The Retrait of Rhetoric

Diane Davis

What is universalizable about differance with regard to differences is

that it allows one to think the process of differentiation beyond every
kind of limit: whether it is a matter of cultural, national, linguistic,
or even human limits. There is differance (with an ‘a’) as soon as
there is a living trace, a relation of life/death or presence/absence.1
But if one reinscribes language in a network of possibilities that do
not merely encompass it but mark it irreducibly from the inside,
everything changes.2

In the 1971 essay ‘White Mythology: Metaphor in the Text of

Philosophy’, Derrida describes metaphor as a classical philosopheme
that is necessary for the stabilisation of the concept and/but that
philosophy is unable to master or contain.3 His aim is not to
show, with Anatole France’s Polyphilos, that philosophical discourse
consists of a ‘white mythology’, of worn-out metaphors whose
sensual/material (natural) origins have been forgotten; it’s not to flip
the privilege from the intelligible to the sensible while maintaining
the philosophical schema’s simple, oppositional clarity (WM, 215).
Derrida’s aim is instead to demonstrate, through what he later calls
the ‘retrait of metaphor’, a kind of primary ‘metaphoricity’, a ‘tropic
and prephilosophical resource’ that gives the very idea of the proper its
chance, along with its figural ‘detour’ (229).4 He exposes the intra-
relation of metaphor and concept, the sensible and the intelligible,
their co-emergence as differential instantiations of this now generalised
metaphoricity, this pre-proper ‘resource’ of referral and deferral
that exceeds linguistic structuration. A necessarily relational—plural,

The Oxford Literary Review 45.2 (2023): 165–185

DOI: 10.3366/olr.2023.0414
© Edinburgh University Press
www.euppublishing.com/olr
166 Oxford Literary Review

divided—‘origin’, this tropic cradle lets loose an abyssal movement

of turning and swerving that, going all the way down, promises no
proper ground, ‘no unity of meaning’. It’s ‘not even a polysemia’,
Derrida emphasises; it’s ‘a nonmasterable dissemination’ that, granting
the possibility of language, ‘belongs to what is outside language’ and
therefore to what is ‘outside humanity’ (248).
Nearly two decades later, in a rarely cited interview conducted by
Gary A. Olson in 1990, Derrida directly addresses scholars in the
field of rhetoric and writing studies.5 Recalling his argument in ‘White
Mythology’, he notes that there is a ‘rhetoric hidden in philosophy
itself’, both in the form of a persuasive discourse—a ‘rhetoric against
rhetoric, against the sophists’—and ‘in the way concepts or arguments
depend intrinsically on metaphors, tropes, and are in themselves to
some extent metaphors or tropes’. It’s not ‘that all concepts are
essentially metaphors and therefore everything is rhetoric’, he clarifies,
but rather that ‘the concept of metaphor, first, is a metaphor; it’s loaded
with philosophy—a very old philosophy’ (the metaphysics of presence).
For this reason, he would ‘distrust, suspect, the couple concept and
metaphor’, as well as ‘the opposition between philosophy and rhetoric’
(RC, 16).
Derrida insists both on the ‘importance of rhetoric and the limits of
rhetoric—the limits of verbality, formality, figures of speech’. For him,
rhetoric is ‘something decisive in society’ that can be reduced neither
to ‘what is traditionally called figures and tropes’ nor to a ‘technique
exterior to the essential activity’. He cautions again, however, that
this doesn’t mean everything comes down to the ‘formal technique
of speech acts’, to ‘the way you utter words, the way you use tropes,
the way you compose’. Rhetoric is itself dependent on ‘conditions that
are not rhetorical’—which is to say: ‘the possibility of speech acts [. . . ]
depends on conditions and conventions which are not simply verbal.
What I call “writing”, or “text” is not simply verbal. That’s why I’m
very interested in rhetoric but very suspicious of rhetoricism’, of ‘giving
rhetoric all the power, thinking that everything depends on rhetoric as
simply a technique of speech’ (15).
Olson wonders, then, if it’s fair to level the charge of ‘rhetoricism’
at Friedrich Nietzsche, Paul de Man, J. Hillis Miller, Barbara Johnson
‘and others’ who ‘promulgate’ a reductive notion of rhetoric, equating
it with ‘the cognitively disruptive interplay of tropes—the status of
Diane Davis 167

text as an allegory of its ultimate unreadability’ (17). What’s crucial

for Olson seems to be whether ‘these deconstructivists’ render an
‘undue truncation of what appears to be a Western rhetorical tradition’,
including Aristotle’s more expansive understanding of rhetoric as the
observation, in any situation, of the available means of persuasion.
Derrida resists this characterisation, maintaining that each of these
thinkers aims, in their own way, both to disrupt this tradition and
to inherit it otherwise. However, in response to the allegation of
‘rhetoricism’, most specifically in de Man’s case, Derrida waffles: ‘This
is a very delicate question since names are dropped. It’s very difficult.
As you know, I’m very close to the people you mention here, but at
the same time I’m not doing exactly what they’re doing with regard to
rhetoric’, he concedes (17–18).
De Man’s ‘theories precisely of grammar, rhetoric, tropes, and
persuasion are very complex ones’, Derrida insists, contending that
when de Man ‘speaks of “unreadablity”, he is not simply a rhetoricist’.
—Not simply. But then, a parenthetical qualification: ‘(although in
comparison what I’m doing is less rhetoricist than his work)’. There
is, Derrida acknowledges, sometimes a sense in both de Man and
Miller that ‘rhetoric could have the last word [. . . ] especially for
Paul de Man. Then, perhaps, someone could speak of rhetoricism.’
Indeed, he confesses, ‘sometimes I am tempted to say, “There is a
danger of rhetoricism here—that is, of claiming to exhaust the text,
the reading of the text, through the means of rhetorical questions”’
(18, my emphasis). He continues:

What I’m suspicious of under the name ‘rhetoricism’ is the authority

of language. Rhetoric comes from, as you know, a Greek word
meaning speaking. So, the charge of logocentrism or phonocentrism
is, by itself, a charge against rhetoricism—not the narrow field
of what we call rhetoric, but simply the authority of speech, the
authority of speaking. (19)6

By the time he sits down with Maurizio Ferraris in 1994 for a

discussion that will eventually be published in A Taste for the Secret,
Derrida asserts that the ‘crux’ of what he now calls his ‘profound
debate with Paul de Man’ comes down precisely to de Man’s ‘more
“rhetoricist” interpretation of deconstruction’.7 By contrast, what
168 Oxford Literary Review

Derrida calls ‘deconstruction’, he reminds us, was from the start ‘a

protest against the “linguistic turn”’, an attempt to mark ‘the limits
of the linguistic and the limits of the rhetorical’, and to indicate that
‘the authority of final jurisdiction is neither rhetorical nor linguistic,
nor even discursive’ (76).

