasche 's book
leads him to abstract from context nonetheless and to screw down the possi-
bility of detachment from context with his other hand, so here he is led to
gather up this dispersed plurality of strictly limitedly transcendentalising
terms by abstracting from them a single concept called 'infrastructure'
which it is Derrida's contribution to philosophy to have invented. That this
ought not to be possible is ensured by the quasi- itself, the always inter-
rupted movement of transcendence, which Gasche explicates admirably at
several points, and nowhere more so than where he argues that 'the re-mark
is the structure that accounts for the possibility of all transcendental or the-
ological illusions ... ' (218), of which illusions Gasche 's notion ofinfrastruc-
ture as concept would be a particularly re'fined example.
94
Geoffrey Bennington
5. Transcendence and Finitude
We can sharpen up all these points by linking. transcendence and finitude,
as Derrida has since his earliest work. The 'interrupted movement' just
mentioned, for example, appears in 'Signature, Event, Context', in the dis-
cussion of the necessary possibility of the absence of any empirically deter-
mined addressee in general: 'And this absence is not a continuous
modification of presence, but a rupture of presence, the "death" or possi-
bility of the "death" of the addressee inscribed in the structure of the mark
(it is at this point, I note the fact in passing, that the value or the "effect" of
transcendentality is necessarily linked to the possiblity of writing and
"death" thus analysed)'.
48
And the derivation of 'I am mortal' and event-
ually 'I am dead' from 'I am' will be remembered as one of the more strik-
ing feats of La Voix et le
49
The paradox would be that the
movement of transcendence is set in motion by the necessary possibility of
my death, but the effect of that movement, traditionally, is to determine my
being as essentially immortal, subject to death only accidentally.
Gasche links this insistence on finitude to Nietzsche, Dilthey, and the
early Heidegger, only to argue that such an insistence does not really suc-
cessfully break with a philosophy of the subject (84-5). But as usual, ac-
cording to the tension we have already identified, the simplicity of this type
of contextualizing is complicated later. For example, discussing the links
between iterability and idealization, most clearly laid out by Derrida in
'Limited Inc.', Gasche writes, quite correctly, 'Although itcrability as such
is the becoming of intelligibility and ideality, the very possibility of repeti-
tion as the root of truth also prohibits truth from ever becoming itself. Iter-
ability, without which the ideality on which truth is based could not be
achieved, is at once [i.e. simultaneously or, as HaiVey always puts it, with
a preciosity rather touching in view of general brutality of her writing, a la
fois] the death of truth, its finitude' (215). Or again, analysing the structure
of auto-affection, Gasche again gets it right: 'interiority, however success-
fully produced in auto-affection, appears to be dependent on structures of
finitude; a certain outside inhabits this interiority, which is thus preyented
from fully coinciding with itself. Yet without the impurity of this outside,
a self could not even hope to aim at coinciding with itself' (233). Indeed,
everyth.ing we have said about context seems to be to do with this question:
discussing questions of closure and transgression,. Gasche affinns, again
quite rightly, that 'The excess or transgression of philosophy is, therefore,
ceoffrey Bennington 95
decided at the margins of the closure only, in an always strategical-that
is, historically finite-fashion' (169). And yet, in the closing pages of his
book, in the context of the description of metaphoricity as quasitranscend-
ental already quoted, Gasche refers to Derrida's 'persistent critique of the
notion of finitude' (because he wants to distinguish Derrida 's 'quasitran-
scendental' from what he calls Heidegger's 'finite transcendental') with-
out providing any references at all for this apparently startling idea. Finitude
as an essential part of the explication of conditions of (im)possibility, yet
also (only) part of a tradition and the object ofDerrida's 'critique'? Are we
dealing with two different notions of finitude here? But why doesn't Ga-
sche (or at least his index) make that clear? As the understanding not only
of Gasche's book but also of one of the.most enigmatic statements in the
whole of Derrida ('Infinite differance is finite', La Voix et le phenomene,
p. 114 [102]) would seem to depend on some clarification of this question,
we had. better take our time. This is still all about death and transcendence:
and suggests a nexus of Kant, Heidegger and Derrida which may well not
fit easily into Gasche 's historical schemes.
HaiVey is a help here, thematizing the problem as that of death rather
than that of finitude. Her misreading of what is at stake here (for it is de-
monstrably a misreading, as we shall show, and not a matter for differen-
ces of opinion) will clarify the issue. A sense of what is to come is given
early in her book, when she suggests, still working with her general sense
of deconstruction as a teleological activity, that insofar as deconstruction
is to be thought of as the 'origin' of differance (i.e. the 'practice' of decon-
struction in some sense 'produces' differance as its result), then 'beyond
Derrida [as usual, this beyond turns out to be extremely dispiriting], we
suggest, the structure of deconstruction as the origin of differance and as
the telos of deconstruction (with metaphysics playing the role of mediator,
in the Hegelian sense, here) entails a "subjective" process of recovery or
revelation' (87). This idea is then provisionally dropped as still 'pre-Hei-
deggerian', but returns later to dominate the closing parts of HaiVey 's read-
ing. A glimpse of the particular deviation involved here comes in the
explication of the 'necessary possibility of my death' argument which we
have seen to be essentially bound up with the question of transcendentality,
and is driven by a legitimate and necessary concern for what becomes of
the idiomatic in a system which appears to cross out any constituting sub-
jectivity. Harvey begins by weakening somewhat the force of Derrida 's ar-
gument by insisting on the 'my death' part and neglecting the corollary
96
Geoffrey Bennington
involving the necessary possibility of the death of any empirically deter-
mined addressee in general, with. the result that the argument is taken to
imply an essentially public ('for others') quality of meaning (195: 'Sense
or meaning is thus essentially puplic or collective ... '; cf. 223: 'As we have
shown, meaning is for Derrida ultimately a collective affair'; our quarrel is
with the 'essentially' and 'ultimately'): this is not simply false, though al-
ready a slightly humanistic version ofDerrida, for whom, it will be remem-
bered, the trace precedes even such distinctions as human/animal and
animate/inanimate, with the clear implication that there is nothing specifi-
cally 'human' (even if collective) about meaning.
50
Making meaning a
function of intersubjectivity rather than simply subjectivity might be good
enough for sociologists or Habermas, but not for us. Harvey does also rec-
ognize this, by insisting (there is an awful lot of insisting in her book) that
the question of the idiomatic in some sense transgresses a simple notion of
humanity into the realm of animality and/or the sacred. Unfortunately, how-
ever, the discussion of animality is completely vitiated by Harvey's taking
as Derrida 's own statement what is in fact his reconstruction of what he
calls, immediately before the passage she quotes, a 'still living myth' (Har-
vey, 167 and n19, referring to De la grammatologie 344 [242]; Harvey pro-
ceeds to find it 'ironic' that Derrida should hold such a view of animality,
and ties herself in knots for a while getting him back out of the trouble he
was never in: see too on animality pp. 180-1 and 201, and 196 for the con-
vergence with the sacred, still on the basis of the same type of misreading,
which will return).
This question of the idiomatic returns a little later in the context of a
slightly curious argument which moves from the recognition of the 'unna-
mability' of differance and a repeated insistance on the (relative) context-
specificity of Derrida's various 'nicknames' for differance (which, as
Harvey rightly says, is itself already a nickname) (Ibid.), to a sense that all
these nicknames might nonetheless be taken to designate 'a certain concept
of differance which transcends specific situations, circumstances, and texts'
(212). The argument here, rather as in Gasche, appeals to the force of de-
tachment and abstraction from context, indeed recognised by Derrida,
51
as
a way of re-philosophising deconstruction whenever it appears to be escap-
ing: the point being that Derrida does not simply respect the purely nomi-
nal unity of a text as secured by its author's signature, but is happy to find
Platonic moments in, say, Freud. This would imply that the 'context-spe-
cificity' of a particular nickname elaborated while reading Freud must carry
Geoffrey Bennington 97
with it a force de rupture with respect to that context. This structure is of
course what provi.des the pretext for critics of Derrida to complain on the
one hand that he spends inordinate amounts of time picking at the tiniest
textual details while on the other making everything the same 'from Plato
to NATO', as one of the twentieth century's major Marxist aestheticians
puts it, 5
2
wittily. Having made this necessary point (which just is the (in-
terrupted) movement oftranscendentality which Gasche tends to immobi-
lise; this is, rather, though still not satisfactorily thought as such, a
movement or devenir:.transcendental) which would appear to be leading
toward a reconceptualising and proper-naming of the 'concept' which Der-
rida's nicknames never quite name, Harvey suddenly veers back to proper
names of authors, and asks the following question which is unfortunately
syntactically undecipherable:
Nevertheless is it possible, according to Derrida's own 'system of
interpretation', to view the differences between 'supplement', 'dif-
ferance', 'reserve', 'trace' and 'writing', for example, as related only
to the proper name of the author which, as we know, no longer legit-
imately houses its ultimate authority? (212-3)
The argument now proceeds as follows: Derrida suggests that the relation-
ship between the nicknames (or infrastructures, or quasitranscendentals) is
itself ruled by the play of 'differance' (and this Gasche recog-
nises--that one such name can describe the relationship of all is another
example of decapitation, re-mark and interrupted transcendence which will
always prevent Gasche from being correct in calling 'infrastructure' a con-
cept). In the play of differance (or supplementarity), there is an indefinite
substitution of signifiers which never comes to rest in a transcendental sig-
nified. As Derrida says early in De la grammatologie, without a transcend-
. ental signified, the difference between signifier and signified cannot be
rigorous. (33 [20]) A 'signified' is just a signifier positioned by other sig-
nifiers as a signified; the signified is 'always already in the position of a
signifier' (107 [73]); once there is strictly speaking no signified, there is
strictly speaking no signifier either (32n9 [324n9]) (and, still worth recall-
ing, no question of a 'materiality of the signifier': see pp. 20, 45 and 138
[9-10, 29, 91] for the reasons). This argument, which Harvey does not re-
call quite in this form at this point, is briefly invoked by Derrida in the much
later discussion of incest in Rousseau, which is the (largely forgotten) con-
98 Geoffrey Bennington
text of HaiVey's discussion: even the prohibition of incest is not an origin
or fundamental signified, even though, in the economy of Rousseau's text,
it is presented as the unnamable origin, 'the first supplementarity which
allows in general the substitution of signifier for signified, signifiers for
other signifiers' (376). The paradox for a conventional reading ofRousseau
is that this 'origin' is unthinkable in terms of presence; for once he pres-
cribes that the signified or represented (the mother) not be made present
against its usurping and dangerous supplements, signifiers or repre-
sentatives (other women), then the value of presence no longer holds at the
basis of the system which consistently valorises presence nonetheless. This
basic prohibition which is the 'origin' of society and language in Rous-
seau.'s description is thus no origin because already a substitution: all of
Rousseau 's subsequent suspicion about signs and representation, and his
general demand that the thing itself be brought forward in its unmediated
presence, are undermined by this 'first' supplementarity. As Derrida says,
showing how even Rousseau 's system does not escape the general argu-
ment against the transcendental signified, 'there is a point in the system at
which the signifier can no longer be replaced by its signified, which has as
its consequence that no signifier can be, purely and simply' (376).
Harvey does not quite reestablish this context, as is always possible. But
that this be always possible and sometimes recommendable does not of
course mean that any handling of context will do (see De la grammato-
logie, p. 227 [158]). Derrida, in a passage HaiVey quotes, goes on to argue
that this point of non-replacement is also the 'point of orientation' for the
whole system, 'the point where the fundamental signified is promised as
the end point of all the referrals, and escapes as that which would, in the
same blow, destroy the whole system of signs' (376). Language points to
this point and in so doing forbids it: this just is the structure of desire ('our
indestructible and mortal desire') which is thus, following an argument by
now familiar, made possible by that (spacing, supplementarity) which also
makes its fulfilment impossible (cf. the famous statement earlier in De la
grammatologie: 'Without the possibility of dijferance, the desire for
presence as such would not find its breathing-space. Which means simul-
taneously that this desire carries within it the destiny of its non-fulfilment.
Dijferance produces what it forbids, makes possible the very thing it ren-
ders imt)ossible' (206) [Repetitive, isn't it? As HaiVey puts it in the single
funniest sentence in any of these books, 'If one asks it what it means, it just
goes on repeating the same thing over and over again, we repeat' (194: if
Geoffrey Bennington 99
we had the time and competence, we would write a stylistic study of the
(repetitive) use of the closing (first-person plural) performative verb inHar-
vey's sentence structure)] ..
