TOWARDS A DELEUZIAN POETICS
Article written for an unpublished book edited by Frans-Willem Korsten and Vincent Meelberg.
Poetry is often understood to be “located” somewhere on a line running from language to music. Louis Zukofsky, the twentieth century American poet, even defined poetry as an integral with lower limit speech and upper limit music. According to this view, poetry is something like language aspiring towards the condition of music, to use Walter Pater’s phrase. Most radical in this respect is the sound poetry of the historical avant-garde and its project of reconstructing the so-called language of paradise, the prelapsarian Ur-Sprache. But long before Kurt Schwitters and Hugo Ball, philosophers as diverse as Giambattista Vico and Jean-Jacques Rousseau had claimed that the earliest peoples spoke poetry. “Verse, singing, and speech have a common origin,” Rousseau writes. “The first discourses were the first songs” (Rousseau 1986:50).
Jacques Derrida’s deconstruction of Rousseau’s “Essay on the Origin of Languages” is well-known. In fact, one doesn’t have to be as subtle a thinker as Derrida to spot the aporias and contradictions in this curious text. Interestingly enough, Gilles Deleuze and Félix Guattari, in A Thousand Plateaus, praise Rousseau’s essay. If the figure of Rousseau remains at best marginal in Deleuze’s philosophical canon, here we find his predecessor hailed as the theorist of a possible alternative to modern linguistics. Deleuze and Guattari appear to agree, or at least do not seem to disagree, with Rousseau as far as the “original” link between language and music is concerned:
We are not suggesting any correspondence [between music and language]. We keep asking that the issue be left open, that any presupposed distinction be rejected. This especially applies to the language-speech distinction, which is used to relegate all kinds of variables at work within expression and enunciation to a position outside language. The Voice-Music relation proposed by Jean-Jacques Rousseau, on the other hand, could have taken not only phonetics and prosody but all of linguistics in a different direction. (Deleuze and Guattari 1996:96)
What is asked is that any presupposed distinction be rejected. This paper is not concerned with the reasons Deleuze and Guattari they target the distinction between langue and parole, or even what Rousseau could have done for linguistics, which is obviously their main concern on this plateau, “November 20, 1923: Postulates of Linguistics.” I’m interested solely in the “Voice-Music relation,” because this relation – whether it is thought of as analogy, continuity, identity or none of the above – is so often used, implicitly or explicitly, in definitions of the poetic. Deleuze and Guattari do not talk about poetry on this plateau, of course, and indeed they hardly ever do, but I’m interested in the consequences of Deleuzian or Deleuzo-Guattarian thought for the theory of poetry. The guiding question of this paper will be: “What is this Voice-Music or language-music relation, and how does it relate to poetry or poetic texts?” In order to do this I will confront Deleuze and Guattari’s work with that of Rousseau and also Heidegger, and with three poems by Jack Spicer. These various comparisons and contrasts, these double or triple mediations, I feel, are necessary in order to discover the place – and I think there is a place – in Deleuzian thought, of poetry, as distinguished from literature in general.
Rousseau and Heidegger, of course, do talk about poetry in their writings on language. Rousseau equates poetry with song, claiming that the first discourses, the first laws even, were sung. Language, for Rousseau, is first of all an instrument for the communication of feelings and thoughts, of passions, rather than needs. He famously states that “One stalks in silence the prey on which one would feast. But for moving a young heart, or repelling an unjust aggressor, nature dictates accents, cries, lamentations” (Rousseau 1986: 12). Similarly, in humanity’s childhood, only poetry was spoken, that is to say, figurative and passionate language, not reasonable and methodological discourse. The latter is seen as the corruption of the original language, going hand in hand with despotism and tyranny. The best languages are “favorable to liberty,” as Rousseau puts it, because they are “sonorous, prosodic, harmonious” and hence facilitate mutual understanding and brotherhood (Ibid.: 72-3). One notices here a certain resemblance between Rousseau’s corrupted and tyrannical language and Deleuze and Guattari’s description of the so-called signifying regime as despotic. To Rousseau’s original poetry, then, would correspond perhaps pre- and countersignifying regimes, but also the territorial semiotics of animals, as described in the plateau “1837: Of the Refrain.” The most obvious analogy is between Rousseau’s emphasis on song and melody and Deleuze and Guattari’s theory of the ritournelle, the little tune or refrain that constitutes the act of territorialization. Of course, there are crucial differences, which I will come to later.