The notion of the trace or of the text is introduced to mark the limits
of the linguistic turn. This is one more reason why I prefer to speak of
the ‘mark’ rather than of language. In the first place the mark is not
anthropological; it is prelinguistic; it is the possibility of language,
and it is everywhere there is relation to another thing or relation to
another. For such relations, the mark has no need of language. (76)

The ‘mark’, as any inscription of a differential relation, iterates within

an illimitable network of (re)marks to which there could be no outside;
indeed, this boundless text ‘beyond human speech acts’ would be
inclusive even of the ‘flesh and bone’ of corporeal existence.8 At stake
for Derrida in his ‘debate’ with de Man is this prelinguistic and
nonanthropological text—or more specifically, the differentiating force
of this text-ile’s (re)productive events.
What I’ll suggest here, however, is that this différantial force, this
movement of a radically generalised ‘metaphoricity’ that is neither
linguistic nor human is nonetheless irreducibly rhetorical. What’s still at
stake in Derrida’s ‘profound debate’ with de Man is the deconstruction
of rhetoric more broadly, in other words, the tracing of a rhetorical
force in excess of its linguistic (verbal/discursive) instantiation, and
therefore also of its typical elaboration in what Derrida calls the
‘narrow field’ of rhetoric and what Olson calls ‘Western rhetorical
tradition’. What’s still at stake is another inheritance of this tradition,
one that remains attuned to a fundamental rhetoricity that exceeds any
theory or technique of persuasion: an unmasterable affectability and
responsivity—an undeclinable persuadability, we might even say—that
would be productive of the (re)mark, and so of both language and life.
To echo Derrida on metaphor: the retrait of rhetoric as a formal
theory or practice of actual speech acts ‘gives place to an abyssal
generalization’ of the rhetorical that, ‘having become excessive’,
can no longer be contained within its philosophical/metaphysical
designation.9 Or to echo Gasché’s paraphrase of Derrida on
Diane Davis 169

metaphor: Rhetoric as an art of turning (tropic and persuasive)

is itself ‘comprehended within’ a more fundamental turning from
which it ‘turns away, seeking to dominate it’.10 This irreducible
rhetoricity ‘cannot be dominated’, in Derrida’s words, ‘by what it
itself has engendered, has made to grow on its own soil’ (WM,
219). Rhetorical theory will never master the quasi-originary and
ontologising rhetoricity that nonetheless calls for it, and which—like
writing, like metaphoricity, like the ‘yes-yes’ or (en)gage—is no longer
a game played ‘in the world’, as he puts it in Of Grammatology; rather, it
becomes one more name for ‘the game of the world’, the very condition
for the life of the living.11
As a way into the im-possible task of attending to a rhetoricity that
precedes and exceeds the art of rhetoric, we’ll review Derrida’s most
explicit attempt to demonstrate ‘a danger of rhetoricism’ in de Man’s
work, a tendency to allow the linguistic to serve as ‘the authority of
final jurisdiction’. Our goal, however, is not so much to substantiate
Derrida’s suspicions about de Man’s potential rhetoricism as it is to get
on the tail of a rhetorical force that would be irreducible to it. Our aim,
that is, will be to follow the traces of a generalised rhetoricity older than
the authority of (human) language—and so to become like heliotropes
turning toward an also turning sun.

Textual Events
In 1998, Derrida participates in a conference organised by Tom
Cohen, Barbara Cohen, J. Hillis Miller and Andrzej Warminski, titled
‘Culture and Materiality: A post-millenarian conference—apropos of
Paul de Man’s Aesthetic Ideology’. In response to this assignment’s
‘implacable imperatives’, Derrida unleashes the à propos, the border-
crossing ‘reference-to’ that ‘allies chance to necessity, contingency to
obligation, machinelike association to the internal, intentional, organic
link’.12 ‘Typewriter Ribbon’ is a lengthy examination—eventually
apropos the assigned text—of what de Man calls the ‘textual event’
in the final essay of Allegories of Reading: ‘Excuses (Confessions)’.
According to Derrida, de Man’s approach to Rousseau’s Confessions
illegitimately binds the à propos, treating even Augustine’s Confessions
and Rousseau’s Catholicism as if they were ‘hors de propos, extrinsic
to his “propos”’ (TR, 87). In his ‘reading of Rousseau’s Confessions’,
Derrida writes,
170 Oxford Literary Review

de Man never speaks of Augustine and of this Christian history. It

is necessary to make at least some minimal reference to this because
the sedimentation in question forms an interior stratum of the very
structure of Rousseau’s text, of its ‘textual event’. It is not certain that
a purely internal reading can legitimately neglect it, even supposing
that the concept of ‘textual event’, to quote once again these words of
de Man, leaves standing the distinction between internal and external
reading. For my part, I believe that if there is ‘textual event’, this very
border would have to be reconsidered. (80)

