back toJacket2
Ron Silliman, photo Jeff Hurwitz 1998

This piece is about 25 printed pages long.
It is copyright © William Watkin and Jacket magazine 2009. See our [»»] Copyright notice.
The Internet address of this page is
Photo: Ron Silliman, photo Jeff Hurwitz 1998

Back to the Ron Silliman feature Contents list

Ron Silliman Feature

William Watkin

Projective Recursion:

The Structure of Ron Silliman’s «Tjanting»

We can’t avoid structure
a      void      structure

 — (Charles Bernstein, “In Parts”)[1]


As the reader may be more than aware of, but also happy to be given cause to recall, Ron Silliman’s Tjanting (1981), is a book-length prose poem written over three years consisting of just over two hundred pages broken up into only nineteen paragraphs of significantly unequal lengths. The first paragraph, “Not this.” is a mere two words, as is the second, “What then?”,[2] however the last paragraph takes up well over seventy pages and has been estimated to consist of 42% of the book’s total length.[3] The reasons for such divisional asymmetry become apparent when one learns that Tjanting, like so many of Silliman’s works, is based on rule-governed procedures (RGPs) established in advance and designed to be generative of a text.[4] In this instance Silliman uses the Fibonacci number sequence to determine the number of sentences in each paragraph. The Fibonacci sequence is possibly now one of the most famous sequences of numbers, beginning as follows: 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, 987, 1597, 2584, 4181 and so on. We stop at 4181 because that’s where Silliman does. In case you are unfamiliar with this sequence and to save hours of cogitation, each number in the sequence is the sum of the previous two. Silliman’s explanation for this complex procedure is that he wanted a “content-dependant” or “content-centred” work,[5] which emulated the dialectical materialism of the class struggle, allowing him to show what class struggle looked like formally.


That aside, returning to our number sequence, the first sentence for Silliman constitutes the thesis, the second and all subsequently underlined numbers are the antitheses (although confusingly the first thesis begins with a negation).[6] Each paragraph that follows in this bifurcated double sequential model contains all the sentences from the previous paragraph on its side of the dialectical divide, plus an amount of additional sentences equal to the number in the paragraph that directly precedes it. In addition, the repeated sentences nearly always go through some sort of modification to alter their sense or reveal new links with the sentences that surround them within the paragraph—this is an example of what Silliman and others call torquing. As Silliman explains, the kernel of the book is two sentences “each of which expands into a series of paragraphs interspersed with the other”,[7] a particularly attractive schema for him as it begins with two ones, a structure “that not only permitted the parallel articulation of two sequences of paragraphs, but also determined that their development would be uneven, punning back to the general theory of class struggle”.[8] This then is the procedural basis of the poem’s unusual structure.


New Sentence


Due to this choice of procedurality the structural demands of the procedure mean the poem body expands at such a rate that it literally spirals out of ‘control’ if we take control to mean certain taken for granted qualities of structural cohesion. The final paragraph is so long that one’s conception of it as a structure of sense cohesion, which is what after all the paragraph is, becomes hopelessly compromised. Silliman most certainly intends this. Yet the poem is also constructed due to another procedure which cuts across the overall structure dictated by the Fibonacci sequence as asymmetrical dialectic. Composed along the lines of what Silliman calls the new sentence, wherein the assumed sequential development from one sentence to the next is not upheld, within the poem the localised relation between one sentence and the next becomes such a cause for concern that the lack of cohesion in the overall body is often forgotten as one worries over the lack of any clear local cohesion.


The new sentence was devised initially by Silliman to describe what he and his contemporaries were attempting in their prose poetry. The basic idea of new sentence prose, developed in the essay ‘The New Sentence’, is to replace the broken, measured line of the poem with the non-broken, unmeasured sentence as basic unit for composition. However, in replacing the line with the sentence, while a number of presuppositions about poetry are attacked, numerous other assumptions about prose, more powerful in their way, are taken on board. The main problem, as Silliman states it, is that of the integration of the sentences into a larger structure of semantic cohesion, in particular the paragraph and the book as a whole. He notes an automatised syllogistic movement in reading prose that works strenuously to relate proximate sentences in a linear progression to each other by subordinating them under a larger assumed unity. While in poetry too there is an urge for syllogistic cohesion, semantically incoherent elements are allowed if they are dictated by the laws of prosody and indeed one could state that the central essence of poetic structure is to confound syllogism by virtue of semiotics and recurrence.


New sentence prose attempts to remove the “syllogistic leap” as Silliman calls it, by placing sentences together which do not cohere into a unified narrative or argument at the paragraph level. As fellow language poet Bob Perelmen notes:


A new sentence is more or less ordinary in itself, but gains its effect by being placed next to another sentence to which it has tangential relevance: new sentences are not subordinated to a larger narrative frame nor are they thrown together at random. Parataxis is crucial: the autonomous meaning of a sentence is heightened, questioned, changed by the degree of separation or connection that the reader perceives with regard to the surrounding sentences.[9]


It is clear from this that ‘new sentence’ is something of a misnomer. More fitting would be ‘new paragraph’ or, because ostensibly what Silliman is attacking is the means by which sentences integrate through what linguistics call cohesion into a semantic unit defined by the quality of texture, ‘new text.’ That said, as Silliman explains in reading the disjunctive sentences of Coolidge’s “Weathers”, “In other contexts, any of these could become a new sentence, in the sense that any sentence properly posed and staged could”.[10] The point here is that the effect of placing one sentence next to another with which it does not form a sequential cohesive link demands that one look around for meaningful cohesion and in doing so the meaning of the sentence changes and it becomes another sentence. The new sentence then emulates symbolism structurally. Every sentence has its grammatical meaning and an additional referential meaning or set of meanings generated beyond the meaning of the signifiers contained in the sentence yet dependant upon these signs to generate the ‘new’ meaning.




Silliman has noticed, like many of us, that if you give a reader two or three recognisable semantic units in proximity they will begin to construct meaningful ties between the units. He calls it the parsimony principle although stylistics simply calls it cohesion, or a presupposition as to what words are referring to that is satisfied as the text unfolds and relations of interpretability are established within a unified semantic context. Cohesion is a linguistic term to refer, not to how a sentence or grammatical phrase holds together, that is the task of generative grammar, but how sentences and phrases cohere into a structure of semantic sense that we call a text. For a text to cohere the sentences must be tied together based on the following functions: a presupposition about the relation of one sentence to another that is satisfied, the use of internal deixis to produce ties of reference and repetition, resulting in a cohesion of interpretability where, for example, ‘he’ of sentence 2 refers to ‘Adam’ of sentence 1. We must make a point of clarification here in relation to deixis and note that often deixis is not indication of an external world, but the application of an internal matrix of references using what is sometimes called anaphoric deixis. Thus in the example above ‘he’ is anaphoric deixis referring back to the name Adam by substituting ‘he’ for Adam. This process can also work in reverse where ‘he’ comes to be called ‘Adam’ at which point it is called cataphora. The inter-relation between anaphora and cataphora will become essential as this essay progresses. As can been seen from these three modalities, the ties that bind sentences at a local level through presupposition, pointing and interpretation, slowly come to form the texture of the text based on issues of diction, tone, limited subject matter and so on. This occurs by the sentences unfolding in a dynamic, temporal and spatially sequential series, one sentence after another, presented in a wider situation or textual context, what Silliman calls the level of integration.[11]


