It is refreshing to read an essay in the first issue of Papers on Joyce in which J. Carnero González introduces an idea that is "specific and exact" (Carnero González, 21). Although he later qualifies his statement, writing that it "remains imperfect and inconclusive," he announces: "What follows, therefore, is a hypothesis based on evidence of publication and a proposal for reading Finnegans Wake in its full light." González's proposal is genetic in nature: after referring to Joyce's statements about his work on the Wake being like "boring through a mountain from two sides," he claims that the word "wordspiderweb" describes both the method of composition and the resulting book: "the vivid image of such a creative process is that of a spider's web in the making." There is method in the apparent chaos of the Wake's publication history.
Like a spider, Joyce begins in the middle of the mountain. At its centre is "King Roderick O'Conor," the first text Joyce wrote after finishing Ulysses. The day on which Joyce began to write again, 10 March 1923, is the absent (because unpublished) "hypothetical centre" of his mountain. Joyce then moves to the mountain's right side, "with the completion of Chapter II.iv (383-399), after which a move to the mountain's left side results in the publication of all of Book I, although not in the order in which it would ultimately appear in 1939." Book I must again be divided into two parts and then we see that Joyce first published the right-hand side, though again not in the order in which the pieces appear in the final text of 1939. Joyce first published the first and the last, I.v and I.viii. Then "[i]t would seem that, having plumbed and anchored the pillars closest to the nuclear centre, Joyce was then able to move outward from the core and to publish, from April to June of 1927, the first four chapters of what would be Finnegans Wake (I.i-iv, 003-103), concluding, as a culminating flourish, with I.vi in September of this same year." In 1928 Joyce composes a fragment between I.viii and the germinal "Roderick O'Conor" (282-304) and then he moves immediately to the right side of the mountain, publishing, from left to right, the four chapters of Book III. Joyce then wrote II.i and two passages of II.ii. "Finally, two fragments of II.iii (309-331 and 337-355) continue the approximation, within the nuclear centre, to the two initial pages of March 1923."
I have quoted the description of Joyce's methods in some detail because González makes large claims for them: "Joyce himself sought in [the] serial publication to reveal bit by bit the narrative thread and inhering plot of his work in progress." González proposes "that Finnegans Wake be read as Joyce's contemporaries read it over the years, following the sequence of first publications, yet privileging their definitive versions appearing in 1939." If we proceed thus, we observe three stages, the first offering a general view, the second and third a progressive approximation. These stages can be distinguished by means of the date of publication: the first from April 1924 to fall/winter 1925-1926, the second between April and September 1927 and the third from March 1928 to July 1935. Each stage also consists of five steps.
The practical and theoretical implications of the proposal to read Finnegans Wake through Work in Progress do not concern me here. I believe, with the author, "that something, if not much indeed is to be gained by reading Finnegans Wake in the full light of Work in Progress's serial publication." The only problem is that there are serious difficulties with González's presentation of the facts concerning the Wake's genesis.
A first point is the choice of published fragments over first versions. In a footnote González considerably clouds the issue by writing first that he refers to the first moment a chapter or fragment was published and then:
Any student and admirer of Joyce knows all too well that the writer's first versions differ considerably from definitive ones (if indeed a definitive version ever existed). Nevertheless, we must consider that the first public printing of a passage, fragment, or chapter holds its own legitimate value from the very moment nuclear elements are found within it and the author sends it to be printed.Of course González cannot mean that the texts he collectively calls Work in Progress are "first versions." A look at the Finnegans Wake volumes in the Archive or at the more accessible A First-Draft version of "Finnegans Wake" by David Hayman should be enough: there is as much difference between the first versions and those published serially as between these latter and the final text published in 1939.
