Since the beginning of the twenty-first century, the Joycean world has been in a state of excitement: a number of unknown original documents have surfaced unexpectedly one after the other. In less than six years, close to a thousand pages of mostly very important manuscripts have come on the market. What, we might ask, is the excitement all about? What difference does this make when we already have at our disposal all the documents reproduced in the two beautiful volumes of color facsimile of the Rosenbach Ulysses manuscript and the 63 large volumes of facsimiles of the James Joyce Archive (11 volumes for the early works, 16 volumes for Ulysses, 36 volumes for Finnegans Wake, including the 16 volumes representing the 14,000 pages of the Buffalo notebooks)?
Confronted with this overwhelming bulk, we might even be tempted to ask a more basic question. Why should we waste our time studying Joyce’s manuscripts? We sometimes feel that his books are already more than we can handle. Why should we burden ourselves with the additional tasks of deciphering his often almost unreadable handwriting, charting the maze of his proliferating archive and making sense of the maelstrom of his drafts?
One obvious answer is that if we were afraid of difficulty, if we were not somehow attracted by difficulty, we would not be reading Joyce at all. I once wrote that the point of studying the drafts of Finnegans Wake was not that they would solve the obscurities of the book, but on the contrary that they provided us with an inexhaustible store of supplementary obscurity. This was probably going too far: we have to acknowledge that manuscripts do help us to answer many questions, but at the same time they open up many more enigmas, revealing hidden dimensions of Joyce’s works, and more generally of his work.
First of all it should be noted that some of the books currently accepted in the Joyce canon are actually manuscripts, rather than works. Readers of the trade editions should remember that Epiphanies, Stephen Hero, and Giacomo Joyce were never published by Joyce: as far as he was concerned, they never went beyond the stage of manuscripts. Competent editors have made books out of these manuscripts, and we can read them, to a certain extent, as if they were self-contained literary works, but there is something artificial in the exercise. Stephen Hero in particular is so conspicuously unfinished, it is so clearly related to the genesis of A Portrait of the Artist as a Young Man, that it is difficult to consider that it has an autonomous existence. We spontaneously tend to read it genetically, both as an abortive version of the later novel and as a source of biographic data that help us understand the mind that created the subsequent masterpieces. At the same time we feel that this way of reading does not do justice to an interesting original writing project, but we cannot help the fact that this project was never brought to completion and was superseded by a new project that eventually produced A Portrait.
A few years ago, one of the best experts on Joyce’s manuscripts, Danis Rose, expressed his intention to publish a new book, called Finn’s Hotel, made of the sketches that Joyce wrote in 1923 (“Roderick O’Connor,” “St. Kevin,” “St. Patrick,” “Tristan and Isolde,” “Mamalujo,” and so on) and later incorporated into Finnegans Wake. Rose was never allowed to proceed with what would have been a stimulating if scientifically questionable experiment. Joyce’s writing projects at the time, as far as we can know them, seem to have been quite vague and it is dubious that Joyce ever intended to publish these sketches in a collection, as a kind of late counterpart to Dubliners. If he had chosen to publish them, it is probable that he would have added to their number and it is certain that he would have elaborated them beyond the state represented by the 1923 manuscripts. Such a book could only be an editorial artifact, constructed on a very shaky basis, but it would act as a useful reminder that the Wake was not all ready in Joyce’s head when he started writing, that it only gradually took shape along the 16 years of strenuous work. On the other hand it could be misleading to hypostatize in the form of a book what must have been at most a very transitory intention in the course of the long evolution of the project that came to be known as “Work in Progress” and later resulted in the publication of Finnegans Wake.
Publishing a manuscript as an autonomous work (as opposed to simply editing it as a manuscript) is the dream of many editors, as it is a rewarding experience in many respects, but it is not something that can (or should) often be done. Until the moment of final publication, a manuscript is not a work of art, but an instrument towards the elaboration of a projected work, not a literary text but a protocol for the fabrication of a text. Therefore, its most important feature is not its textual content but the direction(s) it indicates. Of course, if the final text is absent, or incomplete, we are glad to retrieve textual matter from the manuscript and to read it as if it was a text. On a similar basis, using manuscripts to improve the text of published works is a very common practice, sanctioned by a secular tradition. A whole branch of philology, known as “textual criticism,” is devoted to establishing the best possible texts, usually by means of emendation of defective versions. In the case of works that antedate the modern period (biblical, classical or medieval), this can only be done by conjecture or by comparing different scribal copies. But when authorial manuscripts are available, they are considered the best expression of the author’s intention and as such the ultimate recourse. Let us take a few examples from Ulysses.
