No More Learning

1 Paul de Man's reading of this moment in the third Critique is no less enigmatic and, if any- thing, even more peculiar, not least of all because the vision of the ocean "as the poets do it"--"merely by what appears to the eye" (bloss . . . nach dem, was der Augenschein zeigt--"merely according to what the appearance to the eye shows," to put it more "literally," or "according to what meets the eye")--is termed by him a "material vi- sion" whose "materiality" is linked to what de Man calls Kant's "ma- terialism" (or "formal materialism"): "The critique of the aesthetic," he writes, "ends up, in Kant, in a formal materialism that runs counter to all values and characteristics associated with aesthetic experience, including the aesthetic experience of the beautiful and of the sublime as described by Kant and Hegel themselves" (AI 83). 2 That it might be better not to assume anything about our understanding of de Man's difficult "materiality" and "materialism" is certainly confirmed by the way the term gets introduced in "Phenomenality and Materiality in Kant. " After characterizing the architectonic vision of the heavens and the ocean--"The heavens are a vault that covers the totality of earthly space as a roof covers a house," writes de Man--as being neither "a trope or a symbol" nor "literal, which would imply its possible figurali- zation or symbolization by an act of judgment," de Man writes that "The only word that comes to mind is that of a material vision, but how this materiality is then to be understood in linguistic terms is not,
3
4 Andrzej Warminski
as yet, clearly intelligible" (AI 82). Since "material" is a word, the only word, that comes to mind here, one can already suspect that its intel- ligibility will indeed have a lot to do with its being understood "in linguistic terms. " We will get to those terms soon enough, but it is al- ready worth remarking that the word material is one that merely "comes to mind," as though on account of the lack of a word, the proper word, to designate the peculiarly unfamiliar nature of this vi- sion. I say "unfamiliar" advisedly, for de Man goes to some pains-- both before and after the word that comes to mind--to explain at length what this material vision is not and is not like. It is a vision en- tirely devoid of teleological interference, it is not a metamorphosis, not a trope or a symbol, heavens and ocean as building are a priori, previ- ous to any understanding, to any exchange or anthropomorphism, there is no room for address in Kant's flat third-person world, this vi- sion of the natural world is in no way solar, it is not the sudden discov- ery of a true world as an unveiling, as the a-letheia of Heidegger's Lichtung, "we are not to think of the stars as suns moving in circles," nor are we to think of them as the constellation that survives at the apocalyptic end of Mallarme? 's Coup de De? s, and so on. The list of what this vision and its materiality are not (and are not like) could be extended; as de Man says, "It is easier to say what the [Kant] passage excludes and how it is different from others than to say what it is. " Indeed, since "no mind is involved in the Kantian vision of ocean and heaven," it is no wonder that the only word to characterize it (appar- ently) nonnegatively can only "come to mind"--as though "by acci- dent," as one says, no doubt simultaneously utterly random and yet completely determined, that is, overdetermined like the nightmarish hypograms of Ferdinand de Saussure. 3
But what is most striking (for the "mind" or the "eye" or whatever) about de Man's elegiac-sounding and yet nonelegiac enumeration of what the poets' material vision of heaven and ocean is not and not like is his going out of his way to insist that it is not like the poet Wordsworth's, for example, apparently similar intuitions in passages like the nest-robbing episode of The Prelude where the destabilized sky is nevertheless still a sheltering sky. "Kant's passage is not like this," asserts de Man, "because the sky does not appear in it as associated in any way with shelter. " Dwelling poetically in Kant's architectonic world would seem to mean precisely not dwelling in the building con- structed of heavens and ocean when it is seen merely as the poets do it, according to what the Augenschein shows:
The poet who sees the heavens as a vault is clearly like the savage [in Kant's Logic],4 and unlike Wordsworth. He does not see prior to dwelling, but merely sees. He does not see in order to shelter himself, for there is no suggestion made that he could in any way be threatened, not even by the storm--since it is pointed out that he remains safely on the shore. The link between seeing and dwelling, sehen and wohnen, is teleological and therefore absent in pure aesthetic vision. (AI 81)
Nor, de Man insists, is the Kantian vision like the "sense sublime" in the famous passage of Wordsworth's "Tintern Abbey," which is "an instance of the constant exchange between mind and nature, of the chi- asmic transfer of properties between the sensory and the intellectual world that characterizes [Wordsworth's] figural diction. " No mind being involved in the Kantian vision, "to the extent that any mind, any judgment intervenes, it is in error. " And since Kant's architectonic world is not a metamorphosis, not a trope, not a symbol, and prior to any exchange or anthropomorphism, it cannot be addressed the way the poet Wordsworth does it in book 5 of The Prelude as "the speaking face of nature. " (Actually, in Wordsworth it is "the speaking face of earth and heaven" [and not the "speaking face of nature"] and it is not, at that moment, addressed! )5 So: not a sheltering sky or earth, not in an economy or tropology of exchange in relation to the mind, and not anthropomorphized or to be addressed. Such would be the materi- ality of what the Augenschein shows in Kant's, for lack of a better word, material vision.
I recapitulate de Man's examples here in order to give some sense of how far he goes in his insistence that what the poets do in Kant is not (like) what the exemplary poet Wordsworth does. What are we to make of this apparently stark divergence between a material vision "as the poets do it," according to Kant, and a figuralized aesthetic vision and a sublime that are everything the material vision is not, as one poet, Wordsworth, does it, according to de Man? And we do not have to know all that much about the special status of "Wordsworth" in de Man's private "canon" to know better than to think that Wordsworth is somehow being given as an example of an insufficient or "inauthen- tic" poet! The fact that Wordsworth comes back still later in the essay to serve quite different purposes--this time as an example of other texts in which there is a "blank" like the "blank" de Man reads be- tween sections 27 and 28, that is, between the accounts of the mathe- matical and the dynamic sublime, in Kant's third Critique--should be
"As the Poets Do It" 5
6 Andrzej Warminski
enough for those who can read. But how Wordsworth comes back here is certainly telling. This time it is not so much what Wordsworth wrote, what is there on the page, as what he did not write but was neverthe- less able to articulate: that is, "the blank between stanzas 1 and 2 of the Lucy poem 'A slumber did my spirit seal . . . ' or between parts 1 and 2 of the Boy of Winander poem. "6 As it happens, what he articu- lates here is an example of a moment when "articulation is threatened by its undoing," when there is "a shift from a tropological to a differ- ent mode of language," as in the case of the "blank" between mathe- matical and dynamic sublimes, where "one could speak of a shift from trope to performance" (AI 89). Given that Wordsworth, of all poets, is able to do this, to do what Kant, or at least Kant's (formal materialist) text, does, it would be worse than premature to relegate him to merely aestheticist status as though he were only another aesthetic ideologist, only another Schiller. 7 It would be more helpful perhaps to recall that de Man's insistence that what the poets do in Kant is not what the poet Wordsworth does is very much like his equally stark declaration in "Anthropomorphism and Trope in the Lyric" that whatever Baudelaire's "Correspondances" may be, "it is, emphatically, not a lyric" but rather something of "an infra-text, a hypogram" underneath lyrics like "Obsession" (or "odes," "idylls," or "elegies") or, for that matter, pseudohistorical period terms such as romanticism or classicism, which are "always terms of resistance and nostalgia, at the furthest re- move from the materiality of actual history" (RR 262). If Wordsworth can be both unlike "the poets" of Kant--in seeing the sky as a shelter- ing sky and nature in terms of phenomenal figures that enter into a tropological system of exchange with the mind or the Imagination and that can be anthropomorphized and addressed--and yet also like them, in being able to articulate, if not to say, the moment of disrup- tion, "the material disarticulation not only of nature but of the body" and thus "the undoing of the aesthetic as a valid category," then "Wordsworth" is very much also "like" the Baudelaire, or one could better say, the Baudelaires of de Man's "Anthropomorphism and Trope in the Lyric. " As the "author"--or rather the signatory--of both the "lyric" "Obsession" and the emphatic nonlyric "Correspondances" that is legible like an infratext or a hypogram "underneath" it, Baudelaire clearly both does and does not do what the poets are sup- posed to do in de Man's account of Kant's material vision. And he does and does not do it because he writes two texts: the lyric "Obsession" and the emphatic nonlyric "Correspondances. " By writing the latter,
Baudelaire writes a text of "true mourning," as de Man puts it at the end of "Anthropomorphism and Trope," that allows for noncompre- hension and enumerates "non-anthropomorphic, non-elegiac, non- celebratory, non-lyrical, non-poetic, that is to say, prosaic, or, better, historical modes of language power" (RR 262). In doing so, "Corre- spondances" constructs something like that architectonic world of Kant's material vision--in this case, not so much a building that is not for dwelling and does not shelter as a temple in which no sacrifice that could transport us from the world of the senses to the world of the spirit takes place. Yet by writing the second text, "Obsession," Baudelaire also writes a lyric of recollection and elegiac mourning that adds remembrance to the flat surface of time in "Correspon- dances" and that engages the full panoply of lyric tropes and devices-- anthropomorphism, apostrophe, exclamation, a je-tu structure, specu- lar symmetry along an axis of assertion and negation, and so on--to result in "the reconciliation of knowledge with phenomenal, aesthetic experience" (RR 258), which, historicized, issues in "the aesthetic ide- ologization of linguistic structures" (RR 253). In writing both texts, Baudelaire is indeed like Wordsworth the phenomenalizing "roman- tic" poet and like Wordsworth the formal materialist who would be as nonlyrical and nonpoetic as those most prosaic poets of Kant. (So: the more "poetic" Wordsworth and Baudelaire, the less they are like "the poets" of Kant; the more "prosaic," the more material and historical. )
