In response to:

The New 'Ulysses': Unanswered Questions from the January 19, 1989 issue

To the Editors:

Philip Gaskell’s letter [Letters, NYR, January 19] endeavours to set right some misleading speculations in Charles Rossman’s “The New ‘Ulysses’: The Hidden Controversy” [NYR, December 8, 1988]. Yet it seems not to have helped Rossman with the questions that trouble him.

When the James Joyce Estate accepted my proposal to edit Ulysses, they also asked three Joyce scholars of their choice to advise them. Richard Ellmann’s, Philip Gaskell’s and Clive Hart’s advice to the Estate was extended as assistance to the editor. Naturally, the edition benefited materially from the scrutiny they gave the editorial work in draft stages. From the very outset, however, I assumed full scholarly responsibility for the edition and made it clear that, in the event of disagreements, I would not relinquish nor accept curtailments of it.

The cooperation with the Estate and their advisers was good. How and why it was that two of the advisers nevertheless, and only at a very late stage, were unprepared to respond to the full implications of the over-all editorial concept, is not for me to say. The concept itself had been there to grasp from the beginning, since it was demonstrated in the trial chapter on which my proposal was based. The advisers made two moves which both deserve respect. One was to try to sway me into altering my procedure in one important, though strictly circumscribed area of Joyce’s text for Ulysses. The other was to stand by their advisory obligation to the Estate and not dissociate themselves, in the end, from the edition which in their own words they regarded as “prepared with a quite exceptional degree of scholarly expertise.” On my part, I upheld my concept. The advisers’ final stance implies their acceptance of an edition which, though not in every detail the one they would have established had they been the editors, is nonetheless a possible and defensible representation of James Joyce’s Ulysses. In no way is the edition a compromise.

The area of dispute is easily specified. The Estate’s advisers did and do not distinguish documents and text. Seeing that certain documents—the author’s fair copies, in particular, of some of the novel’s episodes—are outside the line of transmission, they consider unique readings from them inadmissible to the critical text. This is a possible attitude, and many textual critics would hold it. Yet it is a purely formal one, and I do not subscribe to it in my solution to the problem of editing Ulysses. The documents in question show revisions, demonstrable or critically ascertainable, which only by the accidents of Joyce’s writing habits did not descend to the published text. Even though the documents stand outside the line of transmission, I accept such revisions in direct line of the development of the text. For instance: of the seedcake which Molly passed into Leopold Bloom’s mouth, Joyce writes first that it was “sweet and sour with spittle.” In revision, he makes it “sweetsour of her spittle,” poetically condensing the adjectives into an oxymoronic compound, and adding the personal pronoun to intensify the recall in Bloom’s memory. The revised reading is admitted to the critical text although it happens to be unique to a document outside the line of transmission. Not to admit it were to suspend a critical insight in favour of a strictly formal procedure. Either decision is defensible, and each gives a Joycean reading. Whether an editor, on argued grounds, chooses the one Joycean alternative or the other, he does not corrupt the text. If the “New Ulysses” incurs no editorial compromise, it is also not corrupt.

On our joint objective to free Ulysses of corruptions, Philip Gaskell’s letter is quite clear. In two thirds (or some 320) of 484 readings he would not wish to go along with the edited Ulysses because in these, as he says, it adopts text from outside the line of transmission. This, by his count, leaves 160-odd readings where we would still differ. If Gaskell’s copytext, as I assume, is the first edition, whereas mine is what I have termed a “continuous manuscript text,” it is truly remarkable that we should come so close in the editorial result. For it amounts, on his part, to a very extensive emendation of the first edition—which indeed, as he states, requires editorial changes in close to 1700 “(predominantly) verbal alterations” alone.

Figures, however, are not the answer to the questions—spurious or legitimate—that have arisen around the “New Ulysses.” What I have said concerning the Gaskell-Rossman exchange in your issue of January 19 should suffice to indicate that a knowledgeable argument takes a completely different direction from the one for which Charles Rossman falls. My remarks should also emphasize that the debate around Ulysses must seek levels of distinction and understanding that have been lacking in The New York Review of Books.