Typically, readers seek to integrate the units based on the context of their presentation and own reading habits. The four most common means by which sentences are tied together based on presupposition, deixis and interpretability are reference, ellipsis, conjunction and lexis. A word refers to a previous object or situation; the phrase that links one sentence to another is eradicated but assumed as it has become automatised from continual use allowing a mental jump to occur; a word like ‘and’ is used to tie together two phrases; or a commonality of diction gives the feeling of a cohesive integration of subject matter and context. Poets, including Silliman with his theory of the parsimony principle, seem almost in wonder as to how integrated cohesion occurs when in fact it is a rather simple and common linguistic function. That said the conventions of reading pertaining to sequentiality, deixis and interpretability do alter radically between different text entities, so that three lines of poetry will be read into cohesion significantly differently than three prose sentences, a fact linguists do not always dwell on.


Noticing that in prose semantic cohesion through integration into a larger semantic unit is essential, Silliman negates that larger integration forcing the reader to find semantic cohesion at the localised sentence level. It is a simple act of making the basic unit of prose, the sentence, fully visible to the reader, perhaps for the first time. It also imposes, as we will come to see when we consider torquing, a mode of poetic cohesion upon a prose texture. Thus new sentence poetics confounds poetic grammatological conventions, most particularly lineation, but only by imposing poetic cohesive conventions upon prose. In an odd sequence Silliman’s prose poetry negates lineated poetics via a poeticised prose resulting in something significantly different from either, worthy of a new name such as the new sentence: a type of literature between poetry and prose.


This poeticisation of prose units is further heightened, as I said, by the use of torquing, common to prosody but rather unusual in ordinary prose. Torquing, Silliman explains, is “the projection of the principle of equivalence from the axis of selection into that of combination”,[12] by which he means the writer chooses the word not due to its semantic equivalence, it says what the writer means, but due to its ability to fit into the textual unit. Rhyme is the classic example of this where we can often observe that a word is chosen first because of its sound and only second because of what it means (especially in a rhyme impoverished language like English). In new sentence prose the meaning of a word is determined by its location or placement in proximity to those sentences around it, for these are the only locales of possible cohesion. Perhaps for the first time the reader is forced to look at the sentence’s location in space to find its meaning, as the overall syllogistic presupposition of the text based on the assumption of a unified texture has been omitted.


A Reading of Tjanting


All of this is rather technical and the best and most enjoyable way to understand the new sentence and the effects of torquing remains to read some, so let us return to Tjanting. Take, for example, paragraph 8:[13]


Analogies to quicksand. Nor that either. Burglar’s book. Last week I wrote “I can hardly grip this pen.” White butterfly atop the grey concrete. Not so. Exactly. What then? What it means to “fiddle with” a guitar. I found I’d begun. One orange, one white, two gray. This morning my lip is swollen, in pain. Nothing’s discrete. I straddled an old chair out behind the anise. A bit a part a like. I cld have done it some other way. Pilots & meteorologists disagree about the sky. The figure five figures in, The way new shoots stretch out. Each finger has a separate function, Like choosing the form of one’s execution (Silliman, Tjanting 15).


We can immediately feel the disjunctive effect of new sentence poetics and the opening self-referential “Analogies to quicksand.” seems more than apt certainly as one progresses into the second half of the poem. Yet, based on what I have already presented, this passage is far from unstable ground in fact it is a carefully constructed perfectly coherent prosodic unit. Simple elements immediately spring to mind. The paragraph is 21 sentences long because that is what the Fibonacci sequence dictates. We know therefore that the next paragraph will be 34 sentences. Knowledge of these measures means that for a few paragraphs at least the length of these paragraphs will allow for reading them as coherent wholes, but that within a very short period the number of sentences per paragraph will be in the hundreds then thousands. Thus we stand on the brink of a treacherous bog for meaning, but as yet we are safe.


The second element we rely on is new sentence anti-syllogistic cohesion. While one sentence does not follow the next semantically, meaning is to be grasped from the local environment. Thus quicksand solidifies into “the grey concrete” which, through reference to the previous paragraph seems to suggest the surface of the poet’s “backyard” where he habitually sits and composes poetry. Here he sits on an old chair in the yard watching butterflies land on the concrete. The first is a “White butterfly atop the grey concrete.” The next alight anaphorically in the metonymic form of colours: “One orange, one white, two gray.” At this moment we apply the traditional cohesive techniques of presupposition and deixis to allow simple colour references to supply a butterfly-being to that initial, cataphoric, “One.” In doing so we alter the course of time, for the “one” is initially cataphoric, projecting forwards for future referential specificity, but then in recalling the white and the grey of sentence 5, we transform it into anaphoric retrograde reference. A matrix of temporal cross-reference is thus established, anaphoric butterfly-one, cataphoric one-butterfly, anaphoric butterfly-one, wherein it becomes impossible to ascertain if anaphora or cataphora comes first. Silliman, however, subtly undermines our attempt at referential deictic presupposed cohesion by altering the grammatological essence of the colour grey, so that concrete is grey but my assumed butterflies are ‘gray’. This we will come to see is a clue as to a potential disjuncture of my cohesive assumptions.


The final element we need to take into account of course is sentence recurrence. Some phrases such as “Last week I wrote ‘I can barely grip this pen.’” are modified from their earlier manifestations, in this instance its first appearance in paragraph 4: “Last week I wrote ‘the muscles in my palm so sore from halving the rump roast I cld barely grip the pen’” (Tjanting 15). The content of the sentence is significant of course because it suggests an extension of the poem beyond the opening lines, and also an unreliability in what Freud might have called a human tendency to secondary revision. Every time the poet writes what he wrote last week he writes something different. In contrast to this other recurrent phrases suffer no torque at all: “What then?”, “Nor that either.” These stand out in contrast to similar phrases repeated in odd-numbered paragraphs most particularly “Not this.” which by paragraph 9 becomes “Not not not-this” (Tjanting 16). Thus Silliman presents three forms of recurrence: repetition with modification, repetition without change, and accumulative repetition without change, presenting three laws of recursion in a kind of atonal harmony.


If recursion is a law of poetry, as I believe it is, it matters a great deal as to the detail and nature of recursion and how the different possible recurrences, and there are certainly more than three though not many more, interact or combine together. In Tjanting Silliman presents recurrent torquing as transformation, zero-torquing, and paratactic torquing. By these terms I mean that torquing can radically change the sentence’s morphology, the sentence can be repeated unchanged purposefully to confound the expectation of torquing, or unchanged repetition upon repetition become piled on top of each other paratactically in an example of what Hegel calls “bad infinity” and Socrates “eternal regress”.