All of this to be able to discuss Harvey's deviation around the question
of the idiom. For by not respecting context here, forgetting perhaps that the
calculation of the differential stricture of the movements of attachment and
detachment is always an event to be negotiated in its singularity, too con-
fident in her argument that the 'nicknames' can always break with their
context, she jumps in with the following gloss of the 'point of non-replace-
ment', having preceded the quotation by the suggestion that this is 'the locus
of the signified as such, it would seem', and urgent in its string of rhetori-
cal questions:
What is this 'point of non-replacement' for Derrida, which is also
the 'point of orientation'? The non-replaceable, non-representable,
which allows for representation and replacement; indeed the play of
differance itself. Do we not here have the idiom, the subject, the
writer, the non-authorised autl).ority, the illegitimate father, who in
fact has never left his text? Do we not have the 'madness' which rea-
son thinks it has excluded from the house of being or language it-
self? Do we not have that intrinsically unsayable, the point of the
non-blink of the eyes in that face of the other which is also our own,
which we know has never left us? We might indeed be in the Abyss,
and it may indeed be in us, as Derrida says; and there may indeed be
an intrinsic and inextricable relation of signifier and signified, but
does this not also translate, at this point of non-return, of non-repla-
ceability, non-representability, and non-substitutability, into the re-
lation of author to text? . . . Is it not this that Derrida himSelf has
pointed towards throughout his work? (213-4)
To which the answer is: no. (Or at least, not all of these things.)
But Harvey, not to be deterred, proceeds to claim that this 'point which
does not exist' 'could only be the me of my writing', the 'non-existence' of
which would be the recognition of inscription: 'it is the "me", as the sub-
ject of metaphysics, as the thrown Dasein [which are surely not quite the
same thing], who is tossed into a relation of submission to the very thing
he would command-the text of life itself' (214; the 'he' here reminds us
of Spivak's daft insistence on assigning the structure of necessary-possi-
100 Geoffrey Bennington
bility-of-absence-of-author to male authors, on the grounds that Derrida
has almost exclusively written about men).
53
Again, this is not simply
wrong, but partial and limiting and humanistic.
These problems occur in part because Harvey tends to take commentary
on Rousseau as assertion by Derrida
54
(although we might accept Llewe-
lyn's claim that 'Nowhere in his variations, Schubertian in their length and
subtle modulation, on Rousseau's Essay on the Origin of Languages does
Derrida enunciate a thesis on the origin of languages in propria persona'
( 47) only if we understand that this does not necessarily imply that every-
thing Derrida enunciates is simply in Rousseau 's persona). The same prob-
lem recurs and accentuates the inflection of Harvey 's argument we are
attempting to follow. Glossing Heidegger on being-towards-death, Harvey
recalls the moment of her analysis we have just shown to be incorrect, and
now explicitly calls the point of non-substitution 'the subject' (231). To
shore up this argument, Harvey now says that, according to Derrida, death
is the 'master-name' of the supplementary series (and again finds in strange
that Derrida should appeal to a 'master-name' at all: this recurrent need to
invoke irony or strangeness might have prompted some doubts as to the ac-
curacy of the readings-luckily our literary sensitivity, sharpened up by
having to deal with the intricacies of free indirect discouse in Flaubert, pre-
vents us from getting such things wrong).
55
She goes on to find Derrida
ventriloquising Rousseau on art and music. This is in fact again a recon-
struction of the logic of Rousseau 's thinking rather than a claim made by
Jacques Derrida, for whom there are no master-names in this manner. This
does not mean that all of Harvey 's statements about death are wrong, but
that at the very least she introduces an unnecessary pathos into a thinking
which is not pathetic in this way. Life-death for Derrida is again a question
of differential stricture, and death is not a question of 'the ultimate menace'
as Harvey is forced to think (232), nor is it simply the telos of all life, how-
ever devious life's approach to it (234 ). The detour is already death a ~
much as it is life; if 'life [is] death deferred', this is not to be read as meaning
simply that 'Just as all metaphysical oppositions turn into their opposites
[presumably meaning the terms of all metaphysical oppositions] over time
and space which is writing, so too life becomes death for Derrida' (234)
[gosh]-and Harvey in fact knows and says all this too in these confusing
pages, well enough to undennine completely her own claims about anchor-
ing the system in something called a subject, or Dasein,. or life, or death.
Geoffrey Bennington
All of which is still entirely within the problem of understanding the ap-
parent Kantianism of the appeal to conditions of (im)possibility, and the at-
tendant problems of finitude, death and (quasi)transcendentality. We have
not yet brought in other Kantian words and themes which are quite essen-
tial to what is going on here: this is all difficult and we might always yet
throw over this type of philosophical patience, who knows, in favour of the
immediate seductions of poetry or revolutionary politics. After all, as phil-
osophers go, Kant was really bourgeois, and was more or less d e l e g a t e d ~
his class to tighten up on a nasty individualism under threat, wasn't he?
But Idea of Reason, Transcendental Aesthetic, Transcendental Imagination
are lurking here in these three books, and we cannot deny their importance
easily (especially their importance in poetry and revolutionary politics).
Here, linking on to what we have been worrying away at, is Llewelyn quot-
ing 'The Ends of Man' (written, notoriously, in May 1968, when all the rest
of us were out in the streets, why weren't you?):
It could be shown that at every stage of phenomenology, and notably
each time recourse to the 'Idea in.the Kantian sense' is necessary, the
infinity of the telos, the infinity of the end rules the powers of phe-
nomenology. The end of man (as factual anthroplogicallimit) an-
nounces itself to thought from the end of man (as determined opening
or infinity of a telos). Man is what has a relation to his end, in the
fundamentally equivocal sense of this word. Since always. The tran-
scendental end can appear to itself and be deployed only on condi-
tion of mortality, of ~ relation to finitude as origin of ideality. The
name of man has always been inscribed in metaphysics between
these two ends. It has meaning only in this eschato-teleological situ-
ation. (Marges, p.147; Llewelyn, p. 36; tr. mod.)
This summary of a whole tradition of thinking will help. The transcenden-
tal movement makes sense only in the case of a finite being; my finitude
makes (im)possible my transcendence. I think, therefore I am, and before
you know it I'm essentially an immortal thinking substance, though only
the possibility of my death makes this possible and thus impossible. Again,
my finitude is a condition of my experience, conceived as a basic passiv-
ity with respect to an already-existing world: this situation could not hold
for an infinite being. That experience has traQScendental conditions of
possibility. Finitude and transcendence go hand in hand: this is Heidegger's
102 Geoffrey Bennington
demonstration in Kant and the Problem of Metaphysics. Space and Time,
Transcendental Aesthetic. But Derrida's writing as espacement (a spatial
word with a perfectly current temporal usage in French) precedes space and
time as their condition of possibility. Gasche and Harvey both point to the
possibility, briefly raised in De la grammatologie, of thinking of Derrida
as providing a more rigorous or originary transcendental aesthetic than
either Kant or Husserl. This is an attractively daunting reading for us lite-
rary people for whom these matters (especially time) are almost entirely
mysterious. 5
7
Derrida seems to provide some support for this type of read-
ing at the beginning of the 'Hinge' section of De la grammatologie, for
example. 5
8
But much later, Derrida shows how the space-time 'produced'
by the trace cannot be thought along the lines ofKantian or Husserlian tran-
scendental aesthetics or kinaesthetics ( 410-11 [290-1 ]). Harvey runs
through this argument in the 'Postface' to her book; Gasche first sees in
Derrida 's hint a sign of the temptation to talk still in terms of
in early work which we have noted, and which may be partially what leads
Harvey astray. But later, discussing spacing (which we think he limits too
closely to space), he wants no nonsense at all about .Derrida writing a new
transcendental aesthetic: 'As should be obvious at this point [get that?], this
cannot possibly be the case. Spacing is not a form of (pure) intuition that
structures a subject's experience of the world ... The notion of spacing as
the condition of possibility of sensible and intelligible, or ideal, space
undercuts the very possibility of a self-present subject of intuition or of
universal and absolute experience' (201) (and undercuts equally, he might
have added, quoting these pages from the Grammatology, 'the origin myth
of an uninhabited world' (411 [291])).
Llewelyn is more inclined toward the Kant of the Transcendental ('pro-
ductive') Imagination: still Heidegger's Kant, this. (Gasche discusses the
Transcendental Imagination in the historical part of his book, very much in
the perspective of Hegel's Kant (29ft)). Llewelyn says that 'There is much
to recommend interpreting some of Derrida 's writings as a reworking of
Kant's analysis of the productive imagination a partir de Heidegger's re-
working', and immediately adds, rather mysteriously, 'That Derrida ex-
presses a low opinion of the explanatory power of appeals to the
imagination goes some way to confirm this' (116). Although no reference
is given for this 'low opinion', we assume that Llewelyn means a passage
early on in 'Force and Signification' where Derrida indeed suggests that
critics appeal too easily to a notion of the imagination, and then glosses
Geoffrey Bennington 103
Kant, affirming that in spite of differences, the First and Third Critiques
are talking about the same imagination, and proceeds to track the. 'origin'
of that imagination in what we can now see to be an early sketch of the no-
tion of differance.
59
In Heidegger's reading of Kant, it is this same tran-
scendental imagination which is the groundless, abyssal root of
transcendence and of pure thought in general, the 'origin' even of sensory
intuition. In Heidegger 's reading, the transcendental imagination is essen-
tially temporal, and Llewelyn goes on to suggest that Kant's treatment of
the threefold temporal synthesis (which he is clearly himself reading
through Heidegger) can provide a 'somewhat _similar temporal structure'
to that at work in Derrida, and which Llewelyn has just set apart from Hus-
serlian and Hegelian time and history and linked to Freud's Nachtriiglich-
keit (117). Heidegger's fundamental argument, which Llewelyn recalls
rather allusively, is that the order of syntheses given by Kant, of apprehen-
sion in intuition, then of reproduction in imagination, then of recognition
in the concept, is misleading. Heidegger argues first of all that pure ap-
prehension (which produces the now as such, as opposed to empirical in-
tuition which aims at such and such an entity present now) is already a mode
of the transcendental imagination. The second synthesis, of pure reproduc-
tion, must distinguish the series ofnows and preserve a past now for recall
now. As this is in fact a necessary condition for a experience of any now at
all (Critique of Pure Reason, A101-2), Llewelyn is able to argue that this
would imply a sort of originary delay (the production of the present in some
sense depending on its reproduction): 'In the beginning was the begun, the
always already there of a temporality that cannot be traced as one traces a
line, cannot be traced.' ( 117). At first sight this does not seem quite to be
Heidegger's point, however (not that Llewelyn is necessarily pretending
that it is). Just as previously Llewelyn thought that Derrida was making an
unnecessary fuss about the question of priority of langue and parole, and
that it was sufficient to think of them as equiprimordial, so here it looks as
though we can say.to Llewelyn that the same holds of the first two syn-
theses, in what Heidegger calls their' originary unity'.
6
~ x c e p t that in mov-
ing into the third synthesis, which Heidegger is very keen to demonstrate
as formative of the future, he indeed allows Llewelyn's point: there is, tran-
scendentally speaking, an absolute past prior to any present which alone
ensures the unity of that present.
61
Heidegger's question about the third
synthesis is, however, that of what ensures the unity of the present and that
'past' if not an advance, futural determination towards that unity as ident-
104 Geoffrey Bennington
ity. Heidegger, of course, in a way Derrida would be suspicious of, would
like to make of that futurity an 'original' time which would be of a piece
with his notion of authenticity as analyzed in tenns of Being-towards-death:
Llewelyn does not here insist on this, but moves into the third synthesis as
follows:
The transcendental concepts that condition empirical concepts in-
clude the concepts of the three analogies, substance, cause and reci-
procity. These are respectively analogies of the pennanence of time,
that is, time itself, and of the temporal modalities succession and
coexistence. Inevitably, if temporality and the constructions of the
schematism of imagination that is the milieu between intuitive per-
ception and conception are congenitally contaminated by delay and
relay, if what Kant calls self-affection is infected by the primary sec-
ondariness at which he himself hints, then the concepts of substance
and cause and reciprocity which are for Kant the bounds of sense
will only be theoretical and fictive, 'effects' of an aboriginaltiream-
time where representation is older than presentation and presence is
perpetually postponed. (118)
Which would seem 'to imply something very similar to Levinas's point
about Derrida, namely that Kant's Idea of Reason, which is not a concept
of the understanding because of the impossibility of conceptual validation
through intuition, is already at work in the concept and the understanding,
that the evidence of present intuition, rigorous reliance upon which will
preserve us, insofar as is possible, from transcendental illusion (i.e. treat-
ing the Idea as though it were a concept of the understanding) is itself a
transcendental illusion.
62
So not a new transcendental aesthetic, but a radical appeal to something
like the transcendental .imagination which would account, in an essentially
Heideggerian way, for the (im)possibility of intuition and conceptuality?