In the first edition and in subsequent editions until 1984, the following sentence was to be found in the “Sirens” episode: “Miss Douce huffed and snorted down her nostrils that quivered imperthnthn like a shout in quest.” The ending of this sentence is certainly puzzling, but we are used to being puzzled by the stylistic audacities of this episode. Readers certainly wondered what a “shout in quest” might be, but they took comfort from the fact that on the next line but one, Miss Douce is indeed “shouting.” Or rather this is what they did before Hans Walter Gabler’s edition showed that the last words should read: “imperthnthn like a snout in quest” (U 11 144–6). Looking at the typescript confirms that this is indeed what Joyce had written, but the compositor of the first galley proofs had misread an “n” for an “h” (the typed letters are not very well defined and his eyes may have been attracted by the “shouting” below). The new reading makes very good sense as it confirms that the sentence contains an echo of another passage, earlier in the episode: “Imperthnthn thnthnthn, bootssnout sniffed rudely” (U 11 100).1 We might wonder whether the manuscript was really necessary in this case. The “snout/shout” substitution could have been conjectured by a good editor on the grounds that “shout in quest” is plainly absurd and that snout is called for by the echo of the earlier passage. But it is very dangerous to mend a text like this one on the basis of conjecture. This is what the proofreader of the 1926 edition tried to do in our next example, also from “Sirens.” The first edition (1922) had this passage:
—…. ray of hope….
Beaming. Lydia for Lidwell squeak scarcely hear so ladylike the muse unsqueaked a ray of hopk.
The 1926 edition, followed by all the subsequent ones until 1984, normalized the last word to “hope,” on the assumption that “hopk” was a glaring typo. A look at the archive definitely confirms that it was “hopk” that Joyce wrote here and that the correction was indeed a corruption, but in order to understand what happened, we have to follow the development and radicalization of this segment.
It all begins with two sentences inscribed in the margin of the second draft of the episode: “A ray of hope is beaming. Lydia for Lidwell unsqueaked a cork.”2 We can see that the first portmanteau in the sentence, “unsqueaked,” expressive of the noise produced by the unscrewing of the cork, had already been coined at this early stage. On the fair copy,3 Joyce changed this to
—….ray of hope….
Beaming. Lydia for Lidwell squeak scarcely hear so ladylike the muse unsqueaked a cork.
In this new context, “unsqueak” acquires a different meaning (a suppression or attenuation of the squeaking) and the pseudo-genteel affectation (“so ladylike the muse”’ actually draws our attention to the sexual connotations of the unscrewing noise. (U 15 1975; “Man and woman, love, what is it? A cork and bottle.”) This version also emphasizes the interference between the song in the background and the narrative with the suppression of the word “beaming” from the italicized quotation and its integration in the authorial narration.
It was only on page proofs that Joyce all but replaced the material cork with an echo from the song in the preceding line, creating at the same time an onomatopoeic effect, imitative of the sound of the popping cork (“unsqueaked a ray of hopk”). This new portmanteau is much more disturbing than the first (hence the proofreader’s feeling that it had to be a mistake), not a well-formed hybrid but a linguistic chimera. The final result is an amazingly rich sentence, counterpointing noises and music; spoken, sung or imagined voices; pretentious clichés and salacious innuendos. Thanks to the manuscripts we are able to follow the entry of the voices and the gradual densification of the polyphony.
Our next two examples, from “Circe,” will confirm that the traditional project of textual criticism of using the manuscripts in order to establish the text is often an impossible task. On the contrary, a genetic investigation of the archives of writing tends to have a destabilizing effect on the text. In every edition until 1984, this was the beginning of the “Circe” episode:
(The Mabbot street entrance of nighttown, […] Rows of flimsy houses with gaping doors. […] Round Rabaiotti’s halted ice gondola stunted men and women squabble. They grab wafers between which are wedged lumps of coal and copper snow […].)
