29 January 2002

Audrey Horton

James Joyce's literary masterpiece Ulysses first published as a whole book in 1922 has recently been the subject of litigation once again. The title of this article may call to mind perhaps the obscenity trials of 1921 and 1933. This time the case involved interesting and in some respects untested concepts of the law of copyright and passing off. In the case of Sean Sweeney and Graham Camps (suing as trustees of the Estate of James Joyce) v Macmillan Publishers Ltd (trading as Picador) and Danis Rose, Mr Justice Lloyd was called upon to decide three issues which had come before the English High Court for the first time. In his judgment of 22 November 2001, he granted the Estate an injunction restraining copyright infringement in respect of a work entitled "A Reader's Edition" of Ulysses by James Joyce edited by Danis Rose and published by Macmillan/Picador. Lloyd J held that Mr Rose had infringed the copyright in manuscripts, drafts and other preparatory materials written by Joyce and published during his lifetime by including passages and text from the Ulysses drafts ("the Ulysses drafts") in his "Reader's Edition". Lloyd J also found that Mr Rose could not rely upon the transitional provisions reviving copyright in the UK: he had not made "arrangements" for exploitation of the work before the relevant dates in 1995 i.e. during the legal window when the original 1922 text of Ulysses (published in Joyce's lifetime) was out of copyright. Although the Judge found that the copying of the 1922 edition in "A Reader's Edition" would have been the subject of a compulsory licence under the implementing regulations, this was, on the facts, of no assistance in relation to the copying of "the Ulysses drafts", because these had never been out of copyright.

In addition to the inclusion in "A Reader's Edition" of material taken from "the Ulysses drafts", Mr Rose had also included in his book material which had no precedence whatsoever in either the 1922 text or in any other manuscript, draft or preparatory material written by Joyce. The Estate had also argued that the inclusion of such material not as written or as presented by Joyce meant that Mr Rose's book misrepresented to the public that it was the famous work which had come to be known as "Ulysses by James Joyce", and amounted therefore to passing off. The Judge rejected the Estate's case of passing off, largely on concluding that it was not possible to define objectively a class of editions of Ulysses, outside which "A Reader's Edition" could be alleged to fall.

Each of these three questions will be considered in turn, both as to the particular law and facts, and with a wider view as to more general implications if any, under the following headings:

  1. Construction of the UK Regulations implementing the Copyright Term Directive.
  2. Subsistence and Infringement of Copyright in Literary Drafts
  3. Passing Off in the context of Literary Works.
  1. Construction of the UK Regulations implementing the Copyright Term Directive

    The works of James Joyce, along with other literary figures such as D.H. Lawrence and Virginia Woolf, have exercised a peculiar fascination for copyright lawyers since EC Directive 93/98/EEC of 29 October 1993 (often known as the "Term Directive" required Member States to harmonise the duration of copyright under their laws to life of the author plus seventy years. In the UK this Directive was implemented by the Duration of Copyrights and Rights in Performance Regulations 1995 S.I. 1995/3297 with effect from 1 January 1996, six months later than the Directive required. Since copyright in the UK had previously lasted for fifty years after the death of the author, the effect of the Regulations was both to "extend" existing copyrights and to "revive" copyright in works which had gone out of copyright. James Joyce died in 1941. As a result copyright in the works published in his lifetime, including the 1922 edition of Ulysses, expired 50 years from the end of the year in which he died, i.e. on 1 January 1992. However, as a result of the Directive and implementing Regulations, the 1922 Ulysses has been protected in the UK by "revived" copyright from 1 January 1996.

    In contrast to the published works of Joyce such as the 1922 Ulysses, a large volume of complex preparatory materials, consisting of drafts, revisions, typescripts, proofs, manuscripts and so on, was not published until the 1970s. In particular the James Joyce Archive was not published until 1976-78 in a total of no less than 63 volumes. Copyright in respect of these materials which were not published in Joyce's lifetime lasts for 50 years from the end of the year in which publication takes place. The Defendants denied subsistence and infringement of copyright in such drafts. They admitted copying the 1922 Ulysses but claimed to have done so while it was in the public domain.

    In the case of "revived" copyright, the Directive indicates that Member States' implementing legislation should contain transitional provisions dealing with "the protection of acquired rights and legitimate expectations" of persons exploiting works at a time when they were in the public domain, but it does not dictate the precise form such provisions should take. Regulation 23 of the UK's implementing Regulations protects acts done after 1 January 1996, (such as here the publication of "A Reader's Edition") following "arrangements made" to exploit the work while the work was in the public domain.