We can find some further support for such a reading in a passage from 'La
Phannacie de Platon', arguing that the pharmakon is neither simple nor
composite, but 'rather the prior milieu in which is produced differentiation
in general, and the opposition between the eidos and its other; this milieu
is analogous to the one that later will, subsequent to and according to the
decision of philosophy, be reserved for trao.scendental.imagination, that "art
hidden in the depths of the soul"' (p. ~ 4 4 [126]). This passage is signalled
Geoffrey Bennington 105
by Gaschc (151-2), arguing that his infrastructures are 'instances of an in-
termediary discourse, concerned with a middle in which the differends are
suspended and preserved' (we think 'intermediary' is still misleadingly dia-
lectical here, though Gasche takes due care to insist that these are not 'third
terms that eventually intiate solutions in the form of speculative dialectics'
(Ibid.): Derrida's 'milieu' can safely be translated as 'milieu' in such con-
texts, something like a 'solution' in the liquid sense, from which the 'dif-
ferends' never quite crystallise out. This general milieu, which is that of the
'same' which is not identical, allows for infinitely more subtle degrees and
varieties of difference than dialectical thinkinJj) But if this milieu of dif-
ferance, almost of 'transcendental dispersion', is the deconstructive anal-
o gon of the Transcendental Imagination, what of the Idea of Reason quoted
above that might, it would seem, gather all together again, at least in a tele-
ological perspective, and provide precisely the articulation of possibility
and impossibility we are looking for, in a sort of generalized 'illusion' which
could not be measured against any sort of truth?
The quotation from 'The Ends of Man' above links the Idea to phenom-
enology, and it is indeed in the 'early' analyses of Husserl that Derrida
makes frequent reference to it. This is an extremely delicate question which
will return us to finitude: put bluntly, a very common misconception of
what Derrida is about, which Levinas 's comment runs the risk of reinforc-
ing, could lead to a version of deconstruction which assumes that dijfer-
ance just is the postponement to infinity of the Kantian Idea.
64
This
misconception would be encouraged by the insistence on finitude we have
so far been inclined to accept.
Harvey makes quite persistent use of 'the Idea in the Kantian sense' in
her own account of deconstruction: the remarks about the 'fil conducteur'
lead her almost necessarily to wonder whether the 'goals' or 'intentions'
she ascribes to deconstruction are to be thought of as such Ideas (and there-
by essentially unreachable-so the 'deconstruction of metaphysics' could
never in principle be achieved) (25). She is quite rapidly able to argue that
this is the case with grammatology (which would refute Ulmer,
for example,
6
though we are inclined to think that grammatology is always
already impossible in a more radical manner than an Idea of Reason), and
goes on to assert that this is also true of the 'detachment from metaphysics'
(73). This would appear a reasonable enough description (and is certainly
useful for us when we need to argue against over-hasty sociologisings of
Derrida, which assume that differance or 'play' are supposed to be propo-
106 Geoffrey Bennington
sals for livinJg and then get indignant because they a p p ~ a r unrealistic-this
sort of error makes us sorely tempted to treat Derrida as doing transcend-
ental philosophy, if only to avoid time-consuming argument) were the Idea
not also the object of Derrida 's analyses. When she comes to address these
analyses, there is a necessary complication. On the one hand, a notion she
has been using as an operative concept for her own analysis reappears as
an object for the object of her analysis (and in fact an object for her object's
object). On the other, as this is the point at which the possible complicity
between Derrida and that object (here, Husserl) is addressed (as it was by
Gasche, and we remember the problems raised by a certain 'radical contin-
gency' interfering with a model of filiation and tradition), then we are re-
turned to questions of metalanguage and decapitation. This complication
infiltrates Harvey's syntax and produces, 'The final thrust of the decon-
struction ofHusserl involves a certain recognition by Derrida of that which
he admits Husserl recognized' (85). This 'thrust' is the well-kriown moment
at which Derrida asserts that the whole system of Husserl 's 'essential dis-
tinctions' turns out to be purely teleological. Derrida's argument, it will be
recalled, is that on the one hand phenomenology as a whole is grounded in
the privilege of the 'living present'; but on the other, as what is (purely)
thought under this concept is determined as ideality (i.e. permitting of in-
finite identical repetition) as opposed to factuality, then the living present
is in fact infinitely deferred, and thus itself has only an ideal status. The
possibility, for example, of the objective replacement of subjective express-
ions is for Husserl an ideal, but one which is infinitely distant, and from
which, therefore, we are in fact infinitely removed. 'As the ideal is always
thought by Husserl in the form of the Idea in the Kantian sense, this sub-
stitution of ideality for non-ideality, of objectivity for non-objectivity, is in-
finitely deferred'. This differance, says Derrida,just is that of the difference
between ideality and non-ideality. In this situation, the possibility of mak-
ing the essential distinctions Husserl demands (notably between indication
and expression in the context of La Voix et le phenomene) is also infinite-
ly deferred. Whence the slightly tricky aporia Derrida discovers. In fact,
the distinctions are not respected, because their possibility is infinitely
deferred; en droit, in principle, at the end of the infinite deferral, they dis-
appear because their maintenance depends on the maintenance of a distinc-
tion between fact and principle, which distinction disappears in the Idea,
which promises principle as fact. Once again (we re-repeat), 'their possi-
bility is their impossibility' (La voix .. . , p. 113 [ 101 ]). Derrida suggests that
Geoffrey Bennington 107
this would imply that Husserl never really tried to derivedifference as sec-
ondary to presence, though he nonetheless also does do so.
Harvey takes this recognition to mean that in some sense deconstruction
borrows its telos and origin from phenomenology (85), and to do this she
has to assume that the 'aporia' located by Derrida just is a description of
the 'Idea in the Kantian sense' (86), which it is not, being rather an inter-
ruption of progress to the Idea. This leads her to suggest a 'principle of prin-
ciples' for deconstruction, namely '"no practice is ever totally f a i ~ f u l to
its principles'" which depends, in its banality, on this misreading. (Derrida
scholars who recognize that Harvey's formulation is a (near) quotation from
De la grammatologie p. 59 [39] will again note the effects of decontextuali-
sation.) Even though Harvey goes on to say that only differance makes the
Idea in the Kantian sense possible (89), she here significantly forgets the
corollary of impossibility, and her general attachment to teleological expla-
nation in fact commits her to a view of differance as Idea. Yet Derrida goes
on:
This appearing of the Ideal as infinite differance can only be pro-
duced in a relation to death in general. Only a relation to my-death
can make appear the infinite differance of presence. At the same
stroke, compared to the ideality of the positive infinite, this relation
to my-death becomes an accident of finite empiricity. The appearing
of infinite differance is itself finite. Given this, differance, which is
nothing outside this relation, becomes the finitude of life as essen-
. tial relation to self as to one's death. Infinite differance is finite. It
can therefore no longer be thought of within the opposition between
finitude and infinity, absence and presence, negation and affirma-
tion. (114 [102])
For however useful the stress on finitude, the danger is that it reinforces a
view of infinity which is, precisely, metaphysical. Gasche usefully recalls
(179) Derrida's refusal of a -classical account of the impossibility of total-
ization on the grounds of the empirical finitude of man, faced with the in-
exaustible richness of experience it can never master. The problem with the
insistence on finitude, however ueful it can appear (even for Gasche, as
we have seen) is that it tends to reinforce the claims of the transcendental
and the teleological, via the 'Idea in a Kantian sense' ,_at least read in a cer-
108
Geoffrey Bennington
tain (Husserlian) manner. There are several points in De la grammatologie
to disallow such a reading.
67
Whence, again, Gasche's use of 'quasi-transcendental'. It is now clear
that it would be wrong to consider this as a sort of failed transcendental, as
an unfortunate expedient forced upon us by a sort of sober pragmatism
against the wilder hopes of the metaphysicians, as our own borrowing of
the notion of decapitation from Glas might suggest.
And it is in fact in Glas that some of Derrida's most explicit statements
about the transcendental a ~ e to be found, though none of our three authors
quotes them. Here,, for example, in a discussion of He gel on Antigone and
the role of the sister in general:
And what if the unassimilable, the absolutely undigestable played a
fundamental role in the system, an abyssal role rather, the abyss
playing a quasi -transcendental role and allowing there to be formed
above it, like a sort of effluvium, a dream of apeasement? Is it not
always an element excluded from the system that assures the space
of possibility of the system? The transcendental has always been,
strictly, a transcategorial, what could not be received, formed, fin-
ished in any of the categories internal to the system. The system's
vomit. (171-183a; 210-227a [151-162a])
Even to present this as an interrupted movement towards, as we have tended
to do, is no doubt still too teleological and ideal. Here again, attention to
stricture and striction would help, and Glas is again the place to look:
Striction no longer allows itself to be circumscribed as an ontologi-
cal category, nor even as a category at all, not even a trans-category,
a transcendental. Striction-what serves to think the ontological or
the transcendental-is thus also in the position of a transcendental
trans-category, transcendental of a transcendental. . . There is no
choice to be made here: whenever a discourse is made against the
transcendental, a matrix-striction itself--constrains the discourse
to place the non-transcendental, the outside of the transcendental
field, the excluded, in a structuring position. The matrix in question
constitutes the excluded as the transcendental's transcendental, as
simili-transcendental, as transcendental contra-band. The contra-
band is not yet dialectical contradiction. It necessarily becomes it, to
Geoffrey Bennington 109
be sure, but its not-yet is not-yet teleological anticipation, which
means that it never becomes dialectical contradiction. It remains
something other than that which, necessarily, it is to become. (272a;
340-1 a, [244a])
Gasche suggests all too briefly that Heidegger's 'finite transcendentals'
constitue Being as Being understood and interpreted (316-7), and are still
bound to logos and logic. The quasitranscendentals, however, are not 'situ-
ated within the traditional conceptual space that stretches from the pole of
the finite to that of infinity', but 'are at the border of the space of organized
contamination which they open up'. We dis-locate the traditional space by
a radicalized appeal to finitude, but in the 'second' moment that finitude is
re-inscribed, like all the other terms. Similarly, the quasitranscendentals do
not respect the distribution of empirical and transcendental (see too p. 151),
and
By dislocating the opposition of fact and principle-that is, by ac-
counting for this conceptual difference as difference-the quasitran-
scendentals, instead of being more radical, seem to be characterised
by a certain irreducible erratic contingency. (317)
This contingency (as usual the word will not quite do, but we are wary of
a common tendency around deconstruction to bewail the inadequate or
metaphysically usurped language available, the weary sigh of the decon-
structionist for whom language is too coarse and vulgarised to do the trick:
this would again be to fall too simply into finitude and the Idea.
68
nte old
names with which deconstruction works are not provisional stand-ins in
view of a perfected new language to come: the necessary imperfection of
these words means they are just fine as they are, nothing could do better
this time; paradoxicallJJ enough, there will be no substitute for these incom-
parable supplements) will recall that which Gasche recognised contami-
nating the filiation of Derrida from Husserl, and which we suggested
disrupted any simple historical placing of Derrida as the latest or last phil-
osopher of the tradition. This 'contingency' guarantees that despite Ga-
sche's necessary insistence on deconstruction's 'mimicry' (and mockery)
of philosophical systematicity, Derrida 's work is less a system that a series
of impure 'events'. In the context of his own salutary doubts as to the possi-
bility of placing Derrida in the History of Western Thought (82), Llewelyn
110 Geoffrey Bennington
quotes from La Carte postale an avowal of a longing for presence,
proximity, proper meanings and wonders how we could ever decide
whether 'Derrida' could truly have intended such things. Llewelyn's point
is that such a question seems impertinent, and we could say it is a dramati-
sation of the necessary possibility, entertained elsewhere in La Carte Pos-
tale, that Derrida might always be writing the opposite of what he thinks
or 'believes'. We are more inclined to take such sentiments literally, and
argue against anybody that deconstruction really is a thinking of and for
the present, for now, en ce moment meme?
0
This will help to negotiate the
pathos of past and future which deconstruction also inevitably generates.
71
Insofar as deconstruction scans the rhythm
72
of our decapitation or going
under, then the emergence towards concept or Idea which these three books
unequally espouse is a partial moment of that moment, helping us to get
our breath before the plunge, certainly, but largely leaving us to take that
plunge alone.
6. The Beginning
The problem around the 'Idea in the Kantian sense' inevitably opens the
question of ethics, where, almost by definition, our real interest lies: Der-
rida 's persistent location of' ethico-theoretical' decisions at the root of sup-
posedly purely theoretical concerns (see La Voix et le phenomene) is proof
enough of that. The Idea links pure reason and the faculty of judgement
(via the sublime) to the practical reason (which Heidegger would also want
to ground in the transcendental imagination (op. cit. 30)), but we now
know this would still fall within the structure of finittide. If we are right to
suggest that Derrida interrupts the movement toward the Idea, and para-
doxically plunges us back into the 'present' and the 'empirical', then clear-
ly our ethics and politics are changed too. The free community of rational
beings can no longer simply be invoked, even regulatively, to orient our
ethical and political judgement, nor can its various surrogates. But freedom,
quaintly enough, is nevertheless what Derrida is all about. The paradox of
traditional political thinking of all colours would be that by taking its model
from conceptual thinking, it projects freedom as a state at t h ~ end of a pro-
gress ideally oriented by calculable and programmable laws. Freedom is
ejected from now except in the negative form of unforeseen obstacles.