Gabler remarked that Joyce had actually written “grimy houses” instead of “flimsy houses” and “coral and copper snow” instead of “coal and copper snow”: the typist had misread the very difficult handwriting of the fair copy and the corrupted words persisted in later versions. The 1984 edition (and all subsequent critically established editions) correct the mistakes and restore Joyce’s original intention. They are probably right to do so – and yet some of us cannot help regretting the version of earlier editions. The flimsiness of the houses seemed particularly appropriate to the unreal atmosphere of the episode, suggesting the painted canvas of a stage setting. The grime is of course important too, but some of its associations are picked up and even reinforced by the coal color of the snow eaten by the dwarves in the same version. So the corrupted text could be considered to be, in some respects, superior to the version of the fair copy.4 We cannot reject such considerations as purely subjective, as they may condition the way we interpret what happened after the fair-copy stage. We could wonder why Joyce never caught these mistakes on the typescript, the galley proofs, or any of the page proofs. There can be no question of flagging attention as this is the first paragraph of the episode. Joyce did look carefully, as he added a word and changed some punctuation. Could he have endorsed the changes, consciously or unconsciously? We can never answer such a question with any certainty. Eugene Jolas, who often worked with Joyce on the proofs of Finnegans Wake, confirms that it was not unusual for him to do so.5 (The French translation cannot help us in this case: the translators did not find anything the matter with this version so they had no cause to ask Joyce.) An editor, however, has to make a decision. There is on the one hand the certainty that, up to the fair copy, Joyce intended to have “grimy” and “coral,” and on the other the possibility that Joyce changed his mind, inspired by the mistake, and the fact that all the (numerous) subsequent changes in the episode were made in the context of a first paragraph that included the words “flimsy” and “coal.”
Changes of intention are an editor’s nightmare, as they often result in contradictions and dilemmas. On the contrary, they are the very substance of genetic criticism. The difference of perspective is particularly clear in the case of an addition to the proofs of “Circe” that did not make it into the published text and yet sounds strangely familiar to the reader of Ulysses. Towards the end of the messianic scene (U 15 1913), Joyce had intended to insert the following:
Don Emile Patrizio Franz Rupert Pope Hennessy (in medieval hauberk, two wild geese volant on his helm, appears and, with noble indignation, disowns Bloom) Put down your eyes to footboden, big grand pig of Jude all covered with gravy! (Buffalo V.C.5b, JJA 26 175)
This genetically fascinating passage represents a culmination of Joyce’s expansive creative energies. It has often been said that Joyce’s mode of writing is mostly accretive: he adds much more than he deletes. After all, Ulysses was first conceived as a short story and progressively became the huge novel that we know. The expansive process was at its height in 1921 and it is particularly in evidence in the messianic scene of “Circe.” By way of additions on manuscript and typescript, and several pullings of galley and page proofs, Joyce transformed two lines of dialogue into a 20-page scene. There is no way of knowing if this would ever have stopped or how much bigger Ulysses could have grown if Joyce had not set himself a final deadline: he wanted his novel to be published for his 40th birthday on February 2, 1922. By mid-December 1921 there was still half of “Circe” and most of the three last chapters to finalize, so it was necessary to make an end to this passage: Joyce sent Darantière, his French printer, the signed bon à tirer, the passed-for-press form authorizing its final printing. A week later, however, he sent him a duplicate of the last set of proofs with new additions (including the Don Patrizio passage) and, on top of the first page, the mention “Corrections supplémentaires si encore possibles James Joyce” (“Supplementary corrections if still possible James Joyce”). This sentence is important in several respects. It illustrates Joyce’s creative energy, the intensity of his will to write, of his desire to improve his work, to go on supplementing until the last moment (actually beyond the last moment),6 to keep open the possibilities of writing and to postpone the closure inherent to the written text as a finished product. It is almost as if he had signed the death sentence of his own work as a living creation and was now begging for a reprieve. But the indication is also fascinating because of its conditionality. It shows that a writer’s purpose is not an absolute, but a fluctuating, time-bound transaction between a series of writing events and a series of external constraints: the text of a work cannot be defined with any certainty by the fiction of a unified authorial intention.
The further adventures of this passage prove it even better. As was to be expected, the supplementary proofs arrived too late and were never used by Darantière. Joyce was understandably reluctant to waste the amusing apparition of this composite figure of the Irish exiles. He therefore incorporated it, with a few variations, in the set of proofs for the end of the same chapter that he had at hand at the moment (JJA 26 316). The editorial problem raised by this writing event has no really satisfactory solution. The 1984 edition, answering Joyce’s fervent prayer, incorporated the addition where he had wanted it to go, in the messianic scene (U 15 1914–17). But this reinsertion created an interesting but unintended echo with the very similar passage at the end of the chapter (U 15 4506–9). The 1986 edition (and all subsequent editions) wisely considered that the mid-January 1922 insertion canceled the indication on the mid-December 1921 proofs so that only the second passage was kept (U 15 4506–9). Another, drastic, solution would have been to reinstate the passage in the “Messianic Scene” and to excise the other one, on the grounds that it was only inserted at the end of the chapter as a second-best choice. This would not be a worse compromise than the other solutions, except that we would loose the changes introduced in the second incarnation of the piece.
This is how it was finally incorporated:
DON EMILE PATRIZIO FRANZ RUPERT POPE HENNESSY (in medieval hauberk, two wild geese volant on his helm, with noble indignation points a mailed hand against the privates) Werf those eykes to footboden, big grand porcos of johnyellows todos covered of gravy!