I will leave here the various self-referential references in this section as I think for most these are self-explanatory as is how they tie together the work, and close on the consideration of the final few sentences of the paragraph. “The figure five figures in.” for various reasons. At first the phrase seems random and yet poetic with a nice combination of alliteration and assonance. Semiotics for their own sake are not uncommon in the poem although in Silliman’s work in general this tendency is a lot less common than many of his peers. We may then realise that five is a Fibonacci number and perhaps, inspired by the pellucid prose of one Dan Brown, look for clues in paragraph 5. Unlike Dan, I did not find the messiah hidden in those occult symbols however assiduously I applied myself. More fruitfully, five is soon picked up in reference to the fingers which each have, we are told, a separate function—the fingers of the self-same hand that last week could barely function as it could not grip the pen. The new shoots stretching out, according to cohesion, relate somehow this figure of five, the hand, in that the grammatical link. “The way…” suggests the cohesive quality of presuppositional ellipses, mapping “stretching” and “shoots” onto a human hand. This renders the section reminiscent of Ashbery in offering grammatical and logical cohesion that cannot be sustained. So fingers, for we pre-supposers, stretch like shoots, perhaps getting rid of that pesky cramp picked up last week. If each finger has a separate function, then I suppose an argument could be made that the choice of finger is a choice of function “Like choosing the form of one’s execution.” But again it is the rules of cohesion that ask for this rather than any real logic. We find ourselves in the midst of a tension between two conjunctions, “The figure… the way” and “Each finger… like”, which comes from the power of the parsimony principle but which he also part supports by the new sentence logic. This is further developed in paragraph 10 where the phrases again are located in close proximity, and torqued to develop potential semantic links such as are only hinted at here.[14]


This allows us to view new sentence poetics as not simply or even disjunctive, but rather a rhythmic, spatial and semantic game with the naturalised and assumed conventions of cohesion. This presents a localised or small version of the projective-recurrent cataophoric-anaphoric matrices that habitually bind long textures together through techniques of indication and reference both forward and back. The beautiful thing about Tjanting is the manner in which the RGPs of poem project the torquing of phrases forward, either providing further evidential support for what we might term a semantic cluster or undermining an assumed cluster making the whole element of referential indication a dynamic rather than entropic process. Thus if we now consider our butterfly-on-concrete-colour cluster we discover in paragraph 10 “Blue butterfly atop the green concrete.” and also, “One white, two gray, one orange, two longhair, two not.” (Tjanting 16). The possibility of the colours referring to butterflies now seems more remote as I do not believe there is a long-hair butterfly (Google agrees). Perhaps he is speaking about caterpillars, although I doubt it. Rather this sentence now seems more interested in logical problems of category and belonging: are the two longhair also white and grey; are the two that are not longhair the same two that are gray and orange? How can this be possible, as that would require four of whatever we are talking of, and what are we talking about in any case, and so on.[15] Thus projective torquing can produce semantic cluster confirmation or can confound or at least confuse semantic clustering.


To sum up, paragraph 8 presents the metrical and tabular iteration of the following poetic effects. New sentence poetics determines that initial feelings of disjunction are negated in part by a localised form of cohesion. This localised cohesion is cut across or supported by trans-paragraph cohesion, requiring that one reads backwards to confirm certain semantic clusters, and also forwards. The use of recurrence and torquing produces a matrix of coherent elements, those which will recur, yet the complexity of torquing (transformative, zero- and paratactic) inbuilds instability and the unexpected. Similarly, localised readings of new sentence semantic clusters may play games with a reader’s desire for cohesion by presenting apparent links grammatically that are sometimes supportable and at other times not. Finally, this effect is compounded by the ongoing torquing development of the paragraphs. There is no law to dictate which sentence will follow which in the next paragraphs, so that initial cohesion due to proximity can be rapidly undone. Similarly torquing may add an element which secures a semantic cluster, or elements which place it under question. Placed together these three elements, localised non-sequential coherence, torquing recurrence, and trans-structural non-sequential coherence all operate to produce not disjunction but disjunctive cohesion. They un-hook the train of links and instead open up a Mallarmean constellation of meanings.


That said, because of the Fibonacci sequence determining the measure of these paragraphs, there is a sense of doom hanging over this procedure: there will always be new material organised in unpredictable fashion, and eventually the number of sentences will overwhelm one’s ability to produce any coherent wholes. In this sense the procedure illustrates the most provocative elements of the Derridean concepts trace and difference not unfamiliar origins of much Language poetics. Each permutation of each sentence bares the trace of its previous appearance, which may now have changed, and its previous location. Yet at the same time each sentence also projects forward the future of all potential meanings which, literally in the poem, proliferate beyond human comprehension. All of this is of course due to the effects of spacing: the difference between one sentence and the next, between one paragraph and the next, and between one instance of the sentence and the next.


The Double-Helix Procedural Matrix


Habitually this mode of composition has been termed disjunctive poetics and it is disjunctive of certain assumptions pertaining to cohesion at the local and structural level of prose discourse. Yet disjunction as a term belies the fact that committed readers of the avant-garde know well, which is that the interruption of a convention of cohesion merely lays out the ground rules for the development of new methods of cohesion. Thus the new sentence is also the new cohesion. If we take our three elements of cohesion, sequentiality, deixis and interpretability, we can see them in action in paragraph 8 but modified radically to operate within poeticised prose. The rules of cohesion have essentially been torqued. Sequentiality here is clearly disrupted, yet of course the disruption depends first on the assumption of sequentiality, an assumption perpetuated by the grammatological appearance of the text as prose. In fact it is not prose at all but alinear poetry. Unable to link sentence A to sentence B, we cast around and find that sentence A refers to a sentence we recall from paragraph 6, that sentence B comes to refer to sentence F and so on. Here sequentiality is replaced by what I call tabular-planar projective-recursion. This is a concept I will clarify when I turn to the concept of poetic structure but a few words here would be of use I think. Basically in tabular-planar projective-recursion the poem is taken as a plane of meaning rather than a one-dimensional linear progression. One reads forwards, backwards, down, up and diagonally—even exo-textually. We have already seen how cataphoric projection and anaphoric recursion are essential to the overall structure of cohesion and Tjanting’s varied attacks on cohesion, and these elements form the basis of all poetic and indeed written tabularity.


Similarly new sentence poetry’s demand that one read trans-sequentially to find semantic cohesion adds to this conception of the work as multi-dimensional plane of meaning. Meaning is available all around the sentence, above and below, making it a tabular semantic entity rather than solely a syllogistic, developmental, sequential one. Thus meaning is accessed by looking backwards at sentences that occurred in other paragraphs here repeated with torque, and the expectation, based on this rule, of the recurrence of other sentences. In addition once one becomes familiar with new sentence torquing one realises all sentences find cohesion with other sentences, except those chosen specifically not to plus the 2584 additional sentences of paragraph 19 (a proportion of these are potentially cohesive of course cataphorically). Thus sentences on the whole attain cohesion with sentences that are not contiguous to but are proximate to the sentence in question. These cousin-sentences lie before or after the sentence separated by a double caesuric gap. This gap is filled each time by the one sentence, at least, that lies between cousin sentences, and the grammatological application of full stops to mark this essential spacing.