    Mr Rose's evidence was that he did not start on his project until after 1 January 1992 and that he had completed the bulk (90%) of the work, necessary for "A Reader's Edition" by the end of 1993, i.e. during the time Ulysses was out of copyright. Unfortunately he did not keep a record of the work done as of that date. No contract was entered into between Mr Rose and Macmillan until an option agreement of 6 December 1995 which gave Macmillan the right to negotiate in good faith for terms for publication "until 3 months following ratification by Parliament of the revisions to the Copyright Act". This was followed by a publishing agreement signed on 13 August 1996 providing for delivery of the text by Mr Rose by 2 February 1997.

    To have the protection of Regulation 23 the "arrangements" have to have been made either before 1 January 1995 or 1 July 1995 (depending on which sub-section of Regulation 23 is relied upon). There are no UK cases construing the meaning of "arrangements": the only assistance given by the Regulations is that it means "arrangements for the exploitation of the work in question" (Reg. 23(5)). In essence the Defendants argued that Mr Rose's completion of most of his editing work by the end of 1993 was sufficient to protect him under the Regulation. The Estate argued that "arrangements", while potentially encompassing more than contracts, nevertheless required some kind of bilateral arrangement.

    The Judge accepted the authority of the ECJ in the case of Butterfly Music srl v Carosello Edizioni Musicali & Discographiche srl Case C-68/98 [1999] ECJ 1-3939 [2000] 1 CMLR 587, for general guidance on the underlying policy behind the Term Directive in the context of the adequacy of individual Member States' provisions safeguarding the economic interests of third parties acting in good faith while revived copyright works were in the public domain.

    In particular in his opinion the Advocate-General characterised the transitional provisions adopted by Member States as by their nature having to "be as narrow as possible, since they are equivalent to exceptions inserted into the general system for the protection of related rights which is created by the rules of the Directive. In other words when the national authorities are called on to give effect to the requirement rights of third parties, they must have in mind that that protection is by way of exception". The principal objective of the Directive in extending and reviving copyright was thus to be restricted as little as possible.

    Against this background, Lloyd J held that the very limited contacts Mr Rose made with publishers before 1 January 1995 did not amount to "arrangements for the exploitation of the work". He accepted that "arrangements" was not necessarily limited to "arrangements by way of contract" but he said "nevertheless they must be of some degree of solidity or certainty, such that it can be said that acts done later are done in pursuance of the arrangements". The great problem with a wide construction of "arrangements" such that it would cover the facts as presented here of Mr Rose performing the bulk of his work while Ulysses was out of copyright, was that it would be difficult to draw a distinction between that case and one where only a small amount of work was done before 1 January 1995, but the work was completed afterwards in pursuance of work started before then. In both cases it could be said the works were published in pursuance of arrangements made before 1 January 1995.

    The Judge also pointed out that a narrower construction of Regulation 23 did not sterilise the efforts of those such as Mr Rose made while the work was in the public domain. At least so far as the 1922 edition of Ulysses is concerned, the compulsory licensing (upon payment of a reasonable royalty) provisions of Regulation 24, would permit publications of work done while Ulysses was out of copyright. However "A Reader's Edition", which goes beyond the 1922 edition and infringes the copyright in underlying drafts, could not take advantage of Regulation 24.

    Had it been relevant on the facts, the Judge did find that the notice given by the Defendants which was couched in conditional or contingent terms (i.e. as a fall-back position in case the claim to be free to publish under Regulation 23 was not made out) constituted a valid notice under Regulation 24 entitling publication of the 1922 text of Ulysses upon payment of a reasonable royalty. The Judge puzzled over the requirement for "reasonable notice" in Regulation 24 since the copyright owner cannot stop the intended acts once notice is given, but he rejected the argument that it was necessary to supply a copy of the intended publication with the notice. The Notice requirement simply meant publication must be made openly and not clandestinely and could not be used to justify acts retrospectively.

  2. Subsistence and Infringement of Copyright in Literary Drafts

    James Joyce's Ulysses is in the Judge's words "famously complex and radical in its literary style". Joyce wrote his masterpiece over a long period of time, during a process in which he created very many drafts, revisions, manuscripts, typescripts, proofs and working papers. Many of these have further manuscript alterations and amendments, as Joyce continuously revised his work at each stage of writing, for example on proofs supplied to him by the printers. The result is that there are for each of the 18 episodes of the novel various more or less recognizable layers of earlier and later draft manuscripts (Joyce did not type and wrote his novel by hand), followed by further layers of revised typescripts, with for example further revisions on the corrected typescripts returned to Joyce, (sometimes there would be more than one copy of the typescript each with different alterations), and thereafter further manuscript alterations on proofs prepared by typesetters. Some preparatory materials have been lost: very many remain in existence and those extant were published in the James Joyce Archive from 1976-78. Occasionally even now a previously unknown manuscript is discovered. Joycean scholars actively debate the status and relative authority of each stage in this creative process.