Some of the changes (“points a mailed hand against the privates” instead of “disowns Bloom,” “johnyellows” instead of “Jude”) are directly related to the new location: this archetypal incarnation of the Irish wild geese, who was adding his own very idiosyncratic voice to the general reprobation of Bloom at the end of the messianic scene, is now involved in the confrontation between Stephen and the British soldiers. The anti-Semitic outburst is thus converted, with the necessary adjustments, into anti-British insults. It is a process of interaction or transaction, for the contextual influence is reciprocal. The new environment is modified by the insertion of the element from the earlier background: the rather depressive end of the chapter acquires, through this insertion, something of the sheer exuberance characteristic of the messianic scene.
Retracing the genesis of a passage like this does not provide us with easy editorial solutions, but it is productive in different ways. It can affect our reading of the published text by strengthening some of our interpretations or helping us to develop new ones. For instance the fact that the Don Patrizio passage could be made to fit in two different contexts suggests an equivalence between those contexts, a similarity, also supported by the “Cyclops” episode, between Irish anti-British xenophobia and anti-Semitism. It draws a parallel between the ordeals of Bloom, excluded as a foreign element even by the arch-cosmopolitan Don Emile Patrizio Franz Rupert Pope Hennessy, and of Stephen, integrated by force in a community of hatred from which he cannot escape.
Apart from its value as an auxiliary for the interpretation of the text, the archive of writing reveals the dynamics of invention. The very fact that the insertion was suspended for a couple of weeks has induced an evolution. The passage is rewritten in accordance with a new temporal context (as opposed to the change of spatial context we have just described), reflecting at least a slight change of intention and perhaps even a stylistic evolution. The language of the cosmopolitan exile becomes even more international (“Werf those eykes to footboden,” “big grand porcos”) and seems to have moved one step forward towards Finnegans Wake.
By studying the traces of writing we are able to identify micro-contexts such as those defined by the one-week interval between the signing of the passed-for-press form and the change of mind that prompted new additions, or the four-week interval between the two versions of the insertion (or even the few moments that separate the writing of a word and its crossing out). We can then observe interferences between those micro-contexts and relate them to the macro-context of the general evolution of Joyce’s work (“a book that in some aspects began as a sequel to A Portrait of the Artist as a Young Man ended as a prelude to Finnegans Wake”; Groden 1977: 13) as well as to the circumstances of Joyce’s biography and the general political events of the time. There is, however, a very important intermediary contextual level represented by the sharply distinct phases of writing characteristic of Ulysses.
Any work of art bears the scars of the process of its making, but these scars are usually almost invisible on the smooth surface of the finished product. In the case of Ulysses, they are quite plain: anyone can see that there is very little in common between, for instance, the “Telemachus” and “Calypso” episodes and “Cyclops” or “Ithaca.” The difference is so great that it can only be explained by radical changes in the writing project and any interpretation of the book that does not take into account the temporal dimension of its production is bound to run up against insurmountable difficulties. Michael Groden has shown decisively that Ulysses as we know it is the result of the layering (layering, not juxtaposition) of three large temporal contexts, corresponding to what he calls the early, middle and last stages (Groden 1977: 13). The early phase corresponds to what Joyce himself called the “initial style,” a mixture, in various proportions, of monologues and third-person narration. The recent manuscript discoveries suggest that this phase was itself probably more complex than we imagined and that the “initial style” was actually the result of an evolution that took place around the year 1917. The middle stage represents a shift away from realism and an emphasis on parodies and stylistic pyrotechnics. The late stage exacerbates some of the earlier tendencies towards symbolism and schematism and the book seems to fold upon itself and to feed upon its own substance.
Although the three styles correspond primarily to the early, middle, and last chapters of the book, they are not merely juxtaposed in a linear sequence. Not only were the later chapters composed in continuity with the early ones – the inertia of what is already written is a condition of any writing – but one of the major aspects of the late style of Ulysses is precisely the rewriting of previous episodes: this is something that any reader of “Circe” will notice and that can also be detected in “Ithaca” and even in “Penelope.” Studying the history of composition allows us to specify this process and to understand it as bidirectional. Because of the publication deadline, the printing schedule had been arranged in such a way that Joyce was writing the last episodes at the same time as he was correcting the proofs of the earlier ones. As a consequence, the (extensive) correction of the proofs for “Telemachus,” “Nestor,” “Proteus,” “Calypso,” and “Lotus Eaters” belong to the chronological context of the writing of “Ithaca” and “Penelope.” This means that the early episodes, some of which had been drafted many years before, were actively present on Joyce’s writing table as the end of the book was being written, but also that the revisions of the first part were introduced from the perspective of the late style. We can see also that during a single week of October 1921 Joyce marked proof for “Lotus Eaters,” “Hades,” “Aeolus,” “Lestrygonians,” “Scylla and Charybdis,” “Wandering Rocks” and “Sirens”: all the modifications entered at this moment in this wide array of episodes should be studied in the context of each other.
