But after the publication of a text, the documents that comprise its pretext are understood through the text that they had generated. Almuth Grésillon defines a pretext as "un ensemble constitué par les documents écrits que l'on peut attribuer dans l'après-coup à un projet d'écriture déterminé dont il importe peu qu'il ait abouti ou non à un texte publié" (Grésillon, 109). Thusly defined, the pretext participates and corresponds within a work in progress, but it can only be discernible through the context of its product, whether that is a published text or not. The final text bears the traits of the pretext whence it derives, but what I would like to ask today are the other ways in which a pretext might impact against the final text, the ways in which a text-say for example Finnegans Wake-registers its work in progress. This would be something like what Daniel Ferrer calls the "memory of the context": the traces that each intermediary stage deposits into the retrospective wake of composition, traces that are only available once the "text" is construed into a broader context of pretext, text and paratext (Ferrer, 233).
Before moving to Joyce, I would like to say a few things about Mallarmé and his aborted project of writing Le Livre. From an early age, Mallarmé had planned to write a Total Book in which all his work and all his energies would contribute and participate: a Dantean vision in which the world would be gathered and bound by the spine of The Book. In 1868 he described this still-inchoate plan--one which he envisioned would occupy him for at least 20 years--to his friend Henri Cazalis: "Mon oeuvre est si bien préparé et hiérarchisé, représentant comme il le peut, l'Univers, que je n'aurais su, sans endommager quelqu'une de mes impressions étagées, rien en enlever" (Mallarmé 1965, 99). Unfortunately, Mallarmé did not write this Livre-at least not as such-and indeed he spent most of the 1870s and 1880s stuck in a colossal writer's block, a hideous inability to write anything, a crise de vers. In another letter to Cazalis he described this crise by noting that "le simple acte d'écrire installe l'hysterie dans ma tête" (Mallarmé 1956, 301). (Having just written a dissertation, such a line appeals to me.)
What Mallarmé did write were plans as to what the Livre would be like: some of these are highly refined essays, most of which are collected in the volume Divagations, and others were simply notes (I will talk about these shortly). In a certain very limited sense we are left with a pretext bereft of a final text. Indeed Mallarmé's oeuvre-the one he did manage to write-bears the marks of this groping inability to write the Livre. As Leo Bersani says in his engaging study of Mallarmé's crise: "The very crisis which threatens the writing of poetry sustains poetic composition" (Bersani, 2). If--as Mallarmé famously wrote in the essay Le livre, instrument spirituel--"tout, au monde, existe pour aboutir à un livre" (Mallarmé 1945, 378), then even the crise de vers ends up vers le Livre.
We seem to be quite far from Joyce here. After all Joyce did manage to write Finnegans Wake, despite much complaining and the exhausted near-abandonment of writing in 1927 to James Stephens, a writer with more stamina (see Letters I: 252). But the Mallarméan experience of writing the Livre--or, more properly, of not writing the Livre--can provide some insight into the inscrutable rapports between text and pretext: how pretexts undo the sacrosanct notion of an autotelic text even as they contribute to its eventual manifestation.
First, I would like to turn briefly to the remnants of the notes Mallarmé wrote in preparation for the Livre. Most of these were burned after his death in 1898, according to his wishes, but one notebook survived and was published in 1957 by Jacques Scherer in the volume Le "livre" de Mallarmé (with the word livre within quotation marks). In his introduction, Scherer characterizes these notes as an imperfect record of a thinking of and towards le Livre. Scherer thus defines them by some missing book which would have been their fulfillment.