73
Geoffrey Bennington Ill
Interestingly enough, our three books have little to say here to help us.
Gasche never broaches the question at all, despite the demand .that justice
be done to Derrida. Llewelyn (focussing explicitly on 'sense', of course),
says little more, and although _both briefly mention Levinas in the context
of a questioning of classical ontology, neither points out that in Levinas
what is 'othetwise than being or beyond essence' (Llewelyn, p. 41, refer-
ring of course to the title ofLevinas's 1974 book) is bound up with the ethi-
cal. Harvey does examine the question more closely: first in the course of
her explication ofDerrida 's 'trace', in its 'profound though subtle differen-
ces' with that ofLevinas (pp: 169-74), but more importantly in a return to
the 'point of orientation' of the system, which we have seen to be the pro-
hibition of incest in Rousseau, and which Harvey mistakenly tries to make
the point of the .subject or author in Derrida. Now, however, she links this
point to the (im)possibility of ethics. Although this is still in part an anthro-
pologising description about the socialisation of the child '(in all of us)'
(226), Harvey quite rightly iinks this questi9n to Derrida 's writing 'beyond
good and evil' as the 'non-ethical opening of ethics'. Treading carefully
around the pitfalls of Harvey 's description, we cari pick out a valuable ar-
gument relating back to the idiomatic and the singular: the point being that
the ethical is radically 'particular' (Harvey's term, p. 227; we prefer 'sin-
gular'), and therefore bound to an always_ particular context. (We would
want to argue that this return to the question of context implies that these
books, addressing us explicitly in context, are already thereby engaged in
an 'ethical' relation.) Which implies that any generalisation of an 'ethics in
general' or 'ethics as such' or 'ethics itself' is already unethical. This can
be linked to Derrida's comments on Levi-Strauss in De la grammatologie,
where the giving of proper names, the 'suspension of the absolute vocative'
( 164 [ 112, which gives 'vocative absolute']) is called the originary violence
of language. Whence the familiar (im)possibility. This is presumably basi-
cally the impossibility of a philosophical ethics which is, whether in Kan-
tian or, apparently, Levinasian form, always still subordinate to ontology
determined in terms of presence.
There is no question of going into these questions here, although they
are the questions that matter to us.
74
Let us end with the beginning, the 'lite-
rary', the 'Introduction' to Parages, whose motifs, in-citations and solli-ci-
tations,
112 Geoffrey Bennington
lead back to places where, the criteria of decidability ceasing to be
ensured, a decision can, finally, engage.
And the event take place. More than once. [ ... ]
... ventures beyond the too-well-received divides between perfor-
mative speech and constative speech, into those environs where a
border line begins to tremble. It regularly undecides itself, between
the event of citation, in advance divisible and iterable, and the desire
of the coming itself, before any citation. But the event-meeting, de-
cision, call, nomination, initial incision of a mark-can only come
about from the experience of the undecidable. Not the undecidable
which still belongs to the order of calculation but the other, which
no calculation could ever anticipate. Without this experience, would
there ever be the chance of a step taken ffranchi]? A call for the event
(viens)? a gift, a responsibility? Would there be any other thing, any
other cause than causality? Would not everything be delivered over
to the programme? (pp. 12, 15)
Of course we have been insufficiently (and therefore excessively) philos-
ophical in our reading ofDerrida, and have much to learn from these books,
however excessively (and therefore insufficiently) philosophical they all
are too. It happens in our case that we have learned most of what we un-
derstand of events, citation, doubling (and even decapitation and hilarity)
from 'literary' texts (for the sake of economy, from Le Rouge et le noir
alone)?
5
But there is no linear scale here with a literary end and a philos-
ophical end:
76
at their best, all three of these books know this well enough.
Gasche's point is that in thinking we were being literary (or psychoanaly-
tical, or linguistic, or rhetorical) we were in fact being excessively, and
therefore this is the point of many of Derri-
da's essays, of course. But the sternness of the (which can also
provoke our hilarity, as here), however justified by the extraordinary
amount of rubbish talked for and against Derrida and deconstruction, might
always rephilosophise what, by going through philosophy to its edges (we
are tempted to say its ha-ha), is also, by that fact, the least philosophical
discourse imaginable. Harvey's third 'open letter', to Derrida himself, rec-
ognises the unrealised telos of her own work as a 'philosophy of decon-
struction' that Derrida himself would reject (xiii); Gasche's Introduction
carefully sets at a distance this risk he also acknowledges to be inevitable
and necessary; Llewelyn characteristically ends best, on the double affir-
Geoffrey Bennington 113
mation which is the object of some of Derrida 's most recent work,
78
and
which is also a call to the other, 'we repeat'. And to Llewelyn, adding to
Derrida's 'Amen amen, oui oui' his own 'hear hear' (123), we add, with an
eye on 'Pas' and 'Two words for Joyce', our own perplexed 'now, now',
polite 'come come', and mild 'ha ha'.
Notes
1. Rodolphe Gasche, The Tain of the Mirror: Derrida and the Philosophy of Reflection
(Harvard University Press, 1986); Irene E. Harvey, Derrida and the Economy of Dif-
ferance (Bloomington: Indiana University Press, 1986); John Llewelyn, Derrida on
the Threshold of Sense (London: Macmillan, 1986). Page-references to these three
works will be given in the text. No attempt is made here to do justice to Stephen Mel-
ville's interesting but rather wayward Philosophy Beside Itself: Deconstruction and
Modernism (Minneapolis: Minnesota University Press, 1986), which will, however,
occasionally reappear in the notes. Gillian Rose's Dialectic of Nihilism: Post-Structu-
ralism and Law (Oxford: Blackwell, 1984), and Peter Dews's Logics of Disintegra-
tion: Post-Structuralist Thought and the Claims of Critical Theory (London: Verso,
1987), both of which contain 'philosophical' discussion of Derrida, are not discussed
here, despite some egregious misunderstandings and tendentious misrepresentations:
see our forthcoming review in New Formations. Of course we were not unaware of
other philosophers taking an informed interest in Derrida, but who have not (yet) taken
the risk of writing books about him: this is dedicated to Andrew Benjamin, Robert
Bemasconi, Jay Bemstein, David Krell and David Wood, in the hope that they will
forgive our philosophical naivety. .
114 Geoffrey Bennington
2. See especially 'Philosophy as a Kind of Writing , in Consequences of Pragmatism
(Brighton: Harvester, 1982), pp. 90-109 [in Harvey's bibliography (273), this article
is attributed to a certain Richard Rotry, and called 'Philosophy as a King of Writing'],
and his contribution to the Against Theory 'debate' ('Philosophy Without Principles',
in W.J.T. Mitchell, ed. Against Theory: Literary Studies and the New Pragmatism
(University of Chicago Press, 1985), pp. 132-8), where he seems (p. 135) prepared to
concede to Searle the philosophical points in his exchange with Derrida, which can
only be an unusual reaction, even among analytical philosophers. As a preliminary
index of the changes in Christopher Norris s (and therefore many others ) accounts of
deconstruction. it is worth noting that in his New Accents volume, Deconstruction:
Theory and Practice (London: Methuen, 1982), Norris finds a' fine lucidity in Rorty s
account (pp. 128-9), whereas in his Modem Masters volume, Derrida (London: Fon-
-tana, 1987), he is rightly critical (pp. 150-5): this turn seems to be brought about by
Norris's essay, 'Philosophy as a Kind of Narrative: Rorty on post-modem liberal cul-
ture', in The Contest of Faculties: Philosophy and Theory After Deconstruction (Lon-
don: Methuen, 1985), pp. 139-66. Norris writes copiously, fluently and with every
appearance of plausibility in his regularly-produced books: we think this makes criti-
cism of his errors the more important. though are aware that such criticism will, for
the same reason. appear churlish and ungenerous. Let us say immediately that the Mod-
. em Masters volume is much more secure in every respect than the New Accents, as
will become clearer in these notes: nonetheless, even the Modem Masters book tends
to place Derrida within entirely classical polarities with a philosophical end and a lite-
rary end, for example, or a rational end and an irrational end, or an Enlightenment end
and a post-modem end, and, within the terms of these polarities, which are never re-
ally questioned as such, remain content to argue that Derrida is in each case nearer the
former end than you thifzk, the 'you thus positioned being either the literary critic with
all his/her familiar naiveties and rash enthusiasms, or the philosopher with all his/her
prejudices. The book does at least urge, even if it is very far from demonstrating, the
philosophical 'seriousness' of Derrida's work, and its 'ultimately ethical nature' (p.
230: on a strict reading, none of the words in this last formulation is acceptable).
Readers of the latter part of the book are also, unfortunately,likely to be misled when-
ever Norris himself translates from the French: sometimes he helpfully provides the
French words in question, so that on p. 205, for example, anyone can see that 'struc-
ture destinale' cannot really be translated as 'structural destiny', and, on p. 201, that
'program of indoctrination' is perhaps a little strong for 'institution d'enseignement'.
But no such help is available on, for example, p. 197, where Norris's rendering is 'One
can understand this Declaration as a vibrant act of faith, as an hypocrisy indispensable
to any political, military, or economic coup de force, etc., or, more simply, more econ-
omically, as the deployment of a tautology: insofar as this Declaration has a meaning
and an effect, there must be a final, legitimizing instance'. An accurate translation of
the French text (from Otobiographies: L' Enseignemenl de Nietzsche et la politique du
nom propre (Paris: Galilee, 1984), p. 27 [not p.9, as Norris has it], would give: 'One
can understand this Declaration as a vibrant act of faith, as an hypocrisy indispensable
to a politico-militaro-economic etc. coup de force or, more simply. more economically.
as the analytic and consequent unfolding of a tautology: for this declaration to have a
meaning and an effect, there must be a last instance . Perhaps these are only nuances:
Geoffrey Bennington 115
in the preceding chapter, however, Norris is led to attribute to Derrida a surprising de-
gree of sympathy with 'Oxford philosophy', on the basis of a quotation from La Carte
postale de Socrate a Freud et au-dela (Paris: Aubier-Flammarion, 1980), p. 108, where
he translates the word ingenuite as 'ingenuity' (p. 197)-it does, of course, mean 'inge-
nuousness' or 'naivety'. .
3. For Derrida as antifoundationalist, see Rorty again and Stanley Fish's 'Consequences'
in the Against Theory volume (pp. 106-31 (p.112)); for deconstruction as scepticism,
see Norris, Deconstruction, p. xii: 'Hume saw no way out of his sceptical predicament,
except by soothing the mind with careless distractions( ... ). Deconstruction is likewise
an activity of thought which cannot be consistently acted on-that way madness lies-
but which yet possesses an inescapable rigour of its own'. See too p. 127-8, where Der-
rida' s supposed scepticism is seen as better than others because he had to work hard
to get to it. The whole book has a sense of deconstruction always about to nm off un-
controllably into 'madness', and can give only extrinsic and accidental reasons why it
does not. In The Contest of Faculties, pp. 216-7, Norris again refers to 'the sceptical
rigours of deconstruction' and suggests further that 'Deconstruction is simply the most
hard-pressed and consequent of relativist doctrines applied to questions of meaning,
logic and truth': but it would seem that this would be only a relative relativism, not to
be confused with the 'unbridled relativism', or the 'relativist euphoria' that would ap-
parently characterise 'absolute' (?)relativism. (We imagine entries in a latter-day Dic-
tionnaire des idees 'Relativism: always unbridled (see scepticism)';
'Scepticism: always thoroughgoing (see relativism)'.) The account in his Modern Mas-
ters volume limits itself to Derrida's 'vigilant scepticism as regards deconstructive
"method"' (p. 20), and later suggests (we think) that Derrida is not simply a sceptic
nor a simple opponent of scepticism. But in any case Derrida is neither a sceptic nor
a relativist, nor simply an antifoundationalist, and fallibilist could only be a very par-
tial description.
4. Jonathan Culler, On Deconstruction: Theory and Criticism after Structuralism (Lon-
don: RKP, 1983), p. 213. Norris, Derrida,p. 21, suggests, to the contrary, that as part
of their 'desire to annex deconstruction as a kind of anti-philosophy', literary critics
show a 'strongly marked preference for those texts where the deconstructive ground-
work (so to speak) is very largely taken as read, and where Derrida most thoroughly
exploits the resultant for experiments in style' (the subsequent compli-
cation of the terms of this description do not alter its empirical falsity).
5. Culler, p. 85n. The note reads: 'I will not attempt to discuss the relationship of Derri-
dian deconstruction to the work of Hegel, Nietzsche, Husserl, and Heidegger. Gayatri
Spivak's introduction toOfGrammatologyprovides much useful information. See also
Rodolphe Gasche, "Deconstruction as Criticism"'. See too Culler's footnote 1 to p.
227 for a general reflection on how both defenders (including Gasche) and opponents
of deconstruction tend to make a distinction between 'original' and 'derivative' ver-
sions of deconstruction.
