SOMETHING INTENDED, COMPLETE

1Biographers have addressed the fullness of Yeats’s life, but his self-fashioning and self-mythologizing compel a parallel task, the biography of his books. Biblio-biography is often confined to the biography of a single book,2 but Yeats was a pleromatist. Committed to oeuvre throughout his life, he updated his collected works as a self-image, and his canon-formations involved the relegation of works which did not fit his idea of a Collected Works as a ‘permanent self’. ‘It is myself that I remake’ was his reply to those who regretted this textual husbandry.3

2Yeats’s biblio-biography, then, will be a history of these paradoxical intentions. Ideally, the writing of such a book would depend upon a fully edited Collected Works. Regrettably, the editing of Yeats’s works has yet to address the ‘fullness’, of such an edition, and that is my subject in this present essay. At the heart of his idea of himself as a textual pleroma was a driving awareness of audience. Speaking later in life to Will Rothenstein he

… quoted his brother Jack as saying he painted to please himself and that the public chose to pay him. This was not Yeats’s attitude to poetry: ‘You must remember your audience, it is always there, you cannot write without it’.4

3Both as a young poet and increasingly after the establishment of the Dun Emer Press, his individual assemblages were first tested on coterie audiences whilst having been written or rewritten with this wider ambition, though not without a compelling sense of the primacy of his Irish audience. The two-volume The Poetical Works of William B. Yeats (1906–1907) sought to bring to American audiences some sense of his strategy for growing his Irish audiences by way of the theatre

I am no longer writing for a few friends here and there, but asking my own people to listen, as many as can find their way into the Abbey Theatre in Dublin or some provincial one when our company is on tour. Perhaps one can explain in plays, where one has so much more room than in songs and ballads, even those intricate thoughts, those elaborate emotions, that are oneself.5

4By contrast, The Collected Works in six volumes of 1922–1926 was a reading edition designed for the much greater variety of audiences he perceived himself to have, and which he wished further to foster. The concept of the reading edition, is returned to below, but it is worth saying here that within his audiences, Yeats above all encouraged that coterie of those who had ‘read all that I have written’6 and who would read new works in the light of that whole. Perhaps it is best to write of that ‘whole’ using his word, ‘work’. For Yeats, ‘work’ was both activity and what it realized, beyond individual works and even the collected works—‘“The work is done,” grown old he thought, | “According to my boyish plan”’, ‘Perfection of the life, or of the work’, the spiritual intellect’s great work’(VP 577, 495, 632). The use of the word goes well beyond the writing and staging of plays, and the repeated use of the word in the following passage of 1906 is typical

My work in Ireland has continually set this thought before me: ‘How can I make my work mean something to vigorous and simple men whose attention is not given to art but to a shop, or teaching in a National School, or dispensing medicine?’ I had not wanted to ‘elevate them’ or ‘educate them’, as those words are understood, but to make them understand my vision, and I had not wanted a large audience, certainly not what is called a national audience, but enough people for what is accidental and temporary to lose itself in the lump.7

5Oeuvre, then, is of key importance to the understanding of Yeats’s overall intentionalism as well as for the understanding and application of his authorial intentions. He uses the word ‘intention’ and various forms of it scores of times in critical prose relating to his own work, and in his reviewing. Intentions are articulated or envisaged for poems, plays, books, and works, as well as textual or versional changes, as textual ambitions or (to be more precise) bibliographical foreconceits, and even in terms of intended audiences. Footnoting his obsession with the paradoxes of sudden changes in historical cycles, by an appeal to A Vision’s esoteric audience, Yeats tells us that this difficult book is ‘intended, to use a phrase of Jacob Boehme’ s, for my “schoolmates only”’.8 The remark is not intended to exclude the exoteric audience who are to be teased into that expertise whereby his work can be read in a specialist light: he encourages learning by an appeal for learned feedback.9

6The ‘first principle’ in Yeats’s ‘A General Introduction for my Work’ —arguably designed as his final statement on his own writing—begins from an emphatic distinction.

A poet writes always of his personal life, in his finest work out of its tragedies, whatever it be, remorse, lost love or mere loneliness; he never speaks directly as to someone at the breakfast table, there is always a phantasmagoria. Dante and Milton had mythologies, Shakespeare the characters of English history, of traditional romance; even when the poet seems most himself, when he is Raleigh and gives potentates the lie, or Shelley ‘a nerve o’er which do creep the else unfelt oppressions of mankind’, or Byron when ‘the heart wears out the breast as the sword wears out the sheath’, he is never the bundle of accident and incoherence that sits down to breakfast; he has been re-born as an idea, something intended, complete. A novelist might describe his accidence, his incoherence, he must not, he is more type than man, more passion than type. He is Lear, Romeo, Oedipus, Tiresias; he has stepped out of a play and even the woman he loves is Rosalind, Cleopatra, never The Dark Lady. He is part of his own phantasmagoria and we adore him because nature has grown intelligible, and by so doing a part of our creative power…. The world knows nothing because it has made nothing, we know everything because we have made everything.10

7When asked for up to eight prefaces for the volumes of the planned Dublin Edition, Yeats had demurred. Much as he hated writing prose late in life, he did wish to make a comprehensive statement for his entire oeuvre, and he worked at ‘the’ (or ‘my’) ‘general introduction’11 as he referred to it throughout its composition, which occupied him exclusively for more than two months.12 Remarkably attuned to his idiom, Mrs Yeats and/or Thomas Mark, came up with the title ‘A General Introduction for my Work’ when that essay was finally yielded up by Scribner to Yeats’s primary initiating trade publishers Macmillan in London, who prepared it for publication in the context of Essays and Introductions (1961).13

8Yeats’s professionalism as a writer included a willingness to delegate certain textual decisions to trusted agents, so much so that he could be impatient when his instructions were compromised, e. g., by Elkin Mathews, and even irascible on one famous occasion in 1907 when A. H. Bullen, starved by Yeats for copy, seemed about to print from an unrevised text.14 He was otherwise flexible with such trusted editors as Bullen and, later, the great Thomas Mark (1890–1963) of Macmillan, London, to whom he delegated increasingly.15 As his career drew to a close, he came to rely on the established mechanisms of a publishing process designed to reconcile the making of money with the idiosyncratic needs of a distinguished (and never very profitable) Nobel Prize winner.

