WHEN we speak of an edition succeeding or failing, we normally refer to its success or failure in giving us what the author wrote, or what the author demonstrably intended to appear, but which for some reason did not appear. The editor is of course at the service of the author and the reader: it is not part of the editor's function to second-guess the author, to speculate as to what the author might have written had circumstances been different, or to apply a subjective test as to what the author "might have intended".

Nor is it part of the editor's function to supply knowledge that was not available to the author. The circumstances were as they were, and cannot be retrospectively altered.

When, as is the case with the new Danis Rose "Reader's Edition" of Ulysses, these fairly fundamental criteria are not observed, the question becomes not so much whether the enterprise succeeds or fails, but rather whether the term "edition", already grown rather elastic, has been stretched past breaking point.

In some respects what Rose has produced does resemble a conventional edition: it is based on an "isotext", which is itself an assemblage of all the extant documents which lay behind the first edition of Ulysses. As far as one can judge from the very scanty example given (just one, not very lengthy, passage), Rose's "isotext" is quite similar to that produced by Hans Walter Gabler for his 1984 "Corrected Text", (hardly surprisingly considering that Rose had himself worked on that edition). Behind this assemblage, then, (if Rose actually deigned to give it to us) would lie verifiable, evidential documents which anyone can check to confirm or disprove its accuracy.

No, the radical aspect of Rose's procedure lies in his subsequent operation: what he calls his "copyreading" of the text thus arrived at (although the term "copyreading", as it is used here has little connection with its normal meaning). What Rose has really done is not copyreading, but rather an aggressive sub-editing of Joyce's text.

As a sub-editor myself, I can recognise all these procedures: the attempts to ensure clarity, to smooth the path as much as possible for the reader.

It is one thing, however, to practise this art or craft on journalistic writing (eminent though some of its practitioners undoubedly are), with the needs of a casual and harried reader in mind; it is another to practise it on the presumably immortal words of James Joyce.

Examples of the pitfalls into which Rose's methods lead him are easily to hand: they are provided in his own introduction. He alters the phrase "his neighbour nist not of his wile" to "his neighbour wist not of his wile" on the basis that "nist not" is a double negative and is therefore confusing to the reader. But in the 14th century English of The Travels of Sir John Mandeville, on which this passage is mainly based, double negatives are absolutely normal. The change is without foundation and Joyce's artistic purpose is undermined.

Incidentally, in a discussion with me on BBC Radio 4, Rose claimed that he had not inserted any word into the text that Joyce had not at some stage written, yet "wist" is a clear instance of such an insertion, and others could be cited.

Again, the verb "was" is inserted in the following sentence, which is based on the diary style of Pepys and Evelyn: "the big wind of last February a year ... was a small thing beside this barrenness." Rose states that this verb existed in a "protoversion", and I do not doubt it. But we can be quite sure that Joyce's subsequent dropping of it was intentional, for it is of the essence of the diary style he is parodying that main verbs are omitted, e.g., "Late supper and so to bed." For an edition that asks to be judged on "primarily aesthetic" terms, these decisions do not inspire confidence.

A further example comes from the "Ithaca" section towards the end of the book. The passage (on page 645 in Rose's version) is too long to quote but essentially two sentences have been broken into four. It is true that as it stood, the passage strained at the limits of English grammar, but some of us were under the illusion that this was the whole idea, that, as Joyce once put it, there is a difference between a book like Ulysses and a pound of chops. The whole tendency of the "Ithaca" section is towards long, complicated, unwieldy sentences.

There is, finally, the strange case of the final section "Penelope", where the sublime confidence that has sustained this enterprise suddenly seems to wane, and we are given two versions, one that of the first edition, and the other that of Rose, who puts back apostrophes that Joyce deliberately removed - and does this to the text of a man who nearly came to blows with the printers of the French translation over his insistence that standard French accent marks, cedillas, etc., be taken out.

In a sense, though, to take the edition on its own terms, to argue about the appropriateness or otherwise of one or another alteration, is beside the point: the point is that the changes-introduced by Rose are unwarranted, gratuitous and baseless. And to cite just these instances is to leave unmentioned such matters as the hyphenisation of compound words, the arbitrary capitalisation of words left uncapitilised, (in defiance, sometimes, of Joyce's explicit correction, e.g., Jesuit for "jesuit" on the first page), the gaps introduced into "Oxen of the Sun", and the typographical mess made of "Circe", (compare the clarity and elegance, of the presentation in the much-maligned first edition).

IT is most regrettable that the first edition of Ulysses produced in Ireland should be so seriously flawed. The fact that, as I wrote in these pages recently, there is a problem with the text of Ulysses, that the text may in some respects be inherently unstable, does not give any licence for so radical an overhaul as is undertaken here.

A contemplated legal challenge by the Joyce estate to this edition is apparently not going ahead, the deadline for claiming copyright infringement having now passed. It is as well that this should be so: attempts at suppression, as Joyce's own history shows, tend to be self- defeating.

More important than any legal challenge is the weight of public opinion. I do not believe that there is a massive public demand for a Ulysses made easy, a Ulysses with the crooked straight and the rough places plain. There could be a case for producing a first-draft version of Ulysses, as there is a first-draft version of Finnegans Wake, (in a way it already exists in the Rosenbach Manuscript, on which Rose has done much good work), but that is not what is on offer here. What is on offer here, unfortunately, is based on editorial principles so misguided that the text produced is terminally unreliable and both the reader and the author have been done a disservice.