Much like Rousseau, Heidegger thinks the essence of language as poetry, and poetry as melodic language or song. “Poetry is the unconcealment of beings,” Heidegger stresses again and again. For him, poetry constitutes original saying (sagen), which is also a showing or pointing (zeigen): it points out and brings to presence. Heidegger says in “The Way to Language”:
The saying is the mode in which propriation speaks. Yet mode is meant here not so much in the sense of modus or “kind”; it is meant in the musical sense of the melos, the song that says by singing. For the saying that propriates brings what comes to presence out of its propriety to a kind of radiance; it lauds what comes to presence (Heidegger 1993: 424)
Adam is a pastoral poet, a sort of singing shepherd of being. And yet, we know from Heidegger how being predetermines its own forgetfulness; so here too, as in Rousseau, we come across something of a Fall motif: the original saying, by its very form, by its bringing to presence, is always in danger of becoming information, that is, being subjected to the formalizing logic of the technological enframing of the present, of the manipulable and controllable.
It is notable how both Rousseau and Heidegger represent a secular Fall, narrating it as a process of stratification and formalization. In Rousseau, the original melodic language that expresses passions degenerates into a civilized and economic language articulating needs: consonants are victorious over vowels, and language becomes more and more prosaic, more like writing and more formalized. Similarly, in Heidegger the original saying is lost, because “[e]nframing, the essence of modern technology . . . ordains for itself a formalized language—that kind of informing by virtue of which man is molded and adjusted into the technical-calculative creature, a process by which step-to-step he surrenders his ‘natural language’” (Ibid. 420-21). For both Rousseau and Heidegger, then, the Fall is a Fall from melodic innocence, from the poetic speech of the first men and the original saying of the poet. For Deleuze and Guattari, on the other hand, the refrain, insofar as it constitutes a territorial assemblage, is not innocent at all; the refrain is always territorializing – and territorialization is a Deleuzian name for the Fall. The state of innocence in Deleuze, if we are allowed to speak of such a thing, is of course the virtual, the plane of immanence, which in many passages in his work appears under a paradisiac aspect. The reader must forgive me for putting it so crudely. Obviously, for Deleuze, innocence is always already lost; because the plane of immanence does not exist outside of or before territorial assemblages. The given are the forms of the strata, and innocence is something to be achieved or constructed, rather than retrieved. Nevertheless, the plane of immanence is fully real if virtual; it is the Absolute where indeed any presupposed distinction is rejected, first of all between content and expression, but no doubt also between language and music. Despite the obvious differences, then, we see some striking resemblances between Rousseau, Heidegger and Deleuze and Guattari, as far as their thinking of the language-music relation is concerned. For all of these thinkers, there is an original language-music relation which is lost, following a lapsarian narrative. Now, what are the consequences for their respective conceptions of poetry? I think that if we can formulate a Rousseauist and a Heideggerian poetics with regard to the language-music relation, we can then begin to think of a properly Deleuzian one. This would be too speculative an enterprise, however, if we did not put the theories to the test from the very beginning by bringing in some actual poetry.
The following poem by Jack Spicer is as much about language and its musical qualities, as it about falling. It is called “Duet For A Chair And A Table”:
The sound of words as they fall away from our mouths
Is less important
And yet that chair
take their places
Almost as a kind of music.
Words make things name
Makes the table grumble
In the symphony of God am a table
Makes the chair sing
A little song about the people that will never be sitting on it
Who in the same music
Are almost as easily shifted as the furniture
Can learn our names from our mouths
Name our names
In the middle of the same music.