The notes Mallarmé left behind deal very little with the content of the planned Livre, and instead concentrate a great deal on the form and format the Livre was to take, even dealing with such incidentals as its final cost. Unlike a regular book, Mallarmé planned to have the pages unbound, and so the order in which the Livre would be read would be subject to permutation. Each reading of the Livre would be a performance or séance in which it would adapt itself to its circumstance (cf. Scherer, 58-61). For example, the number of pages in each volume of the Livre would vary according to the number of operators and auditors present at each séance (Scherer 102-3). Verso and recto are to be interchangeable in the multiple possibilities of this volume's binding; and so the Livre would not impose a single direction or vector of reading. Indeed, the notes seem to be experimental jottings concerning this variable ordination of pagination. In a sense, one could consider Mallarmé's plan as an attempt to enact a manual or non-digital hypertext: a hypertext that does not depend on the latest HTML ordinance from Bill Gates or the WWW Consortium.
[Sample page from Mallarmé's "Livre"]
In Le livre à venir, Maurice Blanchot notes that this performative aspect to the Livre--as planned in the notes--would guarantee that the Livre will always be iterated variably, with no original. The Livre is always in progress and "est toujours autre... il n'est jamais là, sans cesse à se défaire tandis qu'il se fait" (Blanchot, 330 n.1) The Livre remains conjugated in the conditional, and this conditionality is what has impacted into a book, which is still, always, a livre à venir. Each single iteration of the Livre is always an imperfect manifestation. The Livre thus oscillates between manifestation and disappearance, a hypothetical disappearance of what never had been.
Following from Blanchot, the Livre then becomes absolutely non-self-contiguous: it is never The Book. Even if Mallarmé had actually managed to produce the damn thing, it would still exist as a virtual book, it would just be a very different virtual book from what he had in fact left behind. The Livre is always in-progress, always provisional, always hypothetical, and always subjunctive. This leads into the question of whether or not Un coup de dés can be construed as either the Livre or as the apologia for not having written it. Blanchot neatly finesses this binarism: Un coup de dés marks the possibility of the Livre as a possibility, and no longer as a necessity. By re-marking itself à l'écart du Livre, it re-marks that the Livre is already à l'écart: the identity of the Livre tel quel is its non existence. The very contingency of Un coup de dés is thus its coincidence with the eternal absence of the Livre:"Il n'est pas moins fermement indiqué, dans Un coup de dés, l'oeuvre même qu'il constitute et qui ne fait pas du poème une réalité présente ou seulement future, mais, sous la double dimension négative d'un passé inaccompli et d'un avenir impossible" (Blanchot, 318). Un coup de dés is the Livre precisely because it is a sort of confession of the impossibility of writing the Livre. Un coup de dés may indeed be an attempt, not just an attempt to present the Livre, but a hypothetical projection of its tentative possibility. In this Mallarmé was--as he wrote in one of the feuillets collected by Scherer--always "fidèle au livre" (Scherer, feuillet 35B). The Livre does not exist, and so Un coup de dés can only remunerate it into a virtual possibility.
[Page 5 from Un coup de dés]
The poem Un coup de dés accepts a variability of reading: it can be pursued from several directions. It consists of a chain of mutual metonyms splattered across several pages which belong within both the multiple series of relationships suggested by the spacings as well as the hierarchies of typography (point-size, capitalization and italicization across pages gathers units that would otherwise remain disassociated). Each unit is a further displaced hypothetical elaboration upon an initial hypothesis which is incorporated into the poem's pages: "UN COUP DE DÉS... JAMAIS... N'ABOLIRA... LE HASARD."
As Malcom Bowie notes, the multiply construable individual units within the poem "exert upon each other an associative pull strong enough to cancel the intricate syntactic patterning which holds them apart; in so doing they become a mosaic of reciprocally explaining fragments, a counter-syntax, a refusal of hierarchy" (Bowie, 145). Amidst the multiple available directions, there is no "unique Nombre qui ne peut pas être un autre," no singular autotelic coincidence of identity. The absence of the Book is thus re-marked by the counter-syntax generated by the poem's word-clusters.
Even if the format of the poem is open to different vectors of reading, Mallarmé spent a great deal of care in ensuring the alignment of the poem's units upon the page. As you can see in the corrected page-proofs for the page 5 of the poem, Mallarmé insists upon a proper alignment between the units on both the verso and recto pages. The indeterminacy of reading emerges only through a great deal of calculation and determination in the writing process.