‘Typewriter Ribbon’ engages an extended reconsideration of this

border, highlighting the implications of its infinite divisibility and
letting loose, within Rousseau’s text, a more radical dissemination of
referral than de Man’s own reading permits.
De Man devotes the first half of ‘Excuses’ to his own meticulous
(if somewhat Lacanian, Derrida notes) reading of the key episode
of Rousseau’s Confessions—to his analysis, that is, of Rousseau’s
motivation for accusing Marion of stealing the ribbon that he
had himself lifted—before then introducing the ‘textual event’ that
displaces that reading. What de Man calls ‘text’ is ‘any entity’
that can be approached as both ‘a generative, open-ended, non-
referential grammatical system and as a figural system closed off by
a transcendental signification that subverts the grammatical code to
which it owes its existence’. This ‘“definition” of the text’, de Man
adds, ‘also states the impossibility of its existence and prefigures the
allegorical narratives of this impossibility’.13 The ‘textual event’ occurs,
he says, when ‘performative rhetoric and cognitive rhetoric, the rhetoric
of tropes, fail to converge’, suddenly revealing ‘the discontinuity
between [these] two rhetorical codes’ (AR, 300). In the ‘exemplary’
case de Man is analysing, this revelation comes when Rousseau
unexpectedly announces that his motivation was not one. It was not,
after all, his desire for Marion that caused him to blurt out her name
to save himself but simply an accident of proximity: he had excused
himself on the first option that offered itself.
The text’s figural line is thereby dis-figured, along with the allegory
of reading it had inaugurated: ‘The act that initiated the entire
[interpretive] chain, the utterance of the sound “Marion”, is truly
without any conceivable motive’, de Man writes; it’s automatic,
Diane Davis 171

machine-like, a metonymic contingency independent of any desire.

The illusion of constative meaning produced by the text’s specular
metaphors is here violently displaced by ‘the total arbitrariness of a
performative excuse’ (288–9). The disfiguring effects of ‘this isolated
textual event’, de Man continues, are then ‘disseminated throughout
the entire text and the anacoluthon is extended over all the points
of the figural line or allegory’ (300). The event, undermining the
entire figural line, demonstrates the impossibility of the text that de
Man’s ‘definition’ had prefigured, thus deconstructing the allegory
of reading.
De Man offers Rousseau’s Confessions as an ‘exemplary’ illustration
of this general theory of the textual event. Here and elsewhere,
his tenacious analyses show that all meaning is rhetorical (meta-
phorical/allegorical) and therefore aberrant; that a grammatical
(metonymic/syntactical) structure effects a machine-like ‘deconst-
ruction of metaphorical totalities’, shattering the illusion of meaning
(260, 298); and that not even the autobiographical subject survives
it. That is to say: rhetoric constitutes (figures) and then threatens
(with dis-figuration) even the reader who seeks to control it—to read
it, to understand it. De Manian deconstruction throws everything
into question, including the tropes of cognition and consciousness.
—Everything, Derrida suspects, except the authority of language, and
therefore, despite it all, a certain humanist privilege.
It’s this apparently ‘more “rhetoricist” interpretation of de-
construction’ that allows de Man to read the Confessions as a
demonstration of the ‘textual event’ without this very reading in turn
overwhelming even what he calls ‘text’ and what he calls ‘event’.
It’s de Man’s attachment to ‘the literality of the letter’, Derrida
worries, that shelters the two isolatable rhetorical codes from an
abyssal metaphoricity (150), that is, from the irreducible destinerrance
necessarily ‘imprinted’ on every mark, ‘phonic or not, linguistic or not’,
as he puts it elsewhere.14 What Derrida calls text, or what he here calls
oeuvre, ‘is an event, to be sure; there is no oeuvre without singular event,
without textual event’, he writes. However, to begin to approach it, one
would first have to ‘enlarge this notion beyond its verbal or discursive
limits’ (133–4).
Warminski argues that Derrida’s extended demonstration is
ungenerous, assuming ‘a certain carping, needling, nit-picking, almost
172 Oxford Literary Review

petty quality’, and that it’s unnecessary anyway because all of the
‘additions, emendations, and supplementations’ he offers are ‘in one
sense or another already included in de Man’s reading’. The apparent
objections, Warminski asserts, stem ‘only from Derrida’s somewhat
perverse refusal to allow de Man to set up his argument’.15 Peggy
Kamuf contends, on the other hand, that Derrida aims precisely to
question this setup, which takes the ‘text’ too literally, he suspects,
binding its disseminating force to the ‘literality of the letter’.16 And
indeed, when approached from the perspective of Derrida’s ‘profound
debate’ with de Man, what Warminski describes as ‘carping’ can be
read instead as Derrida’s claustrophobic reaction to the staging of an
enclosure that he cannot abide—as his attempt to open a window, to
expose a dis-enclosure.
The tight space in question here is, again, the Grand Ballroom
that de Man calls ‘text’. Despite the complexity of this structure,
Derrida resists its oppositional feel and its stifling literality; right
away, he begins mapping out possible ‘exits’—ways ‘out’ through the
‘inside’. He points, for example, to elements of Rousseau’s confessional
narration that are astonishingly similar to Augustine’s: they were
both sixteen, they both steal fruit, they both confess this theft in
Book 2 of their Confessions. So right there where Rousseau claims the
‘veracity of truth’, Derrida suggests the possibility of a ‘supplement of
fiction’, a ‘pure and simple invention of the episode of theft out of a
compositional concern’ that is rooted in a ‘conscious or unconscious
archivation’ (TR, 82–4). It is ‘as if’, Derrida writes,

Rousseau had played at practicing an artifice of composition: he

would have invented an intrigue, a narrative knot, as if to knot
a ribbon around a basket of pears. This fabulous intrigue would
have been but a stratagem, the mēkhanē of a dramaturgy destined
to inscribe itself in the archive of a new, quasi literary genre, the
history of confessions entitled Confessions, autobiographical stories
inaugurated by a theft. (83)