(Spicer 1999: 74-5)
The poem’s beginning is the beginning of a fall: “The sound of words as they fall away from our mouths.” The falling continues in the subsequent lines, iconically enacted by the positioning of some of the shorter lines. Lines 4-6 in their progressive indentation suggest that naming itself is a kind of falling: “And yet that chair / this table / named.” The falling motion continues in lines 8, 11 and 13, where it comes to a stop at “I”, indicating perhaps that the subject is also named, that even as it names itself, it is not the origin of language, of this falling motion, but rather its most curious effect. In any case, words are attributed to things: they fall on a table and a chair, almost imperceptibly, and as a result everything falls elegantly into place, “[a]lmost as a kind of music.” This fall, then, is both a codification and an organisation. We see language coding and identifying, forming and formalizing a world. That is to say, the poem describes an event in the Deleuzian sense, the event par excellence, i.e. the constitution of a territorial assemblage, in my terms: a Fall. The duet is not just between a table and a chair but between words and things, forms of expression and forms of content. Table and chair are incorporeally transformed by being named, they “assume identities / take their places” “in the symphony of God”; “Nothing / Is less important” than words, nothing less substantial, but through them, table and chair become expressive, meaningful; the furniture is rearranged in the house of being, so to speak.
I am deliberately mixing Deleuzian and Heideggerian terms, because I believe that in their conception of the event and its relation to language and territory, these two philosophers are very close. That territorialization is a linguistic affair we know, of course, first of all from Heidegger. Heidegger’s name for territorialization is wohnen, dwelling, which is always a poetic dwelling. For Heidegger, only the human being, Dasein, dwells; only Dasein is a discloser of a meaningful world; the animal is what he calls weltarm, poor in world, precisely because it lacks language, and thus has to sleep outside the house of being. Now, for Deleuze and Guattari, on the other hand, territory is not the prerogative of the human; indeed, the term territory itself has immediate zoological connotations. And yet, for them also, a territorial assemblage is distinguished from mere strata by a regime of signs: “Regimes of signs develop only in the alloplastic or anthropomorphic strata (including territorialized animals)” (Deleuze and Guattari 1996: 504). Territorialization, then, is anthropomorphic, the anthropos including the animal insofar as the latter is territorial, that is to say, has, above all, a semiotic capacity. Animals don’t have language obviously, but they do have their own semiotic systems, their own colors and sounds and gestures that they can make expressive in order to constitute their own little grove of being. Deleuzian territorialization in this respect seems a more inclusive form of Heideggerian dwelling. Moreover, as we have seen, the connection between territorialization and music that is made in Spicer’s poem, and which is so familiar to readers of the plateau on the refrain and the chapter on art in What Is Philosophy?, this is something we also find in Heidegger’s talk of melos or song.
If “Duet For A Table And A Chair” seems quintessentially a Heideggerian poem, it is because for Heidegger a poem is an index, an indexical sign in the Peircean sense, just as true saying (which is by definition poetic saying) is always a pointing, a pointing out or showing. Deleuze and Guattari call the index a territorial sign, and it is indeed impossible to think of a more territorial thinker than Heidegger. The Spicer poem perfectly illustrates this conception of language as pointing, as showing and making present. In After Lorca, his earliest book of poems, Spicer had already declared: “I would like to make poems out of real objects. . . I would like to point to the real, disclose it” (Spicer 1999: 33-4).
It is also possible, however, to see the icon rather than the index as the dominant sign of “Duet.” Above all, “Duet” illustrates territorialization. And the territorialization that it actually performs is completely in the service of the point being made. In other words, it is iconic reterritorialization, a reterritorialization on the face as the substance of expression. That the poem itself has a face we see when, as in lines 4-6, it moves to emphasize what is being said; it imitates the fall mentioned in the first line. Notice also how each line is almost always a semantic unit as well as a breath unit: the poem’s breathing is attuned to the poem’s saying. In this respect it is a very “human” or “humanistic” poem and somewhat sentimental. For the purpose of my argument, I will designate its implied poetic as Rousseauist. For Rousseau, after all, poetry is song and song the ideal expression of passion in language. The poem, the melodic song, is an imitative sign, an icon. “By imitating the inflections of the voice,” Rousseau writes, “melody expresses pity, cries of sorrow and joy, threats and groans. All the vocal signs of passion are within its domain” (Rousseau 1986: 57). In twentieth-century criticism, it is Gérard Genette who has most emphatically defended this position with regard to the poetic text. 