The counterpart of this temporal contextualization of spatially dispersed passages is the superimposition of temporal layers in a single location. We have seen an illustration of this with the “ray of hopk” passage of “Sirens,” with its alterations reflecting a progressively radicalized treatment of language: the layers of revision coexist, as it were, in a single compound word (another anticipation of Finnegans Wake). On a larger scale, the “Aeolus” episode, originally perfectly representative of the initial style, bears the visible traces of its revision in the context of the late style: Joyce did not alter deeply the texture of the episode, but completely changed its external form by injecting the startling newspaper headlines in accordance with the radical modernist aesthetics of the later chapters.
Along the 16 years of the writing of Finnegans Wake, major inflections modified the original project, but the more drastic changes were structural reorganizations and the stylistic alterations were much more gradual, so that we cannot assign major “styles” as convincingly as we do in Ulysses, and given the nature of the published book it is not so easy to detect in it the traces of the process of evolution. However, if we take a bird’s-eye view of the genesis of two of the more accessible passages in the book, the “Anna Livia Plurabelle” chapter (I.8) and the soliloquy of Anna Livia at the end of the book, we can make a few useful general remarks. “Anna Livia Plurabelle” was first drafted in 1924 and was revised at least 17 times during 14 years.7 The main characteristics of the very first draft are still visible in the final text: the oral quality of the exchange between the two washerwomen, the fluidity of the rhythm, the humor. But many other features were added by the successive revisions, reflecting different visions of the work as it took shape between 1924 and 1928. The multi-lingual aspect, absent in the first layer, was introduced at different stages, beginning with an introduction of Scandinavian elements and northern dialects. The number of colloquialisms was then increased and the sound patterns were made more complex. But the most conspicuous feature is the addition of river names: for instance “Go on, go on,” is replaced by “Garonne, Garonne” (FW 205 25). This began with a few scattered allusions, but became a major preoccupation of Joyce, who, for years, harvested hundreds of river names in his notebooks in order to weave them into his sentences. Most of these names are unknown to the majority of readers, so that the allusions cannot be detected without the help of a comprehensive encyclopedia or, even better, of Joyce’s notebooks. It has often been pointed out that these elusive allusions obscure the original meaning of the sentences. Obviously, direct accessibility had ceased to be a concern. On the other hand, the final soliloquy (FW 619–28) was drafted, copied, typed five times and further revised on galley proofs within a few months at the end of 1938. In the process, the passage was considerably expanded – the two last short sentences (“I only hope the heavens sees us. Abit beside the bush and then a walk along the”) were expanded to 87 lines in the printed book (FW 625 36–628 16) – but no spectacular change of direction is taken in this last and concentrated bout of writing, only a development and a progressive intensification. By this time, the stylistic evolution of Finnegans Wake had run its full course. We can even detect a certain regression towards a greater transparency, due perhaps to the fact that the final monologue of Anna Livia was a paradigmatic equivalent of Molly Bloom’s soliloquy, so that the finishing of one book brought memories of the completion of another and the end of Ulysses formed a context for the ending of the Wake.
The layering of successive levels of intentionality can be observed in any work that is produced in time (that is to say, everywhere except perhaps in very short extemporizations), but it is particularly accessible to a genetic approach within the framework of Joyce’s aesthetics of saturation (as opposed, for instance, to an aesthetic model based on deprivation, like Beckett’s). In Joyce, meanings pile up without canceling one another on the model of the portmanteau word (the “ray of hope” does not obliterate the underlying “cork,” “go on” is still narratively and rhythmically perceptible beneath the surface of “Garonne”). Using the manuscripts to follow successive stages of writing produces an effect of ex-plication, a gradual unfolding of the packed riches. And yet it poses specific difficulties, linked with the paradoxes of the notion of overdetermination. The multiplication of levels in a work like Ulysses implies that each element is heavily overdetermined: it is determined by its narrative import, by its various meanings, literal and symbolic, by its intertextual references, and by the tight web of symmetries and echoes that binds it to other parts of the book. It is impossible, however, logically and empirically, that these multiple levels should have been conceived at the same time. We hardly need the confirmation of the manuscripts to know that the infinitely complex structures of Ulysses and Finnegans Wake cannot have sprung fully armed from Joyce’s brains (although Joyce managed to persuade some of his admirers that this had been the case). This implies that the majority of the determinations were discovered retrospectively rather than projected, so that they paradoxically reflect a basic indeterminacy. An open attitude towards the potentialities of writing is required on the part of the critic (reflecting Joyce’s own disposition) to interpret genetically the mechanics of meaning.