9He had long thought of his work in terms of collected editions, and these he saw as canon-forming, even if by definition, no lifetime iteration of such an edition could ever stabilize or complete the canon. Consequent upon canon-formation were acts of relegation. Certain poems, plays, and other writings fell from time to time into a deutero-canon, or were consigned—as he hoped—to oblivion.16 He was firm (and not always popular with his older readers), and suppressed numerous early poems, such as ‘How Ferencz Renyi kept Silent’ which George Russell kept reprinting against Yeats’s wishes, whilst remaining more ambivalent with works such as with John Sherman AND Dhoya.17 On other occasions he would rewrite rapidly and so transformingly that older texts in all genres were supplanted (The Golden Helmet comes to mind as a limit case, a work in all likelihood prematurely published so as to assist in balancing the volume length of the various volumes of the 1908 Collected Works in Verse and Prose, and then entirely recast in verse two years later as The Green Helmet). New texts constantly drove out the old, perturbing—even confounding—the possibility of completeness. Writings such as reviews and numerous pieces of journalism were left in the limbo or oblivion of periodical archives. Beyond the deuterocanon, too, there are the lectures, the earlier writings published since his death, abandoned pieces such as Autobiography—First Draft and The Speckled Bird.18 I refer to the relation between the canon, the deutero-canon, and the various categories of suppressed, abandoned, unpublished and unfinished writings as Yeats’s textual hierarchy. It still presents challenges for editors.

10Realism in respect of textual revision for new bibliographical occasions may be traced in the followings exchanges. In 1901, Lafcadio Hearn protested about the new version of ‘The Host of the Air’:

You have mangled it, maimed it, deformed it, extenuated it—destroyed it totally. … you have really sinned a great sin! Do try to be sorry for it!—reprint the original version,—tell critics to go to perdition, if they don’t like it,—and, above all things, n’y touchez plus!

11Immediately after a (lost) temporizing reply, Yeats wrote to Thomas Hutchinson, defending himself against those who had not liked rewritten versions as found in Poems (1899 and after): ‘One changes for the sake of new readers, not for the sake of old ones’,19 showing that audience was to the forefront in that constant reconstruction of books and texts con sequent upon Yeats’s continual self-construction. ‘Whatever changes I have made are but an attempt to express better what I thought and felt when I was a young man’ he said in 1925 (VP 842), fully aware of the attendant paradoxes. And again, in 1927,

this volume contains what is, I hope, the final text of the poems of my youth; and yet it may not be, seeing that in it are not only the revisions from my ‘Early Poems and Stories’, published last year, but quite new revisions on which my heart is greatly set. One is always cutting out the dead wood. (VP 848)

12Yeats’s shaping at every point offers a unified self-reconstruct with its own chronology. Its profile—the young man old or the old man young?—is where his reading, his anticipation of his audiences, their reading, and their feedback intersect, often in new revision.20 Revision and new writing were not merely interdependent activities, revision of an old poem frequently made the next new poem possible.21

13Such diachronic self-allusions show that his own text had a continuous simultaneity for him which makes oppressive the exclusion of such authorial commentary as I have just quoted. Above all, as he read his Collected Works when assembling the Edition de Luxe in the 1930s, he read it as a single work, cancelling, for example, one passage in Autobiographies which simply replicated another in ‘Dust hath closed Helen’ s Eye’from The Celtic Twilight, and already in proof in Mythologies AND The Irish Dramatic Movement.22 Such selfreading and creative economy silently challenges the editorial avarice of The Poems: A New Edition and its derivatives. These editions stripped out of the plays and the prose various verses never printed as ‘poems’ in their own right by Yeats himself. Adding them to the ‘Additional Poems’ section, which otherwise legitimately collected Yeats’s abandoned poems was a decision widely criticized, but the enduring point here is not merely that by such means what Hugh Kenner called ‘uncanonized scraps and three-line wonder[s]’ were presented as poems, but also that they were thereby printed twice in the Collected Works.23

LE GRAND OEUVRE

14Like Stéphane Mallarmé, in whose tradition he wrote,24 Yeats was committed to the idea of the ‘work’ being externalized in a book, or several volumes of a book, with an overarching architecture. Mallarmé’s ‘tout, au monde, existe pour aboutir à un livre’25 resonated with him via Arthur Symon’s ‘Stéphane Mallarmé, an essay Yeats read carefully and uses for quotes he attributes to Mallarmé. Symons had also alluded to Mallarmé’ s ‘Le Livre, instrument spirituel’ in his own words in that essay, catching Mallarmé’s synaesthesia.

That we are now precisely at the moment of seeking, before that breaking up of the large rhythms of literature, and their scattering in articulate, almost instrumental, nervous waves, an art which shall complete the transposition, into the Book, of the symphony, or simply recapture our own: for, it is not in elementary sonorities of brass, strings, wood, unquestionably, but in the intellectual word at its utmost, that, fully and evidently, we should find, drawing to itself all the correspondences of the universe, the supreme Music.26

15Many of Yeats’s references to his ‘work’ catch the alchemical rapture implicit in Mallarmé’s idea of the Great Work, whereby the artist is a spiritual alchemist and the book a vessel of God: thus Yeats’s references to the ‘supreme art’ and the ‘spiritual intellect’ s great work’.27

‘… le Grand Oeuvre. Quoi? C’est difficile à dire: un livre, tout bonnement, en maints tomes, un livre qui soit un livre architectural et prémédité, et non un recueil des inspirations de hazard [sic], fussent-elles merveilleuses… J’irai plus loin, je dirai: le Livre, persuadé qu’au fond il n’y en a qu’un, tenté à son insu par quiconque a écrit même les Génies. L’explication orphique de la Terre, qui est le seul devoir du pöete et le jeu littéraire par excellence: car le rythme même du livre, alors impersonnel et vivant, jusque dans sa pagination, se juxtapose aux équations de ce réve, ou Ode.28