But, of course, we should not take the doubleness of the two here-- two texts, two Baudelaires, two Wordsworths--too literally, as though these Wordsworths and Baudelaires were Schillerian aesthetic ideolo- gists in some "poetic" poems and Kantian formal materialists in some other, rather "prosaic," poems. No, insists de Man, "whenever we en- counter a text such as 'Obsession'--that is, whenever we read--there always is an infra-text, a hypogram like 'Correspondances' under- neath" (RR 262). In other words, again, "There always are at least two texts, regardless of whether they are actually written out or not; the relationship between the two sonnets, obligingly provided by Baudelaire for the benefit, no doubt, of future teachers invited to speak on the nature of the lyric, is an inherent characteristic of any text" (RR 260-61). This is certainly borne out by de Man's reading of "Correspondances"--a text that turns out to be as thoroughly double and duplicitous as the double register of the articulating (and disarticu- lating) word comme in its function as both a term of comparison and metaphorical transport based on substances and their properties and a
"As the Poets Do It" 7
8 Andrzej Warminski
more metonymical syntactical marker of aimless enumeration--as a "metaphor aspiring to transcendental totality" gets stuck in "an enu- meration that never goes anywhere" (RR 250). In other words, the infratext or hypogram of "Correspondances" has already (and always again) produced the lyric "Obsession"--whether or not "Obsession" were ever actually written out. And, one should quickly add, whether "Correspondances" were ever actually written out or not! Clearly enough, the "materiality" of the infratext, or the hypogram, or of what de Man calls the "prosaic materiality of the letter" or "material inscription" (or, for that matter, "the materiality of actual history"), is not accessible in phenomenal experience and what appears in empiri- cal space and time. Materiality--or the infratext or hypogram or the letter or the inscription or actual history or the prosaic language power of the poets--is not something we are going to put our finger on. It is also not something that we can give more than inadequate, provision- al, names to. Just as the "material" of "material vision" is "the only word that comes to mind," so "In the paraphernalia of literary termi- nology, there is no term available to tell us what 'Correspondances' might be" (RR 261), and the terms infratext and hypogram are clearly also makeshift stand-ins. All the same, this does not mean that de Man's "materiality"--however difficult and even enigmatic it may be--is as mysterious as all that. The various formulations of what it is not and what it is like, both in the Kant essays and in "Anthropo- morphism and Trope in the Lyric" (and in other essays of the 1980s), indicate where to look for it, or at least how to read it. And that it has indeed everything to do with reading should already be plenty clear. For what else is one going to do to understand the "disruption" or the "blank"--whether between stanzas or parts of Wordsworth poems or between the mathematical and dynamic sublimes or in the juxtaposi- tion of seeing according to what the Augenschein shows with an alle- gorical narrative of how the imagination sacrifices itself for reason-- except to try to read them? And how read these moments in Kant (or Wordsworth or Baudelaire or whatever) where "articulation is threat- ened by its undoing" except by making them intelligible "in linguistic terms," as de Man puts it, if at these moments we encounter passages "that could be identified as a shift from a tropological to a different mode of language"? The poets can help us here again--in this case, de Man's compact account of how we are (and, as always, are not) to understand the relation between the always two texts that there always are whenever we encounter a text, that is, whenever we read. Going
over this account should make it easier for us finally to go back to Kant's sublime and to read the poets and their purportedly material vision in the context of de Man's reading of the mathematical, the dynamic, and the--for lack of a better word--"material" sublimes.
As it happens, the relation between the two texts that there always are whenever there is text--between an intelligible lyric like "Obsession" and its infratext or hypogram like the forever unintelligible "Corre- spondances"--is far from simple. And the question of the order of their relation--its reversibility or irreversibility--is especially difficult, which is perhaps not surprising since it is the same question as that of the relation between critical and ideological discourse: in shorthand, like the paradigmatic relation between Kant and Schiller or, in this case, between "Correspondances" and "Obsession" in relation to one another and in relation to themselves (as, say, "Correspondances"/ "Obsession" and "Obsession"/"Correspondances"). How does it work? On the one hand, the relation is clear: whenever we encounter a text like "Obsession," there is always an infratext, a hypogram, like "Correspondances" underneath. The lyric "Obsession" and its entire tropological system of devices--that is nothing so much as the "defen- sive motion of understanding, the possibility of a future hermeneutics" (RR 261)--is a reading, what de Man calls here "a lyrical reading- motion" and "a lyrical reading" of "Correspondances. " "Obsession" would be the Schiller to "Correspondances"'s Kant. De Man spells out the one hand:
We all perfectly and quickly understand "Obsession," and better still the motion that takes us from the earlier to the later text. But no symmetrical reversal of this lyrical reading-motion is conceivable; if Baudelaire, as is eminently possible, were to have written, in empiri- cal time, "Correspondances" after "Obsession," this would change nothing. "Obsession" derives from "Correspondances" but the reverse is not the case. Neither does it account for it as its origin or cause. "Correspondances" implies and explains "Obsession" but "Obsession" leaves "Correspondances" as thoroughly incomprehensible as it always was. (RR 261)
Nevertheless, however irreversible this defensive motion of under- standing and its lyrical reading-motion, it would be an error and, in- deed, a similar phenomenalizing ideologization to understand this order and its irreversibility in phenomenal (spatial or temporal) terms:
"As the Poets Do It" 9
10 Andrzej Warminski
Whenever we encounter a text such as "Obsession"--that is, whenever we read? there always is an infra-text, a hypogram like "Correspon- dances" underneath. Stating this relationship, as we just did, in phe- nomenal, spatial terms or in phenomenal, temporal terms--"Obsession," a text of recollection and elegiac mourning, adds remembrance to the flat surface of time in "Correspondances"--produces at once a herme- neutic, fallacious lyrical reading of the unintelligible. The power that takes one from one text to the other is not just a power of displacement, be it understood as recollection or interiorization or any other "trans- port," but the sheer blind violence that Nietzsche, concerned with the same enigma, domesticated by calling it, metaphorically, an army of tropes. (RR 262)
As far as the materiality of the actual history, that is, whatever it is that happens "between" "Correspondances" and "Obsession," is con- cerned, the spatial or temporal phenomenality of which text is "under- neath" which and which text comes after which does not matter and changes nothing, that is, does not happen--and understandably enough at that, for, as I said, it also does not matter whether the two texts were ever actually written out or not! Indeed, even if the "lyrical" reading-motion can go only from "Correspondances" to "Obsession," it is also the case that a reading-motion like de Man's of "Corre- spondances" goes from the all-too-poetic lyric of historicizing literary history that declares, performs (in its synaesthesia), and values sheer aesthetic ideology to an infratext underneath that threatens to dis- articulate the poem's transcendentalizing tropes and end up in "the stutter, the pie? tinement of aimless enumeration" (RR 254). In other words, de Man's own (material? what shall we call it? ) "reading- motion" goes from trope to another mode of language and thus, in a sense, from the lyric "Obsession" to the hypogram "Correspondances. " This does not mean, of course, that "Correspondances" and "Obses- sion" are in fact, materially, historically reversible. What is reversible is only the order of which precedes which and which follows which in the temporality of reading (whether lyric or otherwise), that is, in the temporality of an act of understanding and its inevitable temporaliza- tion in an allegory that narrates this act (which involves an inevitable phenomenalization--as de Man remarks in his own language when he says that the infratext or hypogram is "underneath" the lyric or that the lyric adds remembrance to the flat surface of time in "Correspon- dances"). What is not reversible, however, is the power "that takes one
from one text to the other" in these reading-motions, whether they go from the saturation and emptying out of tropes as the text moves from a tropological to another mode of language--from trope to perfor- mance, say--or from the material inscription of the hypogram in a defensive lyrical reading-motion to phenomenalizing aesthetic ideolo- gizations of a celebratory or elegiac, apostrophizing and anthropomor- phizing, poetic lyric. Both are inevitable, irreversible, what happens. What happens is the power that, as de Man puts it, "takes one from one text to the other"--whether there are empirically one or two or more or fewer texts, or whether they "exist," that is, were ever actual- ly written out, or not! --the sheer blind violence of the inaugural act that put the tropological system into place in the first place and that gets repeated whenever we necessarily and inevitably go from one text to the other--that is, whenever we read.