Our first example, admittedly an exceptional one, takes us back to the late “Circe” proofs. In the Black Mass scene, just before the intervention of Adonai calling “Gooooooooooood!”, Joyce made the following insertion:
The voice of all the blessed
Alleluia, for the Lord God Omnipotent reigneth!
(JJA 26 305)
What happened afterwards demonstrates in the most graphic way the inertia of the signifier and the dynamics of writing. Because of the poor quality of the paper on which the proofs were printed, Joyce’s corrections, marked in ink, seeped through the page, so that they are clearly readable, in inverted mirror form, on the verso (JJA 26 306). This material accident, in the immediate context of the Black Mass (traditionally supposed to be celebrated as a literal inversion of the Holy Mass) suggested the inversion of the letters of the Alleluia and triggered the following expansion of the scene on the next proofs:
THE VOICE OF ALL THE DAMNED
Htengier Tnetopinmo Dog Drol eht rof, Aiulella!
(From on high the voice of Adonai calls)
ADONAI
Dooooooooooog!
(JJA 26 322–33)
The published version of the scene looks so integrated that it seems difficult to believe that its elements belong to different strata of writing and that its present form is the result of a material accident. Is it possible, for instance, that Adonai’s “Goooooooooood” was not meant from the start to be an inverted echo of “Dooooooooooog”? Manuscripts show that the prolonged call was inscribed months before the writing accident. 8 This proves beyond any doubt that the canine echo, which is probably the most conspicuous element in the scene, is an afterthought, but this does not weaken the relationship established a posteriori between the two elements. Other retrospective links are automatically established with the Protean dog roaming the episode, with Bloom’s “Dog of a Christian” addressed to the Man in the Macintosh during the messianic scene, with the mysterious answer (echo or dog whistle?) to Mulligan’s call in the course of his initial blasphemous Mass, and so on.
The productivity of the accident does not stop there. A few days later, on another batch of proofs, Joyce added the words “from right to left” to the description of the apparition of Bloom’s son Rudy reading a book. This links the inversion of letters to Hebrew script and connects the Black Mass to a network of pre-existent motifs such as the Kabbalah.
The potential for the diffusion of effects is even greater in Finnegans Wake as the next example demonstrates. It begins with a series of notes taken by Joyce in 1924 in notebook VI.B.14 about a particular feature of Celtic historical linguistics, the p/k (or p/q) split. It divides Celtic languages in two groups: Brythonic languages (such as Breton) admit the sound p where Goidelic languages (such as Irish) have the sound k. A consequence of this is pointed out in a note on page 218 of this notebook (“pascha I. casca/purpura Kurkura”), perplexing but clear enough when confronted with its source in a book by Stefan Czarnowski:9
Dans les premiers temps les Irlandais remplaçaient le son p par k (écrit c) dans les mots empruntés, par exemple purpura devint corcur, pascha – casc [At first the Irish replaced the sound p by k (written as c) in words they had borrowed, for example purpura became corcur, pascha – casc] (Czarnowski 1919: 32 n. 2).
This national mispronunciation is a pattern of linguistic deformation that Joyce would exploit more than once in Finnegans Wake. But its meaning was enriched in the following year by another discovery. As Joyce was reading Freud’s Collected Papers, he took note of an infantile phrase used by one of the young patients “do no 1” (from “Hans: ‘Oh, I’ll come up again in the morning to have breakfast and do number one,’ ” in “Analysis of a phobia in a five-year-old,” Freud 1925: 160). Number one and number two are of course nursery euphemisms for urinating and defecating. This is taken up again further in the same notebook: “no 1 or no 2.” Then, on top of the next page, we find the letters “pp or kk” followed by “keykey ahah.” Passing from an English to a French version of the excremental functions (in French pipi et caca are precisely “number one and number two,” with roughly similar childish connotations), Joyce starts playing with the sounds, relishing perhaps this coincidental historicoscatological homophony.
The game is continued on a page of a draft (47482a-23v, JJA 60 44) that is exactly contemporary with this notebook. Apparently, Joyce had written first:
earned ^+forecast+^ by Cain, outflanked by Ham, reordered by Patrick, delivered ^+evaded+^ by Tristan, by Patrick’s dear.10
He then takes up the Irish “kurkura” of notebook VI.B.14 to coin the amusing “kurkle katches,” which gave him the key for a transformation of the passage, prompting a series of additions, substitutions, and overwritings:
^+kurkle katches+^ forecast by /C → K/ain, outflanked by /H → K/am, reordered ^+inklored+^ by Patrick ^+Paw Kawdreg+^, evaded ysold by /T → K/ris/t → k/an ^+Kriskan+^, by Patrick’s dear, by Karnell overagain.