Barbara Johnson translates:

[T]he Great Work. What would it be? … a book, quite simply, in several volumes, a book that would be a real book, architectural and premeditated, and not a collection of chance inspirations, however wonderful… I would go even further and say the Book, convinced as I am that there is only one, unwittingly attempted by anyone who writes, even Geniuses. The orphic explanation of the Earth, which is the poet’s only duty and the literary mechanism par excellence: for the rhythm of the book, then impersonal and alive, right down to its pagination, would line up with the equations of that dream, or Ode.29

16Yeats also conjures out of Mallarmé’s idea of the Book as Spiritual Instrument, a sense of the imminence of some new dispensation to be expressed in a new ‘Sacred Book’ of the Arts. Rosa Alchemica and The Tables of the Law are filled with the idea, and the latter ties the idea of a new aesthetic revival onto the coat-tails of Joachimist prophecy. Yeats’s formulation (citing sometimes Mallarmé, sometimes attributing the idea vaguely to Verhaeren, or Nerval) that ‘our whole age is seeking to bring forth a sacred book’, is a catchphrase recalling a remark he accorded to Owen Aherne in that story, that ‘world only exists to be a tale in the ears of the coming generations’.30 ‘Some of us thought that book near towards the end of last century, but the tide sank again’, said Yeats of the fin de siècle.31

17But while it is fair to say that Yeats’s Collected Works was intended to be a ‘Great Work … a book, quite simply, in several volumes, a book that [is] a real book, architectural and premeditated’, it was never envisaged as the Sacred Book of the fin de siècle. The Thirties offered more ample opportunities for fulfilment at life’s end of that great self-embodiment, with Macmillan offering him a seven volume Edition de Luxe. When the Edition de Luxe had to be shelved during the Great Depression, Charles Scribner’s Sons offered an American alternative, to be named the Dublin Edition, also to be sold in sets. These were, however, turbulent times for publishing, and both editions (the Macmillan Edition de Luxe having been renamed the Coole Edition after his death) were halted by the Second World War. Yeats, however, died with his work incomplete—his oeuvre shelved. Trusted delegates—his wife George, his agents at A. P. Watt, and above all his publisher’s reader at Macmillan—were all in place for when the time came, as it did, but not until well after his death and the War. And when it did, publishing conditions, audience needs and audience expectations, had inevitably moved on.

18In the 1970s the expiry of Yeats’s European copyrights could be realistically envisaged, and academic pressure grew to re-edit his works on a grand scale. The re-establishment of texts is often thought to create new copyrights. Whether and, if so, to what extent such editing can have such an effect has never been legally tested, but commercial pressure from the Yeats Estate and Yeats’s principal publishers, Macmillan (London) began to chime with academic ambition. What became the new Collected Works (projected as fourteen volumes) was commissioned in the 1976 and inaugurated with the publication of the first volume, The Poems: A New Edition, in 1983. Issued—as the six-volume Collected Works had been in the 1920s—in a series of books published as and when they became available, and as yet (2017) unfinished, the individual volumes of the new Collected Works have obscure and greatly divergent relations with the archival remains of Yeats’s plans for collected editions of the 1930s. In other respects, too, it could not be said that that they embody a sustained (or even a sustainable) editorial policy. They vary erratically in the style and quality of textual approaches, decisions and annotational policies.

VAIN GAIETY: LOST ALLUSIONS, BIOGRAPHICAL DELUSIONS, SPECIFIC CONFUSIONS

19The current Collected Works (Macmillan and Charles Scribner’s Sons), the Cornell Yeats Manuscripts Series (which would seem to have abandoned the idea of including the MSS of Yeats’s Prose) and the Collected Letters (Clarendon Press) are all sufficiently far advanced for it to be apparent that the annotation for such series require continuous, essential maintenance.32 While these projects are all very different from each other, it is the difference in their respective annotational policies which is most immediately striking. By focusing below only on annotation of the Poems, it is possibly to get to the heart of these differences.

20Since the concept of a reading edition will be increasingly important to what I have to say about both annotational and textual policy, let me say here that such an edition would be a point of entry for readers unfamiliar with the history of Yeats’s texts. The Variorum Edition (1957 and later) with its collations of all previous published versions of texts against a ‘final’ text, is not that book, though the so-called ‘difficulties’ of using that edition have been absurdly overstated by those with interests vested in undervaluing the abilities of potential readers.33 While a reading edition of Yeats’s works will offer information about and possibly pathways to earlier and discarded versions of texts and the contexts of those texts—perhaps in appendices, notes, or commentary—it must be, in essence, a final text edition established according to patient understanding of the full, distributed publishers’ archives (which in Yeats’s case are very widely distributed), and of what they reveal about the author’s intentions for such reading editions of the whole of his works. In all the turmoil of late twentieth century editorial theory, it was easy to be detained by the intricacies of ‘versioning’ theory, but one cannot lose sight of the necessity of having carefully-compiled, accurate reading editions of final texts, if only as a stable point of departure for more specialized varieties of reading and scholarship.34

21The three principal editors of such reading editions of Yeats’s Poems as we have, Richard Finneran, A. Norman Jeffares and Daniel Albright, are all dead. All were intentionalists, though divided in their approach to usage and interpretation of the archives.35 In annotational policies, they are just as sharply divided. In both texts and their annotation, readers are presented with qualitatively different experiences. Yeats’s ‘fullness’ as a writer is at the heart of the problem which the Finneran edition throws onto its readers, while it offers to Jeffares and Albright the possibilities of vastly superior annotation. The late twentieth century editing of Yeats became something of a test bed for annotation policy, and it is there that we start.

22The stated ‘purpose’ of the notes to The Poems: A New Edition was threefold:

1. to annotate all specific allusions in the poems.36 Annotation of other kinds, as well as interpretive [sic] commentary, has been avoided. Thus, for example, information on Yeats’s sources is given only when Yeats called attention to them (whether in the poem or in a note), as with ‘Imitated from the Japanese’.