The entire affair of the ribbon may have been a fabulation constituted
through ‘quasi-quotations’ drawn from ‘the archival economy of a
palimpsest [. . . ] a clandestine or encrypted lineage, a testamentary
cryptography of confessional narration’ (83). Indeed, Kamuf reminds
Diane Davis 173

us that in Derrida truth-telling depends on this supplement of fiction,

on a performative dimension that ‘allies the oeuvre to a machine
and to a mechanical repetition’; in Derrida, truth requires ‘a kind of
truth-telling machine’.17 The singular event of Rousseau’s Confessions is
simultaneously a quasi-machinelike repetition of this archive, Derrida
suggests, this ‘encrypted lineage’ on and in which it is itself inscribed.
The Confessions can be an event, in Derrida’s terms, only to the extent
that it has its ‘outside’ right there on the ‘inside’.
Likewise, though de Man insists that there is ‘nothing in the text’ to
suggest a ‘concatenation that would allow one to substitute Marion for
Mme de Vercellis in a scene of rejection’ (AR, 285), Derrida reads, in
the text, a ‘topology of sequential juxtaposition’ that is productive of
just this Oedipal concatenation (TR, 92). Derrida describes Rousseau’s
text as ‘a parergonal composition’ proceeding as a temporal succession
of ‘interlocking frames’, which is also a ‘sequence of supplements’:
‘the Catholic Mme de Warens is succeeded by the no less Catholic
Mme de Vercellis, then comes “poor Marion” and the theft-lie of the
ribbon’ (93). This sequence, Derrida writes, weaves a ‘Marial chain
of three women to whom a desire without desire links [Rousseau] as
to the breast of a virgin mother—a Mary’ (97). The succession is itself
enough to suggest, in the text, not only the Oedipal substitution that de
Man denies, but also a buildup of meaning and desire archived in the
sound ‘Marion’, which must also be denied if Rousseau’s Confessions
is to illustrate de Man’s theory. Derrida later adds both Kleist and
Rousseau’s ‘marionettes’ to this archive, noting that ‘despite its alleged
contingency’, the name ‘Marion’ becomes impossible to ‘separate
from either Marie/Mary or marionette’ (105, 158), and therefore
from a persuasive metonymic aggregate that is also de Man’s desire:
Marion/marionette/machine.
According to de Man, the drama of understanding is a ‘linguistic
rather than ontological or hermeneutic’ predicament (AR, 300), which
is why he elsewhere maintains that one can ‘approach the problems
of ideology and by extension the problems of politics only on the
basis of a critical-linguistic analysis’.18 But Derrida wants to show
that human language is itself but one instantiation of a boundless
and quasi-machinelike movement of differentiation, of iterability, of
metaphoricity that is productive of what he calls ‘text’. Or oeuvre.
‘Typewriter Ribbon’ reinscribes the event-machine that is Rousseau’s
174 Oxford Literary Review

oeuvre within this illimitable archive of archiving events that are not
only verbal and not only human. They are, however—we’ll come
back to this—thoroughly rhetorical: both persuasive and persuadable,
affecting and affectable, altering and alterable.

Oeuvre-Events
To be what it is, an event must be absolutely singular, happening one
time only, and/but it must also be legible as such, experienceable:
an event is not an event unless a living being is ‘affected by it’,
Derrida writes (TR, 72). That’s the conundrum. Legibility requires
repeatability, which suggests that the singular happening must also be
iterable, subject to or instituted through ‘the calculable programming
of an automatic repetition (what we call a machine)’, as he puts it
(72). Everything that happens takes place at this paradoxical rendezvous
between the event and the machine, an apparent antinomy that
must somehow be thought together. Thinking the logic of the event
together with the logic of the machine would produce ‘a new logic,
an unheard-of conceptual form’, Derrida writes. Their monstrous
offspring, the ‘event-machine’, would be (productive of) a ‘super-
monster of eventness’ able to survive the instant of its institution,
seizing a ‘virtual future’ (iterability) precisely by ‘cutting itself off from
its presumed responsible signatory’ and from its presumed operation
(73, 75).
Derrida has christened this (irreducible) progeny elsewhere, multiple
times in multiple ways (for example, as mark, text, archive), but in
‘Typewriter Ribbon’, he names it ‘the work, l’oeuvre’ (75). An event,
he explains,

requires not only an operation, an act, a performance, a praxis, but

an oeuvre, that is, at the same time the result and the trace left by
a supposed operation, an oeuvre that survives its supposed operation
and its supposed operator. Surviving it, being destined to this sur-
vival, to this excess over present life, the oeuvre as trace implies from
the outset this structure of sur-vival, that is, what cuts the oeuvre off
from the operation. This cut assures it a sort of archival independence
or autonomy that is quasi-machinelike (not machinelike but quasi-
machinelike), a power of repetition, repeatability, iterability, serial
and prosthetic substitution of self for self. (133)
Diane Davis 175

Derrida emphasises the ‘quasi’ here to distinguish the oeuvre’s structure

of sur-vival from what de Man calls the ‘text-machine’. The latter,
de Man says, is ‘like Kleist’s marionettes’ in that it involves ‘the
anamorphosis of a form detached from its meaning’, the ‘gratuitous
improvisation’ of ‘a preordained pattern’. The machine, de Man
explains, is ‘like the grammar of the text when it is isolated from its
rhetoric, the merely formal element without which no text can be
generated. There can be no use of language which is not, within a
certain perspective thus radically formal’ (AR, 294). Derrida detects, in
this depiction of the text-machine, a danger of rhetoricism: a tendency
to treat an apparently isolatable rhetorical code within an apparently
isolatable object-text as a ‘causa sui’, as if it had the authority of final
jurisdiction (TR, 133). Rather than argue against this description,
he reinscribes it otherwise, in part by highlighting more than once
de Man’s recourse to analogy (‘like a grammar’), which implicates
this text-machine in a network of resemblance, a fabric of irreducibly
tropic relations that testify to its own divided origin (105, 114). Here’s
Derrida:

It is not said that the machine is a grammar of the text. Nor that
the grammar of the text is a machine. One is like the other once
grammar is isolated from rhetoric. [. . . ] The machine is determined
on the basis of grammar and vice versa. Isolated from its rhetoric, as
suspension of reference, grammar is purely formal. This is valid in
general: no text can be produced without this formal, grammatical,
or machinelike element. No text and no language. De Man right
away adds, speaking of language after having spoken of text, and here
they amount to the same thing. (152–3)

What Derrida calls the oeuvre, which is not reducible to (human)

language, survives thanks to the divided origin to which every analogy
already testifies.
The necessary divisibility of the origin (which cancels the very idea
of an ‘origin’) is what Derrida is insisting on when he describes the
oeuvre-machine as both cutting and/but also cut: ‘This cut is not so
much effected by the machine (even though the machine can in fact
cut and repeat the cut in its turn) as it is the condition of production
for the machine. The machine is cut as well as cutting with regard to
176 Oxford Literary Review

the living present of life or of the living body: it is an effect of the cut
as much as it is a cause of the cut’ (133). The cut runs right through
the ‘living present of life’, dividing it within itself, marking it from the
start with the necessary possibility of repetition that both gives it its
chance (makes it experienceable, livable) and ensures that any attempt
at re-appropriation will deliver an ex-appropriation. There is therefore
no ‘present life’ from which the oeuvre would be cut off that is not
already infinitely divisible, iterable, offering itself up quasi-perforated
for endless cutting and scattering.
This cut in the ‘living present’ is (re)produced by the machine and
the condition for the machine’s production. The oeuvre is therefore
both ‘the trace of an event, the name of the trace of the event that will
have instituted it as oeuvre’ and also ‘the institution of this event itself’,
Derrida writes.