The p/k substitution is at work on the name of Irish heroes (Partrick is transformed into Kawdreg and, more surprisingly, Parnell into Karnell, with a hint of the carnality that caused his downfall), spreads to other words (“inklored,” suggesting traditions penned in ink), and seems to be getting out of hand, perhaps due to its newly discovered excremental connotation, with a surplus, an overflow of k (of caca), encroaching over other letters (Kam and Kriskan!?).
Although this passage was never incorporated in Finnegans Wake, we find two occurrences of “kurkle” (FW 95 22 and 296 13). In the first appearance in particular (“Fine feelplay we had of it mid the kissabetts frisking in the kool kurkle dusk of the lushiness”) the excremental connotation is clearly present (pissabed). But we should consider that it is also present in each of the several instances of p/k inversion in Finnegans Wake. Taking this a step further, we become aware that no p or k in the book remains unaffected: each k could be a p and a k may be hiding behind each p and in the shadow of each of these letters lurks a marginal allusion to Celtic philology and to infantile scatology.
Of course Joyce did not write every p and every k with this association in mind, but this is not the point. Joyce knew that the signifying engine he had made up worked autonomously like the printing press gone wild fantasized by Bloom in “Aeolus” (U 7 102–4). It is interesting to compare a remark by Frank Budgen describing Joyce in the early days of writing Ulysses (“Joyce … was a great believer in his luck. What he needed would come to him. That which he collected would prove useful in its time and place”; Budgen 1972: 173–4) and a note inscribed in 1924 in notebook VI.B.14 (p. 32): “JJ no gambler because/his style gambles/infinitely probable.” By this time Joyce knows that he does not need to gamble because his style does it for him, and it cannot lose because its bets saturate the gambling table. Each word, each letter, opens up an indeterminate number of meanings, or, to put it differently, produces retrospectively a multitude of determinations.
Discovering this has a sobering and a liberating effect on the reader. On the one hand, manuscripts reveal a number of meanings and references that are almost undetectable in the published text but that determined its present state. Knowing these opens up new depths of understanding, adds new dimensions to the text. However, the recent discovery of long-lost manuscripts proves that the documents we have are far from complete, and even if we had everything that Joyce ever wrote, it would only give us an incomplete picture of the writing process: we cannot hope to ever recover everything that Joyce intended. On the other hand, manuscripts also reveal that the productivity of the text surpasses Joyce’s intentions at any particular moment. Authorial intention cannot be the last word on the text: it is something that we should not neglect, something that we should pursue through its fluctuations and nuances by a rigorous interpretation of the extant documents, but it cannot be a limitation of our powers of interpretation.
NOTES
1 This echo, by the way, was probably spotted by the translators of the 1929 French version, who produced a translation (“comme un groin qui furète”) which is in this case more authoritative than the original first edition: translators are by nature close readers and these were lucky enough to have Joyce nearby, so that they had the possibility of asking him what he had really meant to say (Joyce 1929).
2 NLI MS 36,639/9, p. [10r].
3 The fair copy of this episode has been lost, but it can be reconstructed with the help of the typescript and the Rosenbach manuscript.
4 This passage has been discussed more than once. Some Joyceans think on the contrary that the “corrected” reading is superior to the previous version. Vicki Mahaffey, in her pioneering essay on intentional error, considers that “coal and copper snow” is a mere corruption that is detrimental to our understanding of the passage and its relation to other parts of the book.
5 ”Joyce would improvise whenever something particularly interesting occurred to him during the reading, and occasionally even allow a coquille – a typographical error – to stand, if it seemed to satisfy his encyclopedic mind, or appeal to his sense of the grotesque hazard” (Jolas 1998: 11).
6 Again, Jolas shows us that Joyce’s attitude had not changed ten years later, although he was working under very different affective and intellectual circumstances: “Once, I remember, I had just finished reading the last page of final proof and had given orders to go ahead with the printing…. As we settled back comfortably, the telephone rang. An excited voice announced that a heavy special-delivery letter from Paris had just arrived for us at the printshop. Joyce wanted to make further additions, one of them probably the longest he had yet invented: an onomatopoeia of over fifty letters expressing collective coughing in a church during a sermon. It was included” (Jolas 1998: 11).
7 This proliferation of version attracted early attention. Even before the manuscripts became available, critics such as Leon Edel and Edmund Wilson discussed the evolution of the chapter based on published versions. It was later studied by Fred H. Higginson, A. Walton Litz (1961), and Claude Jacquet (1985), among others.