Hans Walter Gabler
Institute for English Philology
University of Munich
Munich, West Germany

To the Editors:

John Kidd writes [Letters, NYR, December 8, 1988, p. 62] in his discussion of the Conolly/Connolly question that “the man’s real name was being rejected in favor of the spelling in the Rosenbach Manuscript.” This raises, in the mind of an interested observer, two issues. What was the man’s real name and what authority does the Rosenbach Manuscript have in the transmission of Ulysses?

In previous correspondence [NYR, June 30, 1988, p. 33, and Letters, September 29, 1988, p. 81] Dr. Kidd cited Thom’s Directory as the major source for the single “n” spelling and this has been discussed in their responses by John O’Hanlon and Michael Groden. In this latest stage of the argument, Dr. Kidd cites J.B. Lyons, author of James Joyce and Medicine (1974), as a further authority supporting the single “n,” stating that Lyons “went to considerable trouble to identify the doctors mentioned in Ulysses.” He quotes “for example” a passage from p. 145 of this work which in its one mention of “Conolly” spells it with the single “n.” But this is the only mention of the name in the course of the book, not merely a single example; in the index the name, as Dr. Kidd acknowledges, is spelt “Connolly.” 50/50 choice—hardly firm or definitive proof. Indeed, if Dr. Kidd had examined the latest work by Dr. Lyons, Thrust Syphilis Down to Hell and Other Rejoyceana (1988), he would have found “Connolly” spelt consistently with a double “n” on the four instances it appears in the book in addition to the index. Furthermore, one of those instances is as caption to a portrait of the “real man.” In addition, Lyons goes so far as to correct the spelling of the quotation taken from p. 12 of the 1968 Penguin edition of the novel from “Conolly” to “Connolly.” In other words, Dr. Kidd’s fresh authority makes the same amendment as Professor Gabler. One further source which might also be cited for the “real name” is Who’s He When He’s At Home compiled by Shari and Bernard Benstock (1980) where “Conolly” is acknowledged as the version appearing in Ulysses (Random House, 1961) but “Connolly” is given as the “real name.”

Some readers may by this point ask: does it matter? Yes, because the astringent debate about the status of the 1984 edition revolves round such details and each one when eventually revealed by Dr. Kidd must be dealt with in turn, either to accept or to reject, but in every case to be discussed fully. This may lead to an eclectic editorial method but the circumstances of the composition and publication of Ulysses demand it. (It may also be noted in passing that the Conolly/Connolly decision, and some others of Dr. Kidd’s “errors,” was first questioned by the late Charles Peake at the 1985 Monaco conference, in a paper reprinted in Assessing the 1984 Ulysses edited by George Sandelescu and Clive Hart [1986].)

However, the more substantial issue of editorial principle and textual authority which is raised implicitly by Charles Rossman’s discussion of the correspondence leading to both the 1984 and 1986 editions, “The New ‘Ulysses’: The Hidden Controversy” [NYR, December 8, 1988], demands equal debate. The further reply by Dr. Rossman to Philip Gaskell [Letters, NYR, January 19] underlines a certain naïveté about editorial method that was, I feel, present also in the earlier article. One gains the impression, rightly or wrongly, that Dr. Rossman believes that a perfect, “uncorrupted” text of Ulysses, somehow free of critical judgements, upon which the entire body of Joycean scholars would agree, is a possibility. It is not. All those interested in the text of Ulysses probably possess a copy of some edition covered with annotations and corrections which each regards as the best possible text. There will be, however, no absolute agreement upon either the corrections or the criteria by which they were made. Decisions have to be made; there are balances of probability; all back their own judgement. So Philip Gaskell and Clive Hart disagree with a number of Professor Gabler’s critical judgements and would propose their own changes to his edition but neither, I presume, would champion any pre-existing edition over Gabler’s.