2. Cross-references to Yeats’s other works or passages in his correspondence are not offered, except in some rare instances where such references provide the most concise annotation.

3. Unnamed individuals are normally not identified, except in poems explicitly presented as autobiographical statements, as with the work beginning ‘Pardon, old fathers’. (PNE 613; CW1 623)

23It seems that from the outset, little or no thought was given to the mysteries inherent in the concept of a ‘specific allusion’ —specific to whom and when, and if so, how was such a specificity to be preserved? Readers soon found that their superior knowledge and so confident perceptions of ‘specific allusions’ did not pass muster as ‘specific’ to the editor of Collected Works 1. A telling example is Yeats’s hope to ‘dine at journey’s end | with Landor and with Donne’ (VP 336). No reader looking across Yeats’s repeated references to both poets could be expected to be satisfied by ‘Walter Savage Landor (1775–1864), John Donne (1571 or 1572–1635) English authors’ (CL1 641). From at least 1896, Yeats had known that ‘I shall dine late; but the dining room will be well-lighted, the guests few and select’ was a sentence from Landor’s confession that ‘Poetry was always my amusement; prose my study and business’ in his late ‘Imaginary Conversation’ with Archdeacon Hare.37

24Just when the context of Yeats’s wider reading and lived life can be drawn upon to amplify meanings in the text, the interdiction against ‘[c] ross-references to Yeats’ s other works or to passages from his correspondence’ confers authority solely on the Editor himself. This might account for the highly competitive, ‘finger on the button’, pub-quiz style of the annotation.38

25It ‘did not seem possible to assume a “common body of knowledge” which all readers’ throughout the world ‘would share’, confides Finneran.39 By contrast, Michael Sidnell condemned CW1’s busy and inutile annotations as the ‘dead mice in the bread’ (YA3 229). While much was overly and yet inadequately annotated, more was mysteriously and deliberately not annotated at all. Richard Ellmann was always fascinated by the ‘latent biography’ as he put it to me, of poems and plays. Foreseeing the monopoly which the Finneran edition might come to hold in the US market, its annotation was, as he put it bleakly to me, ‘a blight’.40 Numerous samples of baffling notes are listed in my TLS review of The Poems: A New Edition.41 And audiences who are challenged by the candour of ‘untraced’ face instead the smokescreen of knowingness, or what might be termed the wild goose chase after a red herring, as in the edition’s account of ‘Tulka’ followed, inevitably, in current editions and student guides.42

26Nowhere is this more obvious than in the stricture against identifying ‘[u]nnamed individuals’. Protesting his old New-Critical self-denial whilst in fact denying readers’ needs, Finneran turns to ‘Upon a Dying Lady’ which omits Mabel Beardsley’s name. Ernest Rhys asked Yeats in a letter of 1 June 1934 whether the Dying Lady in fact had been Mabel Beardsley. Yeats’s reply

has not yet been traced, but if it is ever located, I suspect we will discover that Yeats answered, in effect, that although the poem may have been ‘about’ Mabel Beardsley, he preferred to present it as a universal statement about death and dying, much as he told Lady Gregory that ‘To a Wealthy Man…’ was addressed to ‘an imaginary person’ (see note to 114.4).43

27Yeats himself, of course, had profoundly seen that ‘[w]e may come at last to think that nothing exists but a stream of souls, that all knowledge is biography’.44 It is alarming enough when an editor projects upon an author’s intentions and sense of audience his own critical preconceptions.

28But it is doubly disturbing that an editor should forswear what is out there in the textual pleroma because he declines to use cross references. Yeats had written to Lady Gregory before Mabel Beardsley was dead or even the sequence finished,

I have written three little lyrics about Mabel Beardsley dying but I will not send them till I have done a fourth to link them together. I think they are really quaint and touching.45

29His letter to Thomas Sturge Moore of 17 November [1918] asks Sturge Moore to design the cover for The Wild Swans at Coole (1919), commenting that it contains ‘all my recent poems, the Mabel Beardsley poems and so on’ (TSMC 33). A copy of The Wild Swans at Coole (Cuala, 1917) was inscribed to James A. Healy in July 1938 ‘The poems headed “To a dying Lady” are about Mabel Beardsley’ with a touching reminiscence of that ‘heroic person’ and her kindness to visitors when she was dying (YAACTS 8 252. Having edited that volume of Yeats: An Annual of Critical and Textual Studies, Finneran did not revise his passage.

30Daniel Albright condemned with vigour what he called ‘Professor Finneran’s revulsion against biography’

The consequence of this policy is that the protagonist of ‘An Irish Airman foresees his Death’ is not identified as Major Robert Gregory; the statue discussed in ‘A Bronze Head’ is not identified as an image of Maud Gonne; and so forth. Professor Finneran is right to note that Yeats had some purpose in omitting such names from the text of his poems, but if an annotator tells us anything, he should tell us those names. If one assumes that ignorance is helpful to interpretation, then any annotation whatsoever is harmful to the text. Professor Finneran’s and Professor Jeffares’s conflicting speculations on the identities of Yeats’s remote ancestors, to whom Yeats dimly alludes in ‘Pardon, old fathers’, are of some interest; but the intimate, direct affiliation of the poems just mentioned with the lives of his friends is of much greater interest.