[Page-proofs for page 5 of Un coup de dés]
I would now like to turn to a very specific example from the archive of Finnegans Wake to show how these Mallarméan notions of the Livre and of the Livre à venir might be applicable to a reading of Joyce. Firstly, Bowie's characterization of how the elements within Un coup de dés interact in a counter-syntactical refusal of linear hierarchy is not without relevance for a consideration of the operations of Wakean wordplay. David Hayman calls this "superparataxis": a combination of both vertical and horizontal parataxis, elements within coordination yield to many not-necessarily-compatible schemes of patterned subordination without ever being subsumed by one universal logic (Hayman 1985, 194-6; Hayman 1987, 194-6). If the word suvntaxioe means a disciplined arrangement of proper subordination, coordination and agreement for suitable progression, then in the Wake--as in Un coup de dés--there are many different syntaxes which do not coordinate and intersubsume under a single ball and chain of command.
I would like to suggest that Un coup de dés be considered as a possible model for Wakean writing, that the multiplicity of inter-relations between the units of the poem, in that they work against a singular definitive reading might provide an example of how Wakean syntaxes work through the concatenations of polylinguistically polysemous portmanteaux. One consequence of this, at least according to Blanchot's interpretation of Mallarmé, is that amidst differential readings, the Livre remains perpetually virtual: the variability of the vector of reading guarantees that any single iteration of the Book remains a simulation of The Book, The Book which is always forthcoming. Therefore any single book is but an imposture book, or to quote the line from the Finnegans Wake notebook from which I derived my title today: "imposture book through the ages, revered more & more" (VI.B 10: 9).
The plurality of available syntaxes also plays out across the archive of a text's pretexts. The pretexts, as well as aiming towards a final work through which their work will be fulfilled, tear the final work away from the possibility of a definitive foreclosure. Pretexts open the text out onto a broader field of textuality, a field which haunts the pages of the closed book. Following from suggestions I have made elsewhere, the interplay between the various stages within a pretextual archive is almost as malleable and multi-directional as the rapports between the elements within Un coup de dés, and any operation of textual verifiability necessarily rests within an expansive chain of mutually contextualizing fragments.
I would now like to take up a specific example from Finnegans Wake to further illustrate this point. In the account of the missing Letter or mamafesta in chapter I.5, there is a parodic rendering of a pedagogical presentation of manuscript study. Although the Letter is ostensibly missing, many hypotheses exist as to its nature and circumstance. Indeed the Letter could be said to exist as the multiple hypotheses that proliferate around its absence. The document, such as it may be read, is what absorbs its misreadings: "Closer inspection of the bordereau would reveal a multiplicity of personalities inflicted on the documents or document and some prevision of virtual crime or crimes might be made by anyone unwary enough before any suitable occasion for it or them had so far managed to happen along" (FW: 107.23-8). The text is already corrupt and indeterminate (document or documents) and such indeterminacy and confusion is also applied to the inspector(s) of the manuscript (it or them). Here questions concerning textuality cannot be disentangled from corruption. There is thus a double path here: both reader and manuscript intertwine so as to constitute each other within a genealogy of error. "Thus the unfacts, did we possess them are too imprecisely few to warrant our certitude" (FW: 057.16-7). The philological recovery of the text becomes the recovery of an error, which is itself an error added on to the list thereby ensuring its (erroneous) continuity. The document's survival is ensured by philological misinterpretation. The reader and the manuscript exist only to the extent that each is a function of the other within an act of hermeneutic redegeneration:
In fact, under the closed eyes of the inspectors the traits featuring the chiaroscuro coalesce, their contrarieties eliminated, in one stable somebody similarly as by the providential warring of heartshaker with housebreaker and of dramdrinker against freethinker our social something bowls along bumpily, experiencing a jolting series of prearranged disappointments, down the long lane of (it's as semper as oxhousehumper!) generations, more generations and still more generations (107.28-35).As soon as a pattern emerges, the traits disseminate. There is no definitive statement, only a precession of cryptic statements of disappearance. The telling of the fall and desistance recurs shamelessly, perpetually deferring a definitive instance of closure. The text, such as it may have been, is inseparable from its reconstitutions: "it has acquired accretions of terricious matter whilst loitering in the past" (114.28-9). An Ur-text recedes under the weight of palimpsestuously added inscriptions. In this sense, the discourse around the mamafesta recalls the Mediæval commentary tradition which "does not simply 'serve' its 'master' texts; it also rewrites and supplants them" (Copeland, 3). A more contemporary analogy would be the part of the Internet known as Usenet, where a single posting elicits scores of responses, each of which is liable to elicit further response until the responses bear no apparent appurtenance to the original posting, a posting which survives in name only. One posting generates a multitude of commentary and response, a multitude which evolves to disappropriate itself from its source (this phenomenon is also not unknown on the Finnegans Wake e-mail list FWAKE-L, as I'm sure we all know).