In the case of the Conolly/Connolly decision, the single “n” spelling is attested to by the printing of this particular episode in the Little Review, by the galley proofs of its proposed but never executed printing in the Egoist, and by the French printer’s first placards (proofs) of the first edition of the novel. The double “n” exists in the Rosenbach manuscript, a holograph fair copy of the novel prepared by Joyce for sale to supplement his income. The Rosenbach manuscript predates any of the sources for “n.” The judgement has to be made whether the later material represents Joyce’s final thoughts on the matter, either by making a deliberate alteration or by approving by act of omission an alteration made by a third party—typist, printer, whatever—or whether the later material represents a corruption of Joyce’s words. In this particular instance, there seems no reason why Joyce would make or authorise the change from “Connolly” to “Conolly” and there is no extant evidence from letters or postcards to the typist, for instance, confirming his involvement in the change. So the balance of probability swings towards “Connolly” and so Professor Gabler made the emendation. The fact that the real Connolly was spelt thus does not in itself justify the change but it does confirm it.


It was Clive Hart himself who wrote, in his essay in the collection edited by George Sandelescu and himself, of “the perfect Ulysses, totally harmonious, entirely errorfree, the Ulysses that has never existed, can never exist.” The solution is not to abandon Professor Gabler’s work. The Critical and Synoptic Edition of 1984 represents both a work of some scholarship and a source of textual data. The Corrected Edition of 1986, the trade edition, should not, however, stand inviolate through successive printings but Professor Gabler should be given the opportunity by Random House and Penguin to make further changes if he is swayed in his critical judgement by present and future discussions. In the meantime, each of us can continue to annotate and “correct” our copy of Gabler as we did with editions of the past.
Alistair McCleery
Edinburgh, Scotland

Charles Rossman replies:

Readers may well wonder why Dr. Gabler dwells for a paragraph on who advised whom, under what circumstances, to what effects. (The matter surfaces throughout Gabler’s letter, in phrases like “the Estate and their advisers,” “the Estate’s advisers,” and “their advisory obligation to the Estate.”) In “The New ‘Ulysses’: The Hidden Controversy” I cited letters from the estate and the edition’s advisers that indicated that the authority and role of the advisers were, in fact, open to definition and redefinition during the process of editing Gabler’s Ulysses. Most crucially, Peter du Sautoy assured Clive Hart as early as October 1981 that the academic advisers, not the general editor, would “have the last word” in settling disagreements. But by June 1983, when Gaskell and Hart resigned as academic advisers only to rejoin the project two months later, the ground rules had changed. Now Gabler had the last word, rather than the advisers, as evidenced by the fact that the estate proceeded with publication of the 1984 Synoptic and Critical Edition over the objections of two of the advisers.
In retrospect, Gabler regards himself as having had full editorial authority from the outset of the project. But the evidence in Richard Ellmann’s papers at the University of Tulsa suggests instead that his authority grew substantially during the course of editing, while that of the advisers waned. Eventually, he not only published the text that he wanted, but at the time that he wanted. (The 1986 trade edition, in particular, now appears to have been hurried into print despite mounting public criticism that echoed the advisers’ earlier, private criticism.) What had begun as a cooperative editing venture, overseen by advisers who had the ear of the estate while working closely with Gabler about specific editorial problems, ended up with an edition that was the product of one man’s editorial direction.

Gabler twice asserts that his edition is not a compromise. He apparently means by this that he did not modify his “over-all editorial concept” in response to the criticism of Philip Gaskell and Clive Hart. In this regard, Gabler strikes me as incontestably correct. All the compromises were made by Gaskell and Hart (and by the estate). Even Gabler appears puzzled by the advisers’ curious behavior, as when he observes that “how and why” two of the advisers opposed his “over-all editorial concept” at a very late stage “is not for me to say.” Gabler correctly points out that his “concept” had “been there to grasp from the beginning.” Since Gabler wrote these words, a conference on the “new Ulysses” took place on February 2–4 at the University of Miami. In Miami, I myself introduced a panel that included Gaskell and Hart, and I invited a fuller explanation of their opposition to and ultimate acceptance of a text that they continued, even in Miami, to criticize quite boldly. I specifically asked whether editorial principles had been compromised for commercial motives. Unfortunately, my questions were deflected by responses from the floor. As a result, neither Gaskell nor Hart commented on these issues.