Deixis here is used as an indication of indication as such and is thus revealed as a semiotic rather than semantic device. From the opening paragraph’s use of “this”, the poem indicates that deixis with be used to point to language’s ability to speak and yet say nothing specific and will regularly be used anaphorically to imply cohesive ties and to undermine them as we have already seem. I have written extensively about this in several different places so will not dwell on this except to say that deixis in Tjanting does not provide cohesion of reference but cohesion of referentiality as such, and of language as semiotic, meaningless basis for all meaning.[16]


Finally, the rule of interpretability is fairly simple. One simply adopts the conventions and habits of poetic interpretability in place of those of narrative or argumentative prose. Thus in Silliman’s work as a whole, disjuncture follows the another Derridean law, this time pertaining to invention, namely that invention depends on the pre-existence of convention, so that the new sentence can be seen as new in comparison to the old, and that as soon as it is posited as new or singularly inventive, it ceases to be so and instead becomes a convention.[17] Not that new sentence poetics becomes a convention replacing free verse, but rather the inventiveness of Tjanting is a dynamic interaction with convention not simply its negation. Thus while the poem interrupts the three laws of cohesion, in replacing them with three other laws, if first allows us to see the previously naturalised laws as laws, replaces them with three other laws revealing that nomos or the law as such is mere convention, and yet chooses three laws that specifically and permanently disrupt cohesion, even if they are themselves poetically cohesive in a different way. This final point is what we now need to concentrate on. How does Tjanting disrupt cohesion with alternate cohesive devices and not simply replace one convention with another? In other words, why is it that antithesis does not simply overtake thesis and become the new thesis? How is the radical, asymmetrical, spiralling instability of new sentence poetics maintained? What stops the dialectic, the syllogistic structure par excellence, attaining synthesis? How does Silliman avoid structure and thus void structure?


The means by which Tjanting achieves this, and what makes it, to my mind, a masterpiece of twentieth century poetics, is based on the tensile interchange between the two procedural elements chosen. The choice of the Fibonacci sequence means that the paragraph becomes undermined as a structural container or stanza for sense. At what point this occurs is impossible to determine exactly but certainly by paragraph 19 the numerical proliferation of sentences per paragraph means the grammatological cohesive device is rendered useless. Thus confounded at the macro-level, the reader takes some comfort in localised cohesion instead, only to find that the naturalised cohesive concatenating link between one sentence and the next in linear sequence does not pertain. To discover the tangential link that still subsists, as Perelman terms it, one then has to read beyond the localised progression of sentences and view the sentences as a kind of aggregate. Lost within the street plan of the poem, one needs to climb a tall building to get an overall view of the local environment. Thus the structure of the overall poem forces one down to the localised level of the sentence, while the lack of cohesion between sentences calls for a structural overview. Procedure one, Fibonacci, is cut across by procedure two, new sentence, producing a double-helix procedural matrix which constitutes, in fact, certainly the DNA of experimental poetics but also, I believe, of poetry as such.


The two RGPs of the poem are profoundly at odds with one another and yet, in presenting a challenge to syllogism at the overall level and at the local level, they are inter-related in the way that macro-disjuncture calls for micro-cohesion and micro-disjuncture requires a macro-relational cohesion. This is remarkable enough but due to the prevalence of torquing in the poem a third dimensionality is applied which is perhaps rather astonishing. Tjanting, along with Hejinian’s My Life project, is the best example of the conception of torquing that we have. In Tjanting torquing provides a transversal link between the two otherwise heterogeneous elements of the poem’s structure, demonstrating to my mind something much more interesting and central than class struggle, what Jacobson and Agamben would call a shifter. A shifter, Agamben uses deixis as an example of this,[18] is an element that moves between two fields which work together and yet can never touch. Just as for Agamben deixis moves between language’s semiotic and semantico-referential capabilities, so torquing here—or what one might call recurrent, combinatorial, modificatory, torquing—moves repeatedly between sentence and structure at almost every point of the work.


Torquing alone means one rises above the linear sequence of the sentence to consider the work in terms of its planar tabularity. Yet torquing composed under the demand to repeat phrases with modification, something Tjanting shares with My Life, means that the structural development of the work is always held in check by a demand to read backwards and forwards across the whole body of the poem. The repetition of a phrase recalls the previous instances of the phrase which in Tjanting always take one back to the generative source of the poem, its first ten paragraphs that occupy the first three pages. At the same time, as I have already indicated, torquing also means that one can predict the phrases to come. That said this macro-cohesive device has within it a tiny flaw. Each time the phrase recurs it changes so that repetition provides both a point of structural stability and intrinsic instability. This may be described as meta-anti-torquing, in that the phrase’s recurrence is due to the combinatorial rather than selective urges of the poet, and yet as the poet modifies each recurrence he imposes a selection and thus regains a localised form of agency. This interaction between agency and desubjectivization is typical of all procedural forms which determine a set of rules beyond the agency and autonomy of the author. And yet the author chooses these rules and within the rules they are free to select any words and combinations they wish.


In other words the poem does not avoid or void structure, but presents and celebrates it. Of course, what one now means by structure has suffered a profound transformation. The structure of Tjanting consists of the following elements which I will go on to define as fundamental to the conception of poetic structure. It is composed of measured prose, by which I mean the numbers of sentences in each paragraph are counted and determined by a formal set of predetermined rules. I would term this paragraph measure. The numerical sequence chosen however is important for two reasons, first it introduces asymmetrical instability, and second it expands at such a rate that the act of counting itself becomes destabilised. Here two asymmetries contest, that between the number of lines in each sequential paragraph and that between the two limits of the poem: the first three pages contain approximately 50% of all paragraphs, the last paragraph consists of 42% of the total poem. Within each paragraph the cohesive link between sentences is broken and instead, in accordance with new sentence poetics, one must look to the wider context of the sentence usage to discover relations and cohesive links. This expansion from sequence to context called torquing is then expanded by an additional rule which is that every sentence in every other paragraph is repeated with modification.[19] This imposes the rule of localised torquing across the whole body of the poem whilst also revealing that although torquing places combination over selection, modificatory recurrence reintroduces selection at the point of combination.


Overall then the structure of the poem is based on two streams. The structural level determined by one set of rules and the sequential by another. That the rules of both levels are disjunctive rather than cohesive forces the reader to seek for structural cohesion at the sequential level and vice versa. In addition, the use of a modificatory recurrence binds the two levels of the poem together in a midway move between one side and the other that is called shifting. Meta-anti-torquing dramatises the tensions of the poem at the structural and linear level through an intra-structural and translinear projective-recurrence. These elements, which are the result of a radical avant-garde experiment which seemingly negates all elements of poetry through the adoption of prose and then all assumptions of prose through the abandonment of sequential and structural cohesion, Tjanting shares with all poetry. In this sense, the experiment merely confirms the truth of already existent data, which is after all one key function of experimentation.