Every surviving oeuvre keeps the trace of this ambiguity. It keeps

the memory of the present that instituted it, but, in this present,
there was already [. . . ] the essential possibility of this cut—of this
cut in view of leaving a trace, of this cut whose purpose is survival.
[. . . ] This cut is at once a wounding and an opening, the chance of a
respiration, and it was in some way already there at work, à l’oeuvre. It
marked, like a scar, the originary living present of this institution—
as if the machine, the quasi-machine were already operating, even
before being produced in the world, if I can put it that way, in the
vivid experience of the living present. (134)

Like a scar. This cut—this structure of originary metaphoricity—that

has always already divided any apparently isolatable identity or present
presence maintains an opening that assures the necessary possibility of
survival, of living on, of life—or rather of life death. As if l’oeuvre were
irreducible. As if it were oeuvre-events all the way down.

Undeconstructable Rhetoricity
The oeuvre-event is not necessarily a human work, nor is it necessarily
linguistic, verbal or even discursive. I want to highlight just two of the
instances in ‘Typewriter Ribbon’ where Derrida explicitly underscores
this point, confirming the boundlessness of the textile, or ‘the archive
as oeuvre’ in which or on which Rousseau’s oeuvre-event is itself
Diane Davis 177

inscribed (133). And I want to note that both of these pre-

linguistic and nonanthropological oeuvre-events require a supplement
of rhetoricity, a fundamental rhetorical force that Derrida marks
without naming as such.
The first instance, recounted by Rousseau, involves the deathbed
scene of Mme de Vercellis, who lets fly a nonverbal ‘next-to-the-last-
breath’, as Derrida puts it, which calls for a final explication. Rousseau:
after Mme de Vercellis ‘could no longer talk and was already in her
death agony, she broke wind loudly. “Good,” she said, turning over,
“a woman who can fart is not dead.” Those were the last words she
spoke’ (quoted in TR, 95). And Derrida: ‘After her ultimate silence
or her last words have been verified, after it has been said that “she
could no longer talk”, well, there she goes and farts again. She thus
adds a living, surviving gloss to this after-the-last word: a fart’ (95).
The penultimate puff is a nonverbal provocation emphatic enough—
persuasive enough—to raise the nearly dead. This singular yet legible
eruption, the expiring trace of an internal alterity testifying, still, to a
living on within the economy of death, compels a singular rejoinder: a
‘testamentary metalanguage on a next-to-the-last breath’, confirmation
from one who, on her way out, is nonetheless ‘not dead’. A last
discursive expression affirms a last nondiscursive expression; it’s a
‘double expiration’, as Derrida puts it (95).
I emphasise the call and response framing of this tragicomical event,
as it sketches the scene of an irreducible rhetorical engagement that is
bound up in the very structure of survival. For an event to be an event,
Derrida writes, it has to ‘happen to someone, to some living being who
is thus affected by it, consciously or unconsciously’ (72). To be affected
by some happening is (already) to respond to it, to receive it as a sort of
undeclinable call. Now, ‘it may be’, as Jean-Luc Nancy proposes, ‘that
one responds to the call only by repeating the call—as night watchmen
do. It may be that the imperative is not the response, but only the
obligation to respond, which is called responsibility.’19 Response-ability,
before anything else, is this obligation to respond, this quasi-originary
rhetorical imperative to which one finds oneself already responding—
prelusively, quasi-automatically (ça répond)—before there is a chance
to decide, even where this ineluctable response is followed by a refusal
to sign for it, to confirm by following up. A supposed operation that
is not necessarily human leaves a trace of itself, a sign of life, which
178 Oxford Literary Review

calls for a(nother) response, which leaves a trace of itself. . . This scene
of address ‘is not necessarily a dialogue or an interlocution, since
it assumes neither voice nor symmetry’, Derrida writes in ‘Ulysses
Gramophone’, ‘but the haste, in advance, of a response that is already
asking’, calling for a(nother) response.20
The nonlinguistic address that cuts through the silence of Mme de
Vercelli’s ‘death agony’ after her ‘last words had been verified’ effects a
final rally, a response that is signed by a testamentary speech act: ‘Good
[. . . ] not dead’. It’s not an apology—she doesn’t apologise, which is
the speech act one might well expect under less dire circumstances. No,
rather than excusing herself (‘oh, excuse me!’), she responds as if taking
a life-conferring call: she responds by reviving enough to speak, again,
leaving a(nother) sign of life. Before anything else, the response involves
an affirmation—yes, hello!—that is simultaneously an individuation,
the cutting of an I that is already responding, already addressing itself
to some other in response. The yes, as both an affirmation of the other
and a differentiation from the other, connects what it holds apart:
Yes-I. There could be no response without the spacing of individuation
and no individuation without response: the I is allotted, each time,
in a quasi-automatic responsivity that precedes and exceeds it. In ‘A
Number of Yes’, Derrida describes this yes as a ‘sort of pre-engagement
presupposed by every language and every type of speech act’. Belonging
‘without belonging to the whole that it simultaneously institutes and
opens’, the yes ‘causes to be and lets be everything that can be said. [. . . ]
It is without being language.’21 Which is why he also writes: ‘I say the
yes and not the word “yes” because there can be a yes without a word.’22
The ‘pre-performative force’ of this ‘minimal primary yes’, like
a ‘tap through a prison wall, marks, before meaning or signifying:
“I-here”, listen, answer, there is some mark, there is some other’.23
Without which, no I and no response. To be an affectable being is
to have been always already affected: always already une oeuvre, then,
a surviving trace. That is: the undeclinable rhetorical imperative that
(un)grounds the so-called ‘authentic’ response in a quasi-machinelike
reaction also grants (or cuts)—as a work, une oeuvre—both the
unsubstitutable singularity of this here I and the living present of
the life that it calls its own. Death, as Levinas once put it, is the
scandal of ‘no-response [de non-réponse]’.24 Nothing happens for a
living being, no event and no life, outside this undeconstructible
Diane Davis 179