We all know that annotation requires such rigor of selection that one might say that to annotate is to omit. But I cannot approve of the tone that Professor Finneran takes when he says, at the beginning of his notes, that he is not going to mention that ‘Upon a Dying Lady’ was based on the death of Mabel Beardsley (p. 613). It is as if he were pleased to withhold a fact that most readers would consider relevant.46

31Such annotational prejudices would have been slightly less inexcusable if The Poems: A New Edition had freely cross-referred its readers to Jeffares’s A Commentary on the Poems of Yeats, or if Finneran had at least sought to co-ordinate newer issues and settings of his volume with Jeffares’s A New Commentary which appeared a few months after his own book.47 I recall Finneran’s dismay, not only that the Commentary had been updated and expanded. He had, in fact, sought to supplant the Commentary with his own annotation, impoverished by the reasoning set out above. Worse, in his view, Jeffares, in keying it to the arrangement of The Poems: A New Edition, had publicly expressed in his preface his regret at what he saw as the flawed decisions over volume-arrangement and poem order in the Finneran edition.48

RECOVERIES: WIDER STILL, AND WIDER

32Readers had proved quick to notice Yeats’s patterns of cross-allusion throughout his works. When Allan Wade’s edition of The Letters of w. b. Yeats appeared in 1954, after some of the letters to Sturge Moore, John O’Leary and Katharine Tynan had been separately published (all 1953), the field of reference widened, and it did so again when the Macmillan Uniform Edition of the prose works began to appear from 1955 onwards, and it deepened with the appearance of the two Variorum Editions (1957 and 1966).

33In so far as Finneran’s annotational policies applied to the other volumes of the Collected Works (of which he was co-general editor), the embarrassment multiplied. This is too large a question to illustrate here, but if contrasted with the annotation of the Collected Letters, certain conclusions can be drawn about the obligations on editors of works and life-documents in the case of a closely-documented life of reflection, writing, reading, travel in the service of a political and intellectual cause, and theatrical activity.

34In 1994 Ronald Schuchard contrasted the work required of him as co-editor of Volumes 3–5 of the Collected Letters and the annotating he undertook for t. s. Eliot’s Varieties of Metaphysical Poetry: the Clark and Turnbull Lectures.49 Reviewing recent debates over annotation, and offering examples from the work of others and his own annotation (including exposure of his own occasional mistakes), Schuchard contrasted his two projects as follows: ‘Yeats requires more notes of recovery, Eliot of explanation… the virtue of brevity may be a dishonorable excuse for avoiding the detailed attention that the text requires’.50 In this—a then challenging view—Schuchard followed the lead of Wilmarth Lewis, w. j. b. Owen and Donald H. Reiman, who had also bravely asserted against the conventional wisdom of the day, that annotation was ‘commentary as a means to a knowledge of the poet’ s knowledge, so that we may the better understand his art and his wisdom’(Owen) and that the annotator’s duty is ‘to track down particular information that has eluded his predecessors but also to raise new questions and problems that, if solved by researchers in the future, will result in fuller understanding not only of the actions of poets but also of their unstated reasonings as well’(Reiman).51

35The contrast between the results of the ‘recovery’ approach and those found in the annotation of The Poems: a New Edition all too often resembles Lady Gregory’s distinction between the candlestick-maker who ‘holds up the light and hands it on from generation to generation, taking it from under the bushel that it may search the dark corners of the house’, and the butcher, whose ‘trade is in dead meat, it is to that his scales are adjusted, and the stirrings of life disturb his calculations; his business and his duty … to destroy life wherever it appears’.52 Questions which form themselves when we consult Finneran’s annotation are choked off by the notes. (In some later volumes, such as Later Essays, one can be surprised by joy when reading the notes while some few editors chafe at the series policy.) But where annotation solidifies into factoids as illustrated above, a seeming knowingness can beguile some readers into an unquiet trust that the editor must know best. Such knowingness does not know—or pretends not to know—that it deprives readers of that which might help them as distinct from controlling them from a shop-worn postulate of the old New Criticism, bankrupted by the archives and records of lived lives and by historical bibliography.

THE WORD-HOARD, SELF-ANNOTATION, SELF-ALLUSION, COMMENTARY

36Yeats’s instinctive awareness of implied and diverse audiences—Elkin Mathews’s Vigo Street poets and their readers was a ‘special public I would be glad to get’53—ensured that he was a self-annotating writer. From The Countess Kathleen and Various Legends and Lyrics (1892), Yeats had been predisposed to ‘poems with notes’ to help the reader, a practice which is at its most highly developed in The Wind Among the Reeds (1899), i.e., nearly thirty years before Hope Mirrlees’s Paris (1919) or Eliot’s The Waste Land (1922), for which poem Yeats’s volume had provided a clear exemplum. After Yeats’s death, he was well served, too, by Jeffares’s 1945 Trinity College thesis and later editions of its substance as A Commentary on the Collected Poems of w. b. Yeats (1968) and A New Commentary on the Poems of w. b. Yeats (1984).54 Jeffares was a classicist, and approached the writing of commentary very much as if he had been not merely an annotator of Catullus. A schoolboy-editor at Yeats’s old school, he had even commissioned and published a poem from Yeats in 1937 in The Erasmian55), Jeffares had the unique opportunity to meet numerous of Yeats’s friends, to collect the table-talk and anecdotage, and to read the poems for their ‘latent biography’, to use Ellmann’s phrase. But his task was deeply enriched by Yeats’s own prefaces, and notes, and ranges widely into the hinterland of his plays and fiction, his critical and expository prose, and his letters, lectures, speeches, broadcasts, interviews and table-talk. Jeffares annotated Yeats by tracing the connexions between poems and other parts of the Yeats canon, from which connexions Yeats’s strategies of self-allusion may be studied.56

37The Jeffares commentaries thus read Yeats’s poems by means of Yeats—his associated writings, such as notes, prefaces, passages from the speculative prose, offering a wonderfully implicit cross-referencing system. Let us, for the moment, try to hold all these writings as ‘Yeats’s word-hoard’, with his oeuvre at its core. That word-hoard was in every sense, including Yeats’s special sense, enunciated by a product of a voice, the ‘speech of a man’ during a lived life (YT 74). Then as now its mode of existence is in language housed in original print media—a negligible amount is also in recorded sound. When reproduced in new paper and digital forms, attempts must be made to respect the means by which Yeats allowed the part to be read by the whole, the new by the old.