We now have two semiotic conceptions of poetry: the poem as index and the poem as icon; the former could be called “Heideggerian”, the latter “Rousseauist”. This leaves Peirce’s third type of sign: the symbol, which he defines as having an arbitrary relation to its referent. Deleuze and Guattari write that the symbol is a relatively deterritorialized sign, making it the paradigmatically linguistic sign. But how should we conceive of a symbolic poem? This type of poem would be read simply as conventional, communicative text, in the way one reads an e-mail or a newspaper for example, from which it does not differ in any special way. This is perhaps not a very “poetic” conception, but it is nonetheless something a poet can reflect on, as in the next Spicer poem I want to discuss. “A Red Wheelbarrow” is a ruthless parody on the well-known poem by William Carlos Williams with the same title: 
Rest and look at this goddamned wheelbarrow. Whatever
It is. Dogs and crocodiles, sunlamps. Not
For their significance.
For their significance. For being human
The signs escape you. You, who aren’t very bright
Are a signal for them. Not,
I mean, the dogs and crocodiles, sunlamps. Not
(Spicer 1999: 103)
The poem starts off with a command: “Rest and look.” This command is only implicit in Williams’ original, but no doubt it is there also, as it is in all such imagist or imagistic works, and indeed in all poetry whose content is preformed in order to be seen, either literally or mystically, i.e. epiphanically. This particular form of content I want to call “presentation” or the “it is”: things give themselves to be seen. In this very Heideggerian form, everything is a showing, a pointing out, and we are forced to be astonished about the appearance of even the most trivial thing, like a wheelbarrow or a pair of peasant shoes.
If “Duet” had strong Heideggerian, indexical elements (subsumed under the dominance of Rousseauist iconicity), this time Spicer’s poem is as anti-Heideggerian as it gets. The form of content is indeed the “it is”, but what is is “whatever. “Dogs and crocodiles, sunlamps”: these are certainly not the rain water and white chickens we know from Williams’ poem. They seem out of place and we have to assume that Spicer introduces a metaphor, whereas Williams stuck to “things themselves.” Yet as comparisons to the wheelbarrow neither dogs, crocodiles nor sunlamps, make much sense, let alone all three together. If a comparison is a form of expression, it is here deterritorialized by the deterritorializing content: the comparison is sent off on an indefinite line of vehicles (dogs, crocodiles, sunlamps, . . . whatever), a line which suggests that the wheelbarrow is like anything else, that it is precisely whatever it is, not as a concrete thing which organizes a concrete world around itself – the way Williams had presented it – but at best a meaningless cliché from recent literary history.
It is easy enough to interpret this poem as an instance of poetic or even linguistic pessimism. The poem is not much older than “Duet Between A Chair And A Table” and Spicer’s remark about wanting to point to the real and disclose it, yet the difference in attitude could not be greater. When this poem addresses us and says: “Rest and look,” rest your eyes, that’s exactly what we cannot do; if our eyes would rest, we would stop reading and the signs would escape us. The poem expresses a disbelief not only in language as an instrument of communication but as a regime of signification. The word “wheelbarrow” is just a word, meaningless. There is no event here, no consistent territorial assemblage. “You, the reader,” the poem addresses us, “You are not very bright,” or to put it in Heideggerese, “a pretty lousy world-discloser; your Lichtung or clearing doesn’t clear up anything at all. But your faint radiance is a signal, perhaps a distress signal that represents you in your impotence to other signs.” Already at the end of “Duet For A Chair And A Table” the subject, the “we” was presented as a shifter, a mere sign that is “almost as easily shifted as the furniture,” as the poem put it. But the lightness and ease of naming in “Duet”, which is summed up in the language-as-music simile and expressed in the smooth, natural flow of the prosody, are missing in “A Red Wheelbarrow.” The latter is aggressive, insulting, disjointed, absurd, and desperate. It is as such that this poem expresses the possibility of the poem as symbol, a symbol in a strange language in fact, an arbitrary and dysfunctional sign. And yet, again, this idea is iconically expressed; there are three lines that end with the word “not” for instance, which stresses the abstract negativity that is the poem’s theme. Again, the poem appears to have a face, not a happy one this time, but a desperate one, and again, the icon seems to be victorious: Rousseau wins, representation and imitation win, humanism wins. If the poem expresses great scepticism about language, its iconic form of expression seems ready to console us.