The "text" thus becomes the disseminative commentary that takes the place of what might have been an Ur-text. On the one hand this proliferation of supplementary commentaries implies an economy of the obscure (if not obscene), but on the other hand in Finnegans Wake it can no longer be construed solely on an archæological level but rather as a play of superparatactic differentiations. The Wake proffers its own glosses and comes to be inseparable from them as if it were a text composed solely of footnotes: glosses of a passage that remains separate and apart. Its text is already paratextualizing, paratactically prone to the paratextual. That into which everything in the world ends up is not a book, but a footnote, just a footnote.
Like the clusters in Un coup de dés, the diverse accounts of the Letter in I.5 explicate and elaborate an absent center around which they appear to orbit. The Letter is subject to addition and accretion, and the ways in which matter is added displaces the Letter away from stability. One particular account of this process of addition and revision in the Wake is exemplary of this from both the perspective of its own genetic development and from its final iteration in Finnegans Wake (pages 119-121). Joyce's work-in-progress worked its own progress into itself and a passage describing a "revise mark" came to be greatly marked by revisions.
In the second draft of what became I.5, Joyce wrote: "the fretful eff (used as a revise mark) stalks all over the page, broods amid the verbiage, gaunt, stands in the window margin, paces jerkily to & fro, flinging phrases here, there, or returns with some half-suggestion, dragging its shoestring" (JJA 46: 310). This draft, typical with all early drafts, is a mess of additions and deletions; even this phrase I have just quoted is the product of two discrete additions. It seems that Joyce is characterizing his own process of addition and accretion on his page since the "fretful eff" describes the "F" mark which Joyce frequently--but not exclusively--used in his manuscripts as an insertion pointer. This allusion is elaborated in subsequent drafts.
[Second-draft of FW 120-121]
In the next draft-the first fair copy-Joyce qualified the account of the fretful eff with some material derived from Sir Edward Sullivan's introduction to The Book of Kells. Such a reference is appropriate since Joyce borrowed many characteristics from The Book of Kells for I.5. You can see an example of the fretful eff on this page, indicating the addition of the following passage: "the curious warning sign before our protoparent's ipsissima verba (a very pure nondescript, by the way, sometimes a palmtailed otter, often the arbutus fruitflowerleaf of the cainapple) which paleographers call a leak in the thatch or the Aranman whispering through the hole of his hat, indicating that the words which follow may be taken in any order desired" (JJA 46: 318).
As James Atherton has noted, in The Book of Kells the symbol known as "head under the wing" and "turn under the path" indicates that the words immediately following are to be read after the end of the following line (Atherton, 200-1; Sullivan, 7). Normally this symbol resembles a capital C, but Sullivan notes that in one instance it appears as "a small man apparently in the act of jumping, with one of his legs cocked up and the other turned down towards the following line of the text" (Sullivan, 8; this charming image is not reproduced in his edition). It seems that the passage Joyce added into the text at this stage was inspired by this passage from Sullivan even if there is no direct citation of Sullivan's account. Indeed, the quixotic character and function of the "leak in the thatch" recalls The Book of Kells's "turn under the path," but this specific phrase apparently comes from elsewhere as it is recorded in notebook B.11 in a very different context: "she saw him in the flesh / a leak in his thatch" (VI.B. 11: 91).