De Man's account of the always two texts of Baudelaire and of reading takes us back to his reading of Kant and helps us to under- stand, in particular, the itinerary, the order, of that reading--that is, the reading of the mathematical and dynamic sublimes and their issuing in the "material sublime" of the poets. Needless to say, understanding this reading, its order, and how the "materiality" of "material vision" emerges from it depends a great deal on making it intelligible "in lin- guistic terms. " "Linguistic" because it turns out that all three moments of Kant's sublime--mathematical, dynamic, and, for lack of a word, material--are to be understood not as philosophical (transcendental or even metaphysical) principles but as what de Man calls a "linguistic principle. " In order: the mathematical sublime becomes intelligible-- all too intelligible (like Baudelaire's "Obsession")--and can "work," but to a formal extent only, as a linguistic principle. The "linguistic model" of this principle is that of discourse as a tropological system-- a very familiar metaphorico-metonymical system of subsitution and exchange on the axes of selection and combination, paradigmatic and syntagmatic. In brief, this system would articulate "the infinity of num- ber" with "the totality of extension"--which is the burden of "proving" the mathematical sublime--in terms of two acts of the imagination: apprehension and comprehension, Auffassung and Zusammenfassung. "Apprehension proceeds successively, as a syntagmatic, consecutive motion along an axis, and it can proceed ad infinitum without difficulty. Comprehension, however, which is a paradigmatic totalization of the apprehended trajectory, grows increasingly difficult as the space cov- ered by apprehension grows larger" (AI 77). This amounts to a system
"As the Poets Do It" 11
12 Andrzej Warminski
of exchange and substitution: "As the paradigmatic simultaneity sub- stitutes for the syntagmatic succession, an economy of loss and gain is put in place which functions with predictable efficacy" but, adds de Man, "only within certain well-defined limits" (AI 77). The limits are clear. Although the power of number can indeed progress to infinity on the level of apprehension--that is, logically, in terms of numerical concepts--the imagination which is to totalize this infinity in one com- prehension soon reaches a point at which it is saturated and can no longer make additional apprehensions: "it cannot progress beyond a certain magnitude which marks the limit of the imagination. " It is at this privileged point which "avoids both excessive comprehension and excessive apprehension" that the imagination makes its stand, as it were, and takes as a trope an impossible trope that is in fact not a metaphor but a catachresis, of a totalized, bordered-off infinity, as though it could comprehend it in one intuition. (Kant's example of Savary's account of one's experience of pyramids is well known. ) What this means is that the mathematical sublime, as such a tropological sys- tem of substitution, is in fact not a judgment of the "absolutely large" but rather a somewhat subrepetitious displacement, transposition (Kant's German in fact says versetzen here), and substitution of the "almost too large" that is not yet "the too large"--in Kant's terms, of the "colossal" that is not yet the "monstrous"--for the "absolutely large. "8 It would be an impossible phenomenal trope of infinity, of that which is, by definition, not susceptible to being exhibited (dargestellt) in one sensory intuition. (In terms of de Man's reading of the zero in Pascal, this would be once again the substitution of one as a trope of the zero, in that case a substitution of number as trope for that which marks the limit of number, that is the beyond-number, the zero as pure sign. )9 It is right for de Man to say that this certain magnitude "marks" the limit of the imagination, for what is going on here is in- deed the phenomenalization of a mere marker of infinity (like, say, a zero) in a perceptible, imaginable, conceivable trope (like, say, a one). If the articulation of number and extension seems to take place, it does so as a tropological system of substitutions that are impossible except in terms of such a purely formal system. De Man summarizes:
The desired articulation of the sublime takes place, with suitable reser- vations and restrictions, within such a purely formal system. It follows, however, that it is conceivable only within the limits of such a system, that is, as pure discourse rather than as a faculty of the mind. When the
sublime is translated back, so to speak, from language into cognition, from formal description into philosophical argument, it loses all inher- ent coherence and dissolves in the aporias of intellectual and sensory appearance. It is also established that, even within the confines of language, the sublime can occur only as a single and particular point of view, a privileged place that avoids both excessive comprehension and excessive apprehension, and that this place is only formally, and not transcendentally, determined. The sublime cannot be grounded as a philosophical (transcendental or metaphysical) principle, but only as a linguistic principle. Consequently, the section on the mathematical sub- lime cannot be closed off in a satisfactory manner and another chapter on the dynamics of the sublime is needed. (AI 78)
So: if the mathematical sublime is "possible" only within the confines of such a purely formal tropological system, it is no wonder that the epistemological and the eudaemonic proofs of the mathematical sublime--that de Man treats before his discussion of Auffassung and Zusammenfassung and Kant treats after--end up in the assertions of the possibility of the sublime by dint of its impossibility and failure: "The sublime cannot be defined as the failure of the sublime, for this failure deprives it of its identifying principle" (AI 75). The "section on the mathematical sublime cannot be closed off in a satisfactory man- ner" because its (linguistic) principle of discourse as a tropological sys- tem cannot itself be closed off. For what happens is this: in its purely positional transposition of number into extension, of inscribed mark- ers into phenomenal tropes, of catachreses into impossible metaphors, the tropological system of the mathematical sublime introduces into itself an excess or a lack that cannot be mastered or controlled or accounted for by the resources--by the principles of substitution and combination--of that system and therefore prevents itself from ever being able to close itself off as a system. (This is an excess--of mark- ing, of substitutions other than trope, purely differential relations and entities; and a lack--of the one metaphor that could complete the tropological system and allow it to close itself off. )10 De Man's way of putting it is that "the transition from the mathematical to the dynamic sublime, a transition for which the justification is conspicuously lack- ing in the text, . . . marks [again, marks] the saturation of the tropo- logical field as language frees itself of its constraints and discovers within itself a power no longer dependent on the restrictions of cogni- tion" (AI 79). In other words, it is precisely the impossible tropes of
"As the Poets Do It" 13
14 Andrzej Warminski
the infinite--of that which is overdeterminately exterior to the tropo- logical system of Auffassung and Zusammenfassung--that prohibit the tropological system of the mathematical sublime to close itself off, that is, prevent it from being able to account for its own principles of substitution and exchange in terms of principles internal to its (tropo- logical) system; meaning that the one thing this tropological system cannot account for is its own production, the "principle" according to which it was put into place in the first place. Hence this system opens up radically, and empties out in the force, violence, and power of the dynamic sublime--which force, violence, and power in the end (as at the beginning) are only the repetition of the inaugural act that put the tropological system into place "in the first place. " According to de Man, this is "the only way to account for . . . the extension of the lin- guistic model beyond its definition as a system of tropes": "From the pseudocognition of tropes, language has to expand to the activity of performance, something of which language has been known to be ca- pable well before Austin reminded us of it" (AI 79). Hence the "lin- guistic model" of the dynamic sublime--where the mind overpowers the might of nature and discovers itself independent of nature--would be that of discourse as performative.