Operating at both local and structural disjunctive levels, the beauty of Tjanting is the manner in which it undermines sentence cohesion whilst simultaneously attacking the rule of total textual cohesion, which Silliman equates with the logical form of deduction called syllogism. Quoting various linguists in The New Sentence for example, Silliman considers the problem of sentence integration or how sequential sentences become integrated into “higher units of meaning?”—here for the sake of argument paragraphs and the work as a whole as these are the two levels of integration under attack in Tjanting. The role of syllogistic structural cohesion is of central importance and indeed four of the eight rules of new sentence poetics concern precisely this issue in relation to syllogism. The fundamental rule of syllogism in the new sentence is that it must be limited and controlled. At the same time is must be bifurcated into primary syllogism, “between the preceding and following sentence”, and secondary syllogism where “movement is toward the paragraph as a whole, or the total work”. The two different levels of syllogism explain the contesting RGPs within Tjanting but also speak to the essential nature of poetic rhythm which occurs in the tensile traversal of localised and trans-structural elements of cohesion and torquing. Silliman concludes that the dual action of syllogistic limitation and control “keeps the reader’s attention at or very close to the level of language, that is, most often at the sentence level or below” (New Sentence 91).


On the whole, then, the convention of new sentence poetics seeks to confound structural or secondary syllogism, yet in Tjanting the use of the Fibonacci sequence actually demands one consider structural as well as cohesive disruptions. As in Hejinian’s My Life, it would seem the greatest new sentence poems are those that add a structural level of syllogistic disruption to cut across the effect of limited and controlled primary syllogistic disjuncture. Taking this to be the case syllogism relates not just to localised cohesion but is the name given to overall structural cohesion providing a movement up and out of the sequential linearity of prose into higher level integration. If, for example, we take Aristotle’s famous definition of syllogism as “speech (logos) in which, certain things having been supposed, something different from those supposed results of necessity because of their being so”,[20] we can see that as Silliman himself states, there is an assumed gestalt structure to syllogism. Syllogism is usually defined as an argument where a third term or proposition (conclusion) is inferred from two others (premises). The third term, the syllogeme if I may refer to it as such, does not progress from the other two sequentially but is created out of them in an ecstatic moment of logical poiesis. This third term, the “something different”, is then often the basis of a new syllogism and so the work progresses forwards due to the sequentiality of the first two terms, and upwards by virtue of the ecstatic gestalt nature of the third.


In a way syllogism answers Aristotle’s own question in Metaphysics as to what constitutes structure: what is it in a thing that makes it the thing that it is or “what-it-was-to-be-that-thing”. Speaking of the difference between the ontological question, what it means to be, and the more complex question, how to say that something ‘is’ that is a compound of elements, Aristotle significantly uses the example of a syllable. A syllable is made up of two elements, vowel and consonant, yet “is not just its elements… So the syllable is something in its own right, not just a heap of vowel and consonant but something different”.[21] Aristotle concedes that this “different thing must in turn be an element or composed of elements”. This raises two essential problems. One side of the argument contends the “something different”, which allows a compound of things to fuse into the thing that it is not just a heap of the elements of which it is composed, is an essential element within the thing. The other argument, the one Aristotle sides with, is that this different element is the causal combination of elements that, in a gesture of gestalt, propels out of the work its essence or substance.


Aristotelian deductive syllogism argues that the something different is first of all deduced from the elements included in the thing, so emerges from that thing, yet emerges as alterity. The conclusion of the syllogistic deduction depends on the two premises, yet it concludes something that could not be found in either premise alone. This thing is the substantial cause of the two premises as premises and not just statements, enacting that almost magical transformation of any old thing, any statement you might make, to a thing as such, statement as premise. Interestingly the logic in play here is that the conclusion results sequentially from the two premises and yet is that which, a posteriori, confers upon them the fact that they are combining to form a conclusion. This is the complex temporality of all structure: something that arises out of a combination of things that transforms them from being just a heap [sorites] into being a compound of elements that can be said to have common substance and cause, a structure.


We must keep in mind this reference in Aristotle to the syllable as not being merely a heap of but also the sum of its parts. For example in reading the work of Stein, Silliman argues first that literature and especially prose fiction depends on a syllogistic progressive ascent where the third time is omitted but implied. Here the work builds up by a series of equivalent units rather than the structural development of an argument where the third term becomes the building block for a new logical progression. Although Silliman does not mention it the name for this kind of open-ended syllogism is sorites after the Greek for a heap. The classic example of a syllogism of weak predicates, sorites, is a heap of sand. If a thousand grains of sand makes a heap, and a heap minus one grain of sand remains a heap, then repeated application of premise two would lead one to conclude that one grain of sand can be a heap, even no grains at all. The issue here is the disjuncture between a strong subject, grain of sand, and weak predicate, what precisely is a heap. We all know what a heap is or should be in experience, but it is impossible to objectively number the grains that first constitute the heap. Aristotle touches on this in Metaphysics when he explains that the element that makes a thing what it is cannot be contained in the thing because then one needs a new element to explain what a thing is plus the old element, resulting in “infinite regress”.


Sorites or heap names the necessity of structure and its impossibility. One needs to step out of sorites’ circular and open-ended logic to define a compound of elements as a thing as such, and yet in doing so one confounds the basic law of all deductive logic, all logic essentially, namely that the logical third term can be deduced out of is preceding two premises. Perhaps it is the fact that Aristotle uses syllables as his example that has fated sorites to become a central poetic category. It is a term that, as far as I am aware, first entered aesthetics in Coleridge’s Biographia Literaria where he differentiates fancy and the imagination as the difference between a sorites mode of aggregating accumulation, and the gestalt shift of what we are here calling syllogism, and which he calls imagination. [22] Fancy left to its own devices results in a mere paratactic accumulation of elements, but imagination transforms those elements into something more than the parts. In effect this names the two forms of logic to be found in Aristotle’s work, inductive and deductive. Sorites logic produces the general from the particular and is the basis of empirical science, while deductive logic produces a new conclusion from already existing elements. Fancy, sorites and inductive thinking must always contend with the problem of interior structural regress: at what point do we have enough particulars to calculate the general? Imaginative deduction however applies an ecstatic moment that jettisons the problem of the compound in favour of overall substantial cause. Fiction is fancy while poetry is syllogistic imagination.


One might here then present a sister law to Silliman’s law of fiction. If fiction is syllogism where the third term is removed but implied, poetry is syllogism in which the whole rather than being more than the parts (gestalt-argumentative syllogism) is totally coextensive with the parts. Returning to Silliman’s four laws we can see that limiting the progress of syllogism to a local, contextual, yet trans-sequential environment, means that the ecstatic break of syllogism can never become the basis of a proceeding and vertically inclined argument, but is only ever raised to the point that it collapses back into the torque of a poetic environment. The syllogistic limit means that the two parts determine the whole as merely the possibility of a relation between the two parts. In traditional logic an element of statement one must be a part of statement two allowing for a deduction which is statement three. In new sentence syllogism statement one and statement two share no discernible commonality, but come to determine a cohesive or semantic relation. A third term is not deduced from the relation of two preceding terms, but their potential relation becomes the deduction.


