The publication history of Ulysses is disputed and obscure. There have been at least 18 editions, and variations in different impressions of each edition. Joyce's handwritten manuscripts were typed by a number of amateur typists (one of whom was Robert McAlmon).
According to Joyce scholar Jack Dalton, the first edition of Ulysses contained over two thousand errors but was still the most accurate edition published. As each subsequent edition attempted to correct these mistakes, it incorporated more of its own, a task made more difficult by deliberate errors (See "Episode 17, Ithaca" above) devised by Joyce to challenge the reader.
Notable editions include the first edition published in Paris on 2 February 1922 by Sylvia Beach at Shakespeare and Company (only 1000 copies printed) and printed by Darantiere in Dijon; the pirated Roth edition, published in New York in 1929; the Odyssey Press edition of 1932 (including some revisions generally attributed to Stuart Gilbert, and therefore sometimes considered the most accurate edition); the 1934 Random House US edition; the first British edition, of Bodley Head in 1936; the revised Bodley Head edition of 1960; the revised Modern Library edition of 1961 (reset from the Bodley Head 1960 edition); and the Gabler critical and synoptic edition of 1984.
Gabler's "corrected edition"
Hans Walter Gabler's 1984 edition was the most sustained attempt to produce a corrected text, but it received much criticism, most notably from John Kidd. Kidd's main theoretical criticism is of Gabler's choice of a patchwork of manuscripts as his copy-text (the base edition with which the editor compares each variant), but this fault stems from an assumption of the Anglo-American tradition of scholarly editing rather than the blend of French and German editorial theories that actually lay behind Gabler's reasoning. The choice of a multiple copy-text is seen to be problematic in the eyes of some American editors, who generally favour the first edition of any particular work as copy-text. Less subject to differing national editorial theories, however, is the claim that for hundreds of pages—about half the episodes of Ulysses—the extant manuscript is purported to be a "fair copy" which Joyce made for sale to a potential patron. (As it turned out, John Quinn, the Irish-American lawyer and collector, purchased the manuscript.) Diluting this charge somewhat is the fact that the theory of (now lost) final working drafts is Gabler's own. For the suspect episodes, the existing typescript is the last witness. Gabler attempted to reconstruct what he called "the continuous manuscript text", which had never physically existed, by adding together all of Joyce's accretions from the various sources. This allowed Gabler to produce a "synoptic text" indicating the stage at which each addition was inserted. Kidd and even some of Gabler's own advisers believe this method meant losing Joyce's final changes in about two thousand places. Far from being "continuous", the manuscripts seem to be opposite. Jerome McGann describes in detail the editorial principles of Gabler in his article for the journal Criticism, issue 27, 1985. In the wake of the controversy, still other commentators charged that Gabler's changes were motivated by a desire to secure a fresh copyright and another seventy-five years of royalties beyond a looming expiration date.
In June 1988 John Kidd published "The Scandal of Ulysses" in the New York Review of Books, charging that not only did Gabler's changes overturn Joyce's last revisions, but in another four hundred places Gabler failed to follow any manuscript whatever, making nonsense of his own premises. Kidd accused Gabler of unnecessarily changing Joyce's spelling, punctuation, use of accents, and all the small details he claimed to have been restoring. Instead, Gabler was actually following printed editions such as that of 1932, not the manuscripts. More sensationally, Gabler was found to have made genuine blunders, the most famous being his changing the name of the real-life Dubliner Harry Thrift to 'Shrift' and cricketer Captain Buller to 'Culler' on the basis of handwriting irregularities in the extant manuscript. (These "corrections" were undone by Gabler in 1986.) Kidd stated that many of Gabler's errors resulted from Gabler's use of facsimiles rather than original manuscripts.
In December 1988, Charles Rossman's "The New Ulysses: The Hidden Controversy" for the New York Review revealed that Gabler's own advisers felt too many changes were being made, but that the publishers were pushing for as many alterations as possible. Then Kidd produced a 174-page critique that filled an entire issue of the Papers of the Bibliographical Society of America, dated the same month. This "Inquiry into Ulysses: The Corrected Text" was the next year published in book format and on floppy disk by Kidd's James Joyce Research Center at Boston University. Gabler and others rejected Kidd's critique, and the scholarly community remains divided.
Gabler edition dropped; publishers revert to 1960/61 editions
In 1990, Gabler's American publisher Random House, after consulting a committee of scholars, replaced the Gabler edition with its 1961 version, and in the United Kingdom the Bodley Head press revived its 1960 version. In both the UK and USA, Everyman's Library also republished the 1960 Ulysses. In 1992, Penguin dropped Gabler and reprinted the 1960 text. The Gabler version remained available from Vintage International. Reprints of the 1922 first edition are also now widely available, largely due to the expiration of the copyright for that edition in the United States.
While much ink has been spilt over the faults and theoretical underpinnings of the Gabler edition, the long-awaited Kidd edition has yet to be published, as of 2015. In 1992 W. W. Norton announced that a Kidd edition of Ulysses was to be published as part of a series called "The Dublin Edition of the Works of James Joyce". This book had to be withdrawn, however, when the Joyce estate objected. The estate refused to authorise any further editions of Joyce's work for the immediate future, but signed a deal with Wordsworth Editions to bring out a bargain version of the novel in January 2010, ahead of copyright expiration in 2012.