Although the passage, the transition--which is in fact a "break" and a "discontinuity" and hence not a transition at all--from mathe- matical to dynamic sublime, from cognition to act, from trope to per- formative, is called "irreversible" in de Man's sense (as he elaborates at the beginning of "Kant and Schiller"), there is no doubt that to the ex- tent that this passage is something that happens, an event, and thus truly (and, as we know, materially) historical, it is indeed also a "repe- tition," as I have already put it, a repetition "in the Kierkegaardian sense," as de Man might put it, of the inaugural act that put the tropo- logical system into place, again, "in the first place. " This is most vivid- ly legible right away at the outset of the discussion of number, of numerical concepts, as Kant writes that "the power of numbers pro- gresses to infinity"--die Macht [the same word that abruptly begins section 28 on the dynamic sublime: "Macht ist ein Vermo? gen, welches grossen Hindernissen u? berlegen ist (Power is an ability that is superior to great obstacles")] der Zahlen geht ins Unendliche. 11 If numbers have this power, then there was something of a "dynamic sublime" al- ways already (and always not yet) there in the mathematical sublime and its attempt to border off and exhibit this unimaginable and non- phenomenal power in one intuition (which it cannot do except in im-
possible, catachrestic tropes that are more markers than metaphors). And, of course, that "power" of number to progress to infinity is its entirely mechanical, automatic ability to "designate" the infinite by writing it, inscribing it, in an arbitrary differential mark. In short, the mathematical sublime too has at its "origin" a power that is itself put into place by an inaugural act of material inscription--minimally, the (aesthetic reflective) judgment that determines the magnitude of the measure by, say, dividing up the extension of a ruler into inches by marking and inscribing them; but, again of course, that the three "lin- guistic models" of the sublime--tropological, performative, and, call it, inscriptional--are intricated together and in a sense already "there" at the outset becomes legible only if de Man's (and, indeed, Kant's own) "reading-motion" and its narration in what can only be called an allegory (of reading and unreadability, yes) are allowed to unfold in order. It is telling that the order of de Man's presentation is not ex- actly, not quite, the same as Kant's. Indeed, there is something like a "logic of the sublime"--or, better, a sublime program, pro-gramma-- at work in de Man's own presentation as he first recounts the episte- mological and the eudaemonic (failed) proofs of the sublime, identifies them as "subreptions" in which a metaphysical principle mistakes it- self for a transcendental principle, and summarizes the difficulty by ref- erence to the passage on "thinking" (denken) the impossibility of an exhibition of ideas in section 29 of the Critique (i. e. , the section that contains the passage on material vision)--and all this before going back to the opening paragraphs of section 26 and the discussion of Auffassung and Zusammenfassung. No wonder that de Man's transi- tion reads a little oddly; the first sentence of the paragraph begins: "Still in the mathematical sublime, in section 26, next to the epistemolo- gy and the eudaemony of the sublime, appears another description" (AI 77; emphasis added), which sounds like "Meanwhile, back in the mathematical sublime. . . . " This is odd because Kant's description of Auffassung and Zusammenfassung and the tropological system they constitute is not "next to" but rather before the epistemology and the eudaemony of the sublime. De Man's getting to it only after he has dis- cussed them as well as denken and thus reordering Kant's presentation follows a certain "logic" of the sublime in that it provides a certain "privileged place" that itself allows an easier "comprehension" of his own reading-motion's difficult apprehensions and renders his reading of the mathematical sublime "intelligible in linguistic terms"; that is, de Man's passage on the tropological system itself serves as something
"As the Poets Do It" 15
16 Andrzej Warminski
like a "metaphor" of comprehension that makes what precedes and follows in his reading of Kant easier to understand. That the one figure of the double operation of Auffassung and Zusammenfassung de Man provides should be what he calls a "simple phenomenology of read- ing" is, as one says, no accident:
The model reminds one of a simple phenomenology of reading, in which one has to make constant syntheses to comprehend the successive unfolding of the text: the eye moves horizontally in succession whereas the mind has to combine vertically the cumulative understanding of what has been apprehended. The comprehension will soon reach a point at which it is saturated and will no longer be able to take in addi- tional apprehensions: it cannot progress beyond a certain magnitude which marks the limit of the imagination. (AI 77)
Once this "simple phenomenology of reading" is understood in lin- guistic terms as precisely a tropological system that cannot close itself off--that is, that can account for everything except its own principles of constitution and therefore cannot read itself--the phenomenology of reading turns into a veritable allegory of reading. And when this happens, the "eye" that cooperates with the "mind" so readily in the phenomenology of reading turns out to be completely dis-junct from any mind whatsoever and not unlike the eye of the savage or the poets who see only according to the pure optics of what the Augenschein shows or what only meets the eye.
I linger with the rhetorical structure of de Man's own essay only to indicate how deep the "deep, perhaps fatal, break or discontinuity" that de Man reads at the center of the third Critique runs. Its depend- ing "on a linguistic structure (language as a performative as well as a cognitive system) that is not itself accessible to the powers of transcen- dental philosophy" (AI 79) is just one such "break. " It recurs (in sec- tion 29) in the stark juxtaposition of the passage on material vision with a story of how the imagination sacrifices itself for the reason, and, indeed, has always already occurred (as recurrence) whenever articula- tion is threatened by its undoing. The break or discontinuity, the disrup- tion or disarticulation, gets repeated, happens, occurs--and is legible, in the order of reading, as "a shift from a tropological to a different mode of language"--whether it be the disarticulation of Kant's sub- lime (as an aesthetic reflexive judgment), or of aesthetic judgment as such, or of the category of the aesthetic (as philosophical category), or of the articulating project of the third Critique to serve as a "bridge"
between the supersensuous underlying nature and the supersensuous underlying freedom, or, ultimately, the disarticulation of the critical philosophy itself when it turns out that the transcendental discourse, and thus the critical subject itself, cannot ground itself transcendental- ly (which is the ultimate project of the mere "appendix" (Anhang) on the sublime, according to de Man). 12 In any event, all this is at stake in the sublime and in de Man's reading of the sublime as not a transcen- dental but rather a "linguistic" principle. And this means that what happens in this reading is not at all a "reduction" of Kant's analytic to "language" or "linguistic models. " For these models turn out to be not models at all, as each one fails to account either for itself or for its other--as cognition (and its tropological system) can never account for the act (and least of all for the act that put its tropological system of substitutions and exchanges of meaning into place in the first place), and the power of the act can never be strong enough to verify (i. e. , to make true) that it took place, happened, was in fact an event. The point is rather that the transcendental discourse needs to have recourse to (always defective) linguistic models precisely at the moment when it would claim to be able to ground itself transcendentally--and thereby complete and close off the critical philosophy--and that this self- grounding project therefore fails, and has to fail, like any and every attempt to define and determine "language" as a theoretical object of study. So: perhaps we are now in a better position to go back to what the poets do when they see only that which the Augenschein shows. Let's look again at what de Man calls "our question":
Our question, then, becomes whether and where this disruption, this disarticulation, becomes apparent in the text, at a moment when the aporia of the sublime is no longer stated, as was the case in the mathe- matical sublime and in the ensuing general definitions of the concept, as an explicit paradox, but as the apparently tranquil, because entirely un- reflected, juxtaposition of incompatibles. Such a moment occurs in the general remark or recapitulation (section 29) that concludes the analyt- ics of the sublime. (AI 79)
At first glance, what de Man has in mind by "such a moment" seems relatively straightforward: namely, the curious and unexpected passage on "material vision" that occurs in section 29. The "purely formal" and thus "purely material" vision of heaven and ocean would indeed be the "apparently tranquil, because entirely unreflected, juxta- position of incompatibles" insofar as it would be the tranquil vision
"As the Poets Do It" 17
18 Andrzej Warminski