rhetoricity, this irreducible response-ability that has no need for

the specifically linguistic interventions it might also institute
and open.25
The second instance I want to highlight seems at first to come from
out of nowhere. Derrida is analysing Rousseau’s guilty jouissance over
his ‘quasi-incest’ with Maman when he abruptly swerves:

A few years ago, when I was rereading these pages of Rousseau

for a seminar [. . . ] a prodigious archive had just been exhumed,
in Picardy, and then deciphered. In layers of fauna and flora were
found, protected in amber, some animal or other (which would be
nothing new), but also the cadaver of an insect surprised by death, in
an instant, by a geological or geothermal catastrophe, at the moment
at which it was sucking the blood of another insect, some fifty-four
million years before humans appeared on earth. (TR, 130)

After an effusive description of this archive’s significance, to which we

will return, Derrida admits: ‘I don’t know why I’m telling you this’
(130). And/but then, two pages later: ‘But yes, I think I remember now’
(132). What he remembers and repeats is something he had already
relayed in ‘Fourmis’ in 1994: the first proofs for Of Grammatology had
come back with an overdetermined typo that rendered, in the epigraph
he devotes to Rousseau, the word ‘incest’ as ‘insect’. ‘I was tempted, for
a moment, not to correct it’, Derrida confides. ‘The compositor in fact
had set: “J’étois comme si j’avais commis un insecte,” “I was as if I had
committed an insect”’ (132).
His (unconscious) motivation for having just told us about the
Picardy archive, he suggests, was to introduce a discussion about
‘effacement and prostheses, about falsifications of the letter, about
the mutilation of texts, of bodies of writing exposed to cutting no
less than insects are (and “insect”, insectum, as you know, means
“cut”, “sectioned”, and, like “sex”, sexus, sectus, it connotes section,
separation, and so forth)’ (132).26 His motivation was to highlight the
fragility of the oeuvre, which is necessary for its survival: the
oeuvre sur-vives through repetitions that each time threaten its
effacement, falsification, mutilation. Derrida points up de Man’s
various mutilations of Rousseau’s oeuvre—the addition of this
word, the snipping of those two—not to ‘nit-pick’ but to confirm
180 Oxford Literary Review

that the trace of what he calls event is effaceable and erasable—

affectable—and that no performative speech act can protect against
this threat.
There are other reasons to welcome the intrusion of fifty-four-
million-year-old insects into ‘Typewriter Ribbon’, and our interest is
in this one: what’s discovered in Picardy is an archive of prehuman
events, the one-time-only experiences of living beings the size of a
pinhead, the legible traces of which, frozen in ‘an amber coffin’, are
still arriving/happening to ‘us’ today. Fifty-four million years before
humans showed up, living beings experiencing a singular event were
archived in the act of it. It’s one thing to be able to refer to ‘some
animal or other’ fossilised in amber millions of years ago; it’s another
to be able to point to ‘what took place one time, one time only [. . . ]
like that animal surprised by catastrophe at the moment [. . . ] it was in
the process of taking its pleasure sucking the blood of another animal,
just as it could have taken it in some other way’, Derrida emphasises.
‘For there is also a report of two midges immobilized in amber [. . . ]
when they were surprised by death as they made love: fifty-four million
years before humans appeared on earth, a jouissance took place whose
archive we preserve. It arrives/happens to us again, it is still arriving
to us’ (130–1).
It is extraordinary, Derrida continues, to be able to refer to ‘the
archive of a singular event and, what is more, of an event that happened
to some living being, affecting a kind of organized individual, already
endowed with a kind of memory, with project, need, desire, pleasure,
jouissance, and aptitude to retain traces’ (131). What he calls an
event, it’s worth repeating, requires some living being, ‘human or
not’, to whom it happens, ‘who is thus affected by it, consciously
or unconsciously. No event without experience (and this is basically
what ‘experience’ means)’, he writes (72). Therefore, no ‘experience’
without event, either: to have an aptitude to retain (archive) traces,
to be affectable, is already to be an affected/inscribed work, an oeuvre
that produces oeuvres, an archive that produces archives. What is
experienced as unique or personal—project, need, desire, pleasure,
jouissance—cannot not be relational, already the effect of a preliminary
response-ability, an undeclinable rhetorical imperative to respond to
the trace of alterity. There could be no outside to this network of
traces, no experience or response that is not another expression of this
Diane Davis 181

text-ile that (re)produces itself through the ontologising force of an

undeconstructable rhetoricity.