6. Rodolphe Gasche, 'Deconstruction as Criticism', Glyph 6 (1979), 177-216: we are
naive and ridiculous on p. 178 and 'untrained' on p. 183.
7. Edward Said, 'The Problem ofTextuality: Two Exemplary Positions', Critical Inquiry,
4 (1978), 673-714; Fredric Jameson, The Prison-House of Language: A Critical Ac-
count of Structuralism and Russian Formalism; Terry Eagleton' s remarks on Derrida,
116 Geoffrey Bennington
which it seems safest to describe as in a constant process of historical transformation,
are spread across his Waiter Benjamin or Towards a Revolutionary Criticism (Lon-
don: Verso, 1981 ), and especially the chapter, 'Marxism and Deconstruction', 131-42;
Literary Theory: An Introduction (Oxford: Blackwell, 1983) and Against the G r a i n . ~
Essays 1975-1985 (London: Verso, 1986). In a note to her 'Love Me, Love my Ombre,
Elle', Diacritics 14:4 (1984), 19-36 (p. 21n4), Spivak refers to these three authors and
links them to M.H. Abrams in more than one way: she also suggests that what she calls
a 'powerful continuist misreading of Derrida' (whatever that means) in Perry Ander-
son's In the Tracks of Historical Materialism (London: Verso, 1983) is 'worth more
careful scrutiny'. However, only quite brief scrutiny is required to show that Ander-
son 's comments about Derrida are all wrong-see for example his claim on p. 42 that
Derrida, somehow both still on a common path and 'marking the post-structuralist
break', 'rejected the notion of language as a stable system of objectification, but radi-
calized its pretensions as a universal suzerain of the modem world, with the truly im-
perial decree, "there is nothing outside of the text"', "nothing before the text, no pretext
that is. not already a text"'. The Book of the World that the Renaissance, in its naivete,
took to be a metaphor, becomes the last, literal word of a philosophy that would shake
all metaphysics': it would be hard to believe Anderson had read the books he is quot-
ing from, were it not for the fact that he gives careful references. On p. 46 he assumes
that Derrida unproblematically espouses the Nietzsche option at the end of 'Structure,
Sign and Play', despite Derrida's careful statement that no such choice is to be made
(L' Ecriture et la difference (Paris: Seuil, 1967), pp. 427-8; tr. Alan Bass (London:
RKP, 1978), p. 293: 'I do not believe that today there is any question of choosing':
Norris has a brief discussion of this point in Derrida, pp. 139-141, but omits to quote
this clear statement) [see too note 26, below]; on p. 50 Anderson is unable to under-
stand the complexity of Derrida's statements on history and historicity other than by
forcing them into a schema whereby 'A total initial determinism paradoxically ends
in the reinstatement of an absolute final contingency', and then calling this an 'irony';
on p. 54, intent on presenting Derrida as destroyer, he claims that 'Structurality, for
Derrida, is little more than a ceremonious gesture to the prestige of his immediate
predecessors', and then quotes the same bit of 'Structure, Sign and Play' to the same
effect as before (compare the discussion of structurality by Gasche (pp. 145-7) to get
a sense of what's actually going on). In the 'Foreward' to his book, Anderson recog-
nises a debt to Peter Dews, and says that the latter's forthcoming work 'will soon ren-
der these pages more or less obsolete' (p. 8): as Dews 's book makes all the same
mistakes, we assume that it is also the source of Anderson's careful references, and
deserves priority. Dews's work had made Anderson's obsolete before being published,
and Anderson returns the favour by repeating concisely in advance the errors it takes
Dews much longer to get through. As each book thus pre-empts the other and renders
it obsolete, it is strictly speaking unnecessary to read either. Even Eagleton feels ob-
liged to take Anderson to task for what he politely calls his 'seriously one-sided ap-
proach to his topic' (Against the Grain, pp 94-5), but does so with a view to making
a very simplistic distinction between 'left' and 'right' deconstruction, whereas we do
not think that the 'politics of deconstruction' can be dominated in this manner. The
very prevalent approach which is prepared. to countenance deconstruction only inso-
far as it might be 'useful' within a previously defined political strategy (for bringing
Geoffrey B.ennington 117
about socialism, say) is already denying itself any real Wlderstanding of Derrida's
work, the politics of which is much more challenging and complex.
8. The standard case would be that presented by Barbara Johnson 's justly ubiquitous 'The
Frame of Reference: Poe, Lacan, Derrida', Yale French Studies, 52 (1977), 457-505;
reprinted in various forms in her volume The Critical Difference: Essays in the Con-
temporary Rhetoric of Reading (Baltimore: Johns Hopkins University Press, 1980), in
Geoffrey Hartman, ed.Psychoanalysis and the Question of the Text (Baltimore: Johns
Hopkins University Press, 1978); Robert Yowtg, ed., Untying the Text: A Post-Struc-
turalist Reader (London: RKP, 1981). See Derrida's reaction to Johnson's attempt to
find Lacan already saying what Derrida was supposedly trying and failing to say inLa
Carte Postale de Socrate a Freud et au-deia (Paris: Flammarion, 1980), pp. 1624 (tr.
Alan Bass (University of Chicago Press, 1987), pp. 149-51), and especially p. 164
[151] on the status of 'dissemination' as 'one of the words, a m o n g ~ many others, to
lead on beyond any "last wor:d" [in English in the text]', and as not a 'master-word'.
For this type of operation, see too Shoshana Felman, Le Scandale du corps parlant:
Don Juan avec Austin ou la seduction en deux langues (Paris: Seuil, 1978), and Cul-
ler's judicious remarks in On Deconstruction, pp. 118n4 and 261. Gregory Ulmer's
essay Sounding the Unconscious' in Glassary (Lincoln and London: University of Ne-
braska Press, 1986) takes Glas to be in some sense about the Unconscious, and aims
to show 'how psychoanalysis in general, and Lacan specifically, make possible the
kind of critique of self-presence mowtted by Derrida' (p. 37). This he attempts via a
parallel between Derrida on inner speech and Lacan's discussion of Schreber. we can
do no more here than add a reference, which Ulmer seems to forget, to Derrida's own
parallel between 'hearing-oneself-speak' and Schreber, in 'Qual QueUe: Les sources
de Valery', inMarges: de laphilosophie (Paris: Minuit, 1972), pp. 327-63 (p. 354) [tr.
Alan Bass (Brighton: Harvester, 1982), pp. 273-306 (298)], and to suggest that what-
ever the interest of the parallel, to claim that 'what Lacan learns from Schreber about
the operations of the Unconscious in language is the same lesson that Derrida puts to
work in the formation of the theory and practice of the text' (p. 43) is, apart from the
repetition of the theory/practice error, to elide the most interesting questions. Jac-
queline Rose has some suggestive remarks ori Derrida in the Introduction to her Sex-
uality in the Field ofVision (London: Verso, 1986), pp. 18-23, but we think misreads
Derrida: on the 'pulsion du propre' (p. 20) at least by taking commentary of Freud for
assertion of a thesis by Derrida; by assuming that Derrida' s 'endless dispersal of sub-
jectivity' (ibid.) is an attempt simply to get rid of the subject once and for all; and by
assuming a 'psychic' account must be given of the 'logos's' 'forgetting' of differance.
It is striking that none of our three philosophical books fmds it necessary to devote
more than the briefest of passing references to Lacan. Melville's book does go into
these problems at length, and suggests interesting criticisms of Johnson and Felman-
this falls outside the limits of our discussion.
9. I.e., Hegel, Husserl, Heidegger, said by Descombes to dominate French philosophy
from 1933-60, then to be replaced by the three 'masters of suspicion', i.e. Marx, Nietz-
sche and Freud. (Le M erne et l' autre: Quarante-cinq ans de philosophie fra11faise
( 1933-1978) (Paris: Minuit, 1979), p. 13 [tr. Modern French Philosophy (Cambridge
University Press, 1980), p. 3].
118
Geoffrey Bennington
10. Gregory L. Ulmer, Applied Grammatology: Post( e)-Pedagogy from Jacques Derrida
to Joseph Beuys (Baltimore: Johns Hopkins University Press, 1985). Ulmer's basic
premise, which he supports by quoting Culler (see note 4, above) on the tendency of
literary critics to rely on Derrida's 'philosophical' texts (p. xi), is that Derrida 'decon-
structs' philosophical texts and 'mimes' literary and artistic ones (x), which is false,
and that the term 'Grammatology' can be used to cover both of these aspects of Der-
rida, which is also false. Ulmer is confident that 'grammatology' is somehow 'beyond'
deconstruction, and he also suggests that, for example, La Carte postale is 'an anti-
book awaiting relief by a Writing beyond the book. Or, to put it another way, it is a
work of theoretical grammatology which contains the script for an applied gramma-
tology' (xiii). (Norris ventures to use 'applied deconstruction' to suggest a general shift
in Derrida'swork towards 'wider political bearings', in his concern with question of
teaching and the institution of the University (p.13).) Although this approach, and the
peculiar violence it works on Derrida's own usage of the term 'grammatology', do not
of course prohibit Ulmer from saying some interesting things, we are sure these open-
ing remarks are quite mistaken. See too the concise delimitation of these problems in
David Carroll's excellentParaesthetics: Foucault, Lyotard, Derrida (New York and
London: Methuen, 1987), pp. 199n5 and 202n18.
11. Jean Maurel, recorded in Philippe Lacoue-Labarthe and Jean-Luc N ancy, eds., Les Fins
de l'homme: apartir du travail delacques Derrida (Paris: Galilee, 1981), p. 695. See
too in the same volume, apart from many excellent papers, the opening discussion as
to the validity of the 'philosophical' status Nancy and Lacoue-Labarthe want to give
the conference. (pp. 16-7)
12. Jonathan Culler, 'Jacques Derrida', in John Sturrock, ed., Structuralism and Since (Ox-
ford University Press, 1979), 154-80 (p. 155).
13. 'Restitutions of the Truth in Pointing [pointure]', in The Truth in Painting, tr. Geoff
Bennington and lan McLeod (University of Chicago Press, 1987), pp. 255-382.
14. Norris, Derrida, p. 245. This idea of a sudden 'decided shift' betrays a somewhat Len-
tricchian view of history.
15. Jacques Derrida, Dissemination, tr. Barbara Johnson (London: Athlone Press, 1981),
pp. viii and xv. See note 40, below, for explicit statements to the contrary.
16. Culler, pp. 86-8.
17. Sigmund F r e u d ~ Jokes and their Relation to the Unconscious (Pelican Freud Library,
Vol. 6) (Harmondsworth: Penguin Books, 1976), p. 240.
18. Norris, Derrida, p. 243.
19. De la grammatologie (Paris: Editions de Minuit, 1967), p. 233 [tr. Gayatri Chakravorty
Spivak as OfGrammatology (Baltimore and London: The Johns Hopkins University
Press, 1976), p. 162]. All subsequent references will be given in the text, in the form
233 [162], and we follow this convention for references other than the first to other
works by Derrida. We have consistently retranslated, not always because of inaccu-
racies in the published versions, but also because of a general desirability for different
possible translations to be suggested.
20. See Ulmer pp. 55-7 and 94-7: Ulmer's whole approach is attractively off-beat, but it is
rather ill].probable to suggest that 'Grammatology ... is a strategy of cognitive evoca-
tion, modeled on the effect of olfaction' (p. 97). See too Llewelyn, pp. 119-20, also
picking up on Derrida's discussion of disgust in Kant, in 'Economimesis', inS. Aga-
Geoffrey Bennington 119
cinski, et al., Mimesis des articulations (Paris: Aubier-Flammarion, 1975), pp. 57-93;
tr. R. Klein, Diacritics, 11:2 (1981), 3-25.
21. We do not suggest that Derrida never uses or exploits this term: to take only a recent
example, in De I' esprit: Heidegger et la question (Paris: Galilee, 1987) (t:r. GeoffBen-
nington and Rachel Bowl by, University of Chicago Press, forthcoming), p. 23, Derri-
da refers to an earlier piece of work on Heidegger in which 'I distinguished four fils
conducteurs'. See too De la grammatologie, p. 228 [159, tranSlated as 'guiding line'].