If we want to go beyond the sentimentality of Rousseau, without falling for the provincial poetics of Heidegger, and we still want there to be a specific, non-arbitrary (i.e. non-symbolic) function of poetry,  we could do worse than to turn to the fourth type of sign that Deleuze and Guattari have added to Peirce’s inventory: the diagram. If the index is a territorial sign, the icon a reterritorialized sign, and the symbol a relatively deterritorialized sign, then the diagram is the sign of absolute deterritorialization. A truly Deleuzian poetics would have to start with the primacy of the diagram. The essential importance of deterritorialization is precisely what is missing in both Rousseau and Heidegger. Indeed, for Rousseau and Heidegger it is the territory that is always already lost and which should then be retrieved; whereas for Deleuze and Guattari, the territory is what we are always already inhabiting, so that every becoming (the “innocence” of becoming, that Deleuze speaks so fondly of) can only be a move away from the territory.
Deleuze and Guattari’s privileged strategy of deterritorialization in A Thousand Plateaus is art, especially art music, which they define as the deterritorialization of the refrain. Music has the greatest powers of deterritorialization, they claim. A Spicer poem from the same collection as “Duet For A Chair And A Table” appears to agree. The poem is called “Improvisations On A Sentence By Poe” and somewhat of an ars poetica:
“Indefiniteness is an element of the true music.”
The grand concord of what
Does not stoop to definition. The seagull
Alone on the pier cawing its head off
Over no fish, no other seagull,
No ocean. As absolutely devoid of meaning
As a French horn.
It is not even an orchestra. Concord
Alone on a pier. The grand concord of what
Does not stoop to definition. No fish
No other seagull, no ocean—the true
(Spicer 1999: 69)
The ideal of music as the most deterritorialized art, as the most indefinite, is to a certain extent a restatement of the nineteenth century conception of absolute music, of music as pure form. For an aesthete like Walter Pater, music constituted the ideal art, because it seemed least burdened with meaning, with definiteness and definition. This is also what Poe meant by indefiniteness: indefiniteness of reference, meaning, or content. The lack of content in music, or at least in the best or “true” music, was seen to liberate the form for free play, essentially radicalizing the Kantian aesthetic. Yet we know from Deleuze and Guattari that there is no form/content opposition, but that content is itself formal. So the proper distinction should be between form of content and form of expression. Music as a deterritorializing form of expression takes the refrain as its deterritorialized content; it sweeps it up. This is obviously very different from an aestheticist position that privileges music because it is supposed to be the free play of form. Deterritorialization is not a play with form, it is the destruction of form, both of content and expression. This destruction is not nihilistic, because it entails the positive construction of an autonomous bloc of sensations which is in no way formal. The sign of this bloc is the diagram. Music, but also painting and literature, have the simple task of creating diagrams.
Because a diagram is obviously not a linguistic sign, in literature the diagram can only be created by language somehow becoming other than itself. This does not mean compensating for the arbitrary nature of the sign by iconic reterritorialization (a Roussauist conception), but by opening up a foreign language within language, making the major language stutter, and sending it on a line of flight. Deleuze’s examples are usually modernist novels; he certainly does not make a distinction within the literary, between poetry and prose for instance, or between the lyric and the epic. I do not think that Deleuze’s thought will ever make it possible to make an essential difference between poetry and non-poetry, and I do not think it necessary. I do believe, however, that Deleuze is quite a bit closer to the poetry-critical tradition than appears to be the case.
For how do we make language stutter? Deleuze says, in “Literature and Life”, that “a foreign language cannot be hollowed out in one language without language as a whole in turn being toppled or pushed to a limit, to an outside or reverse side that consists of Visions and Auditions that no longer belong to any language” (Deleuze 1997: 5). These Visions and Auditions, it is said, are the results of minorisation; the true process of minorisation is the creation of syntax, style; the Visions and Auditions then happen “in turn.” Yet, I want to argue, syntax is not enough. Or rather, what is the creation of syntax, of a style, but an operation on rhythm and imagery? That is to say, auditory and visionary qualities are an immanent part of every language, not just literary language; there is no language without rhythm, without sonorous patterning, and there is no language without figures. And yet, rhythm and figure are not language – they’re not in themselves signifying, they constitute immanent outsides of language. Literature, then, intensifies these outsides of language, pushes against them, turning them into Visions with a capital V and Auditions with a capital A. What Deleuze calls Visions and Auditions, then, are the two aspects that must be included in every definition of poetry: namely, that it is rhythmic or, more generally, “musical”; and that it uses imagery or, more generally, figurative language. Different poems and different theories of poetry have usually emphasized one or the other, but quite obviously both are essential. Perhaps there are other outsides than these two – the visible and the audible – but these seem the most important.