If we return to Silliman’s stated intention as regards the structure of Tjanting we can see that he wished to stage within the poem structure what he terms class struggle. The irony of this is quite profound. If we assume Silliman is being serious here, something I have struggled to accept over several years, then like many a poet before him he shows a woeful misunderstanding of the true power of his own work. The dialectic is the most nefarious of all syllogistic structures after all. The Hegelian aufhebung is the archetype of the progressive, deductive, ascending gestalt structure of alterity. Synthesis is the equivalent of what Coleridge calls imagination and indeed is the very source of the term in Romantic poetics. Yet as we know from Perelman and others, the fundamental structure of new sentence poetics is in fact the sorites or paratactic method in which the two terms of the syllogistic triad do not then form a third, integrated conclusion but are rather endlessly folded back on themselves. This is indeed what Silliman has enacted in his proliferating structural asymmetry. The result is not the grammatological performance of dialectical materialism as Silliman contends, but rather the proof of an entirely different conception of temporality namely poetic structure as such. This is best illustrated by a thinker who was intimately involved in Marxist politics in his home country of Italy, and is widely seen as the key revolutionary political thinker of our age: Giorgio Agamben. Yet it is not Agamben’s famous political philosophy that pertains here, but his career long engagement with a consideration of the technicalities of poetics, in this instance poetic structure.




Agamben addresses poetic rhythmic structure at three key moments in his career thus far. He first raises the issue in his first book The Man Without Content (1970) where he presents Aristotle’s ideas on structure as we have just considered them. After Aristotle Agamben defines rhythm as that which negotiates between the very principle of presence, what is outside of a work and makes it what it is (Form), and measure as such as a calculable number. One can already see here the value of such a definition of structure as rhythm at least for the art work. Rhythm is the unquantifiable ‘extra’ element that makes a thing a work of art, yet at the same time rhythm is directly dependant on the elements that make up the work of art. A logical formulation we can now see also as directly related to syllogism as gestalt and the problem of sorites syllogism or weakly cohesive parataxis. Parataxis here being defined as the reproduction endlessly of one rule of cohesion, conjunction, rather than the intertwining development of cohesion across all elements.


This centrality is further perpetuated when Agamben attempts to define the essential and original definition of rhythm by explaining how the interruption of flow in art is an ecstatic arrest of rhythm which, at the same time, inaugurates and announces the very existence of rhythm: “The word ‘rhythm’ comes from the Greek… to flow, as in the case of water. That which flows does so in a temporal dimension: it flows in time… Yet this rhythm—as we commonly understand it—appears to introduce into this eternal flow a split and a stop”.[23] At the point that the rhythm stops we are launched, for a moment, outside the work of art into the place of ecstasy, Agamben comes to calls this the caesura, and are gifted with a view of what art is before falling back into the incessant procession of the rhythm of the work below. This, Agamben argues, is the tantalising gift and reserve of art, the very structure of art “that is at once as Gestalt and number” (Agamben, Man 100).


The conception of structural rhythm between universality and singularity allows Agamben to return to rhythm in the seventh section of Language and Death (1982), a book now recognised as his first major entry into a sustained critique of modern, negative metaphysics. Here he considers the metrics, rhyme-schemes, referential matrices and overall structure of various poems. In so doing he comes upon what he conceives to be the universal law of poetic structure which I have termed projective recursion. In projective recursion the formal rules of poetic rhythm—for example in Tjanting that determine the number of sentences in a paragraph, the new sentence, and torquing—mean that one is always reading and projecting forward both in terms of what follows next, as one does in syllogism, but also in terms of what has to happen semiotically. At the same time forms recur, one looks back at previous lines to confirm formal patterns, structures of reference and symbolism repeat and develop, and one reads forwards through the poem always glancing backwards to see where one has come from. This reading forwards and backwards is typical of reading poetry although not confined to this art form. Agamben concludes from this: “Poetic language takes place in such a way that its advent always already escapes both toward the future and toward the past. The place of poetry is therefore always a place of memory and repetition”[24]


Agamben calls this the metrical-musical element of a work of art or that which determines its overall structural rhythm as we have come to define it, namely the dual act of reading sequentially within flow, and ecstatically in terms of thoughtful commentary on how the work is coming to be a unified structural entity we call art. Tying these ideas together Agamben explains:


The metrical-musical element demonstrates first of all the verse as a place of memory and repetition. The verse (versus, from verto, the act of turning, to return, as opposed to prorsus, to proceed directly, as in prose) signals for a reader that these words have always already come to be, that they will return again, and that the instance of the word that takes place in a poem is, for this reason, ungraspable. (Language 78).


The metrical-musical element relates to the specifics of poetic cohesion that I have been calling the anaphora-cataphora structural matrix. As I have already detailed, anaphora and cataphora are both regularly used in cohesion as examples of presupposition often in conjunction with deixis. The sentences “You were looking for a file. Is this it?” demonstrate how the anaphoric replacement of “it” for “file”, can combine with the deictic “this” thus confirming that “this” and “it” refer directly back to “the file” and not the innumerable other objects that occur say in a long prose narrative. Cataphoric cohesive devices work in the opposite way with initially ambiguous predicative phrases and sentences being brought into coherent focus by the imposition of a subject at the end of the sentence or paragraph: “Is this it, that file you were looking for?” Anaphora and cataphora are useful elements of cohesion that usually operate at a local level although anaphora is commonly used across a whole textual body for variance rather than cohesion, for example the use of euphemisms and ellipses when using the main character’s name in a novel. In poetry, as Agamben’s comments show, the anaphora-cataphora matrix operates across the whole of the poem’s structure, indeed defines its structure. It is the poem’s ability to project forward and turn back on itself, not to develop cohesion but as the very basis of its essence, that transforms anaphora and cataphora into fundamental structural poetic rules in the form of the metrical-musical element. In addition, as Agamben’s words show here, the role of this matrix is semiotic rather than semantic, thus it does not aim for semantic cohesion but what one might call poetic coherence.


Agamben’s most recent comments on poetic rhythm are to be found in his major reconsideration of the temporal-spatial matrix in The Time That Remains (2000). In this work Agamben juxtaposes two conceptions of time: that of movement as such, Aristotelian time as chronos, and movement towards an end point, the Christian tradition of eschaton. Interposed between these two times Agamben locates a third time, the time of the occasion/event often called kairos. Kairatic time is possible due to certain conventions assumed as regards chronos and eschaton within our culture. As time moves forward in a line, chronos, we also see that time is directional towards an end point, our death, the fulfilment of historic progress and so on. In an inverse manner, at the end of a time sequence one tends to reflect backwards over the time that lead up to this point, for example proving that historical destiny has been fulfilled. In the messianic tradition that concerns Agamben in this text, this is called the typos/antitypos matrix, which is the essence of kairatic time. Kairatic time, or the time it takes for time to end as Agamben defines it, dictates that all chronos is end-directed and all eschaton is retrospective.


