"The concept of the 'definitive text' corresponds only to religion
--Jorge Luis Borges(1)
In 1966 Jack Dalton suggested that the text of Finnegans Wake might require as many as 7000 emendations.(2) Indeed, after an investigation of the manuscripts, no-one would doubt that errors creeped into Finnegans Wake during its genesis and gestation. Perhaps Dalton's estimate--which handsomely exceeds the number of emendations made by Gabler--underestimates the degree of textual corruption present in Finnegans Wake. The problem then is how to respond to errors in the Wake once they have been identified and traced out. This is not a trivial issue; no-one would welcome a "Scandal of Finnegans Wake" as a successor to the increasingly tedious and unnecessary "Scandal of Ulysses."
This present note is intended as a warning against producing a "corrected" Finnegans Wake. By this I do not mean that no editorial work should be performed upon the text, rather that such work--which will be valuable--should not lead up to a new reading text of Finnegans Wake. In essence, I advocate the production of something like the verso side of Gabler's synoptic edition--preferably rendered in a hypertext--but bereft of a corrected recto. Bill Cadbury sanely notes that while a critical edition of Finnegans Wake may be desirable as a record of its genesis, such a text could not pretend to be definitive in any possible way.(3) Danis Rose has edited a new edition of Finnegans Wake,(4) and while I welcome its arrival I would not wish to see it replace the flawed edition we have now. I will now provide a few examples of the pit-falls that will face a prospective editor of Finnegans Wake.
In general, the typescripts for Finnegans Wake are fecund with transmissional departures. Any stage not directly produced by Joyce is of course liable to error and, even though Joyce's memory of his writing was very good, it was hardly infallible. Therefore, many errors of transcription eluded Joyce's eye (the errors Dalton bemoans in his "Advertisement" in the Kevin passage were all perpetrated by typists).
The two typescripts for the introductory section of II.1 (FW: 219.01-222.21) are no exception. While by no means the most egregious examples of departure, they do show some interesting classes of erroneous transmission. Although a fair copy is missing for this passage, there are numerous small, demonstrable mistakes made by the typist of the first typescript. The typist failed to consistently capitalize the characters' names on the list of characters and mistyped Kate's role: "Varianoekeand" instead of "Varian) cook and" (JJA 51: 9) However, in the absence of the preceding draft, there are a few departures that are impossible to unequivocally gauge. Joyce's "approached" from the first draft (JJA 51: 6) now appears as "appreached," certainly a plausible pun (perhaps made on the missing fair copy) but also a possible typo. In any case, the typist for the second typescript, which also contains numerous departures, missed this word (JJA 51: 11) and so it remained as "approached" through the final text (FW: 220.08). As this is not an authorial departure, it can be clearly rejected from consideration as a possible emendation.
The second typescript contains only four modifications by Joyce as well as one small proofreading correction. It is notable primarily for the number of mistakes that are made. Several of Joyce's overlay corrections on the preceding typescript were missed entirely. Almost all the mistakes made at this level were missed entirely by Joyce in preparing subsequent drafts. Except for "appreached," all the errors at this level involve a failure to decipher Joyce's overlay additions on the first typescript. Usually these involve omitting Joyce's additions, but in one case the typist actually adds a word. On the first typescript, Chuff is described as "the <fine> fairhaired fellow" (JJA 51: 9); the second typescript reads "the fine frank fairhaired fellow" (JJA 51: 11)--the word "frank" appears without prompting and without further comment or correction by Joyce. Although this word may have been gratuitously inserted, it enhances the alliteration already present in that phrase. This is the sort of departure that makes editors cringe, grown men weep, and certified accountants howl at the Moon like the restless dogs they are. No theory of "passive authorization," no matter how finely nuanced it may be, can accommodate all instances of departures wrought by other hands but retained, somehow and for whatever reason, by Joyce.