That a poem can actually never make anything visible – as “A Red Wheelbarrow” suggests –, that vision remains forever its outside, does not at all mean that this outside is not exactly what the poem is moving towards. “A Red Wheelbarrow” as an attempt to point, an attempt to show or see, is necessarily a failure; but is it not this failure that interests us? Is a poem not exactly a carefully programmed failure of language, a tiny catastrophe, or a machine of words that works by breaking down? This is ultimately what makes “A Red Wheelbarrow” interesting, I think, not the way its failure is iconically reterritorialized. Even the seemingly Heideggerian, territorializing theme of “Duet For A Chair And A Table” and its iconic expression, are they not, in the end, the clothes of rhythm, this deterritorialization of language that can only be conceived of as a “Practice of the Outside” (to use Robin Blaser’s characterization of Spicer’s poetic method)? Spicer believed in dictation, in an outside force writing his books and poems through him:
As I said, if a Martian comes into a room and sees a baby’s alphabet blocks, he’ll obviously use them to communicate. He won’t understand what they’re for or anything else. He’ll simply rearrange them into an order which very good sense Martian-wise, and doesn’t too much Earthman-wise, and he’ll just use them. (Spicer 1998: 131)
Spicer, then, conceived of the Outside as a radically foreign (alien) language intruding on the linguistic competence of the poet. I contend that the Outside of language is in fact nothing linguistic, and that the Martian is in fact a Vision or an Audition. Nevertheless, “Martian” is not an unfitting name for this poetics.
To conclude, I wish to focus on music as an outside of language in the poem “Improvisations On A Sentence By Poe”. Even if “improvisation” is above all a musical term, this does not mean that the ideal poem – again, I’m reading the text as an ars poetica – aspires towards the condition of music qua pure form, as in the aestheticist ideology. If the poem aspires towards the condition of music, it is not in the sense that it takes the form of a musical composition; not at all. Indeed, the poem makes its “point” by the use of an image: a bird alone on a pier, “cawing its head off / Over no fish, no other seagull, / No ocean.” In the middle of all this negation, the seagull itself and the pier are affirmed, but isolated and therefore “devoid of meaning.” In that respect they resemble the red wheelbarrow: if the latter was metaphorically compared to anything, the bird and pier are metonymically related to nothing; they’re not limited by any context and hence indefinite and undefined, not at all a concrete particular, not a territory, but an abstraction. The bird’s cawing marks a refrain that is not, or no longer territorial, a deterritorialized refrain that relates to no fish, no other seagull, no ocean. In that sense, the cawing is already musical; it does not represent or resemble the true music, it is not a metaphor, not an icon. Yet it is also not a synecdoche, an instance or index of the “true music.” In fact, it is a becoming-music without ever becoming actual music, without ever being music, in the sense of having the actual form of expression of music. Similarly, the poem improvises on a sentence by Poe, and performs rhythmic operations on it. Notice how there is no narrative, no argument; the only progress the poem makes is rhythmical, mostly established by the repetition of noun phrases: “an element,” “the true music,” “the grand concord,” “the seagull,” “the pier,” “no fish,” and so on. Despite all these nouns – not a few of which are modified by a definite article – nothing becomes much clearer. On the contrary, things become more indefinite, precisely because their juxtaposition is less grammatical or semantic than it is rhythmic. The poem takes a statement praising indefiniteness, but which is itself clearly a definition, and undefines it, sweeps it up towards music as an outside of language. In this sense, the poem is at the same time a radicalization, a criticism and a destruction of Poe’s definition, or of the form of the definition.