While reference to messianic temporality may seem obscure, Agamben is merely trying to add a new dimension to the twentieth century philosophical consideration of what Husserl calls protention-retention. The same structure is to be found in Heidegger’s definition of human beings as thrown-projectors and most notably in his example of being-towards-death. Being-towards-death structurally emulates the messianic typos/antitypos. The subject moves through time towards its end, yet at the point of realisation of end, which occurs in the subject before they die, the reality of death comes to be the determinant factor of human authentic life. Thus the human projects forward to a death that is argued back through the while of their existence. I typically refer to this as ‘ontological time’. Finally, it is this structure that is the basis for Derrida’s critical development of différance and trace. Différance contains within it the historical trace of elements beyond a subject’s control, understanding and knowledge, and this realisation projects the subject forward into a future of possible excessive elements. Thus the subject always differs from itself and its coming to presence is endlessly deferred into the future but also into the past.


In each of these examples one can see two elements that we have been arguing for poetry. The first is the projective-recurrent tabular structure. The second is an element that occurs at the end of a sequence which, in occurring, comes to define the structure of that sequence qua sequence (ontological retrograde temporality). Commenting on twelfth century European poetics Agamben is amazed to discover the two laws combined into a theory of poetic structure that our ancient forbears termed cruciform retrogradation. Cruciform because the poem becomes two-dimensional, both horizontal and vertical, and retrogradation because effects are agued backwards through recurrence.


As we are now back within the confines of The Time That Remains we can see that everything about messianic time in this work recalls the figurality of the poetic, the temporality of poetry, and the structure of the poem. A structure such as the kairatic oment depends on the precise mix of occurrence and reiteration, anaphora and cataphora that is the basis of any poem structure. As soon as Agamben describes the poem one can begin to see how wonderfully this analogy works, although to describe it as analogy, example, or model is, in each case, insufficient. Thus he says of the closed rhyming lyric form, for example the sonnet, which necessarily will come to an end as determined by the rule of the form: “The poem is therefore an organism or a temporal machine that, from the very start, strains towards its end. A kind of eschatology occurs within the poem itself. But for the more or less brief time that the poem lasts, it has a specific and unmistakable temporality, it has its own time”.[25] This is especially true, he argues, in the case of rhyme, a fact made most apparent in that rather rare stanzaic form the sestina.


A sestina is made up of seven stanzas. The first six stanzas are each six lines long and the six end words are always the same in each stanza only organised in a different combinations. The final stanza or tornada is then only three lines long but repeats all six end-words placing two per line and always ending on at least one of these. Agamben’s example is taken from twelfth century poet Arnaut Daniel but I have also written some years ago about the use of sestina in John Ashbery.[26] The form that still operates on occasion in modern poetry in other words. Agamben’s analysis of the rhythm of the sestina whilst most apparent in this poem form is, I have argued in my own study of this phenomena in modern experimental poetry, a foundational quality of all poetic structure. Put simply, every poem unfolds in linear time semiotically marking this out with great clarity by using artificially ended lines which graphically demonstrate chronos much more adeptly than in any other art form. That said every poem is also a recursive or reiterative structure. Thus in the sestina as one moves toward the predictability of the end the closed form means that in every line the end is prefigured. At the same ‘hermeneutic’ time one also picks up on the interplay, repetition, and variance of the use of homologous rhyming end words. You begin to recognise the pattern, look to how the next stanza will recombine the six fixed elements and thus one is always reading both forwards and backwards. For example in the penultimate stanza one can predict the distribution of the final end words without reading the stanza simply by looking back at their distribution in the previous six stanzas.


This reading back however comes most to the fore in the tornada where, effectively, all usages of the words thus far are recalled in their final combination. This matrix is a particular example of the anaphoric-cataphoric matrix of every poem, converting the poem from a linear-horizontal entity to a tabular planar form. Agamben concludes thus on the sestina:


The sestina—and, in this sense, every poem—is a soteriological device which, through the sophisticated mechane of the announcement and retrieval of rhyming end words (which correspond to typological relations between past and present), transforms chronological time into messianic time… the time of the end, the time that the poem takes to come to an end’ (TTR,83).


This temporality of cruciform retrogradation, projective recurrence, is as apparent in Silliman’s Tjanting as in the work of Arnaut Daniel. Projective recurrence therefore is what I take to be the cause of Tjanting’s lasting brilliance, the essence of poetic structure as a whole, and thus the basis of all poetry. It names not a poetic convention, but poetic being as such. As omnipresent in the poetics of invention as in those of convention, it presents a forward projection of a backwards re-appreciation of the commonality that exists between the history of poetry in the west, which is the history of literature, and the present moment of poetic experimentalism. A commonality that was forgotten during the long century of free verse poetry, which seemed to operate precisely to obfuscate this essential link between poetry, recurrence, and structural rhythmic constraint.


Conclusion, choose one of the following


a) Tjanting is a masterpiece of avant-garde poetics.


b) Tjanting is a masterpiece of traditional prosody.


In attempting to disrupt the conventions of mainstream poetics, and in fact in succeeding in doing so, Tjanting simultaneously presents the essence of poetic structure and adheres to this presence. Here then we may make an important differentiation between poetic conventional rules, and the rule of poetry as such. As far as I am concerned, there is no difference structurally between the work of Arnaut Daniel and Ron Silliman. The combination in the poem of new sentence and Fibonacci RGPs in Tjanting results in a rhythmic interchange between localised sequential flow and ecstatic gestalt tabularity. Although the intention was to interrupt local and trans-structural syllogism, the result is the presentation of the true structure of poetry which remains untouched by the work. The structural rhythm of poetry as Agamben defines it is based on the tensile temporal and spatial interplay between arrested sequentiality and retrograde syllogism. Structure therefore is not a choice between infinite regress and causal substance as the Aristotelian inheritance presents it, but is in fact a dynamic rhythmic interchange between local sequence and meta-structure. Sequence disallows structure from becoming that thing extra, the essence confirming transcendental other, in other words meta-structure. Yet at the same time structure transforms syllogistic prose into disruptive-productive poiesis. Structure is neither the sum of parts or the whole as more than the sum, but the metrical tabular superimposition of one atop and across the other.


This cruciform retrogradation is, significantly, a temporal-spatial category or, more importantly for twentieth century thought, the collapse of this clear distinction. The sequentiality of sentences in Tjanting unfold in time and space. Indeed the use of new sentence techniques mean that linearity, the ideal model of our conception of time since Aristotle, is never secure because cohesion, as we saw, can only be attained vertically or in two dimensions. Similarly the structure of the poem is both spatial, structure after all is an ecstatic moment of overview and so atemporal, and yet also temporal: an open-ended proliferation of number. Here the use of number is essential. Number is the essence of the sorites, unity-confounding paratactic nature of much experimental poetry. Yet number is also measure. Choosing a paragraph measure of the order of Fibonacci produces a structural slow-burn. Initially paragraphs are cohesive, but built into the system is the projection of the inevitable collapse of measure into numbers not immeasurable and infinite, but simply too large to be made to cohere. This is, after all, why prose utilises syllogism locally and structurally: to make cohesive sense of very large text units. Thus number in the poem is both structural and undermining of structure, echoing the means by which sequence is both sequential and tabular.