    Joyce also prepared what has come to be known as the "Rosenbach manuscript". This was created by Joyce episode by episode as a fair copy of the text of Ulysses as it then stood, in order to be sold to a collector named John Quinn in the USA who had offered to buy the manuscript of Ulysses in instalments between 1920 and 1922. Because Joyce continued to revise the text of episodes after the fair copy was made, the Rosenbach manuscript is not the same as the 1922 first published edition of Ulysses. The use by Mr Rose in his "Reader's Edition" of material contained in the Rosenbach but not in the 1922 edition was critical in establishing that the Ulysses drafts had been copied.

    Subsistence of Copyright

    Against the factual background of the complex creative process described above, the legal question of whether copyright subsists in the Ulysses drafts fell to be considered. The Joyce case required the Court to consider the threshold of originality for literary drafts. Copyright subsists in a "literary work" if it is "original" in the sense that it is the result of expenditure by the author of skill, judgment and labour. Clearly copyright subsists in Ulysses as a whole. The Judge also held that the Rosenbach manuscript, once complete, comprised a copyright work in itself, and he further found that the successive typescripts and proof stages each constituted new copyright works, and the earlier drafts before Rosenbach, including notebooks setting out early versions of particular passages or sections, also constituted separate copyright works. Only the final proof with Joyce's amendments on it, identical with the text published in 1922, was the same copyright work as the first published edition of Ulysses.

    The case law cited in the judgment mostly concerned whether copyright subsists in successive versions of a particular design or drawing. No relevant modern authority was found concerning copyright in the earlier drafts of a literary work. Perhaps the closest authority on point was Robin Ray v Classic FM Ltd [1998] FSR 622. In that case five documents and a catalogue had been prepared as preliminary stages to the creation of a database for programming selection. The defendants in Robin Ray argued that the copyright in those preparatory works was "subsumed in" and "merged in" the copyright jointly owned by the plaintiff and the defendant in the database. Lightman J. found this argument to be "wholly repugnant to the basic principles of the law of copyright" and said: "Where a work goes through successive stages in writing in the course of its creation or development, the copyright in the earlier of the writing subsists whatever the use later made of the material contained in it".

    Interlego AG v Tyco Industries Inc [1988] RPC 343 is authority for the proposition, in the case of copyright in engineering drawings for Lego bricks, that minimal alteration and very slight amendments might mean that later drawings would not enjoy a separate copyright because they would not be sufficiently original. However cases such as LA Gear Inc v Hi Tec Sports [1922] FSR 121 concerning copyright in designs for shoes, make it clear that each preliminary version of a finished drawing may be an original work. Nourse LJ presciently commented: "It would be an extraordinary state of affairs if the law were otherwise. Indeed it might have far-reaching consequences on other literary and artistic works, for example on the manuscripts of books and plays". In LA Gear the Court of Appeal thought the principle that copyright may subsist in preliminary versions of a work was so obvious it needed no authority to support it. The Joyce case has now provided such an authority specifically in relation to literary drafts, but in doing so is merely following long-established and basic principles of copyright law.

    Infringement of Copyright

    Having held that copyright subsisted in "the Ulysses drafts" and that (except in so far as text in those drafts was in fact published in the 1922 edition of Ulysses) these materials were not published during Joyce's lifetime, the Judge had to consider whether what Mr Rose had done infringed the copyright in those materials. The Judge accepted Mr Rose's evidence that he had started with the 1922 text and made alterations to that based on the unpublished drafts. He rejected the Estate's argument that by using the Rosenbach manuscript, in the manner Mr Rose admittedly did when he produced "A Reader's Edition", Mr Rose had systematically compared the whole of its text against the whole of the 1922 edition and therefore had used and in effect copied the whole of the text of the Rosenbach (not just the passages in fact transposed into "A Reader's Edition").

    As a result of an earlier interlocutory ruling by Jacob J, the issue of infringement in "the Ulysses drafts" was to be decided on the basis of 8 sample pages, 4 selected by the Estate and 4 by the Defendants. (Ulysses is often said to be notoriously difficult to read: perhaps the Judge felt apprehensive about the whole of the novel being possible required reading, as well of course as wishing to save time and costs).