Retrait of the Trait

In 1992—between Derrida’s interviews with Olson (1990) and Ferraris
(1994)—renowned classical rhetoric scholar George A. Kennedy
published, in a top journal of rhetorical theory, ‘A Hoot in the
Dark: The Evolution of a General Rhetoric’.27 As its title suggests,
it was an untimely essay that proposed, scandalously at the time,
that rhetoric in its most generalised form is both pre-linguistic and
nonanthropological. It’s not a strictly human art of speaking but
a preverbal ‘energy’ that is prior to meaning, intention and belief;
that is ‘manifest in all animal life’; and that ‘existed long before
the evolution of human beings’ (HD, 4). In its essence, Kennedy
suggests, rhetoric ‘may perhaps be identified with the energy inherent
in communication’—physical, emotional, and mental—making it the
precondition not only for spoken language but also for ‘physical
actions, facial expressions, gestures, and signs generally’ (4). As the
absolute motor of any communicative activity, rhetoric is the name
for a force of responsivity that, according to him, human beings
share with every other life form capable of generating/responding
to signals, including plants.28 Every signal sent or received—a
frown, a warning call, a word, a scent—is itself, therefore, an ex-
pression of rhetorical energy. And Kennedy describes as rhetorical
the fascinating communication practices of numerous animal species
(crows, deer stags, octopuses, rattlesnakes), spotlighting their often
quite sophisticated, even Aristotelian, means of persuasion.
He cites Of Grammatology, and he is clearly inspired by Derrida’s
generalisation of writing. However, Kennedy’s ‘general rhetoric’
assumes a naturalist frame informed by evolutionary biology and
studies in animal communication. When he asserts that all signalling
life forms ‘share a “deep” universal rhetoric’, he means it’s biological,
it’s in the DNA (6). It’s because rhetoric is for him ‘something
found in nature’ that he thinks it may be possible ‘to identify some
quantitative unit of rhetorical energy—call it the “rheme”—analogous
to an erg or volt’ (1–2). According to him, while the intensity of this
energy adjusts to immediate circumstances, the faculty for it is innate,
genetically hardwired: ‘rhetoric, as energy, has to exist in the speaker
182 Oxford Literary Review

before speech can take place. It’s prior in biological evolution and prior
psychologically in any specific instance’ (4). Indeed, ‘rhetoric’s function
is the survival of the fittest’, he flatly contends, and ‘the rhetorical code
evolves by selective variation’ (10).
It’s this purely naturalist perspective that allows Kennedy to describe
as indivisible entities both this rhetorical energy and the ‘individual’
whose ‘integrity’ it putatively ‘expresses’ (11). As if this force were
reducible to form, and as if the differentiation of a ‘self’ from an
‘other’ were not already thanks to it. The naturalist frame also leads
Kennedy to mis-take arche-writing for the inscription of empirical
marks, to argue that though ‘writing, or “Grammatology” as described
by Derrida, is prior to speech in that a kind of marking is prior in
evolutionary development [. . . ] marking does not seem to be prior to
rhetoric’, since ‘the impulse and the expenditure of energy required
in marking necessarily must exist before the marking, or “writing”,
can take place.’ Whereas both rhetoric and ‘marking’ are for Kennedy
‘mechanisms for survival’, he insists that ‘the most primitive form
of marking is a vehicle for rhetoric’ (11–13). As if this rhetorical
‘essence’ were, in Derrida’s terms, ‘rigorously independent of that
which transports it’ (WM, 229).
In contrast to Kennedy’s notion of a general rhetoric, Derrida
repeatedly depicts without naming a kind of generalised rhetoricity that
lines up, as nonsynonmous substitution, with what he calls writing,
trace, différance, supplementarity, metaphoricity. Though irreducible,
this rhetoricity could not be innate because it could not not be
relational, both cutting (Yes-I) and already an effect of the cut. An
ineluctable affectability or response-ability that implies a structure of
sur-vival, this rhetoricity is like the Lucretian-Epicurean clinamen, the
unpredictable swerve or inclination that, Derrida writes, ‘exacerbates
an initial gap and produces the concentration of material (systrophē)
that gives birth to the worlds and the things they contain’.29 But this
‘initial gap’ must be understood also to run right through Epicurus’s
atom. That is, this putatively indivisible atom—as ‘trait, letter, seminal
mark’—must itself be understood to be continually ‘multiplying and
dividing itself internally’, scattering itself like superfluous seeds.30 It’s
necessary to recognise, in other words, that any entity’s phenomenal
appearance owes itself to this essential divisibility and to a rhetoricity
that ontologises without ontological ground. There could be no
Diane Davis 183

individual entity, no social relation, no speech act or signifying practice

that is not an expression of this rhetoricity that comprehends it.
The drama of understanding is rhetorical in this sense. Any approach
to ethico-political problems that does not attend to this ontologising
rhetoricity risks compounding them by reducing ‘likeness and turning
to identity’, as Michael Naas puts it, ‘presencing to presence’ and
‘resolute turning to internal resolve’.31 I don’t have to tell you what’s
at stake in this reduction, what’s on the line today in this fragile
distinction between ‘presencing as turning and unturning presence’—
on the left as well as the right.32 And rhetorical studies is in a unique
position to insist on it, to put this rhetoricity ‘on the scene’, to study
it—to turn rhetoric toward the study of this fundamental rhetoricity,
emphasising what Naas describes as ‘the tragic necessity of having to
forge an existence’, each time, ‘from within turning’.33

Notes
1
Jacques Derrida and Elisabeth Roudinesco, For What Tomorrow, translated by Jeff
Fort (Stanford: Stanford University Press, 2004), 22.
2
Derrida, ‘“Eating Well”, or the Calculation of the Subject: An Interview with
Jacques Derrida’, in Who Comes After the Subject? edited by Eduardo Cadava, Peter
Connor and Jean-Luc Nancy (New York: Routledge, 1991), 96–119 (116).
3
Derrida, ‘White Mythology: Metaphor in the Text of Philosophy’, in Margins of
Philosophy, translated by Alan Bass (Chicago: The University of Chicago Press,
1982), 207–71. Hereafter abbreviated as WM.
4
Geoffrey Bennington explains that ‘retrait’ marks ‘both metaphor’s retreat or
retracing from the scene—it has no contrary and knows no bounds, and thus
escapes restrictive definition—and its retracing as something other than it always
was.’ See his ‘Metaphor and Analogy in Derrida’, in A Companion to Derrida, edited
by Ed Leonard Lawlor and Zenep Direk (Oxford: Wiley Blackwell, 2014), 96.
5
Gary Olson, ‘Jacques Derrida on Rhetoric and Composition: A Conversation’,
Journal of Advanced Composition, 10.1 (1990), 1–21. Hereafter abbreviated as RC.
6
Derrida’s reluctance to cop to this temptation is worth noting, as it comes on
the heels of his nuanced response to de Man’s wartime journalism. See Derrida,
‘Like the Sound of the Sea Deep within a Shell: Paul de Man’s War’, translated
by Peggy Kamuf, Critical Inquiry, 14.3 (1988), 590–652, and his biting reply
to ‘tedious’ critiques of that response, ‘Biodegradables: Seven Diary Fragments’,
Critical Inquiry, 15.4 (1989), 812–73. In the former, Derrida suggests that the
discovery of de Man’s Belgian writings obliges ‘all of us’ to reread de Man very
184 Oxford Literary Review