Gaskell and Hart have announced that they are publishing a list in which they differ with Gabler on 484 specific readings. Gabler argues in his letter that some 320 of these arise from his editorial decision to accept readings from the Rosenbach Manuscript even when it is outside the main line of transmission. In contrast, Gaskell and Hart regard such readings as inadmissible. Gabler’s desire here is to lump all 320 such readings together, as though they were really only a single disputed variation. But there are serious problems with this reasoning.

In the first place, that the entire category of 320 readings exists as a result of an editorial decision hardly means that each instance within the category is equally convincing. Rather, each instance must still be assessed on its own merits for accuracy of decipherment and transcription. For example, in the single illustration that he offers, “sweetsour of her spittle,” Gabler’s account of the archival basis for this particular emendation badly misrepresents the facts. Gabler says that Joyce first wrote (in the Rosenbach Manuscript) “sweet and sour with spittle.” Then, Gabler continues, Joyce revised (still in the Rosenbach) to read “sweetsour of her spittle,” this being the version that Gabler accepted into his text. I invite anyone to check the Rosenbach facsimile, “Lestrygonians,” leaf 21, line 27. The words “sweet and sour with spittle” are nowhere to be found in the Rosenbach Manuscript, nor anywhere else in Joyce’s hand. Instead, Joyce wrote “sweetsour spittle,” and then added the two words “of her.” (Nor is it certain that the added words were a revision—they might well have been an oversight in copying that Joyce corrected on the spot.) The phrase “sweet and sour with spittle” does occur, however, in the 1922 first edition (p. 167)—at the end of a sequence of revision, not the beginning. Gabler has preferred, then, an earlier and not a later reading of Joyce’s. And his account of Joyce “poetically condensing the adjectives into an oxymoronic compound” during the process of revision is pure invention.

Quite apart from the need to examine each of Gabler’s 320 readings individually, the disagreement over editorial principle cannot be brushed aside so lightly. That disagreement between Gabler and Gaskell and Hart is deep and, as became clear in Miami, enduring. There Gaskell and Hart frankly admitted that they disapprove on similar editorial grounds of the most famous single emendation made by Gabler, the insertion of the so-called “love passage.” In Miami they also repudiated Gabler’s flush left-hand dashes to indicate speech, his bland typeface for the headlines in the “Aeolus” section of the novel, and much of the new punctuation in “Eumaeus” and “Circe”—all of which are categorical matters affecting over one thousand specific points of punctuation not included on their list of 484 disagreements of wording with Gabler.

Gabler writes that “figures…are not the answer to the questions…that have arisen around the ‘New Ulysses.’ ” Yet they will continue to tantalize anyone who is interested in the forces that led to Gabler’s Ulysses. We recall, for example, that in the early stages of editing, when the edition’s reputation as a scholarly miracle was being generated, Gabler and Michael Groden spoke of eight thousand or more “errors” in the 1961 edition, and Kenner spoke of six thousand. Upon publication of the new edition, Gabler claimed “only” five thousand new readings—some three thousand less than he and Groden estimated when it served their interests to have larger numbers. (Both the estate and publishers expected substantial changes in any new edition, in order to justify copyrights and publication.) Now Gaskell and Hart declare that they recommend a mere 1,584 corrections for the 1961 edition, or only 1,100 more than they recommend for Gabler’s own. My point is that the very category of “error,” like so many other terms of this debate, is open to definition. Various definitions of “error” serve various ends, and all the figures tossed about in this controversy must be scrutinized with this in mind.