Applying Hans Walter Gabler's rule of invariant context, the pre-corrupted reading should be restored unless a passage containing a transmissional departure was subsequently modified by Joyce in a substantial way.(5) Here is an example of an error that should not be corrected, following from Gabler's rule: on the first typescript, Joyce added the following qualification to the Customers: "representatives <civics, each of whom is a jacktitative>" (JJA 51: 9). The second typist entered "civics" but left out the rest of this addition, so, at the second typescript level, the phrase reads simply "representative civics." On multiple subsequent drafts, Joyce modified this line through several discrete additions, so that it ultimately reads "representative locomotive civics, each inn quest of outings" (FW: 221.03-4).(6) Clearly, restoring the phrase lost on the second typescript would be unacceptable since it would now no longer fit into the sentence.
Here is a more complex example. On the first typescript, Joyce added an interesting qualification to "the Radium Wedding of Night and Morning <arranged as the daughter of Tyre and the son of Ausonius> and the Dawn of Peace, Pure, Perfect and Perpetual, Waking the Weary of the World" (JJA 51: 10). This overlay was completely overlooked by the typist for the second typescript. This neglected addition subtly reinforces the pretensions of this universal "Pageant of History" by invoking a union between two great rival sea-ports in the Classical world, Tyre and Ausonia. With the exception of two minor modifications--"Night" became "Neid" and "Morning" "Moorning"(7)--this passage was unchanged through the final text (FW: 222.17-20).
Strictly speaking, Gabler's rule of invariant context would prescribe restoring the missing phrase. But, the context is not invariant here. The missing phrase actually modifies the context in which it was supposed to have appeared. Without this phrase, the passage invokes historical grandeur in the most abstract of terms: "the Dawn of Peace, Pure, Perfect and Perpetual, Waking the Weary of the World." Although the missing phrase builds upon the pretensions of this phrase, it does so by compromising the generality with historical specificity. This phrase changes the logopoeic effect of this passage and so restoring it becomes questionable.
Philip Gaskell and Clive Hart faulted Gabler for a "preference for readings, derived from the evidence, usually manuscript, of Joyce's first thoughts, however strong the case for later versions."(8) A crux like the one above obviously represents a comparatively early thought which was elided, but this absence was seen by Joyce through multiple subsequent drafts without correction. The absence of this brief passage does not affect the passage and, I believe, its re-insertion would modify the passage in ways which might not be valid or desirable. In cases like this, conservatism is probably the best path to take and the passage should be left unchanged.
There are, of course, transmissional departures which detrimentally affect the text. Dalton indicates a few in his discussion of the Kevin passage and these provide ammunition for the argument towards correction. One such error I have found lies in the Ondt and the Gracehoper section of III.1.
Or, if he was always striking up funny funereels with Besterfarther Zeuts, the Aged One, with all his wigeared corollas, albedinous and oldbuoyant, inscythe his elytrical wormcasket and Dehlia and Peonia, his druping nymphs, bewheedling him, compound eyes on hornitosehead, and Auld Letty Plussiboots to scratch his cacumen and cackle his tramsitus, diva deborah (seven bolls of sapo, a lick of lime, two spurts of fussfor, threefurts of sulph, a shake o'shouker, doze grains of migniss and a mesfull of midcap pitchies. The whool of the whaal in the wheel of the whorl of the Boubou from Bourneum has thus come to taon!), and with tambarins and cantoridettes soturning around his eggshill rockcoach their dance McCaper in retrophoebia, beck from bulk, like fantastic disossed and jenny aprils, to the ra, the ra, the ra, the ra, langsome heels and langsome toesis, attended to by a mutter and doffer duffmatt baxingmotch and a myrmidins of pszozlers pszinging Satyr's Caudledayed Nice and Hombly, Dombly Sod We Awhile but Ho, Time Timeagen, Wake! (FW: 414.35-415.15).
This is a complex sentence, rendered all the more inscrutable by the lengthy parenthesis which comprises two sentences. Perhaps I am less-than-astute, but I must have read this passage dozens of times over the years without noticing a very basic syntactic flaw here. The form of this sentence is a conditional sentence yet, as rendered in the final text, the conditional phrasing is incomplete. Nowhere does an apodosis follow from this inordinately lengthy protasis. This sentence postulates a certain conditional outcome to a hypothetical action, yet that possible outcome is not explicitly stated. The inordinate length and extreme hypotaxis of the sentence obscures this fact, but, once this fact is realized, the sentence becomes impossible to parse. This breakdown in syntactic regularity was apparently caused by transmission error. In its initial draft, this sentence was itself the apodosis: "If he was not doing that he was always getting up funny funeralls with Besterfather Zeuts…" (JJA 57: 294). The sentence remained in this format through the publication of Tales Told of Shem and Shaun in 1932. When the galleys were prepared for Finnegans Wake in 1938, the protasis was omitted (JJA 62: 17). On the pages of Tales Told of Shem and Shaun, the second "he" lies almost directly under the first (JJA 57: 325). Evidently, the typesetter skipped the clause following the first "he" and jumped to the second. Scholars of Mediæval manuscripts call this kind of copyist's error haplography.