And of course there is much play onfil andfils in Dissemination. But we should be
wary of making it an important or vital Derridean concept, if only because in 'Limited
Inc.' Derrida suggests it is a term for the 'classical philosopher' (Supplement to Glyph
2 (1977), p. 73; tr. Samuel Weber, Glyph, 2 (1977), pp. 162-254 (p. 244). See tooGlas
(Paris: Galilee, 1974, p. 26a, rpt. Denoel Gonthier, 1981, I, 26a; tr. John P. Leavey, Jr.
and Richard Rand (Lincoln and London: University of Nebraska Press, 1986), 19a: 'I
shall also try not to transform love and the contradiction of familial affect into a privi-
legedfil conducteur, or even into a telos or regulating ideal. It is the test which inter-
ests me, not success or failure'. We take it that Gasche's use offil conducteur in the
context of his discussion of the Platonic symploke (p. 97) is a witticism. For a Kantian
usage, see for example his 'Idea for a Universal History from a Cosmopolitan Point of
View', tr. B. Nisbet. in H. Reiss, ed., Kant' s PoliticalWritings(Cambridge University
Press, 1970), p. 42, and 5 of the Introduction to the Critique of Judgement, where the
essentially teleological nature of the guiding thread is clear. See Lyotard's commen-
tary in Le Differend (Paris: Minuit, 1983), pp. 234-5. In similar vein, Melville uses the
term 'project' (pp. 4 and 60); but see Derrida's reply to our question using the unfor-
tunate formula 'deconstructive enterprise': 'I will insist that there is no such thing as a
deconstructive enterprise-the idea of a project is incompatible with deconstruction.
Deconstruction is a situation. Of course sometimes it takes the shape of a project, of a
text signed by somebody and so on. But what we call deconstruction in its academic
or in its editorial form is also a symptom of deconstruction at work everywhere in so-
ciety and the world; so the "enterprise" is not the essential thing in deconstruction'
('Of Colleges and Philosophy', in Postmodernism, ICA Documents 4&5 (1986), 66-
71 (p.69).) Reinscribing thefil conducteur according to these strictures would take us
back to the notion of flair, on to the notion offrayage, and involve a passing refuta-
tion of Harvey's idea that 'deconstruction, if properly understood, walks through the
text and, if improperly understood, leaves its tracks behind it' (p. 35): it is not difficult
to fmd evidence for the idea that leaving tracks is precisely what deconstruction tries
to do-see for example De la grammatologie, p. 90 [61], on the leaving of a track as
condition of not falling back into pre-critical naivety. See too, tying up these threads,
the opening reflections of 'La Pharmacie de Platon'.
22. 'Why does one still call strategic an operation which in the last instance refuses to be
commanded by a teleo-eschatological horizon? To what point is this refusal possible
and how does it negotiate its effects? Why must it negotiate them, up to and including
this why itself? Why would strategy refer to the play'of the stratagem rather than to
the hierarchical organisation of means and ends? etc. These questions will not be re-
duced so quickly' (Positions (Paris: Minuit, 1972), pp. 95-6; tr. Alan Bass (University
of Chicago Press, 1981), p. 70). There are certainly affinities to be sought in Kant's
Zweckmiissigkeit ohne Zweck of the third Critique, usually translated into French as
120 Geoffrey Bennington
'fmalite sans fm': but before you take this as confirmation of what is taken to be a culp-
able 'aestheticism', please take the time to read Derrida's discussion in 'The Sans of
the PUre Cut', in The Truth in Painting, pp. 83-118. This probably relates to our dis-
cussion of 'the Idea in the Kantian sense' later in the text.
23. C.S. Peirce, Collected Papers (Cambridge, Mass. 1934), Vol. 5, p. 416); quoted in Sa-
muel Weber, 'The Limits of Professionalism', Oxford Literary Review, VolS. Nos. 1-
2 (1982). pp. 58-74 (p. 65). .
24. As,_ for example, in 'The Principle of Reason: The University in the Eyes of its Pupils'.
in Diacritics, 13:3 (1983), 3-20, p. 5 on state of mind while preparing a lecture. On the
'auto' or 'otobiographical' in Derrida, see Llewelyn, p. 89, and Gasche, p. 331nl8 for
brief comment, and Claude Levesque and Christie V. McDonald, eds. L'Oreille de
I' autre: otobiographies, transferts, traductions; T extes et debats avec J acques Derri-
da (Montreal: VLB, 1982) [a revised form of Derrida' s major contribution appears as
partofhisOtobiographies:L'EnseignementdeNietzscheetlapolitiquedunompropre
(Paris: Galitee, 1984)].
25. Glas, p. 132b; 162b [115b] Also quoted by Leavey in his piece 'This (then) will not
have been a book ... ', in Glassary, p. 36. See too, a little later in Glas, the argwnent
ending, 'The death-agony of metalanguage is thus structurally interminable. But as ef-
fort and effect. Metalanguage is the life of language: it always flaps like a bird caught
ina subtle lime' (p. 148b; 182b [130b]; see too 186b;231b [165b]). The notion of de-
capitation also appears in Positions (p. 62 [45]); cf. too the end of 'Hors-Livre' and
the beginning of 'La Double seance' inLa Dissemination ['Out work' and 'The Double
Session' in Dissemination].
26. Norris, Derrida, p. 243; see too p. 46 for a comparison of Glas and Finnegans Wake.
Compare the more cautious and helpful comment made by Vincent Leitch (Decon-
structive Criticism: An Advanced Introduction (London: Hutchinson, 1983), p. 205):
'Glas bears to critical discourse a relation like that which Finnegans Wake holds with
the novel'. In the New Accents book, Norris suggests that 'Within Derrida's writing
there runs a theme of utopian longing for the textual "free play" which would finally
break with the instituted wisdom of language. It is a theme that emerges to anarchic
effect in some of his later texts' (p. 49). Apart from the fact that this could not be a
theme, it is clear that by definition what Norris here misleadingly calls 'free play' can-
not be an object of desire as such, but is what makes desire possible. Ulmer also makes
a mistake about this: 'while Derrida has always seen the difference between literary
and philosophical styles as representing alternative theories of language--ever since
his first book, in which he opposed Husserl to Joyce and declared his preference for
the latter-his decision to Write in a fully experimental style himself came after 1968'
(p. xi): apart from the tendentious before/after division which is of extremely limited
relevance to Derrida' s work, as Llewelyn' s impressive ranging across the whole range
of texts shows (cf. too Gasche, p. 4 ), this repeats the type of mistake Anderson makes
about choosing: although Demda does, in the Origin of Geometry Introduction, pose
the appearance of a choice between a 'Husserlian' and a 'Joycean' option (p. 103
[ 1 02]), he goes on to show that these are not at all exclusive alternatives. If it were at
all possible to find the expression of a 'preference' here, it could in fact only be for
Husserl (seep. 107 [104]), although the impression of a ' p r e f e r ~ c e ' for JQyce (as if
preference were what was at stake here) might just have been gathered from Derrida's
Geoffrey Bennington 121
later summary in 'Two Words for Joyce', tr. GeoffBennington, in Derek Attridge and
Daniel Ferrer, eds., Post-Structuralist Joyce: Essays from the French (Cambridge
University Press, 1984), pp. 145-59 (p. 149) (the French text is to be found in Ulysse
Gramophone: Deux Mots pour Joyce (Paris: Galilee, 1986), pp. 15-53). See too Ma-
rian Hobson's refutation of Ulmer on this point, in 'History Traces' (in Attridge, Ben-
nington and Young, eds. Post ..Structuralism and the Question of History (Cambridge
University Press, 1987), pp. 101-15 (p. 115, n9). We assume that the 'literary', 'Joy-
cean' reading ofGlas was encouraged by Geoffrey Hartman'sSaving the Text (Balti-
more and London: The Johns Hopkins University Press, 1981): see our brief review
in French Studies, Vol. 36, No. 3 ( 1982), 365-6. We should probably add that it is not
at all our aim here to endorse Hartman's claim that 'Derrida tells literary people only
what they have always known and repressed' (p. 23).
27. See note 7 above, and Samuel Weber's commentary in 'Closure and Exclusion', Dia-
critics, 10:2 (1980), 35-46.
28. 'Speculer-sur "Freud"' .In 'Love me, Love my Ombre, Elle' (see note 7 above), Spi-
vak takes the translator (lan McLeod) of an early fragment of this as 'Speculations-
on Freud' (OLR, 3:2 (1978), 78-97) to task for 'miss[ing] the point of the infinitive
and of giving Freud as a citation' (p. 23n7). We can now reveal that the typescript from
which the translation was made, before the publication of La Carte postale, bore the
title 'Speculations-sur Freud'. On stricture, see Marian Hobson, 'History Traces', p.
108.
29. See especially Marges, p. 381 [320].
30. The notion even merits an entry in Culler's index: neither active nor reflexive uses of
the verb 'to deconstruct' will really do, as Melville half-realizes at the end of his book
(pp. 152-3: Melville accepts these strictures when applied to the literary text, but not
when applied to what he calls criticism', which he takes Derrida to be espousing as
the task of philosophy (p. xxvii): his claim that 'the impulse to criticism is the impulse
to make of or find infinitude a positive achievement' (p. 153) betrays a basic human-
ism which limits Melville' s understanding ofDerrida). Llewelyn usefully invokes Hei-
degger's Gelassenheit and Seinlassen to get out of simple active/passive oppositions
(p. 93).1n Gasche's 'Deconstruction and Criticism', De Man is made to look guilty of
the same mistake; in his later '"Setzung and 'Obersetzung': Notes on Paul de Man'
(Diacritics,11:4 (1981), 36-57), however, it appears that De Man's notion of allegory
would prevent this type of accusation. As none of the three books here explicitly dis-
cusses the relationship of Derrida and De Man (except for a brief comment by Gasche
to the effect that 'no-one was more aware than Paul de Man' of the fundamental 'dis-
crepancies' between 'deconstructionistcriticism' and 'Derrida's philosophical enter-
prise' (p. 3)), I shall not go into the question here (except to point out that no such
sense of a nidical discrepancy emerges from Derrida' s M emoires: for Paul de M an, tr.
Cecile Lindsay, Jonathan Culler, and Eduardo Cadava (New York: Columbia Univer-
sity Press, 1986)). Again, Melville goes into the question quite fully.
31. See 'Limited Inc.', p. 38 [205]. Gasche would quite rightly immediately retort with Po-
sitions, p. 45 [33]: 'The representation of language as "expression" is not an acciden-
tal prejudice, but a sort of structural lure, what Kant would have called a transcendental
illusion'. '
122
Geoffrey Bennington
32. Gasche has some slightly confusing remarks on this historical context in his Introduc-
tion, which might be taken to suggest that we are being a little unfair: 'I discuss Der-
rida's philosophy in terms of the criticism to which the philosophical concept of
reflection and reflexivity has been subjected. The reasons for this choice are clearly
circumstantial. Indeed the dominant misconception of Derrida is based on the confu-
sion by many literary critics of deconstruction with reflexivity ... Yet my history of the
critique of reflection ... is not a straightforward history ... Hegel' s speculative criticism
of the philosophy of reflection is given what some may consider inordinate import-
ance. But Part I is intended not as a total history of that problem, but merely as an
oriented history that serves as a theoretical prelude to the systematic exposition of Der-
rida's thought ... In spite of my contention that Derrida's philosophy must be related
to the modern history of the concept of reflection and to the criticism it has drawn, I
seek primarily to bring into view Derrida's debate with the traditional paradigms of
philosophy in general' (pp 5-6, my emphasis). One feels the 'circumstantial' slipping
away.
33. Apart from 'Signature, evenement, contexte' and 'Limited Inc.', see especially Signe-
ponge!Signsponge, tr. Richard Rand (New York: Columbia University Press, 1984),
and, on dates, Schibboleth: pour Paul Celan (Paris: Galilee, 1986), tr. in G. Hartman
and Sanford Budick, eds., M idrash and Literature (New Haven: Yale University Press,
1986), pp. 307-47 ..
34. Descombes takes Derrida to think that history is already over, and that there is nothing
more to say, and no more philosophy to do, only reading and re-reading (Le meme et
l' autre, p. 162 [137:.8]: this is clearly linked to Descombes's argument, in the Intro-
duction to his Objects of all Sorts: A Philosophical Grammar, tr. L. Scott-Fox and J.
Harding (Oxford: Blackwell, 1986: the Introduction does not appear in the original
French version (Paris: Minuit, 1983)), that French philosophy in general has become
dominated by the method of immanent critique (pp. 3-4 ), and to his claim,in 'Les Mots
de la Tribu' (Critique, 456 (1985), 418-444; this text was written to presentto a French
audience the 1984 Johns Hopkins Franco-American 'Case of the Humanities' con-
ference) that 'the textual model has been overexploited ... The textual model is ex-
hausted' (p. 443). These and other arguments by Descombes will be the object of
analysis and refutation in the next issue of OLR.
35. De la grammatologie, pp. 67-8 [45-6], where the argument is that in general analysis
of the treatmel)t of writing by the tradition will ensure the possibility of 'broaching the
de-construction' of this 'greatest totality', 'the concept of episteme and logocentric
metaphysics . Harvey quotes this on p. 87 of her book, but cuts the quotation to omit
the 'broaching', thus turning 'deconstruction of metaphysics' too simply into a telos.
36. Jacques Derrida, 'Pas'. Gramma 3-4 (1976), 111-215 (pp.l18-9); reprinted inParages
(Paris: Galilee, 1986), pp. 19-116 (pp. 26-7).