"devoid of any reflexive or intellectual complication" in which "no mind" at all is involved. The judgment of the sublime here would be precisely nonreflective and nonaesthetic (or other than reflective and other than aesthetic). And it would be the juxtaposition of incom- patibles at least in the sense that the architectonic vision of nature as a building--the heavens as a vault and the ocean as bounded by the hori- zon as by the walls of a building--that is not for dwelling and that does not shelter would be the mere juxtaposition (and utter disjunc- tion) of nature and its purposiveness, as though the eye that sees only according to what the Augenschein shows were reading a figure or a trope (i. e. , nature as a building) completely emptied out of its meaning. "No mind is involved in the Kantian vision of ocean and heaven," de Man writes. "To the extent that any mind, that any judgment, inter- venes, it is in error--for it is not the case that heaven is a vault or that the horizon bounds the ocean like the walls of a building. That is how things are to the eye, in the redundancy of their appearance to the eye and not to the mind" (AI 82). But perhaps the "unreflected juxtaposi- tion of incompatibles" refers most directly not so much to the themat- ics of the passage on material vision as to the juxtaposition of the pas- sage itself with the allegorical tale of how the imagination sacrifices itself for reason--where we deal, says de Man, not "with mental cate- gories but with tropes" (AI 87). The diction of de Man's summary would suggest that this is "such a moment" in section 29: "What makes this intrusion of linguistic tropes particularly remarkable is that it occurs in close proximity, almost in juxtaposition to the passage on the material architectonics of vision, in the poetic evocation of heaven and ocean, with which it is entirely incompatible" (AI 87; em- phasis added). This would indeed be another version of the break or discontinuity, disruption or disarticulation, where there is "a shift from a tropological to a different mode of language. " Still, perhaps one should not hurry quite so much to accept de Man's characteriza- tion of this vision as purely formal, purely material, devoid of intellec- tual complication and semantic depth, and utterly nontropological. After all, as a number of commentators have pointed out, Kant's evo- cations of the heavens as a vault that encompasses everything (alles be- fasst) and borders off (begrenzt) the ocean, and the ocean as an abyss that threatens to swallow up everything (including, presumably, the sky) are clearly figures, tropes. Tropes, first of all, for the mathematical and dynamic sublimes, respectively, with the bordered-off infinitude of the starry sky an apt figure for the mathematical sublime and the over-
powering natural force of the turbulent ocean (that needs to be over- powered in turn by the power of the mind) an appropriate figure of the dynamic sublime. And the passage's proliferating tropology does not stop there. As more than one commentator has also pointed out, the sky as a bow- or arch-shaped "vault" (Gewo? lbe, from wo? lben) is a kind of bridge, in this case a bridge over an abyss figured by the ocean, and thus a strangely allegorical sign for the project of the critical phi- losophy and its dominant architectonic figures: the immense gulf be- tween the domains of the concept of nature and the concept of free- dom that is to be, that must be, bridged and articulated so that the latter can have, as it should, an influence on the former, and that there be a "ground of unity" (Grund der Einheit)--and not an abyss--for the supersensible that underlies nature and the supersensible that underlies freedom. 13
However neat this tropology, it does leave out the ocean when it is at rest and seen, according to what the Augenschein shows, as a clear water-mirror (als einen klaren Wasserspiegel). Between the all-framing starry sky and the all-engulfing abyss of ocean, there is the flat, placid, sheer surface of a mirror without depth. "The sea is called a mirror," writes de Man, "not because it is supposed to reflect anything, but to stress a flatness devoid of any suggestion of depth" (AI 83). This placid flatness does not fit so easily into the tropologies that can account for sky and sea as mathematical and dynamic sublimes or as the bridge of the third Critique over the abyss between the first and second Critiques. But it does indeed provide a nice figure for the mere juxtapo- sition of incompatibles--like the mathematical and the dynamic sub- lime or the understanding and reason, or first and second Critique's, and so on--the purely formal, purely material, vision of what the Augenschein shows, or, even better, the phlegmatic, a-pathetic vision of a calculating, counting Dutchman. In other words, legible here are de Man's three linguistic "models" of Kant's sublime, with the vaulted sky a figure of the mathematical sublime as tropological system (that would border off infinity), the abyssal ocean a figure of the dynamic sublime as performative force, and the clear water-mirror a figure of the "material sublime" whose model would be that of language as ma- terial inscription. But, needless to say, this is all too figural, too tropo- logical; there is all too much purposiveness and too much mind in such a reading. Such a reading would not be how the poets do it. If we ask, in the spirit of de Man's reading, what is the equivalence on the level of language, in linguistic terms, of this placid, flat water-mirror as seen by
"As the Poets Do It" 19
20 Andrzej Warminski
the apathetic Dutchman--"described as a phlegmatized kind of German interested only in the dreariest of commercial and moneymaking activi- ties" (AI 85) in Kant's precritical (1764) "Observations on the Feeling of the Beautiful and the Sublime"14--we get some direction from de Man's own account of how "meaning-producing tropes are replaced by the fragmentation of sentences and propositions into discrete words, or the fragmentation of words into syllables or finally letters. " Where to find, how to read, such a "dismemberment of language" in Kant's text? Another hint from de Man helps: "But just try to translate one single somewhat complex sentence of Kant, or just consider what the efforts of entirely competent translators have produced, and you will soon notice how decisively determining the play of the letter and the syllable . . . is in this most unconspicuous of stylists" (AI 89; em- phasis added). And, indeed, if we go back one more time to the sen- tence on the poets and try to translate it, we find very quickly that it does not in fact say what we and all the translators I have--Bernard, Pluhar, Philonenko--want to see there. For the sentence does not say "we must be able to view the ocean as poets do . . .
and yet find it sub- lime" (Pluhar), nor does it say, "To call the ocean sublime we must re- gard it as poets do" (Bernard), nor does it say, "il faut parvenir a` voir l'oce? an seulement, comme le font les poe`tes, selon le spectacle qu'il donne a` l'oeil, soit, lorsqu'il est contemple? au repos tel un clair miroir d'eau qui n'est limite? que par le ciel et, lorsqu'il est agite? , soit comme un abi^me menac? ant de tout engloutir, qu'il nous est quand me^me pos- sible de trouver sublime" (Philonenko). 15 Without exception, the trans- lators want to link what we must do to seeing--we must see as the poets do--and invariably relegate our nevertheless being able to find the ocean sublime to secondary, subordinate status by supplying a link- ing or a transitional word: Pluhar an "and," Bernard a "to" (in the sense of "in order to"), and Philonenko the relative pronoun que. In doing so, the translations link what is in fact not bridged in the German--that is, must and seeing according to what the Augenschein shows--and conversely dis-join (by means of their supplementary link- ing words) what in fact is linked in the German: namely, must and nevertheless be able to find sublime. Stripped of the subordinate clauses and phrases, the sentence actually reads as follows: "rather, one must nevertheless be able to find sublime [that is, find the ocean sublime] only, as the poets do it, according to what meets the eye, for instance. " In short, one, we, must not see (as the poets do it, etc. etc. ) but rather must be able to find sublime. The link between what we must do--that
is, be able to find sublime--and seeing only according to the Augen- schein may indeed be there, as it were "understood," in the sentence, but, if so, it is there only in subordinated, mediated form. Indeed, the sentence never even says that we must do what we must do as the poets see it but rather as the poets do it, that is, only according to what the Augenschein shows (and not what they or we see). The only actual, explicit seeing in the passage is in the sub-subordinate phrase "for ex- ample, when it [the ocean] is regarded at rest" (etwa, wenn er in Ruhe betrachtet wird)! The shift and, indeed, slippage from "must be able to find sublime" to "must see"--and its concomitant relegation of "be able to find sublime," grammatically the main verb of the sentence, to a mere adjunct, a mere appendix--may appear slight. After all, isn't this what the passage means, and aren't the translators just helping Kant out a bit? Not quite and not just. For in linking seeing to the must, the translators are making things far too easy for us and helping out Kant by turning him into something of a Schiller! That is, they in- troduce the figures of the poets, of the Augenschein, and of the ocean precisely as figures, as phenomenalizing tropes that can make the diffi- cult task easier: that is, having, "must-ing," as it were, to nevertheless find sublime, having to have the "faculty," as it were, of judgments of the sublime. In doing so, the trans-lators, as is their job, carry over and throw up a bridge where there isn't one in the Kant. In the Kant, what we must do is to be able to find sublime despite, whatever, the Augenschein shows, and the bridge between our must and our being able to find sublime is indeed a purely formal, only prosthetic bridge. This would mean, or, better, only mark or inscribe, that what the poets do is not even so much to see according to the Augenschein as to read an inscription, dismembered sentences, words, syllables, letters--like the illegible letter (or all too interpretable hieroglyph? ) of the arching line of the sky on top of the straight or squiggly line of the ocean. Indeed, it would perhaps not be too perverse to suspend Kant's sen- tence in the middle and identify the antecedent of "it" in "as the poets do it" as neither seeing nor being able to find sublime but rather "must": one must (only) as the poets must (nevertheless be able to find sublime) as one must as the poets must. (I've tried out the German: "Man muss bloss, wie die Dichter es tun, mu? ssen"; "Man muss mu? ssen"; "One, we, must must. ") Which amounts to saying that what one must do to be able to find sublime is, above all, introduce, inter-ject, "the poets" between the moral imperative and the sublime judgment. The supplying of the poets, as in Dichter or dictare--the only word that
"As the Poets Do It" 21
22 Andrzej Warminski
comes to mind, as it were--would be the always necessary and always impossible grammatical, gramma-tical, bridge, the bottom line of the prosaic materiality of the letter.