[Plate III from Sullivan's edition of The Book of Kells]
There were, however, verifiable citations from Sullivan added into the text at this level; the attitude of the revise mark has now been revised and "stands dejectedly in the diapered window margin." This is a reference to the diapered designs in The Book of Kells which were used to "brighten small spaces lying between the larger designs" (Sullivan, 28). The entry "diaper"--along with other terms from Sullivan which were used in I.5 and elsewhere, such as "nondescript" (VI.B. 6: 67; Sullivan, 29)--is found in notebook B.6 (VI.B. 6: 62). The fretful eff stands in what is ornamentally designated as the margins as what introduces the marginal matter into the text. It is an ornament with a guiding purpose.
The fretful eff is a mark that directs the vector of reading, the order in which the words are to be construed. And à propos of the farrago of Joyce's manuscripts, such markers are sorely needed. These are then, in a sense, syntactic markers placed before the words (ipsissima verba) that indicate the proper arrangement and coordination of the words through the morass of addition. The markers redistribute sense and sensibility. By being colored with traits deriving from The Book of Kells's ornate addition marks, the character of the revise mark becomes polymorphous and polyvalent. But such superabundance of traits ultimately renders it "a very pure nondescript." Its own shape becomes amorphous in the palimpsest of morphological descriptions through which it comes to be articulated in the evolving text.
Furthermore, in this account, these marks are construed as meaning "that the words which follow may be taken in any order desired." The strict symbolism of these directional markers is taken--in the case of the Letter--of exacerbating rather than mitigating the frenzy of parataxis: parataxis without order. Of course if Joyce and his various assistants had been good readers of this passage and had seriously heeded this advice in the preparation of the Work in Progress, then the text would be quite chaotic. Indeed there are numerous instances during the composition of Finnegans Wake where the hapless typists, confused by Joyce's conventions, misinterpolated additions into the text. But considering the state of some pages, it's amazing that this didn't happen more often.
In the next draft level--the first typescript--Joyce adds a further description of the fretful eff: "the digamma of your barbarian form" (JJA 46: 333). The digamma (F) is the "missing letter" of the Greek alphabet (with a conjectured sound-value of w), it existed in prehistoric Ionic Greek but vanished long before Greek became a written language. Its existence was inferred through consistent gaps in the meter of Homeric epic. These gaps indicated that a letter had been there but has since vanished; the great Classical philologist Milman Parry wrote that "Homer's language has traces of the digamma, but not the digamma itself" (Parry, 131). To remunerate this alphabetical deficiency, philologists devised the digamma which resembles the fretful eff of Joyce's manuscripts; indeed the insertion mark Joyce used to indicate the addition of this passage almost perfectly coincides with the digamma. The digamma lettershape is thus an invention or conceit of barbarian philologists (barbaros being the Greek word for one who does not speak Greek) to compensate for a missing letter inferred only through its traces in the written language. Like the digamma, Joyce's fretful eff is a mark that is not itself to be carried over into the subsequent writing stage: it is a mark that retracts itself once its work of designating the sequence of addition is to be carried out. It is a mark that can only bee re-marked through the traces that it had left.
In the revisions of the Criterion undertaken prior to publication in transition 5, Joyce added two inverted-and perhaps fretful-Fs into this passage (JJA 46: 426). He added into the book the self-erasing revise mark that "facilitated" (I use this word slightly ironically) the incorporation of additions. Note that elsewhere these pages include the fretful eff in its standard role of designating additions, and so Joyce took some pains to ensure that these two effs would be construed as elements to be incorporated into the text rather than as elements that designate the incorporation of other material. In the first instance Joyce indicates the addition with two carats and a line; of course the carat resembles the Shaun siglum (used elsewhere in this chapter), and so that type of demarcation could have been easily misinterpreted. In the second instance, in addition to the carats and the line, Joyce also puts the eff within quotation marks. Note that here the first eff is upside-down and the second eff is upside-down and mirrored.