38By contrast, the critical tradition following after that ‘Death of the Author’ popularized in French theory by Roland Barthes and Michel Foucault (of whom more below) implicitly discouraged the readers it claimed to empower from considering an author’s self-expository writing. In terms analogous to the ‘hermeneutics of suspicion’ popularized after Paul Ricoeur’s account of suspicion as a defining characteristic of the ‘school’ of Marx, Freud and Nietzsche,57 such writings are not to be trusted to provide another text with elucidation. Instead, meanings are to be decoded from within the confines of each play, poem, or novel. This has grievous results for such writers as Yeats who ask that their works be considered ‘in the mass’ rather than in ‘fragments’. In his case, ‘reader empowerment’ provided by such a critical approach would implicitly disavow the poet’s own strategies of self-allusion across ‘the mass’ of his work, whilst further discouraging readers from reflecting on the undeniable and extensive record of authorial change, authorial intention, and authorial meditation upon the very distinction between the author himself and his implied author. To follow this line of thought would be to discountenance the whole genre of authorly commentary, from obiter dicta to autobiographia literaria.

VAIN BATTLE: COLLECTED EDITIONS, INTENTION AS EXPECTATION

39The editorial approach most suited to this author’s work is that refined intentionalism which grasps the individualities and particularities of authors as found in their publishing archives. In Yeats’s case, the scattered A. P. Watt, Macmillan and Scribner archives of proofs, correspondence, TSS and MSS towards published and unpublished books offer traceable details of acts done through authorial delegation which, over the years of association between a writer and a trusted delegate, fall into patterns of ‘authorial expectation’.58 An intentionalist writer, then, ought to present no general challenge to intentionalist editorial theory, and Yeats only began to cited as a ‘notorious’ case when Anglo-American intentionalist editing found itself challenged by influential Franco-German editorial approaches in the last three decades of the twentieth century. What had brought Yeats into editorial problematics was not the quality of the editing of his emerging Collected Works, but turbulence within the world of textual editorial theory itself, to which the editing of Yeats became something of a test case, barely understood if frequently cited by those who sought to generalize.59 I turn now to that turbulence.

BLOWING THE DOORS OFF

40Prefiguring (and in a sense instigating) the crisis in editorial theory of the 1980s, Roland Barthes had pondered the general problem of oeuvre in 1971 thus

… l’œuvre est un fragment en substance, elle occupe une portion de l’espace des livres (par exemple dans une bibliothèque.) Le Texte, lui, est un champ méthodologique… l’œuvre se voit (chez les libraires, dans les fichiers, dans les programmes d’examen), le texte se démontre, se parle selon certaines règles (ou contre certaines règles); l’œuvre se tient dans la main, le Texte tient dans le langage: il n’existe que pris dans un discours (ou plutôt il est Texte par cela même qu’il le sait); le Texte n’est pas la décomposition de l’œuvre, c’est l’œuvre qui est la queue imaginaire du Texte. Ou encore: le Texte ne s’éprouve que dans un travail, une production. Il s’ensuit que le Texte ne peut s’arrêter (par exemple, à un rayon de bibliothèque); son mouvement constitutif est la traversée (il peut notamment traverser l’œuvre, plusieurs œuvres).60

41The Barthes distinction between l’oeuvre and le Texte seems adequately to theorize two ways of thinking about the work, depending on what one is doing with it. There can be little to quarrel over in a generalization so very abstract, but then Barthes, as Italo Calvino is reported to have commented, was a pretty plausible generalizer, in the ‘science of the single object’.61

42Editorial commentary is not itself le Texte, through it nourishes and directs the tasks of experiencing le Texte in Barthes’s term. In uncomplicatedly annotating allusions to penumbral writings outside an author’s own ‘word hoard’ as well as those within it, editorial commentary also accounts for events, images, music, and individuals in history and myth, as well as images. But the methodology of commentary must do justice not merely to l’oeuvre and le Texte, but also account for how the work stands, in any particular textual instantiation, in the book or books which it occupies, has occupied, and will occupy (to pick up Barthes’s point about the ‘traversability’ of text). For the text is experienced not merely in language, discourse, activity or production, it is experienced in books and was designed by an author to be experienced in the context of a book or books. It is frequently said that Barthes’s theories stand in a tradition which stems from Mallarmé.62 It is not always remembered by theoreticians that for Mallarmé, ‘… tout, au monde, existe pour aboutir à un livre’.63

43Michel Foucault, with whom Barthes had, I think, an uneasy relationship, had also pondered the mode of existence of works in language. ‘A theory of the work does not exist, and the empirical task of those who naively undertake the editing of works often suffers in the absence of such a theory’, he claimed. The immediate context of the remark (a hypothetical edition of Nietzsche’s notebooks, with the untidy intrusion of references, diary jottings, or a laundry list) troubled him in a manner that seems captious, and for a reason.

Mais quand, à l’intérieur d’un carnet rempli d’aphorismes, on trouve une référence, l’indication d’un rendez-vous ou d’une adresse, une note de blanchisserie: oeuvre, ou pas oeuvre? Mais pourquoi pas? Et cela indéfiniment. Parmi les millions de traces laissées par quelqu’un après sa mort, comment peut-on définir une oeuvre? La théorie de l’oeuvre n’existe pas, et ceux qui, ingénument, entreprennent d’éditer des oeuvres manquent d’une telle théorie et leur travail empirique s’en trouve bien vite paralysé.64

44The question itself had followed logically enough from Foucault’s starting point, but is a wilful (or self-serving) failure of critical imagination. The Author was dead—for Foucault’s purposes he or she had died in 1967, when Roland Barthes had published La Mort de l’auteur.65 Foucault felt able to go further. ‘Writing’, he claimed, ‘was now

effacement volontaire qui n’a pas à être représenté dans les livres, puisqu’il est accompli dans l’existence même de l’écrivain. L’oeuvre qui avait le devoir d’apporter l’immortalité a reçu maintenant le droit de tuer, d’être meurtrière de son auteur. Voyez Flaubert, Proust, Kafka. Mais il y a autre chose: ce rapport de l’écriture à la mort se manifeste aussi dans l’effacement des caractères individuels du sujet écrivant; par toutes les chicanes qu’il établit entre lui et ce qu’il écrit, le sujet écrivant déroute tous les signes de son individualité particulière; la marque de l’écrivain n’est plus que la singularité de son absence; il lui faut tenir le rôle du mort dans le jeu de l’écriture. Tout cela est connu; et il y a beau temps que la critique et la philosophie ont pris acte de cette disparition ou de cette mort de l’auteur.66