    Mr Rose had clearly had access to "the Ulysses drafts": he had assisted in particular in the preparation of the controversial Gabler edition of Ulysses which had, under authority from the Estate, made use of the unpublished manuscripts and drafts. However there was no real issue about whether Mr Rose had used the drafts: he clearly admitted consulting them in order to produce the text by "A Reader's Edition" in the Introduction to his book. Indeed the main selling point used by the Defendants to market their new edition of Ulysses was that Mr Rose had used all the various earlier "Ulysses drafts" to edit the text, and to eliminate mistakes in the 1922 text based on his selection of variants drawn from the manuscripts and other draft materials.

    The only question was whether what was taken by Mr Rose from the Rosenbach manuscript or other preparatory materials amounted to a substantial part of the text in which the prior copyright subsisted. Here again the Joyce case exemplifies another basic and well-established tenet of copyright law, in Lloyd J's words: "Whether part of a copyright work which has been copied is a substantial part depends as much on qualitative as much as quantitative factors".

    The fact that the passages taken by Mr Rose, for example lines taken from the Rosenbach which were omitted from the 1922 edition, are only a few lines, or less than a line, here and there, did not exonerate Mr Rose. The importance of the text being restored was the very fact of its being restored to the text since its inclusion was one of the things said to make "A Reader's Edition" special and distinct. Despite the relative brevity of the passages taken in relation to the 8 selected pages, Lloyd J held that they did amount to a substantial part of the text of the Rosenbach manuscript, and therefore their copying from that source infringed the copyright in that work.

    The Judge might also have added that in some cases what was taken was less than a line, or even less than a word: sometimes a variation in punctuation might be significant in the context of a novel by Joyce.

    However small the percentage might be in numerical terms, what was taken was a substantial part because it was the most significant or important part of that copyright work.

  3. Passing Off in the context of Literary Works

    The Estate did not succeed in establishing that publication of "A Reader's Edition" amounted to passing off. The alterations to the 1922 text of Ulysses made by Mr Rose fell into two broad categories: firstly, those resulting from Mr Rose's use of the unpublished Ulysses drafts (the subject of the copyright claim); and also a second category of alterations which had no precedence in any source materials as written or presented by James Joyce, but rather emanated from Mr Rose's unusual editing style whereby Mr Rose interposed himself as the copy-editor Joyce might have had if he had been alive today. The latter highly controversial changes made by Mr Rose included, for example, the correction of "factual mistakes" made by Joyce, the elimination of Joyce's characteristic use of compound words, changes in syntax, spelling, grammar, capitalisation, format and, very importantly, changes to Joyce's unorthodox punctuation. For example Mr Rose inserted punctuation including apostrophes into the Penelope episode, commonly known as "Molly Bloom's soliloquy", her famous "stream of consciousness" monologue representing her unstructured rambling thoughts. It is the second category of alterations to the text of Ulysses which gave rise to the passing off claim.

    The Estate argued that the changes and alterations to the text of Ulysses made by Mr Rose were so fundamental and made it so different from what had previously been published under the title "Ulysses by James Joyce" that "A Reader's Edition" could not properly be described as "Ulysses by James Joyce". The term "Reader's Edition" had no accepted meaning and did not convey to potential buyers and readers of the book that what they were getting was different from the original text of Joyce as written and presented by him.

    Lloyd J clearly had difficulty fitting the facts of the case into the boundaries of what he described as more conventional passing off cases. The Alan Clark diaries case (Clark v Associated Newspapers [1998] 1 WCR 1558) was an example of a case where the goodwill of a living writer in his work was protected by a passing off action but was considered by the Judge to be distinguishable from the case of an Estate protecting the goodwill of a deceased author.

    Lloyd J applied the passing off principles as formulated in the Advocaat case [1980] RPC 93. He considered that this was a "class passing off case" but different from the usual case of this kind because there was not a class of traders, all entitled to share the goodwill of Joyce's works, as in the champagne, advocaat and Swiss chocolate cases. While on the facts this may have been so, the Judge may have been wrong to conclude that the goodwill of publishers of Joyce's works was not the same as the goodwill held by Joyce's Estate because they "do not sell the product at all". He did, concede (without deciding the point) that the Estate's commercial interest, as copyright owner, in sales of Ulysses which attract a continuing royalty from licensing, might be sufficient to enable it to satisfy the first Advocaat principle, namely that to show that its "business consists of or includes selling in England a class of goods to which the particular trade name applies".

    However, the Judge found the Estate failed at the second Advocaat hurdle, because they could not show "that the class of goods is clearly defined, and that in the minds of the public or a section of the public, in England, the trade name distinguishes that class from other similar goods" (emphasis added).