These contesting elements are brought together in the poem by use of what Silliman calls torquing and Agamben names the semiotic shifter. The semiotic in the poem, according to Agamben, relates to artificially imposed line-breaks due to syllabic counting. In measured prosody semiotic units impose themselves at the expense of semantic units: line overturns sentence revealing the sensuous law of all art. Torquing in Tjanting cuts across the sequential-structural matrix joining the two elements and disrupting them. Torquing predicts recurrence thus aiding meaning. Yet it is recurrence with modification which is disruptive of stable meaning. It is, in effect, a new form of spatially shifting, temporally predictable, anti-anaphora. In anaphoric deixis, central as we saw to cohesion, the recurrence of something through pronouns, deictic indicators and so on, means that while the signifier of the thing changes, the signified is assumed to remain static. In contrast in torquing, the signifier and the signified change over time, yet the combinatorial element of the sentence, what Agamben also terms the semiotic indication of language as such without a single definable referential meaning, remains static. Thus torquing refers to the semiotic singularity of language as such, away from the assumption of fixed meaning either in terms of words or their meanings.


Torquing thus extends sequentiality beyond and across paragraphs, yet it interrupts sequentiality in that the rule determines sentences will recur but not in which order. At the same time it insists on a concentration on the semiotic nature of language as such, that it happens or occurs at all, in advance of any specific meaning, both because the change of sentence location changes its potential meaning, and the modification of sentences also presents an alteration of meaning. Where the sentence is, and where the words are in the sentence, become more important than what the sentence as a whole says, or what the words alone may be taken to mean. On a structural level torquing provides a rule of cohesion, but an anti-syllogistic one that unfolds in time rather than existing transcendentally in ecstatic space. At the same time its predictably unpredictable recurrence, its Husserlian protention, requires backwards glances to previous occurrences, retention, proving once and for all the fundamental rule of contemporary metaphysics, the temporal-spatial projective recurrent temporality of ontology as such.


Torquing, then, is another name for what might be termed poetic being. It is a temporal entity that unfolds in and across space. It is a sequence that is tabular and disruptive of sequence. It is a projection forwards of a recurrence. And also a recurrence of the law of projection as such. Torquing, therefore, in being transimmanent as Jean-Luc Nancy calls it,[27] and messianic according to Agamben, defines the essence of poetry precisely as that which formally disrupts and presents the very basis of being: a time that is a space, a sequence that is a structure, and projection which is recurrent. The way in which Tjantingcombines new sentence sequential disruption, Fibonacci asymmetrical proliferant structural measure, and a torquing, involuted, diagonal, projective-recurrent shift or passage between, defines it as the most advanced projection of the retention of the most ancient element of all art: structure.


[1] Charles Bernstein, Girly Man (Chicago: University of Chicago Press, 2006) 99.

[2] Ron Silliman, Tjanting (Cambridge: Salt, 2002) 15.

[3] I am grateful to Thomas Fink for much of this factual detail provided by his article “No Other Sentence Could Have Followed but This,”: Ron Silliman’s Tjanting published in Titanic Operas: Poetry and New Materialities at www. emilydickinson. org/titanic/material/finksilliman. html

[4] For a full consideration of RGPs across a range of Silliman’s work see William Watkin, “‘Systematic rule-governed violations of convention’: Ron Silliman’s Poetic Procedures”, Contemporary Literature 48.4 (2007): 499–529.

[5] Silliman Interview, The Difficulties 2.2 (1985) 34.

[6] For more on this negative form of deixis see William Watkin, “The Materialization of Prose: Poiesis versus Dianoia in the work of Godzich & Kittay, Schklovsky, Silliman and Agamben”, Paragraph 31.3 (2008): 344–364.

[7] Silliman, ‘Interview’ 35.

[8] Ibid.

[9] Bob Perelman. The Marginalization of Poetry. (Princeton N. J: Princeton University Press, 1996) 61.

[10] Ron Silliman The New Sentence (New York: roof books, 1987) 88.

[11] This definition is a compound taken from M. A. K Halliday and R. Hasan. Cohesion in English (London: Longman, 1976) and Michael Toolan. Language in Literature: An Introduction to Stylistics (London: Arnold, 1998).

[12] Silliman, New Sentence 89.

[13] For a reading of paragraph 9 to accompany this reading see Watkin, “Materialization of Prose”.

[14] For example in paragraph 10, the stretching shoots are echoed in a “thin black strap”, the often stretching shadow’s, other foliage in the garden such as willows and ferns, and finally seemingly recalled by the fine filament hairs in the human, or at least poet’s, nose (see Tjanting18).

[15] In a later permutation: “Two long-hairs, on gray, one white, one orange.” (Tjanting 18) it seems more likely the poet is considering his own hair. He does this by placing long-hair at the sentence incipit and introducing a dash, suddenly opening up another realm of semantic possibility and indicating, through torquing, how grammatical sequencing and grammatological spacing produce meaning.

[16] See for example the chapter “Projection: There is Language” in William Watkin, The Literary Agamben: Adventures in Logopoiesis (London: Continuum, 2010) 5–42.

[17] This is a shockingly brief but I believe accurate summary of the arguments put forward in Jacques Derrida, “‘This Strange Institution Called Literature’: An Interview with Jacques Derrida” in Acts of Literature (London: Routledge, 1992) 33–74.

[18] See Giorgio Agamben, Language and Death, trans. Karen Pinkus with Michael Hardt (Minneapolis: University of Minnesota Press, 1991) 25.

[19] Those sentences that do not change are still torqued under the rule of what I term zero-torquing.

[20] Aristotle, Prior Analytics I.2, 24b18–20 (

[21] Aristotle, The Metaphysics, trans. Hugh Lawson-Tancred (London: Penguin Classics, 1998)229.

[22] See chapters 12 and 13 of Samuel Taylor Coleridge, Biographia Literaria in The Collected Works of Samuel Taylor Coleridge 7, ed. Kathleen Coburn (Princeton: Princeton University Press, 1983)

[23] Giorgio Agamben, The Man Without Content, trans. Georgia Albert (Stanford: Stanford University Press, 1999) 99.

[24] Agamben, Language and Death 76.

[25] Giorgio Agamben, The Time That Remains, trans. Patricia Dailey (Stanford: Stanford University Press, 2005)79.

[26] William Watkin, In the Process of Poetry: The New York School and the Avant-Garde (Lewisburg, Penn: Bucknell University Press, 2001), 203–6.

[27] See Jean-Luc Nancy, The Sense of the World. Trans. Jeffrey S. Librett (Minneapolis: University of Minnesota Press, 1997) 18–19.

Copyright Notice: Please respect the fact that all material in Jacket magazine is copyright © Jacket magazine and the individual authors and copyright owners 1997–2010; it is made available here without charge for personal use only, and it may not be stored, displayed, published, reproduced, or used for any other purpose.