8 See Rosenbach manuscript p. 79a.
9 The source was discovered by Geert Lernout. See notebook VI.B.14 (Deane, Ferrer, and Lernout 2002).
10 A simplified transcription system is used here: deletion, ^+addition+^, /A → B/ (B written over A).
BIBLIOGRAPHY
Budgen, F. (1972) James Joyce and the Making of “Ulysses.” London: Oxford University Press.
Crispi, L. and S. Slote (eds.) (2007) How Joyce Wrote Finnegans Wake. A Chapter-by-Chapter Genetic Guide. Madison: University of Wisconsin Press.
Czarnowski, Stefan Zygmunt (1919) Le Culte des héros et ses conditions sociales: Saint Patrick, héros national de l’Irlande. Paris: Félix Alcan.
Deane, V., D. Ferrer, and G. Lernout (eds.) (2001 – 4) James Joyce, The Finnegans Wake Notebooks at Buffalo, vols 1, 3, 5, 6, 10, 14, 16, 25, 29, 32, 33, and 47. Turnhout, Belgium: Brepols.
Deppman, J., D. Ferrer, and M. Groden (eds.) (2004) Genetic Criticism: Texts and Avant-Textes. Philadelphia: Pennsylvania University Press.
Ferrer, D. and C. Jacquet (eds.) (1998) Writing Its Own Wrunes for Ever, Essais de génétique joycenne/ Essays in Joycean Genetics. Tusson: Éditions du Lérot.
Freud, Sigmund (1925) Collected Papers, Vol. 3. London: The Hogarth Press.
Gabler, H.-W. (1988) “Joyce’s text in progress.” In Texte 7, “Ecriture – Réécriture – La Genèse du Texte,” pp. 227–47; also in D. Attridge (ed.) The Cambridge Companion to James Joyce, 1st edn., pp. 213–36. Cambridge: Cambridge University Press.
Gabler, H.-W. (2005) “The rocky road to Ulysses,” Joyce Studies 2004, 15, Dublin: National Library of Ireland.
Groden, M. (1977) Ulysses in Progress. Princeton: Princeton University Press.
Groden, M. (1980) James Joyce’s Manuscripts: An Index. New York and London: Garland.
Groden, M. (2007) “Joyce at work on Cyclops’: toward a biography of Ulysses.” James Joyce Quarterly 44 (2).
Herring, P. F. (ed.) (1972) Joyce’s “Ulysses” Notesheets in the British Museum. Charlottesville: University Press of Virginia.
Herring, P. F. (ed.) (1977) Joyce’s Notes and Early Drafts for “Ulysses.” Charlottesville: University Press of Virginia.
Jacquet, C. (ed.) (1985) Genèses de Babel: James Joyce et la création de Finnegans Wake. Paris: Editions du CNRS.
Jolas, E. (1998) “Remembering James Joyce,” Modernism/Modernity 5 (2): 2–29.
Joyce, J. (1929) Ulysse, traduit de l’anglais par M. Auguste Morel, assisté par M. Stuart Gilbert. Traduction entièrement revue par M. Valery Larbaud, avec la collaboration de l’auteur. Paris: La Maison des amis du livre.
Joyce, J. (1975) Ulysses by James Joyce: A Facsimile of the Manuscript, ed. C. Driver. New York: Octagon Books, and Philadelphia: Philip H. and A. S. W. Rosenbach Foundation.
Joyce, J. (1984) Ulysses: A Critical and Synoptic Edition, ed. H. W. Gabler with W. Steppe and C. Melchior. New York: Garland.
Litz, A. W. (1961) The Art of James Joyce. New York: Oxford University Press.
Hayman, D. (1963) A First-Draft Version of “Finnegans Wake.” Austin: University of Texas Press.
Hayman, D. (1990) The “Wake” in Transit. Ithaca, NY: Cornell University Press.
Hayman, D. and S. Slote (eds.) (1995) Probes: Genetic Studies in Joyce. Amsterdam: Rodopi.
Mahaffey, V. (1991) “Intentional error: the paradox of editing Joyce’s Ulysses.” In G. Bornstein (ed.) Representing Modernist Texts: Editing as Interpretation, pp. 171–91. Ann Arbor: Michigan.
Rose, D. (1995) The Textual Diaries of James Joyce. Dublin: Lilliput Press.
Rose, D. and J. O’Hanlon (1989) The Lost Notebook. Edinburgh: Split Pea Press.
Scholes R. and R. M. Kain (1965) The Workshop of Daedalus. Evanston, IL: Northwestern University Press.