Finally, I want to address briefly Alistair McCleery’s concern that I harbor “a certain naïveté about editorial method.” Mr. McCleery fears that I believe in “a perfect, ‘uncorrupted’ text of Ulysses, somehow free of critical judgements.” How Mr. McCleery gained this fanciful impression of my views is beyond me. I do not believe in a single definitive edition, and nothing that I have written suggests that I do. Indeed, I think that several different editions could, and probably should, be prepared. I would especially welcome a critical edition based on the 1922 edition—something that Hans Gabler said in his 1984 Foreword would not be feasible. (In Miami, Gabler retracted this view, and granted at least that the 1922 edition might be used as a copytext for a critical edition.)

Of course no single text will compel the assent of all Joyce scholars. Of course “critical judgments” will enter at every stage. The crucial issue is not universal agreement but that any critical edition must be based on sound editorial theory, faithfully and consistently applied. But as I demonstrated above with the example of “sweetsour of her spittle,” and as John Kidd has demonstrated with a growing list of examples, such as misspellings of the names “Conolly” and “Thrift,” some “critical judgments” are little more than blunders. Gabler himself conceded in Miami that he was mistaken to alter “Thrift” to “Shrift.” Mr. McCleery would do well to acknowledge such mistakes unambiguously.

In Miami, much was vented; much was evaded; nothing was solved. Accordingly, the report of the Tanselle committee becomes an increasingly important event, to which the entire Joyce world looks forward eagerly. Yet even the Tanselle committee cannot resolve the deepest editorial problems that have arisen with Gabler’s edition. No matter what text Random House chooses to continue in print, from now on no edition can be definitive. We have lost our bibliographical innocence forever.

John Kidd replies:

Alistair McCleery is the cocompiler of The ‘Ulysses’ Pagefinder (Edinburgh: Split Pea Press, 1988), a companion to the Gabler Ulysses. A collaborator in the Gabler edition, A. Walton Litz, introduced the Pagefinder as a guide to “the standard text for decades to come.” Mr. McCleery has decided to follow the example of Ian Gunn, the other compiler, who, without mentioning a single textual crux of Ulysses, in the TLS last November called me “a bragging schoolboy.” Mr. Gunn, too, neglected to mention his stake in the controversy. Like the efforts of Michael Groden and John O’Hanlon before him, Mr. McCleery’s letter about Gabler’s misspelling of Conolly Norman’s name does nothing to inspire confidence in the research talents of Gabler or those who support him.

Before me are photographs of the gravestone marked “Conolly Norman” in Mount Jerome Cemetery (see photograph above) and a memorial plaque in the north aisle of Saint Patrick’s Cathedral, Dublin, with the same spelling. Also on my desk is The Dictionary of National Biography, second supplement (1912, reprinted 1920), where Joyce could have read about “NORMAN, CONOLLY (1853–1908), alienist.” The two columns on Norman in the DNB were an important source for J.B. Lyons’s James Joyce and Medicine (1974), which spelled Conolly correctly in the body of the text but incorrectly in the index, as I previously pointed out. Mr. McCleery’s letter notwithstanding, nowhere in his 1988 study does Dr. Lyons comment on the accuracy of either spelling; I suspect his English printers regularized the name from its Irish form. The four occurrences of the erroncous “Cannoily,” including the misquotation from Ulysses and the photo caption, are found on pages 133–134 of Thrust Syphilis down to Hell and other Rejoyceana and are not scattered through the book as Mr. McCleery implies. Nowhere does Dr. Lyons endorse the spelling that has crept into his book and I’m certain he regrets the error. But why is Mr. McCleery quoting against me Joyceans of the 1980s when he could have found turn-of-the-century sources?