Postscriptum: On the Super-performative
The Triumph of Life warns us that nothing, whether deed, word, thought, or text, ever happens in relation, positive or negative, to anything that precedes, follows, or exists elsewhere, but only as a random event whose power, like the power of death, is due to the randomness of its occurrence. It also warns us why and how these events then have to be reintegrated in a historical and aesthetic system of recuperation that re- peats itself regardless of the exposure of its fallacy. 16
As is legible in several places, Paul de Man's title for what turned out to be his last book was Aesthetics, Rhetoric, Ideology. How and why the book ultimately came to be called Aesthetic Ideology is a long and, at times, comical story. In the end, and as always, the matter was de- cided by a combination of contingency and necessity: the "random event" of de Man's death and the (quite legitimate) preferences of "marketing" at the University of Minnesota Press. The difference be- tween the two titles, however, does invite a question: what difference would it make? Would the (re)insertion of the word rhetoric between aesthetics and ideology make any difference at all? Would it not be, at worst, trivial, and would it not, at best, merely reconfirm the suspicion or assumption that de Man's notions of ideology and of the political never get beyond the analysis of purely linguistic phenomena and their reduction to rhetorical structures?
Even beginning an answer to this question--and explaining the dif- ference that rhetoric makes--is not a simple task, but it is always worth noting that de Man was certainly very aware of the question and in response always maintains that "one could approach the prob- lems of ideology and by extension the problems of politics only on the basis of critical-linguistic analysis," which has to be done in its own terms, and that such analysis is "truer" to Marx's own procedures (for example and exemplarily, in The German Ideology) than what gener- ally passes for "critique of ideology. "17 Rather than repeating or sum- marizing arguments made elsewhere,18 let us instead focus on just one moment of de Man's project and his "critical-linguistic" readings--the moment when and the sense in which something, an event, an occur-
rence, something happens, something occurs and, as an event, is genu- inely historical with a "materiality" all its own. As always in the case of de Man, it turns out that rhetoric, the rhetoric of tropes, tropologi- cal systems and their attempt and inability to close themselves off, is what makes all the difference.
After de Man's readings--after reading tout court--what always happens and is thus predictable and inevitable (like death's "random event" and its inevitable reintegration and recuperation? ) is some ver- sion of the question "What now? " or "What next? "--"Now that we know the text is unreadable, its meaning indeterminate if not unde- cidable, what do we do? How do we take the next step, the step be- yond merely linguistic analysis of merely linguistic phenomena, to what really matters, to political stands and political programs and po- litical power, to what really matters out there, beyond the confines of text and language, to us? " This is, of course, the wrong question. And it is wrong not only because it presumes to know ahead of time what "language" and "linguistic" mean, as though the reference of these words were stable and knowable above and beyond all other words-- as though, in short, the referent of "language" and "linguistic" could be phenomenalized, could appear, as an object of consciousness and its phenomeno-logic without the inevitable interference of the rhetorical dimension of "language," without its being turned into a trope. It is the wrong question above all because it is (always already) inscribed within the workings of reading and de Man's "critical-linguistic" analy- ses, for these are precisely analyses of how it is that something can, does, happen, how the "next step" actually occurs. But a word of pre- caution is necessary here: those who have read de Man (even a little) should not anticipate too much, for de Man's next step, what actually occurs in (and as) de Man is not the performative, it is not the perfor- mative speech act or the "performative rhetoric" that seems to be the issue of so many of de Man's readings (from Allegories of Reading on) and their reception and use in the work of others. It is true that a cor- rect enough but ultimately untrue or at least not "true enough" ac- count of the typical "de Manian" reading and what it does with the re- lation between knowledge and act, the cognitive and the performative dimensions of a text--that is, trope and performative--would run as follows: de Man's readings start out by first setting up, reconstructing, the text as trope, as a tropological system (of substitutions and trans- formations of meaning)--or, most directly put, by interpreting the text as to be understood on the basis of (and as) a tropological system that
"As the Poets Do It" 23
24 Andrzej Warminski
would be closed, in the sense that its intelligibility is grounded in some ultimately stable meaning, an ultimately stable hermeneutic horizon of meaning. (In such a setup, the rhetoric of tropes would be continuous with, homogeneous with, logic--the possibility of universal and hence extratextual [and hence extralinguistic] meaning. ) All this means is: de Man begins by interpreting the meaning of the text, figuring out what the text means and how its figural language works to produce that meaning (once one takes even a small step beyond sheer literal- mindedness). De Man's readings, in this account, proceed by, second, demonstrating how it is that the text as tropological system, as system of tropes, in fact cannot close itself off and remains "open. " The rea- son this happens, most directly and succinctly put, is that the tropo- logical system of the text (i. e. , that is the text) cannot close itself off (in a final stable meaning) because that system cannot account for its own production, that is, cannot account for the inaugural act that put it into place in the first place in its own terms, that is, according to prin- ciples internal to itself as system. Hence, third, the text makes a sort of jump--it stutters, as it were--into another textual and linguistic model, that of the performative, of text as act, a model that diverges from the text as trope, as cognitive rhetoric, indeed, disrupts the cogni- tive dimension of the text. The upshot being that the text issues in the performative and that the text as performative disrupts the text as cog- nitive, as trope.