[Revised Criterion pages]
Fortunately the typesetter properly understood Joyce's request and incorporated the inverted effs into the page-proofs (JJA 46: 435). Unfortunately both effs are upside-down and mirrored. To ensure complementarity between the two figures, Joyce restored the upside-down but unmirrored eff to the page. But the order is now reversed: the first eff is upside-down and mirrored, and the second is upside-down. This is how they appeared in transition 5 (JJA 46: 465) and in the final text.
The final text, the product of many other revisions and additions, reads as follows:
that fretful fidget eff, the hornful digamma of your bornabarbar, rarely heard now save when falling from the unfashionable lipsus of some hetarosexual (used always in two boldfaced print types--one of them as wrongheaded as his Claudian brother, is it worth while interrupting to say?--throughout the papyrus as the revise mark) stalks all over the page broods sensationseeking an idea, amid the verbiage, gaunt, stands dejectedly in the diapered window margin, with its basque of bayleaves all aflutter about its forksfrogs, paces with a frown, jerking to and fro, flinging phrases here, there, or returns inhibited, with some half-halted suggestion, , dragging its shoestring; the curious warning sign before our protoparent's ipsissima verba (a very pure nondescript, by the way, sometimes a palmtailed otter, more often the arbutus fruitflowerleaf of the cainapple) which paleographers call a leak in the thatch or the aranman ingperwhis through the hole of his hat, indicating that the words which follow may be taken in any order desired (FW: 120.33-121.13).Over the course of its draft evolution, the account has thus incorporated many additions, including the addition of the addition-mark. All of these additions interrupt an account which is already an interruption: further qualifications spun out over the absence of what was already qualified. Even as Joyce's own insertion mark literally becomes incorporated into the text--amidst elaborations of additions and additions of elaborations--the elaborations serve to render the Letter subjunctive and hypothetical. All of these elements that Joyce's fretful eff bring into the text, including a form of itself, are interruptions, "is it worth while interrupting to say?"
The additions thus tend to render the Letter more and more belated through the force of their accumulation. This mirrors what happened to the text of the Letter during the composition of this chapter. The Revered Letter itself--along with a brief introduction describing its characteristics--was to have been the sequel to I.4. Presumably during the revisions to the typescript in January 1924, Joyce decided to take the Letter out of chapter I.5 altogether. The Letter was then put aside until 1938 when it was incorporated into Book IV. Therefore what is now the text of I.5 began as only the contextualizing matter for that chapter. Furthermore the planned postface to the Letter was also dropped out of the chapter and instead provided the conceptual foundation for Book III, the delivery of the Letter. Both David Hayman and Laurent Milesi have provided much more detail concerning the rearrangement of the Letter (see Hayman's "Introduction" to JJA 46, xiii-xviii and also Hayman 1990, 166-8 and Milesi, 79-107) .
The development of the contextualizing matter for the Letter, then, in its evolution discards the Letter itself. Like the account of the Letter in I.5, the development of the draft of that chapter supplements the very text that it is describing. The description no longer needs its source. Of course the source is, shall we say, "preserved" in the traces that it leaves in the final text: traces which paratactically disvelop on their own initiatives and momentums. And of course the source is preserved in that it appears in a different part of the text, within a different contextualizing apparatus. In Book IV the Letter is thus contextualized differently. The source then--the Letter in Book IV--becomes just another text, and not the object of the philological research in I.5. During the drafting-in-progress, the Letter was, apparently, "taken in any order desired": spat out of I.5 by the force of matter accrued by that fretful eff.