    Lloyd J considered that the Estate had failed to define the class of goods with sufficient clarity and in particular that there was no "objectively definable class" (emphasis added).

    At this point it is necessary to refer again to the background facts relating to and creation and publication of Ulysses in order to understand the special difficulty in this case of defining a category or class of editions of "Ulysses by James Joyce" into which "A Reader's Edition" did not fall. When publication of the first intended editions of Ulysses was frustrated in the UK and the USA by the obscenity laws, Ulysses was published in France by a French printer who did not understand English. The 1922 first edition suffered as a result from a larger than average number of typographical and printer's errors, a problem compounded by Joyce's tendency (as explained earlier) to make late changes even to text previously approved by him at an earlier proof stage, and in the latter stages also by his failing eyesight.

    In 1984 the "Gabler edition" of Ulysses was published, with the approval of the Estate, in an attempt to produce a corrected text of Ulysses by reference to Joyce's manuscripts and other preparatory materials. It was published in two forms: a "corrected version" as a reading text on its own, and also as a scholarly edition with the right hand page showing the reading text and the left hand page showing all relevant variants from which the reading text was derived.

    However, Joycean scholars vehemently disagreed about the variants selected by Gabler for the "corrected" text. As a result of the ensuing controversy two members of the Gabler editorial committee, Dr. Gaskell and Professor Hart, later published their own list of corrections to the 1922 edition of Ulysses. In the judgment this is reformed to as "the Hart/Gaskell repair kit" and, like the Gabler edition, it drew upon the Rosenbach manuscript and other material published for the first time after Joyce's lifetime in the James Joyce Archive.

    As a result there were in existence different versions of the text of Ulysses where corrections had been attempted to the 1922 text, and clearly there was continuing academic dispute and debate about the selection of which variants from the Ulysses drafts should be incorporated into the corrected version.

    Against this background the Judge rejected all of the following alternative definitions of the class put forward by the Estate.

  1. Ulysses by James Joyce as characterised by certain distinguishing features viz. narrative technique, unconventional syntax, motifs and internal echoes, errors of fact, allusions and formatting.
  2. Ulysses as written by James Joyce and as presented in works published prior to 1997.
  3. Ulysses as written by James Joyce and published in his lifetime with his approval or as published thereafter with the approval of his Estate.

The Judge rejected the first formulation as inherently uncertain on the basis that the characteristics of the work, or absence thereof, could only be determined by reference to academic opinion at any given time which would be "an impossible way to define a class of products which, in the mind of the public, has acquired a collective goodwill such that one work falls within it and another can clearly be seen to fall outside it".

The Judge also rejected the third formulation, on the basis that any defining characteristics which involved or allowed for the consent of the Estate to be given or withheld imported a degree of subjectivity which conflicted with the requirement of an objectively definable class which the Judge perceived to be required from the formulation of the law in Advocaat.

The second formulation is also rejected presumably on the basis of inherent uncertainty and/or subjectivity. While the same objection that only fluctuating academic opinion could determine which particular variant from the Ulysses drafts can be said to be "Ulysses as written and presented by James Joyce", there can surely be no uncertainty or subjectivity where material is added to the text which has no precedence at all in any of the underlying Ulysses drafts. The objectively definable class would be editions of Ulysses which contain only material written by Joyce (in any of its variant forms) such as the Gabler edition. An "edition" which contained extraneous material such as "A Reader's Edition" would fall outside this class.

In the most recent Court of Appeal authority Chocosuisse Union des Fabricants Suisse de Chocolat v Cadbury Limited [1998] RPC 117 (Laddie J) [1999] RPC 826 (Court of Appeal). The Court confirmed that the fact that the boundaries of the class are amorphous and cannot be precisely drawn does not mean that the class of goods cannot be defined.

The Judge left open the possibility that there could in theory be a publication passed off as Ulysses by James Joyce which did fall outside some clearly and objectively definable class but he did not see how any of the definitions proposed could encompass the various editions of Ulysses but not "A Reader's Edition". If, however, "written by Joyce" is understood to mean "excluding extraneous material not written by Joyce" perhaps such an objectively definable class could have been formulated. The law of passing off could then have been used to ensure that consumers of famous works of literature would be entitled to rely on names or labels applied to such famous works, to know that they have something unchanged and unadulterated, in the same way as consumers of champagne, advocaat or Swiss chocolate can be assured that what they have is "the real thing".

Also published in the European Intellectual Property Review (Issue 5) in May 2002.

Important - The information in this article is provided subject to the disclaimer. The law may have changed since first publication and the reader is cautioned accordingly.