carefully, ‘to analyse the traps and the stakes—past, present, and especially future’.
‘I know that I am going to reread him’, Derrida adds. ‘He will always interest me
more than those who are in a hurry to judge’ (650).
7
Jacques Derrida and Maurizio Ferraris, A Taste for the Secret, translated by Giacomo
Donis (Boston: Polity, 2001).
8
Derrida, Limited Inc. (Evanston: Northwestern University Press, 1988), 134;
Of Grammatology, translated by Gayatri Chakravorty Spivak (Baltimore: Johns
Hopkins University Press, 1997), 158–9. Derrida makes this point very explicitly in
his 1975–76 seminar on Life Death. See Derrida, Life Death, translated by Pascale-
Anne Brault and Michael Naas (Chicago: The University of Chicago Press, 2020),
77: ‘The modern biologist or geneticist [. . . ] does not write a text on something
that would be outside the text [hors texte], something that would be a-textual, that
would form a referent whose nature would be to be foreign, in its being or in its
structure, to textuality, but, quite to the contrary, he writes a text on a text, a text
on text, in order to demonstrate, recall, or write that his object has the structure
of a text, and that there is no longer anything in the object of his science or of his
research, nothing as scientific object, that is meta-textual.’
9
Derrida, ‘The Retrait of Metaphor’, in Psyche: Inventions of the Other, Vol. 1,
edited by Peggy Kamuf and Elizabeth Rottenberg, translated by Kamuf (Stanford:
Stanford University Press, 2007), 48–80, see especially 22–4.
10
Rodolphe Gasché, ‘The Eve of Philosophy: On “Tropic” Movements and Syntactic
Resistance in Derrida’s White Mythology’, International Yearbook for Hermeneutics,
13 (2014), 1–22 (12–13).
11
Derrida, Of Grammatology, 50.
12
Derrida, ‘Typewriter Ribbon: Limited Inc (2)’, in Without Alibi, translated by
Peggy Kamuf (Stanford: Stanford University Press, 2002), 71–160 (75, 77).
Hereafter abbreviated as TR.
13
Paul de Man, Allegories of Reading: Figural Language in Rousseau, Nietzsche, Rilke,
and Proust (New Haven: Yale University Press, 1979), 270. Hereafter abbreviated
as AR.
14
Derrida, ‘My Chances’, in Psyche: Inventions of the Other, Vol. 1, 344–76 (359).
15
Andrzej Warminski, ‘Machinal Effects: Derrida With and Without de Man’, MLN,
124.5 (2009), 1072–90.
16
Peggy Kamuf, ‘Introduction: Event of Resistance’, in Without Alibi, 1–27 (8–9).
17
Peggy Kamuf, ‘Introduction’, 11.
18
de Man, The Resistance to Theory (Minneapolis: University of Minnesota Press,
1986), 121.
19
Jean-Luc Nancy, The Birth to Presence, translated by Brian Holmes (Stanford:
Stanford University Press, 1993), 323.
Diane Davis 185
20
Derrida, ‘Ulysses Gramophone: Hear Say Yes in Joyce’, in Acts of Literature, edited
by Derek Attridge (New York: Routledge, 1992), 253–309 (299).
21
Derrida, ‘A Number of Yes’, in Psyche: Inventions of the Other, Vol. 2, edited by
Peggy Kamuf and Elizabeth Rottenberg, translated by Brian Holmes (Stanford:
Stanford University Press, 2008), 231–40 (236–7).
22
Derrida, ‘Ulysses Gramophone’, 296.
23
Derrida, ‘Ulysses Gramophone’, 298.
24
Emmanuel Levinas, God, Death, and Time, translated by Bettina Bergo, edited by
Jacques Rolland (Stanford: Stanford University Press, 2000), 37. Levinas describes
the subject as the eruption of an I within the ‘self’ for the purpose of response. This
unity of the corporeal self and the I-for-the-other, Levinas also calls a ‘creature’,
responsibility’s creation. The Levinasian subject is therefore an oeuvre, and/but it’s
not likely that Levinas would call this provocation le visage d’autrui.
25
By contrast, de Man famously writes in The Resistance to Theory that ‘death is a
displaced name for a linguistic predicament’ (81). I doubt Derrida would be caught
dead inscribing that sentence.
26
For a brilliant analysis of connexions Derrida draws here and in ‘Fourmis’ between
insects and sexual difference, see Elissa Marder, ‘Insex’, Parallax, 25.2 (2019),
228–39.
27
George A. Kennedy, ‘A Hoot in the Dark: The Evolution of a General Rhetoric’,
Philosophy & Rhetoric, 25.1 (1992), 1–21. Hereafter abbreviated as HD.
28
In ‘Typewriter Ribbon’, Derrida uncharacteristically situates plants among the
nonliving, in the ‘timeless time’ of ‘sediments and rocks’ (TR, 130)—there, even
he appears to turn a heliotrope into ‘a dried flower in a book’ (WM, 271). By
contrast, in ‘Eating Well’, he had left the flower in the garden: ‘The difference
between “animal” and “vegetal” also remains problematic. Of course the relation to
self in ex-appropriation is radically different (and that’s why it requires a thinking
of différance and not of opposition) in the case of what one calls the “nonliving”,
the “vegetal”, the “animal”, “man”, or “God”’ (106).
29
Derrida, ‘My Chances’, 350.
30
Derrida, ‘My Chances’, 360. For Derrida’s discussion of what Plato considers the
‘superfluous seeds’ of rhetoric, see Dissemination, translated by Barbara Johnson
(London: Athlone Press, 1981), 150.
31
Michael Naas, Turning: From Persuasion to Philosophy: A Reading of Homer’s Iliad
(New Jersey: Humanities Press, 1995), 207.
32
Naas, Turning, 208.
33
Naas, Turning, 220.