45Here I cheerfully confess to losing Foucault and his meaning in that rapture with which his paradox had evidently seized him. The writer’s disappearance into ‘le sujet écrivant’, the writing subject or ‘the subject writing’: ‘[u] sing all the contrivances that he sets up between himself and what he writes, the writing subject cancels out the signs of his particular individuality’ strikes me as legerdemain, which incidentally renders invisible the author’s book as well. To try to explain how the Author has been dehumanized and yet become his own ‘writing subject’, it is necessary to backtrack to Barthes’s 1967 lecture, because the critical tradition which ensued pressed heavily upon editors, the gatekeepers who negotiate between authors and their future audiences.

46Barthes had sought determinedly to liberate French pedagogy by ‘blowing the doors off’ the Lansonian life-and-work approach, what Auden had called the ‘shilling life’ approach to teaching literature, by emphasizing that

L’œuvre, la méthode, l’esprit de Lanson, lui-même prototype du professeur français, règle depuis une cinquantaine d’années, à travers d’innombrables épigones, toute la critique universitaire. Comme les principes de cette critique, du moins déclarativement, sont ceux de la rigueur et de l’objectivité dans l’établissement des faits, on pourrait croire qu’il n’y a aucune incompatibilité entre le lansonisme et les critiques idéologiques, qui sont toutes des critiques d’interprétation. Cependant … il y a une certaine tension entre la critique d’interprétation et la critique positiviste (universitaire). C’est qu’en fait, le lansonisme est lui-même une idéologie; il ne se contente pas d’exiger l’application des règles objectives de toute recherche scientifique, il implique des convictions générales sur l’homme, l’histoire, la littérature, les rapports de l’auteur et de l’œuvre; par exemple, la psychologie du lansonisme est parfaitement datée, consistant essentiellement en une sorte de déterminisme analogique, selon lequel les détails d’une œuvre doivent ressembler aux détails d’une vie, l’âme d’un personnage à l’âme de l’auteur, etc., idéologie très particulière puisque précisément depuis, la psychanalyse, par exemple, a imaginé des rapports contraires de dénégation entre une œuvre et son auteur. En fait, bien sûr, les postulats philosophiques sont inevitables; ce ne sont donc pas ses partis pris que l’on peut reprocher au lansonisme, c’est de les taire, de les couvrir du drapé moral de la rigueur et de l’objectivité: l’idéologie est ici glissée, comme une marchandise de contrebande, dans les bagages du scientisme.67

47The French insistence on what Roland Barthes called La mort de l’auteur was said by both Roland Barthes and, after him, Michel Foucault (who claimed that ‘La disparition de l’auteur, qui depuis Mallarmé est un événement qui ne cesse pas …’68), to take its origins in Mallarmé’s work quoted above. Barthes’s concluding words rousingly insisted ‘nous savons que, pour rendre a l’ecriture son avenir, il faut en renverser le mythe: la naissance du lecteur doit se payer de la mort de l’Auteur’.69 With the author safely killed off and the reader thereby empowered by Barthes, Michel Foucault’s lecture at the Collège de France on 22 February 1969, Qu’est-ce qu’un auteur? proposed to replace the dead, external author with the implied, internal ‘la fonction «auteur»’which, I suppose, is the ‘writing subject’.70 His endorsement by allusion to and rather envious qualification of Barthes’s (uncited) essay concludes:

l’auteur… est un certain principe fonctionnel par lequel, dans notre culture, on délimite, on exclut, on sélectionne… L’auteur est donc la figure idéologique par laquelle on conjure la prolifération du sens.71

48‘Tout cela est connu’ Foucault claimed in the lecture, ‘il y a beau temps que la critique et la philosophie ont pris acte de cette disparition ou de cette mort de l’auteur’.72

On peut dire d’abord que l’écriture d’aujourd’hui s’est affranchie du thème de l’expression: elle n’est référée qu’à elle-même, et pourtant, elle n’est pas prise dans la forme de l’intériorité; elle s’identifie à sa propre extériorité déployée.73

49Remarkably, just as Foucault sought to contract ‘la fonction «auteur»’ to the limits of the text itself, so the text was limited to ‘poetic or fictional texts’74—an even deeper subversion than Barthes’. Happy to see what he called ‘cette catégorie fondamentale de la critique «l’homme-et-l’oeuvre»’75 (i.e., of the Lansonian approach) blown away, he also triumphantly claimed that ‘[a] theory of the work does not exist’. From the perspective offered by authors such as Eliot and Yeats, who themselves had a sense of their ‘work’ as larger than their ‘works’, and who wrote intimately about their own relation with their ‘author-function’, Foucault’s theorizing seems not engage with the record. It is still something of a surprise that neither Foucault nor those who immediately took him up paused to test the newly-minted distinction between author and author function against the record left by authors themselves. As Sean Burke remarked ‘one must be deeply auteurist to call for the “Death of the Author”’.76

LIVING TESTIMONIES FROM DEAD AUTHORS

50Some hardy perennials from Yeats, t. s. Eliot and Seamus Heaney stand in contrast to the once fashionable and now distinctly passé news from France. Eliot and Heaney followed Yeats in discussing their own intentions in the light of their own praxis, and did so in numerous contexts conscious of those they saw as their forerunners. In 1919, T. S. Eliot had offered perhaps the most precisely-thought and impersonally expressed account of the poet’s relationship with his implied author. ‘Tradition and the Individual Talent’ starts by considering the relation to the tradition into which new work interposes itself. Eliot finds both what we might call his ‘poet’ and his ‘implied poet’ when he insists that

[n]o poet, no artist of any art, has his complete meaning alone. His significance, his appreciation is the appreciation of his relation to the dead poets and artists. You cannot value him alone; you must set him, for contrast and comparison, among the dead.77

51If ‘[h]onest criticism and sensitive appreciation’ are to be ‘directed not upon the poet but upon the poetry’, the activity demands the widest of contexts. The ‘importance of the relation of the poem to other poems by other authors’ is well-nigh limitless. Its scope, Eliot declared, is ‘the conception of poetry as a living whole of all the poetry that has ever been written’. Eliot might be thought to have been very much on the side of what I have called the ‘empowered reader’, but his thought is in fact focused on that ‘aspect of this Impersonal theory of poetry … the relation of the poem to its author’ and so the gravity of his discourse is addressed to poets rather than to readers.