This account is correct enough, and many of de Man's readings-- from the early 1970s to the early 1980s--would seem to authorize it. For instance, the end of the famous (or infamous) concluding essay of Allegories of Reading--"Excuses (Confessions)"--would certainly seem to fit: "the linguistic model cannot be reduced to a mere system of tropes," writes de Man, since its "(negative) cognitions fail to make the performative function of the discourse predictable" (AR 300)19 and thus we find that "we are restating the disjunction of the performative from the cognitive" (AR 299-300). Or, for another example, one could adduce de Man's reading of the Kantian sublime in "Phenomenality and Materiality in Kant": in the end, it turns out that the mathematical sublime is grounded not as a transcendental (or even a metaphysical) principle but rather as a "linguistic principle" whose model is that of a familiar metaphorico-metonymical tropological system which, because it (is purely formal and) cannot close itself off, issues in the dynamic sublime whose linguistic model is that of language as performative. Nevertheless, even a cursory look at what actually happens in de Man's
readings cannot help but notice that something else, something more difficult, is going on and that the account above is so partial and so se- lective as to constitute a misreading of de Man. Indeed, it is a misread- ing that leads to all kinds of predictable aberrations, in particular a certain inflation and overvaluation of the performative--as though one could go to the text as act directly, immediately, and while bypassing the moment in the reading when the text's tropological system gets reconstructed, in short, while bypassing the actual act of understand- ing the text, in other words, the text itself! In the case of de Man's reading of the Kantian sublime, for instance, the correct enough focus on the disjunction between trope and performative as the "linguistic principle" underneath the mathematical and dynamic sublimes over- looks one rather prominent fact: namely, de Man's reading of the mathematical and dynamic sublimes takes up only and exactly one- half of his essay! After a typographical break, the entire second half of "Phenomenality and Materiality in Kant" is devoted to an attempt to identify whether and where the disruption or disarticulation at the cen- ter of the third Critique--between cognitive and performative and thus, by extension, between pure and practical reason ultimately-- "becomes apparent in the text . . . as the apparently tranquil, because entirely unreflected, juxtaposition of incompatibles" (AI 79). And such a moment occurs, according to de Man, in the uncanny "material vision" of the sky and the ocean "as the poets do it," a vision utterly devoid of reflection, internality, or mind, a purely formal "vision" re- ducible to the formal mathematization or geometrization of pure optics. This means, in short, that the radical "formal materialism" of Kant's text and its strange "materiality"--a "materiality," Derrida writes, "without materialism and even perhaps without matter"--as an event, an occurrence, what happens, is very explicitly not to be identified with the performative or the performative dimension or "model" of the text. Rather, whatever it is that happens in, and as, "Kant" hap- pens at the point of the "transition" or the "intersection" of the disarti- culation of two divergent systems, two divergent models, cognitive and performative.
The same is true of "Excuses (Confessions)" and its complicated reading of Rousseau. The fact that Rousseau's Confessions is not pri- marily a confessional text (i. e. , the overcoming of guilt and shame in the name of truth and thus "an epistemological use of language" [AR 279]), but also and rather a text of excuse (and thus "a complex in- stance of what [Austin] termed performative utterances" [281-82]),
"As the Poets Do It" 25
26 Andrzej Warminski
does not disrupt the text's intelligibility because both knowledge and action, cognitive and performative, are incorporated in "a general economy of human affectivity, in a theory of desire, repression, and self-analyzing discourse in which excuse and knowledge converge" (AR 287). Or, as de Man underlines, "Knowledge, morality, posses- sion, exposure, affectivity (shame as the synthesis of pleasure and pain), and the performative excuse are all ultimately part of one system that is epistemologically as well as ethically grounded and therefore avail- able as meaning, in the mode of understanding" (AR 287; emphasis added). In short, rather than interfering with or disrupting the figural logic of the text, the "performative excuse" confirms it and is in fact part of it. But what does disrupt this system because it is outside of, foreign and heterogeneous to, the system of intelligibility and under- standing is the radicalization of the excuse that takes place in Rousseau's utterly random, contingent, utterance of the name "Marion"--an ana- coluthon that "stands entirely out of the system of truth, virtue, and understanding (or of deceit, evil, and error) that gives meaning to the passage" (AR 289). It's this "foreign element," continues de Man, "that disrupts the meaning, the readability of the apologetic discourse, and reopens what the excuse seemed to have closed off" (AR 289-90). If this truly disruptive random utterance of the name "Marion" is still to be taken as an "excuse," then it would have to be an "excuse" in a way radically different from "the performative excuse" that was, ac- cording to de Man, still within the system of causes and effects, desires and repressions, hiding and revealing, and so on. And, in any case, it would not be its "performativity" that makes it foreign, radically exte- rior to and disruptive of the system of understanding. Or, if one still wants to speak of "performative" at all in relation to the random ut- terance "Marion," then one would have to think of it as something of a "super-performative"--that is, not one that functions within an es- tablished juridico-political system (within which it can come off or not), but rather one that itself is the inaugural act of positing that puts such a system into place in the first place. In any case, what disrupts the figural chain and the text as system of tropes is not the performa- tive dimension, not language as act, but rather the (impossible and yet necessary) moment of radical excuse, radical "fiction" (as de Man will call it after reading the Fourth Re^verie), at which two "systems" hetero- geneous to one another--like meaning and grammar--"intersect. " It's at the point of the intersection that the text as system of intelligibility and understanding gets disrupted. But, once this "textual event" hap-
pens, occurs, it inevitably gets disseminated throughout the text, all along the narrative line, and turns into a permanent parabasis that de Man, following Friedrich Schlegel, calls irony--"the systematic un- doing, in other words, of understanding" (AR 301). In other words, a certain radicalization of the disjunction or divergence between cog- nitive and performative, trope and performative, takes place in the course of de Man's reading--which suggests that already in the case of "the performative excuse" that would be continuous with and part of the system of intelligibility, there was (always already) a trace of the radicalized "performative," the pure positing power of language whose position--as in the case of the random utterance "Marion"-- as an "excuse" is radically disjunct from, has nothing to do with, the "excuse" as linked to the affective feeling of shame and the under- standing it makes possible.
That what happens is not the performative is very explicitly and di- rectly corroborated by de Man's remarks at the beginning of his spo- ken lecture "Kant and Schiller. " Using his Kant reading to articulate what he means by history as event, as occurrence, as what happens, de Man says that the model for such "historicity a priori" is
not the performative in itself . . . but the transition, the passage from a conception of language as a system, perhaps a closed system, of tropes, that totalizes itself as a series of transformations which can be reduced to tropological systems, and then the fact that you pass from that concep- tion of language to another conception of language in which language is no longer cognitive but in which language is performative. (AI 132)
And this is important enough for him to repeat it and insist on it: "and I insist on the necessity of this, so the model is not the performative, the model is the passage from trope to performative--this passage occurs always, and can only occur, by ways of an epistemological critique of trope" (AI 133). In other words, there is no passage, no occurrence, no event, no history--nothing happens--except as (or "by ways of") an epistemological critique of trope. What happens--if it happens--does so thanks to the (self-)critical power of the text as tropological system that would want to account for its own production (the only thing worth knowing, as de Man says at the end of "Excuses")20 in terms in- ternal to its system. Because the text cannot do this, cannot account for its own production, for the inaugural instituting act that put it into place, what happens instead is the "passage" to the performative, to lan- guage not as cognition but as act. In this emergence of a language of
"As the Poets Do It" 27
28 Andrzej Warminski
power out of a language of cognition, what emerges is in fact the very "origin" of the text, the material trace or the material inscription that would be the condition of possibility and the condition of impossibility of the text "itself. " In Kant's "Analytic of the Sublime" the attempt to ground the critical discourse, to found the very subject of the critical philosophy and transcendental method, instead un-grounds, un-founds, itself in the disarticulation of tropological and performative linguistic models by, ultimately, the "last" linguistic "model": the prosaic materi- ality of the letter, material inscription. In Rousseau's autobiographical project, the attempt to ground the confessional/apologetic discourse, to found the confessional subject, instead disarticulates itself and founders on the random utterance "Marion"--which, of course, is the material trace at the very "origin" of Rousseau's autobiography, the reason, as he says explicitly, for his writing the Confessions in the first place (i. e. , to confess the shameful act). 21 Among other things, such an account helps to put the performative into better perspective. For what happens when the text "passes" from trope to performative--which is not a tem- poral progression but an event, an occurrence (as in "comes to pass")-- is a certain "repetition" of the violent, groundless and ungrounded, in- augural act that, again, put it into place in the first place. The event of this repetition is what gets disseminated all along the narrative line and thus renders the text an allegory of its inability to account for its own production (an allegory of unreadability, to coin a phrase)--with Rousseau's autobiographer doomed to mindlessly, mechanically, repeat- ing "Marion" over and over again, and Kant's critical philosopher "I must be able to bridge pure reason and practical reason," "I must ex- hibit the ideas of reason," "I must be able to find sublime," "I must must," "Ich muss mu? ssen, muss mu? ssen, muss mu? ssen . . . "22
So: that's the difference the reinsertion of rhetoric between aesthetics and ideology makes. Without rhetoric, without the epistemological cri- tique of trope, as de Man puts it, nothing happens. There is no direct, immediate, royal road to the performative, to action and the act, po- litical or otherwise. Pretending that one can go to it directly is sheer delusion and a guarantee that nothing can happen, nothing will ever happen. 23
NOTES
1. It may be helpful to provide the passage from section 29 of Kant's third Critique that de Man reads in the second half of his "Phenomenality and Materiality in Kant. " In Werner Pluhar's uncorrected (a point that is dicussed toward the end
of my essay) translation, it reads: "Therefore, when we call the sight of the starry sky sublime, we must not base our judgment upon any concepts of worlds that are inhabited by rational beings, and then [conceive of] the bright dots that we see oc- cupying the space above us as being these worlds' suns, moved in orbits prescribed for them with great purposiveness; but we must base our judgment regarding it merely on how we see it, as a vast vault encompassing everything, and merely under this presentation may we posit the sublimity that a pure aesthetic judgment attrib- utes to this object. In the same way, when we judge the sight of the ocean we must not do so on the basis of how we think it, enriched with all sorts of knowledge which we possess (but which is not contained in the direct intuition), e. g. , as a vast realm of aquatic creatures, or as the great reservoir supplying the water for the va- pors that impregnate the air with clouds for the benefit of the land, or again as an element that, while separating continents from one another, yet makes possible the greatest communication among them; for all such judgments will be teleological. Instead we must be able to view the ocean as poets do, merely in terms of what manifests itself to the eye--e. g. , if we observe it while it is calm, as a clear mirror of water bounded only by the sky; or, if it is turbulent, as being like an abyss threaten- ing to engulf everything--and yet find it sublime" (Immanuel Kant, Critique of Judgment, trans. Werner S. Pluhar [Indianapolis/Cambridge: Hackett Publishing Company, 1987], 130).