All the writings of Conolly Norman, such as The Clinical Features of Beri-beri (1899) in the library of the Yale Medical School, bear on their title pages what Mr. McCleery calls a misspelling of the author’s name. Volume 421 of our National Union Catalogue reports that American libraries hold three works by Conolly Norman but none by the man Mr. McCleery refers to. A call to Dublin confirmed that the National Library in Kildare Street knows only the spelling found in Ulysses until Gabler “corrected” the text. Norman’s contributions to The Journal of Mental Science, of which he was a coeditor at his death, and his five-page obituary in The Medical Press, March 4, 1908, have the spelling Mr. McCleery calls inaccurate. The obituary was reprinted as In Memoriam. Conolly Norman. “So passed away,” the memorial keepsake concludes, “a man whom there is none to replace, and whom the specialty, the profession, and the country could ill afford to lose.”

Mr. McCleery assumes that Gabler weighed the probabilities and opted for “the real Connolly.” Yet nowhere in his 1984 edition does Gabler cite contemporary sources such as Thom’s Directory or The Dictionary of National Biography for the spelling of names. There is no evidence whatever for Gabler having investigated the names in Ulysses. At the James Joyce Birthday Conference in Miami devoted entirely to the controversy around his edition, Gabler was confronted with the question, What was your policy on the spelling of names? At the end of three days and eight hours on the platform expounding his theories of “bibliographical thinking” Gabler never got around to explaining why he changed so many names in Ulysses without verifying who’s who and where’s where. As I have previously shown in print, Gabler frequently altered names Joyce wrote on the assumption that the author slipped but had intended the historical spelling. Asked why he overruled Joyce’s manuscripts so often, Gabler would not answer in Miami. Nor would he say why he left so many unhistorical spellings uncorrected and unannotated. Asked if he had any policy at all Gabler would not answer.

Mr. McCleery claims that the Rosenbach Manuscript is a superior source for the spelling of the disputed name. But the relation of the Rosenbach Manuscript to the 1922 edition takes on an entirely different character when it is recognized that only Joyce and his typist Claud Sykes ever had their hands on all three copies of the typescript. (Ezra Pound saw only two of the three.) It is unlikely that the American Sykes would know to alter “Connolly” to the historical “Conolly,” so the change can be attributed to Joyce, who had Thom’s at hand as he revised the typescripts.

That Joyce was revising his typescript of Telemachus has implications for Ulysses that go far beyond the spelling of one name. Gabler has substituted three words on the first page of the book on the assumption that whenever the descendants of the three typescripts agree, the change must be Sykes’s and not Joyce’s. Thus the change from “land” to “country” was attributed to Sykes, as well as “out” to “up” and “slow” to “low.” Once we conclude that “Conolly” is Joyce’s correction of his earlier slip and not a typist’s error, the presumption shifts in favor of “country,” “up,” and “low”—the words of all editions prior to Gabler. The last of these revisions is so patently Joycean that even without the correction to “Conolly” an editor should have detected the hand of Joyce on the lost typescripts. Contrast the Rosenbach and 1922 versions of the first page of Ulysses:

He peered sideways up and gave a long slow whistle of call [Rosenbach, 1984]

He peered sideways up and gave a long low whistle of call [1922–1983]

An attentive reader of Ulysses will notice an echo of this later in the Sirens episode:

Lenehan’s lips over the counter lisped a low whistle of decoy. [1984 11.328, 1961 p. 264]

The Sirens episode is literally the great echo chamber of Ulysses. In all editions before Gabler, Buck Mulligan’s “low whistle” prepared us for Lenehan’s “low whistle.” The motif is now gone, and the first page of Ulysses is spoiled by Gabler’s assumption that the change from “slow whistle” of the Rosenbach Manuscript to “low whistle” of the book could not have been Joyce’s. The historical truth of Joyce’s change from “Connolly” to “Conolly,” and the certainty that Joyce alone could have known the real man’s name. should have led Gabler to the reality of the late revisions to the typescripts. The only edition that misspells a famous physician’s name also mars the first page of Ulysses by three changes of wording from Joyce’s final revisions.

The failure to undertake elementary research into the names of Dubliners Joyce knew and Gabler’s insufficient grasp of allusive detail spoil the entire editorial enterprise. Ulysses: The Corrected Text is far worse than what we had before.

This Issue

March 30, 1989