In the qualifying second sentence the author gives a reason for the legitimacy of the first printed versions that is not very convincing. What does the phrase "nuclear elements" mean in this (or any other) context? Are there first versions which do not contain nuclear elements of final versions? In genetic matters there seem to me two different ways of approaching the problem, a progressive and a regressive one. In a regressive genetic analysis we read earlier versions in the often blinding light of the final text. I have elsewhere argued the limitations of this strategy if it is our aim to represent the growth of the text, but it is here that González's argument makes more sense. Most readers of Finnegans Wake are regressive in a wider meaning of the term: for them the final text is enough. Their arguments are those of literary criticism against what used to be called "bibliography" and they have been well-known since at least New Criticism. It is only for these readers that González's argument holds. These texts have at least a certain legitimacy, because they share in the finality of the 1939 final version by having been sent to the printer by Joyce himself.
Progressive readers look at the genesis of Finnegans Wake from the other end: they start at the beginning and they try to establish how the text came into existence. For them, it is obvious that early versions of the text are relevant. Radical exponents will even argue that since Joyce's work-in-progress is a machine continuously producing new versions, there is no basis for privileging the final printed text: all versions are created equal.
My own position is not that extreme: I believe Joyce made it quite clear that the text that was published on 2 February 1939 was the final version, and that all previous versions only existed as that text's pre-texts. In the genetic history, the earlier printed versions were treated in a manner by no means different from other drafts: they were stations along the way, not in any way destinations in their own right. At the same time, Joyce did welcome having to hand his different publishers copy, because it enabled him to work under pressure and he had shown in the final stages of his work on Ulysses that pressure was something he needed in order to concentrate his powers.
So let us assume for a moment that these preliminary publications were as important as González claims, and even that the order in which they appeared tells us something about the nature of the final work. A first problem then becomes the absent centre, "King Roderick O'Conor," which occupies an empty slot. This brief text is a fine example of an early Wake text. There is as yet little morphological distortion, the syntax remains simple, the story (as far as there is any) is simple. Late at night King Roderick O'Conor finishes the drinks left on the table by the departed honourable guests. Apart from the fact that this is, if anything, a story of closure, little is clear about the text's significance. We only know that it took Joyce nearly fifteen years of concentrated work before he incorporated the text into the final version of Finnegans Wake.
The second text is problematic too: "Mamalujo" was published in the first issue of the Transatlantic Review in April 1924, but it is highly misleading to claim that this represents "the completion of Chapter II.iv." In reality the published text is only a fragment of what would become later, and much much later, the last chapter of Book II. The same is true for the next text published: the opening pages of the second chapter of Book I. The first stage of five texts is therefore a much more haphazard bunch than González seems to suggest, especially if we realize that Joyce seems to have gone through some type of crisis about the whole structure of his new work even before he published the first excerpt. We do not have accept Danis Rose's hypothesis that at the end of 1923 Joyce abandoned a collection of short stories for work on a grander project to realize that something did take place at the time. When Joyce gave "Mamalujo" to Ford Maddox Ford for Transatlantic Review, he described it as "a sidepiece."
The next "stage" is much more straightforward: the next five parts of "Work in Progress" were published in transition. Joyce published all of Book I in the proper order in the eight issues of Eugene Jolas's magazine, beginning in April 1927 and ending in November of the same year with chapter 8. Strangely enough, this neatness cannot be found in González's scheme which omits the August publication of I.5, so that I.vi takes up the final slot in the second stage. Instead of opening his third stage with the seventh and eighth chapters of Book I, González inexplicably moves to Book III, while putting the core of II.2 which Joyce had published in February 1928 in a separate category. In the rest of the presentation, similar omissions occur which cause a complete breakdown of correspondences at the end of the third stage. The reason for the omissions is simply that González only considers the first publication of a fragment or a chapter.
The argument in the essay is seriously flawed. For one thing not all of the Finnegans Wake's text is accounted for: there are only fifteen categories (sixteen if we include the absent centre or fourteen when we realize that I.ii appears twice) and the Wake has seventeen chapters. But González's version of the Wake's genesis is quite misleading and it is therefore a great pity that he does not give us the evidence on which he bases himself. Since Ellmann's biography is quoted, I assume that he used the footnote in that book in which Ellmann copies the third appendix in A. Walton Litz's The Art of James Joyce. But Litz gives two sets of data, on the right-hand side the publication (place and time) and on the left the time of actual writing. González would have done well to consider the left hand side, because he would then have seen that there is a much more complex relationship between publication and actual drafting work.