2. All references to Paul de Man, Aesthetic Ideology, ed. Andrzej Warminski (Minneapolis: University of Minnesota Press, 1996), will be indicated by AI fol- lowed by the page number. All references to de Man's "Anthropomorphism and Trope in the Lyric," in The Rhetoric of Romanticism (New York: Columbia University Press, 1984), will be indicated by RR followed by the page number.
3. See de Man's brief but packed reading of Saussure's ana-(and para- and hypo-)grams in "Hypogram and Inscription," in The Resistance to Theory (Minneapolis: University of Minnesota Press, 1986), 36-38.
4. De Man quotes Kant on the "savage" in the preceding sentence: "In a lesser- known passage from the Logic Kant speaks of 'a wild man who, from a distance, sees a house of which he does not know the use. He certainly observes the same ob- ject as does another, who knows it to be definitely built and arranged to serve as a dwelling for human beings. Yet in formal terms this knowledge of the selfsame ob- ject differs in both cases. For the first it is mere intuition [blosse Anschauung], for the other both intuition and concept'" (AI 81).
5. This is no doubt an overdetermined misquotation. See de Man's many texts on (faces in) Wordsworth now in The Rhetoric of Romanticism and Romanti- cism and Contemporary Criticism, ed. E. S. Burt, Kevin Newmark, and Andrzej Warminski (Baltimore: Johns Hopkins University Press, 1993). See also my "Facing Language: Wordsworth's First Poetic Spirits," Diacritics 17:4 (winter 1987): 18-31; reprinted in Romantic Revolutions, ed. Kenneth R. Johnston, Gilbert Chaitin, Karen Hanson, and Herbert Marks (Bloomington: Indiana University Press, 1990), 26-49.
6. It is worth remembering that the "blank" between stanzas 1 and 2 of the Lucy poem and between parts 1 and 2 of "The Boy of Winander" marks the transi- tion from living Lucy and living Boy to dead Lucy and dead Boy. For de Man on "A Slumber Did My Spirit Heal," see his "The Rhetoric of Temporality," now in the
"As the Poets Do It" 29
30 Andrzej Warminski
second edition of Blindness and Insight (Minneapolis: University of Minnesota Press, 1983), 223-25. De Man's most extensive reading of "The Boy of Winander" is in his "Time and History in Wordsworth," in Romanticism and Contemporary Criticism, but see also the discussions in "Heaven and Earth in Wordsworth and Ho? lderlin" in the same volume and "Wordsworth and Ho? lderlin" in The Rhetoric of Romanticism.
7. For de Man on "Schillerizing" and "re-Kantizing," see "Kant and Schiller" in Aesthetic Ideology.
8. Although the reading of Kant's mathematical sublime in terms of such a subreptitious substitution--calling "sublime" what is in fact only "colossal"--is Derrida's (in "Le colossal," in La Ve? rite? en peinture [Paris: Flammarion, 1978], 136-68), de Man's own reading is very close to Derrida's here. That de Man had read Derrida's "Le colossal" is clear in the earlier "Kant's Materialism," also in Aesthetic Ideology.
9. On de Man's reading of Pascal's zero, see my Introduction, "Allegories of Reference," in Aesthetic Ideology, 1-33.
10. The locus classicus for understanding such "economies of the supplement" is, of course, Jacques Derrida, "La mythologie blanche," in Marges (Paris: Minuit, 1972). See also my reading of Derrida and catachresis as the "syntax of tropes" in "Prefatory Postscript: Interpretation and Reading," in Readings in Interpretation: Ho? lderlin, Hegel, Heidegger (Minneapolis: University of Minnesota Press, 1987), liv-lxi.
11. Immanuel Kant, Kritik der Urteilskraft (Frankfurt am Main: Suhrkamp, 1974), 184.
12. And not just according to de Man. There are remarkable similarities be- tween de Man's understanding of the stakes of Kant's "Analytic of the Sublime" and Jean-Franc? ois Lyotard's. Indeed, however different their terms, de Man's and Lyotard's readings coincide in many respects. See Jean-Franc? ois Lyotard, Lessons on the Analytic of the Sublime, trans. Elizabeth Rottenberg (Stanford, Calif. : Stanford University Press, 1994).
13. Although he does not explicitly read the vaulted sky as a figure for the bridge between the supersensuous underlying nature and the supersensuous under- lying freedom, Derrida does link the ocean in this passage to the abyss between them. See La Ve? rite? en peinture, 148.
14. But whose phlegmaticity is then judged sublime in the Critique of Judgment! For more on de Man and the Dutchman, see "Kant's Materialism" in Aesthetic Ideology, 124-25. It is worth noting that de Man's joke in "Kant's Materialism" about Kant's characterization of the Dutch in the precritical text--"I have never felt more grateful for the hundred or so kilometers that separate Antwerp from Rotterdam" (AI 125)--undergoes a slight arithmetical transposition in the later "Phenomenality and Materiality in Kant": "I have never felt more grateful for the fifty or so kilometers that separate the Flemish city of Antwerp from the Dutch city of Rotterdam" (AI 85; emphasis added). The Dutch--those "phlegmatized Germans"--seem to have moved closer to Antwerp by the time of the later essay!
15. J. H. Bernard's and Alexis Philonenko's translations are: Critique of Judg- ment (New York: Hafner Press, 1951), 110-11; Critique de la faculte? de juger (Paris: Vrin, 1984), 107.
16. Paul de Man, "Shelley Disfigured," in RR 122.
17. See Stefano Rosso's interview with de Man in Paul de Man, The Resistance to Theory, 121.
18. See my "Introduction: Allegories of Reference" to Aesthetic Ideology; "Ending Up/Taking Back (With Two Postscripts on Paul de Man's Historical Materialism," in Critical Encounters, ed. Cathy Caruth and Deborah Esch (New Brunswick, N. J. : Rutgers University Press, 1995), 11-41; and "'As the Poets Do It': On the Material Sublime" in this volume.
19. AR with a page number refers to Allegories of Reading (New Haven: Yale University Press, 1979).
20.