37. For a concise statement by Derrida, see Positions, p. 55 [40-1]: 'If there were a defmi-
tion of differance, it would be, precisely. the limit, the interruption, the destruction of
Hegelian sublation everywhere it operates': see too 59-61 and n6 [ 43-4 and n13], La
Dissemination, p. 12n5 [6n] and 280n45 [248n], also referring to De la grammato-
logie, p. 40 [25]. Melville also takes Hegel's Phenomenology to provide a 'primary
context.' for Derrida (p. 37), but at several other points in his book acknowledges the
importance of a Kantian dimension (e.g. pp. x ~ v i i and 164n13). The resistance to the
Geoffrey Bennington 123
determination of difference as contradiction also informs Derrida's discussion of sex-
ual difference in Eperons: les styles deN ietzsche (Paris: Flammarion, 1978), especially
p. 75, n1 [tr. Barbara Harlow as Spurs (University of chicago Press, f979)J, in Glas,
especially 127-131a; 155-61a [110-14a], and 'Geschlecht: difference sexuelle, dif-
ference ontologique' (in Psyche. Inventions de 1' autre (Paris: Galilee, 1987), 395-414,
tr. in Research in Phenomenology, 13 (1983), 65-83). See too the concise statement in
reply to Jacqueline Rose in 'Qn Colleges and Philosophy', p. 71: 'My point is not
against sexual difference. It's against the transformation, the identification of sexual
difference with sexual binary opposition.' None of the three books goes into this ques-
tion (despite the importance accorded it by Derrida), apart from the briefest of com-
mentaries by Llewelyn (pp. 86-7) in a section rather misleadingly entitled 'The Gender
of Truth' (pp. 83-70); Harvey briefly shows some interest in Kant's sexual metaphors
(pp. 16-19), but then drops the question, so we shall too.
38. If we were concerned here to articulate Derrida and analytical philosophy, then the
necessary possibility of not succeeding in referring might well be the point of contact.
Derrida' s argument would say that as failure to refer is a structural or necessary possi-
bility, then that failure inhabits and infects every act of reference, however apparent-
ly successful. Llewelyn has some interesting reflection on the questions analytic
philosophy might ask of this argument (64-6); and we are grateful to Michael Morris
for suggesting that an analytical philosopher's answer to Derrida' s (rhetorical?) ques-
tion, 'What is a success when the possibility of failure continues to constitute its struc-
ture?' (Marges, 385 [324]) is simply, 'It's a success'. This logic of the necessary
possibility is easy to get wrong: Norris's discussion of the 'postal' writing in La Carte
postale suggests two different systems, one in which letters arrive, another in which
they do or might not (Derrida, pp. 192-3)-the point is rather that this is the same sys-
tem, and missing that point violently separates conditions of possibility and conditions
of impossibility: but they are the same conditions. We wonder whether some confu-
sion around the difference between two systems and one double system is not at the
root of some of David Wood's perplexities about Derrida: in 'Following Derrida' (in
John Sallis, ed., Deconstruction and Philosophy (Chicago University Press, 1987),
143-60), he takes Derrida to be exploiting two strategies around the question of the
possibility of the escape from metaphysics: one involves staying in, the other stepping
out (seep. 149)-we think this is the same duplicitous strategy, rather than two dif-
ferent ones. The whole problem of the value of the non-identical same-in-differance
in Derrida would be worth careful study and elucidation.
39. In L' Ecriture et la difference, pp. 293-5 [pp. 196-8].
40. For digging and foundations, see for example Kant' s preface to the frrst edition (1790)
of the third Critique, tr. James Creed Meredith (Oxford: Clarendon Press, 1928), p. 5 .
. See too Derrida' s commentary in La V erite en peinture, pp. 47-8 [39-40]. The diffi-
culty of accurately describing deconstruction as critique can perhaps be measured by
Derrida's comment in 'Lettre a un amijaponais' (in Psyche, 387-93 [tr. in Bemasco-
ni and Wood, eds., Derrida and Differance (Warwick: Parousia Press, 1985), 1-8]), p.
390 [pp.4-5]: 'No more is [deconstruction] a critique, in a general sense or in a Kan-
tian sense. The instance of krinein or of krisis (decision, choice, judgement, discern-
ment) is itself, as is all the apparatus of transcendental critique, one of the essential
"themes" or "objects" of deconstruction'. Or, 'deconstruction is not a critical oper-
124 Geoffrey Bennington
ation. the critical is its object; deconstruction always bears, at one moment or another,
on the confidence given the critical, critico-theoretical, that is to say, deciding instance,
the ultimate possibility of the decidable; deconstruction is deconstruction of critical
dogmatics' ('Ja, ou le faux-bond', Digraphe, 11 (1977), 83-121 (p. 103); quoted in
Leitch, p. 205); the reference to the decidable here suggests thatNorris's reading (in
the context of 'the principle of Reason') of deconstruction's being a 'certain interpre-
tation of undecidability' which 'continues to operate within the problematics of rep-
resentation and of the subject-object relation' (Derrida, p. 162), is wrong.
41. There can be no strictly Derridean justification for the exhorbitant privilege Gasche
gives to the term 'infrastructure' which, despite all his precautions, carries precisely
the sedimentations of ground and foundation he is anxious to avoid. Derrida's most
obvious use of the term, in the third of the Positions interviews, would suggest that it
is engaged in an immediate context to an even greater extent than his other terms: 'But
conversely, what is perhaps being currently reconsidered, is the form of closure that
used to be called "ideology" (a concept no doubt to be analysed in its function, its his-
tory, its provenance, its transformations), the form of the relations between a trans-
formed concept of "infrastructure", if you like, of which the general text would no
longer be the "effect" or ''reflection"*, and the transformed concept of the "ideologi-
cal'" (p. 125 [90]; my emphasis on 'if you like', which Gasche excises from his quo-
tation of the passage (p. 147)). It is hard to see in this a very strong invitation, and
G asche' s rather proud claim that 'The notion of infrastructures has not yet been picked
up by any of those who have written on Derrida' (7) might not signal such perspicac-
ity on his part after all. In Positions, the note added (by the editors of Promesse for the
first appearance of this intetview) to the asterisk in the above quote quotes De la gram-
matologie saying simply, 'Now we know that these exchanges pass only through lan-
guage and text, in the infrastructural sense we now accord this word' (p. 234 [ 164 ]):
butthis adjectival usage scarcely justifies Gasche's nominalisation of such a loaded
term.
42. See also note 25, above.
43. De lagrammatologie, p. 16 [6-7]: 'Not that the word "writing" ceases to designate the
signifier of the signifier, but it appears in a strange light that "signifier of the signifier"
ceases to defme accidental redoubling and fallen secondariness. To the contrary, "sig-
nifier of the signifier" describes the movement of language: in its origin, certainly, but
one already has the premonition that an origin whose structure is spelled out thus-
signifier of signifier---carries itself off and effaces itself in its own production. The
signified here functions always already as a signifier ... '.See too pp. 46 [30], 63 [43],
and 65 [ 44] for further hints.
44. Norris 's New Accents volume gives the puzzled reader no sense at all of why Derrida
should have kept the term 'writing', except out of provocation and loose analogy, in-
forriling misleading approximations such as, 'What is repressed [in Saussure], along
with 'writing' in its common or restricted sense, is the idea of language as a signifying
system which exceeds all the bounds of individual 'presence' and speech' (p. 27: the
second part of this sentence is really quite a novel reading of Saussure), or' ... show-
ing that writing cannot be reduced to its normal (i.e. graphic or inscriptional) sense.
As Derrida deploys it, the term is closely related [but how?] to that element of signi-
fying difference which Saussure thought essential to the workings of language. Writ-
Geoffrey Bennington 125
ing, for Derrida, is [but why??] the "free play" or element of undecidability within
every system of communication' (p. 28); these claims will never make anyone under-
stand why 'Writing is that which exceeds-and has the power to dismantle-the whole
traditional edifice of Western attitudes to thought and language' (p. 29). 1bis precise
moment in the Modem Masters volume is equally obscure: 'It may then become clear
how the individual speech-act (parole) presupposes a grasp of those signifying con-
trasts and relationships which make up the structure oflanguage as a whole (la langue ).
Speech, that is to say, is already inscribed in a differential system which must always
be in place before communication begins. And this system is very like writing, in the
sense that written signs have traditionally been thought of as marks of difference, sup-
plementarity and non-self -present meaning' (p. 92), leaving the student in the position
of Hjelmslev, of whom Derrida says in De la grammatologie (p. 89 [ 60]), 'He would
not have understood why the name writing remained to that X which becomes so dif-
ferent from what one has always called "writing".
45. We are indebted for our understanding of Norris's early misunderstandmgs to Nick
excellent review, 'Nor is Deconstruction', OLR, 5:1-2 (1982), 170-7. In a re-
view of this whole issue ofOLR (Literature and History, 10:1 (1984), 134-6), in the
course of which he confides in the reader, as a serious-minded and above all concerned
person, that he failed to get very far with the extracts of Signsponge also published in
it. because he 'kept thinking about the women of Greenham Common, and decided
life was too short', Terry Eagleton (himself thoroughly upbraided inNorris's book),
attacks Royle's 'esoteric possessiveness' and general elitism in what he describes as
a 'rather shabby review' (p .. l35). We admitto the suspicion that the (very common)
equation of complexity and elitism betrays the real (and, in every sense, pathetic) elit-
ism of equating the popular and the simple.
46. This 'rhetoricizing' view is in fact the general matrix of all the mistakes in Norris's ear-
lier book (and the basis for his critique of Eagleton), and also appears in Norris's The
DeconstructiveTurn: Essays in the Rhetoric of Philosophy (London: Methuen,1983),
specifically on p. 35: ' ... rhetoric (or 'writing' in Derrida's terminology)': for a clear
statement of different intent in the later book (which could almost be a quotation from
Gasche), see Derrida, p. 22: see too pp. 82, 108 and 170 for comments which can be
taken as criticism of his own earlier work. This again marks a shift from his earlier
thinking: The Deconstructive Turn has a 'Methodological postscript: deconstruction
versus interpretation' (pp. 163-73), which is precisely concerned to argue against Ga-
sche' s 'purism'.
47. For the question of technology, see Del' esprit, p. 26. This type of insistence should
make it clear that 'quasi' here affecting transcendentals is not to be confused with
Habermas's usage inKnowledgeandHwnan/nterests, tr. J. Schapiro (London: Heine-
mann, 1972), pp. 194-5, discussed in Garbis Kortian, Metacritique: The Philosophi-
cal Argument of Jurgen Habermas, tr. John Raffan (Cambridge University Press,
1980), pp. 1 05-8, and devastated by Gillian Rose, H egel Contra Sociology (London:
Athlone, 1981), pp. 34-6, who shows that Habermas's 'quasi-' marks a relativising,
naturalistic limitation on the transcendental: this is quite unlike Derrida's thinking, as
could be shown be tracking the similarities of Habermas's argument with Foucault's
'historical a priori', which escapes none of the problems raised by Derrida's 'Cogito
et histoire de la folie' (L' Ecriture et la difference, pp. 51-97 [31-63]). A starker sense
126
Geoffrey Bennington
of how little Habermas can understand Derrida is given by the former's ill-informed
and loose arguments in The Philosophical Discowse of Modernity (Cambridge: Polity
Press, 1988).
48. See Marges, p. 375 [316]
49. For 'I am mortal', see La Voix et lephenomene (Paris: PUF, 1967) (tr., David B. Alli-
son (Evanston: Northwestern University Press, 1973) ), pp. 60-1 [ 54-5]; for 'I am dead',
see Ibid., pp. 107-8 [96-7].
50. See De la grammatologie pp. 68-9 [47]. Cf. too La Dissemination, p. 91 [80-1].
51. This is the problem raised in the 'Question de methode' section of De la grammato-
logie with respect to Rousseau (pp. 230-1 [160-1]), and more especially in the earlier
'Introduction a l'"epoque de Rousseau"' (145-8 [97-100]).
52. Terry Eagleton includes himself in a list of the century's major Marxist aestheticians
in Waiter Benjamin ... , p.96, and uses 'Plato to NATO', among other occurrences, on
p. 141 of the same book.
53. Spivak, 'Love me ... ', p. 19 and n.l. This sort of thing makes us glad for the transcend-
entalising Gasche. We would always recommend preference of the transcendental over
the anthropologising reading of Derrida, were it always possible to tell them apart.
54. We only insist on those moments of misreading which affect the general argument as
we are following it: there are others; on p. 70 Harvey takes the phrase 'In fact, but also
for reasons of essence' to imply an equation or coincidence of fact and essence; on p.