González could have avoided this confusion if he had taken into account some of the scholarship that has been devoted to a study of the genesis of Finnegans Wake since Litz's book. If he had looked at David Hayman's A First-Draft Version of "Finnegans Wake" for example, or his more recent study, The "Wake" in Transit. Or if González had looked at the Archive reproductions of the different stages of development of Finnegans Wake, or at the discussion of the book's genesis in the introductions of the Wake volumes by David Hayman and Danis Rose.
The first problem with this kind of scholarship is that it is not all that accessible. Few university libraries can afford the Archive, my own university does not have a copy of A First-Draft Version of "Finnegans Wake". A second problem is the sheer difficulty of the task of reconstructing the history of Finnegans Wake on the basis of the material presented in the Archive, and Joyce's hand-writing is only one of the problems, and a minor one at that. All of these problems would be solved if Danis Rose could be permitted to publish his synoptic edition of Finnegans Wake, preferably in computer format too. Everybody would then be able to establish an exact chronology for the writing (and publishing) of the Wake.
In the meantime we have only two studies on the whole of the Wake's genesis that are accessible in the two meanings discussed above, David Hayman's The "Wake" in Transit and Danis Rose's more recent The Textual Diaries of James Joyce. The difference between the two books can be expressed in the terms I have introduced above: Hayman is a regressive reader, Rose is progressive.
As the title of his book indicates, Rose does not primarily focus on the drafts and versions of Finnegans Wake but on the so-called Buffalo Notebooks. These are the workbooks Joyce used between 1922 and 1939 in writing Finnegans Wake and as such they form the best record we have of the genesis of Joyce's last book. In these diaries (of the text, not of the writer) Joyce made notes that were destined for his last book. Rose's new chronology of the notebooks not only revolutionizes the study of Finnegans Wake, his book offers a somewhat dry but supremely accessible history of the writing of the Wake.
But this is not a review of Danis Rose's book. Suffice it to say that The Textual Diaries of James Joyce enables us to disqualify both González's premises and his conclusions. Not only does Rose fail to show any relevance to González's conjectures, his book shows that far from trying "to establish 10 March 1923 as the hypothetical centre" of Finnegans Wake because that was the day he wrote "King Roderick O'Conor," Joyce put it and the other early pieces away in the autumn of 1923 and rediscovered them as late as July 1938.
Whether we want to share Rose's conclusion that there was a separate work called Finn's Hotel between Ulysses and Finnegans Wake is not important, what the evidence presented proves beyond all doubt is that at least some of the early pieces, except the first section of I,2, were completely left aside and even forgotten when in November 1923 Joyce began to work in the large red notebook which contains the first drafts of most of Book I. If we are looking for stages in the writing of what later would become Finnegans Wake, it is here that we have to find the first break.
In conclusion I want to congratulate J. Carnero González on his decision to present his findings in the way he did and the editors of Papers on Joyce for publishing his essay in their first issue. Joyce studies would benefit greatly if more critics would lay their cards on the table, if an article would begin with a falsifiable hypothesis which would then be demonstrated. That they can then be proven mistaken is not a deficiency, on the contrary. What would be the point of arguing against a position which cannot be proven wrong because it is not clearly demonstrated and because it does not present a hypothesis? And what, indeed, would be the point of reading it?
[Back to Recent Articles] [Back
to Articles on-line] [Back to Homepage]
J. Carnero González, "Notes toward a Reading Proposal: WP and FW," Papers on Joyce 1 (1995): 21-28.
David Hayman, The "Wake" in Transit, Ithaca: Cornell UP, 1990.
Danis Rose, The Textual Diaries of James Joyce, Dublin: Lilliput Press, 1995.
[Back to Recent Articles] [Back
to Articles on-line] [Back to Homepage]