Having examined this bit of text genetically, we can see that in its evolution the text is constantly producing auto-deviations; in its evolution the text differentiates itself from itself and the traces of these deviations deviously come to be re-marked within the text (cf. Milesi, 103-4). The genetic axis, with its up- and downstream motions (cf. Ferrer, 223-5), offers an additional cluster of syntaxes to the manifold syntaxation of Finnegans Wake.
That's the point of eschatology our book of kills reaches for now in soandso many counterpoint words. What can't be coded can be decorded if an ear aye sieze what no eye ere grieved for. Now, the doctrine obtains, we have occasioning cause causing effects and affects occasionally recausing altereffects (FW: 482.33-483.01).Genetic criticism examines the occasioning causes (the pretexts), not in order that the final text (the effect) may be decoded, but instead so that the effect may be reconfigured or recaused into altereffects: effects that always differ from each other. Genetic criticism confirms that there is no stable text, that the book is always already an imposture book. And so, rather than simplify Finnegans Wake with the promise of information, genetic investigation complicates Joyce's work in ways which are potentially instructive to reading but deleterious, and rightly so, to understanding. Finnegans Wake can never be read by someone looking for answers and for proof. Beyond the words in the book in all their superparatactics of play, there is a writing exterior to sensibility that is perpetually "signing the page away" (FW: 111.21), "A way a lone a last a loved a long the" (628.15-6). And so, as much with Dante at the end of the Paradiso and Mallarmé with the disaster of the unwritten Livre, there is nothing there but the withdrawal of a "locative enigma" (135.26).
James S. Atherton, The Books at the Wake, Carbondale: Southern Illinois UP, 1959.
Leo Bersani, The Death of Stéphane Mallarmé, Cambridge: Cambridge UP, 1982.
Maurice Blanchot, Le livre à venir, Paris: Gallimard, 1959, 1990.
Malcom Bowie, "The Question of Un Coup de Dés," Baudelaire, Mallarmé, Valéry: New Essays in Honour of Lloyd Austin, eds. Malcom Bowie, Alison Fairlie and Alison Finch, Cambridge: Cambridge UP, 1982. 142-150.
Rita Copeland, Rhetoric, Hermeneutics, and Translation in the Middle Ages, Cambridge: Cambridge UP, 1991.
Daniel Ferrer, "Clementis's Cap: Retroaction and Persistence in the Genetic Process," trans. Marlena G. Corcoran, Yale French Studies 89 (1996): 223-236.
Almuth Grésillon, Élements de critique génétique, Paris: PUF, 1994.
David Hayman, "James Joyce, Paratactitian," Contemporary Literature
26.2 (1985): 155-78.
--. Re-Forming the Narrative, Ithaca: Cornell UP, 1987.
--. The "Wake" in Transit, Ithaca: Cornell UP, 1990.
Stéphane Mallarmé, ‘uvres complètes,
Henri Mondor and G. Jean-Aubry, Paris: Gallimard, 1945.
--. Correspondances I, eds. Henri Mondor and Jean-Pierre Richard, Paris: Gallimard, 1956.
--. Correspondances II, eds. Henri Mondor and Lloyd James Austin, Paris: Galimard, 1965.
--. Un coup de dés jamais n'abolira le hasard, eds. Mitsou Ronat et al, Paris: Change/d'Atelier, 1980.
Laurent Milesi, "Metaphors of the Quest in Finnegans Wake," "Finnegans Wake": Fifty Years, ed. Geert Lernout, Amsterdam: Rodopi, 1990. 79-107.
Milman Parry, "The traces of the digamma in Ionic and Lesbian," Language: Journal of the Linguistic Society of America 10 (1934): 130-144.
Danis Rose, "Kells-Dublin-Rome-Trieste-Zürich-Paris," A "Finnegans Wake" Circular 2.1 (1986): 1-13.
Jacques Scherer, Le "livre" de Mallarmé, nouvelle édition, Paris: Gallimard, 1957, 1977.
Edward Sullivan, The Book of Kells, London: Studio Press, 1924.