This is the sort of crux that, once revealed, seemingly screams for correction (or, if not scream, then seductively beckon, with a Gitane languidly dangling beneath come-hither eyes). The published version of the sentence needlessly makes no sense in a particularly cumbersome way, and a simple emendation could return it to a less cumbersome mode of making no sense by restoring a smidgen of syntactic normalcy. The error is needlessly deleterious and could be easily rectified. However, I would argue that the error should remain. If this error is corrected then, obviously, other comparable errors should be emended as well. The problem then would be to establish criteria by which such errors should be corrected. One could start with a principle of strictness and probity: only the most malevolent errors which can be definitively identified should be modified. However, even this seemingly simple prescription will, in practice, be impossible to implement. Not all departures can be unequivocally gauged as being purely accidental. Furthermore, I doubt that anyone could ever develop a rigid set of criteria to unquestionably determine that a given departure is ruinous to the sense of a passage. One man's poison is another man's Irish Stew. Between the missing protasis above and the elided "daughter of Tyre and the son of Ausonius" from II.1 lies a diverse range of errors, each with its own peculiarly gnarled history. For these reasons I advocate rigid conservatism: hands off Finnegans Wake. Obviously, these departures need to be recorded--they should be allowed to enter the Wake's paratext (ideally in a hypertextual variorum)--but they should be barred from the text. Let us leave ill enough alone.
3. "Especially if a critical edition is accompanied by a reading text, the tempting notion that the presentation of chronological changes is mainly for the purpose of giving the evidence for decisions about the final product will--we have seen it happen before--obscure the fact that Joyce's intentions have a chronological dimension which can't be flattened out. What we really want to have, I think, is not an artifact which we can reify as the embodiment of Joyce's final intentions, but discussion of and insight into the full, detailed history of his writing project and hence of the imaginative development of his book. For this, there is no substitute for inference about the imaginative process on the basis of careful and fresh consideration of the documents level by level, for each and every segment of the book. An ideal recreation, in its magnitude, for our ideal insomnia" (Bill Cadbury, "The Development of the 'Eye, Ear, Nose and Throat Witness' Testimony in I.4," Probes: Genetic Studies in Joyce, eds. David Hayman and Sam Slote, Amsterdam: Rodopi, 1995. 203-254. 246). At the conference Genetic Networks in Antwerp in 1998, both Cadbury and Nathan Tenny presented a scintillating demonstration of the possibilities of a hypertextual representation of the Wake's genesis.
5. Hans Walter Gabler, "Afterword," James Joyce, Ulysses: A Critical and Synoptic Edition, eds. Hans Walter Gabler with Wolfhard Steppe and Claus Melchior, New York: Garland, 1984, 1986. 1858-1911. 1895-1900. This is obviously one of the more contentious aspects of Gabler's editorial procedure.
6. When this passage appeared in transition, it read "representative locomotive civics inn quest of outings" (James Joyce, "Continuation of a Work in Progress," transition 22 [February 1933]: 49-76. 51; the pre- and post-transition drafts for this chapter are missing). On an overlay to the first set of marked pages of The Mime of Mick, Nick and the Maggies, Joyce added "each": "representative locomotive civics inn <each> quest of outings" (JJA 51: 204; the comma was added on the typed set of corrections to these pages, JJA 51: 288-89). This word recalls the earlier, lapsed unit "each of whom is a jacktitative."
7. These changes must have been made on the revised transition pages since they do not appear in transition (transition 22: 51), but do appear in the set pages for The Mime of Mick, Nick and the Maggies (JJA 51: 206).