94 she goes into a Heideggerian-toned 'interrogation' of Derrida' s question from 'The
Supplement of Copula', 'Is philosophic discourse ruled', providing the French, 'Le
discours philosophique est-il regie?' to make sure we've got it right: but the immedi-
ate context of this in fact shows Derrida asking 'the classical question: is philosophi-
cal discourse ruled-to what point and according to what modalities-by the
constraints oflanguage?' (Marges, p. 211 [177]), which is hardly the same question,
and shows an exploitation of detachability from context which might have been pointed
out in a 'philosophical' presentation; on pp. 119-20 she blows an excellent presenta-
tion of the argument for a certain ideality of the signifier by concluding, falsely, that
'the essence of the "pure" signifier is always already a signified'; we have seen her at-
tribute Derrida's gloss on Rousseau's view of animality to Derrida himself, and find
it 'ironic' that Derrida should think this; she does the same with childhood on p. 223,
and with history and evil on p. 224. We do not mention examples of misreading in-
duced by reliance on inaccurate translations, which can be found especially on pp. 59
(relying on translation of 'I' on serail en droit' as 'it is right') and 159.
55. Derrida's essay 'Une Idee de Flaubert: La lettre de Platon' ,(Psyche, pp. 305-25) hap-
pily recognises an essential link between philosophy and Flaubertian stupidity.
56. This is the substance of Eagleton's argument directed against Lyotard's recent use of
Kant, in a paper delivered to the IAPL 'Writing the Future' conference, University of
Warwick, 1986.
57. For a notable exception, see T.J.A. Clark, 'Time After Time: Temporality, Temporali-
zation', OLR, 9 (1987), 119-35.
58. 'Origin o( the experience of space and time, this writing of difference, this tissue of the
trace allows the difference between space and time to be articulated, to appear as such
in the unity of an experience (of a "same" lived experience on the basis of a "same"
body proper).', De la grammatologie, p. 96 [65.:6].
Geoffrey Bennington 127
59. L'Ecriture et la difference, pp. 15-6 [7]. For the imagination as a complex point of
proximity between Kant and Hegel, see 'Le Puits et la pyramide: in.troduction a la
semiologie de Hegel' (Marges, pp. 81-127 [69-108]), pp. 89-91 [78-9]. See too Mel-
ville, pp. 51-2. Harvey, (pp. 142-6) having duly quoted Kant. again rather perilously
uses the analysis of imagination in Rousseau out of context to stand for Detrida's
thought. Derrida discusses the imagination in Husserl in La Voix et le pherwmene, pp.
48-9 [43-5].
60. Martin Heidegger, Kant and the Problem of Metaphysics, tr. James S. Churchill
(Bloomington: Indiana University Press, 1962), 33, point b).
61. This is also explicitly Derrida's point about the 'originary passivity' of parole, in De
la grammatologie, p. 97 [66]
62. Emmanuel Levinas, 'Jacques Derrida: Tout Autrement', in Noms Propres (Montpel-
lier: Fata Morgana, 1976; rpt. Paris: Le Livre de Poche, 1987),.65-75. This text is the
object of an extensive unpublished commentary by Simon Critchley. The question of
the future in Derrida is complex, and we fear that Norris (Derrida, pp. 140-1, on the
end of' La Structure, le signe et le jeu ')mistakes the problem here. See too David Krell,
Intimations of Mortality (Penn State University Press, 1986), and our own forthcom-
ing 'Towards a Criticism of the Future'.
63. See 'Geschlecht', p. 412 [81]
64. It may well be that. along with their readings of Levinas, their uses of 'the Idea in the
Kantian sense' would be a promising place to start a reading of Derrida with Lyotard.
We cannot begin that here.
65. Ulmer (see note 10, above) thinks that Grammatology is applicable and thus can and
does already exist
66. See, for a typically robust example of this widespread type of reaction, Allon White,
'Bakhtin, Sociolingusitics and Deconstruction , in Frank Gloversmith, ed., The The-
ory of Reading (Brighton: Harvester, 1984), pp. 123-46 (p. 140): 'The luxury of"end-
less deferral" is only available to those who play with themselves in the abstract idealist
realm of "langue". The real social difference of heteroglossia in fact puts a swift end
to the unmotivated "differance" of Derridean discourse: Derridean "differance" evap-
orates once you move from "langue" to "parole", or from competence to performan-
ces'. See Robert Young's criticisms in 'BacktoBakhtin'.CulturalCritique, 2 (Winter
1985-6), 71-92 (pp. 80-3). Derrida would reply in fairly straightforwardly phenom-
enological terms to this type of criticism: 'One often sees the descriptive practice of
the "human sciences" mix up, in the most seductive confusion (in all senses of seduc-
tive), empirical inquiry, inductive hypothesis and intuition of essence, without any pre-
caution being taken as to the origin and function of the propositions put forward'
(L' Ecriture et la difference, p. 189n1 [316n46]). The same type of argument rules out
in principle the rather simplistic appeal to political or 'historical' 'grounds' which in-
forms all of Eagle ton's discussions, and, for example, Colin MacCabe' s comments on
Derrida in his Foreword to Gayatri Chakravorty Spivak, In Other Worlds: Essays in
Cultural Politics (New York and London:.Methuen, 1987), p. xi, where, after an un-
intelligible half-sentence summary ofDerrida's 'rigorous forms of analysis', MacCabe
goes on: 'Derrida elaborated this work in the context of Heidegger's meditation on
Being and in an attempt to recapture the revolutionary potential of a series of the key
texts of literary modemism-Mallarme, Artaud, Joyce, a project which found its ra-
128 Geoffrey Bennington
tionale in the situation of France in the 1960's ... that period ... was in large part a re-
action both to the sudden advent of consumer capitalism under De G aulle and the wide-
ly perceived exhaustion within the French Communist Party.' The extent to which the
confusion evident in such a description has informed most non- 'deconstructionist' ac-
counts of deconstruction is one of the reasons for our delight in these philosophical
books; and if this is anything like what 'historicizing' deconstruction means, give us
'American deconstruction' any time. .
67. The difficulties here are played out between the assertion on the one hand that 'Death
is the movement of differance in so far as it is necessarily finite' (p. 206 [143]), and,
on the other, that if the 'power' of differance 'were to become infinite-which its es-
sence excludes a priori-life itself would be reduced to an impassive, intangible and
eternal presence: infinite differance, God or death' ( 191 [ 131] ). See too pp. 99 and 108
[ 68 and 73]. litfmite differance is the same as no differance, and this is why both ends
of this 'scale' can can be called both God and death. Dijferance thus exceeds descrip-
tion in terms of finite and infinite and necessarily generates paradoxes and aporias
when approached in the language of metaphysics. This also.implies the necessary in-
terruption of the Idea in the Kantian sense and its relation to the transcendental, and
suggests that the most important thing about the conclusion to Derrida's Origin of Ge-
ometry Introduction, 'Transcendentale serait la Difference' ( 171 [ 153] ), is the tense of
the verb.
68. 'The notion of totality always refers to the essent [I' etant]. It is always "metaphysical"
or "theological'' and it is in relation to it that the notions of finite and infinite get their
meaning' (L' Ecriture et la difference, p. 207 [141]). In a note appended to this state-
ment, Derrida refers to Henri Birault's demonstration that Heidegger progressively
abandons the appeal to finitude as too onto-theological, and adds the following, which
leaves the question of the relationship between Derrida and Heidegger as mysterious
as ever: 'A thinking which wants to go to its own furthest extent [au bout d' elle-meme]
in its language, to the furthest extent of what it is aiming at under the name of origin-
ary finitude or the finitude of Being, ought therefore to abandon not only the words
and themes of the fmite and the infinite but, what is no doubt impossible, all they com-
mand in language (in the profoundest sense of this word). This last impossibility does
not mean that the beyond of metaphysics and onto-theology is barred to all access [im-
praticable]; it confirms the necessity that this incommensurable overflowing push off
from within metaphysics. A necessity clearly recognized by Heidegger. It marks clear-
ly that only difference is fundamental and that Being is nothing outside the essent.'
(pp. 208-9n2 [317 -8n70]). Compare the slightly mysterious avatars of the reference to
Birault (in the absence of reference to this note ofDerrida's), in Gasche's 'Joining the
Text: From Heidegger to Derrida' (in Arac, Godzich, and Martin, eds., The Yale
Critics: Deconstruction in America (Minneapolis: University of Minnesota Press,
1983), 156-175 (174n4)), and in his book, pp. 31.6-7. It is unclear to us, in view ofDer-
rida's note, how Gasche can take Derrida's 'persistent critique of the notion of fini-
tude' (317) as 'proof' that his quasitranscendentals are not finite whereas Heidegger' s
transcendentals are. It seems a mark of final impatience that the term 'critique' should
reappear here on the penultimate page of Gasche's book. In both the article and the
book, Gasche pins the essential difference between Heidegger and Derrida to the lat-
ter's 'passage to the order of discourse of philosophy' ('Joining the Text', p. 173), or
Geoffrey Bennington 129
'inquiry into the conditions of the possibility and impossibility of the logic of philos-
ophy as a discursive enterprise' (The Tain of the Mirror ... , p. 317). Even though Ga-
sche includes in his notion of discourse 'the conceptual, rhetorical, argumentative, and
textual order of philosophy as the thought of unity' (Ibid.), this appeal to discourse
would appear to fall back short of all the earlier insistence, against the literary appro-
priation of Derrida, on the non-empirical and non-phenomenologizable nature of Der-
rida's 'text'. We are far from sure that the difference between Heidegger and Derrida
can be specified so rapidly.
69. There may of course be strategic or economic grounds for preference among terms: see
in 'Cartouches' Derrida's dropping of (what just happens to be) the term 'contingent'
for the term 'lot' (La Verite enpeinture, pp. 230 and 275 [200 and 239], the second of
these words provoking a reference to chance (contingency?) and necessity (lot?). Be-
ware of the common mistake of assuming that Derrida shifts his terms simply because
of a sense that they are going stale and that a change would be good. Supplement, for
example, is as good a term today for the reading of Rousseau as it was in 1967: the
temporality of its rightness is that of the text, not of simple chronological time.
70. See Derrida's second essay on Levinas, 'En ce moment meme dans cet ouvrage me
voici', in Psyche ... , pp. 159-202 (translation Ruben Berezdevin forthcoming).
71. On this question of pathos, see the exchange between Derrida and Lyotard transcribed
inLes Fins de I' homme, pp. 311-3., ,.
72. For a discussion of Being as rhythm in Heidegger and Mallarme, see Gasche, 'Joining
the text ... ', p. 162, and see also Derrida, for example La dissemination, pp. 204-5 and
n3, 293, 312 and n64 [178 and n4, 251, 276 and n76] and, implicitly at least, L' Ecri-
ture et la difference, pp. 334-5 [226], and in the discussion of the fort/da game in Spe-
culer .... One of the many 'chances' of Derrida's extraordinary 'Mes Chances'
(Tijdschrift voor Filosojie, 45e Jaargang, Nummer 1 (1983), 3-40; tr. in Smith and
Kerrigan, eds., Taking Chances: Derrida,Psychoanalysis, and Literature (Baltimore:
Johns Hopkins University Press, 1984), 1-32) is that it refers to Heidegger'sreference
to Democritus'srhythmos(p. 14 [9]). Rhythm, inseparable from stricture, is also every-
where in Glas: see, provisionally, 174b; 216b [154b].
73. See 'Mes Chances' for perhaps the most extended discussion of this question. We might
risk the provisional claim that for Derrida necessity is that there be chance. We are
quite sure that this is where to start addressing questions of freedom and ethics in Der-
rida: these questions also involve questions of the event and of literature: see Mes
Chances', p. 22 [extremely inaccurately translated, 16].
7 4. We think that all of this is compatible with the excellent but still very preliminary piece
by Robert Bemasconi, 'Deconstruction and the Possibility of Ethics', in Deconstruc-
tion and Philosophy, pp. 122-39: see especially the last sentence of the text: 'But we
find the ethical enactment above all in the way deconstruction ultimately refuses to
adopt the standpoint of critique, renouncing the passing of judgements on its own be-
half in its own voice' (p. 136)-here the 'ultimately' is vital.
75. The place of hilarity in Le Rouge et le noir, as 'hilarious performative reading', has
been admirably shown by Ann Jefferson, 'Stendhal and the Uses of Reading: Le Rouge
et le noir', in French Studies, Vol. 37, No. 2 (1983), pp. 168-83.
76. SeeParages, p. 10, saying of the 'places' mentioned in the passage quoted above, 'they
leave no chance, and no right can be recognised in them to any divide between lit-
130 Geoffrey Bennington
erature and philosophy. A proposition which does not forbid, and on the contrary re-
quires new and rigorous distinctions, a whole redistribution of spaces (be it said to
those who would wish to take find their advantage from this proposition, all sorts of
advantages and they are always those of confusion)'.
77. Notably 'Le Facteur de la verite' (in La Carte Post ale), and 'Le Supplement de copule'
and 'LaMythologieblanche' (inMarges).
78. This theme of the double afftrmation is elaborated in a number of recent pieces: see for
example the end of 'Pas', the essay 'Ulysse Gramophone', and the piece in memory
ofMichel de Certeau ('Nombre de Oui', in Psyche ... , 639-50).