In any case, we can see now why Deleuze and Guattari wanted to leave the issue of the Voice-Music or language-music relation open. The question is left open (the French text says it is left en question) because the relation is exactly that: a question, a problem in the positive and productive sense that Deleuze attributes to that concept in Difference and Repetition (Deleuze 1994: 63-4; 140ff). There is no correspondence between music and language, nor is there unity; but their relation is questionable, problematic. Diagrams are problematic signs. A diagram is a question mark, not signifying indefiniteness, but the indefinite itself, the out of focus, the absolutely deterritorialized. And one way – maybe the primary way – to create a diagram, to create a problem, in poetry is to push its linguistic form of expression to the threshold of music. Again, this operation does not produce actual music; in that case, the problem would be solved; it would be reterritorialized, and language and music would relate to each other iconically or indexically, and there would be a correspondence or a continuity. Of course, to a certain extent, solving the problem is inevitable; events are stratified, novelty is formalized – this is the basic condition of the Fall. Things fall into place: poems fall into cannons, into interpretations and representations. Rousseau and Heidegger’s dream of poetry as primordial language, as song, is the ultimate dream of poetic territorializers everywhere: the words falling perfectly on the beat, so to speak, melos as the essentially unproblematic. But we know that a good poem must remain problematic. If it did not, what would interest us in it? A good poem is news that stays news, as Ezra Pound put it, not because it conveys some eternal wisdom by way of a timeless form of expression, but because it shatters form, and makes it possible for us to think the new, and to think ourselves anew.
Gilles Deleuze. Difference and Repetition. Translated by Paul Patton. New York: Columbia University Press, 1994.
——-. Essays Critical and Clinical. Translated by Daniel W. Smith and Michael A. Greco. Minneapolis: University of Minnesota Press, 1997.
Gilles Deleuze and Félix Guattari, A Thousand Plateaus. Translated by Brian Massumi. London: The Athlone Press, 1996.
——-. What Is Philosophy? Translated by Hugh Tomlinson and Graham Burchell III. New York: Columbia University Press, 1996.
Roman Jakobson. “Linguistics and Poetics.” In Language in Literature, 62-94. Eited by Krystyna Pomorska and Stephen Rudy. Cambridge, Massachusetts: Belknap, 1987.
Gérard Genette. “Poetic Language, Poetics of Language.” In Figures of Literary Discourse, 75-102. Translated by Alan Sheridan. New York: Columbia University Press, 1982.
Jean-Jacques Lecercle. Deleuze and Language. New York: Palgrave Macmillan, 2002.
Martin Heidegger. “The Way to Language.” In Basic Writings, 397-426. Edited by David Farrell Krell. San Francisco: HarperSanFrancisco, 1993.
Eugene W. Holland. Baudelaire and Schizoanalysis: The Sociopoetics of Schizoanalysis. Cambridge: Cambridge University Press, 1993.
Jean-Jacques Rousseau. “Essay on the Origin of Languages which treats of Melody and
Musical Imitation.” In Rousseau and Johann Gottfried Herder. On the Origin of Languages: Two Essays. Translated and with Afterwords by John H. Moran and Alexander Gode. Chicago and London: The University of Chicago Press, 1986.
Jack Spicer. The House that Jack Built: The Collected Lectures of Jack Spicer. Edited and with an Afterword by Peter Gizzi. Hanover: Wesleyan University Press, 1998.
——-. The Collected Books of Jack Spicer. Edited and with a commentary by Robin Blaser. Santa Rosa, California: Black Sparrow Press, 1999.
William Carlos Williams. The Collected Poems of William Carlos Williams, Volume I: 1909-1939. Edited by A. Walton Litz and Christopher MacGowan. New York: New Directions, 1986.
 More generally, the primacy of the icon and the principle of equivalence or resemblance, as the basis of the Roman Jakobson’s conception of the poetic function, has been highly influential in structuralist and structuralist inspired literary theory and criticism.
 In the original collection, Spring and All, Williams’s poem is in fact untitled, only numbered. It is poem XXII of the series, and the text reads: “so much depends / upon // a red wheel / barrow // glazed with rain / water // beside the white / chickens” (Williams 1991: 224).
 I am assuming we agree on the dubious merits of Rousseau’s and Heidegger’s respective (re)territorializing agendas. Of course, this is all from obvious, since even the most eminent writers on Deleuze and literature disagree with each other, and apparently with the argument that I set forth in this paper. Jean-Jacques Lecercle, for instance, seems to hold (Rousseauist) iconicity in high regard (Lecercle 2002: 242ff), and in his book on Baudelaire, Eugene Holland makes a case for the decoding powers of metonymy (Holland 1993: 30ff), whereas I think metonymicity is essentially an indexical, territorial (Heideggerian) coding.