52Within this suddenly enlarged field, the poet must strike a startling posture, the very opposite of self-assertion: ‘[t]he progress of an artist is a continual self-sacrifice, a continual extinction of personality’ and in setting himself or herself as an artist in ‘relation to the sense of tradition’, the poet is engaged in a ‘process of depersonalization’ even when he or she ‘exclusively operate(s) upon personal experience.

… the more perfect the artist, the more completely separate in him will be the man who suffers and the mind which creates; the more perfectly will the mind digest and transmute the passions which are its material… Poetry is not a turning loose of emotion, but an escape from emotion; it is not the expression of personality, but an escape from personality. But, of course, only those who have personality and emotions know what it means to want to escape from these things.78

53Halting at that frontier of metaphysics or mysticism (Yeats’s point of departure in Per Amica Silentia Lunae [1918], which almost wholly baffled Eliot when he reviewed it, twice), Eliot confines himself to the ‘practical conclusions as can be applied by the responsible person interested in poetry’. To ‘divert interest from the poet to the poetry is a laudable aim’, he thought, but

very few know when there is expression of significant emotion, emotion which has its life in the poem and not in the history of the poet. The emotion of art is impersonal.79

54Behind this insistent remark, I detect that approval which Eliot had given to the ‘wholly delightful’ sentence he had found (one of three) in Per Amica Silentia Lunae when he had reviewed it for The Egoist:

It is not permitted to a man, who takes up pen or chisel, to seek originality, for passion is his only business.80

55Drawing what he called a ‘suggestive analogy’ from the operation of a catalyst in a chemical process81 Eliot saw ‘the mind of the poet’ as

… the shred of platinum. It may partly or exclusively operate upon the experience of the man himself; but, the more perfect the artist, the more completely separate in him will be the man who suffers and the mind which creates; the more perfectly will the mind digest and transmute the passions which are its material.82

56This catalytic operation allows Eliot to understand just how ‘the mind of the mature poet differs from that of the immature one not precisely in any valuation of “personality”, not being necessarily more interesting, or having “more to say”, but rather by being a more finely perfected medium in which special, or very varied, feelings are at liberty to enter into new combinations’.

And the poet cannot reach this impersonality without surrendering himself wholly to the work to be done. And he is not likely to know what is to be done unless he lives in what is not merely the present, but the present moment of the past, unless he is conscious, not of what is dead, but of what is already living.83

57We have already seen that Yeats’s intentions cover all aspects of his work, and that the ‘intended, complete’ work comes from his profound consciousness of the doubleness of writer who sits down to breakfast and the ‘secret working mind’ (VP 639) within the text, the self ‘reborn as an idea, something intended, complete’ (E & I 509).84 Both Yeats and Eliot had pondered rather profoundly the difference between the author and his or her ‘author-function’ well before Foucault came up with that repugnant abstraction, beguiling as it has proved as a point of reference in modern editorial theory. Each took particular interest in the presence and the limitations of his implied author in the act of revision.

For Eliot: There are two reasons why the writer of poetry must not be thought to have any great advantage. One is that a discussion of poetry such as this takes us far outside the limits within which a poet may speak with authority; the other is that the poet does many things upon instinct, for which he can give no better account than anybody else. A poet can try, of course, to give an honest report of the way in which he himself writes: the result may, if he is a good observer, be illuminating. And in one sense, but a very limited one, he knows better what his poems ‘mean’ than can anyone else; he may know the history of their composition, the material which has gone in and come out in an unrecognisable form, and he knows what he was trying to do and what he was meaning to mean. But what a poem means is as much what it means to others as what it means to the author; and indeed, in the course of time a poet may become merely a reader in respect to his own works, forgetting his original meaning—or without forgetting, merely changing.85

58Authors, says Seamus Heaney, are ‘rewarded by authority’ —and not just in the civic sense, although authorship, to a writer such as Yeats, Eliot, or Heaney is accorded fame and public respect of a kind acknowledged in the Nobel Prize, by no means limited to writing, does impose the civic burdens of a pressing world. Yeats had acknowledged this when he wrote ‘In dreams begins responsibility’ (VP 269). Heaney’s thinking in Stepping Stones (1908—five years before his death—and made up of generous written responses to exhaustive questionnaire-interviews), shows that at every level his ‘authoriality’ (and so his authority) is enriched by the thinking of Yeats and Eliot as quoted above. Passages including the very ones I have quoted from Yeats’s ‘General Introduction’ and Eliot’s ‘Tradition and the Individual Talent’ are, for him, touchstones.86 He writes with authority about his poetic intentions, sometimes in the sense of foreconceits. Heaney and his ‘implied author’ stood in conscious friendly partnership. Foucault’s boast that ‘today’s writing has freed itself from the theme of expression’ was triumphant, but it simply isn’t true. In providing all the raw materials for an Autobiographia Literaria, Stepping Stones epitomized, five years before Heaney’s death, the scandal which the facts of Authorship offer to ‘the death of the Author’and all ‘Theory’ —including editorial theory—which claims kinship with it.

WINDS OF CHANGE: ‘THINGS THOUGHT TOO LONG’

59Once one concedes that writing is also about something other than itself, it becomes easier to see that when new ideas have been imported without context from other cultures their fashionable dominance has been enlarged by their having been reduced to slogans.87 Barthesian and Foucauldian rhetoric dominated Anglo-American critical thinking for a certain season in the assertive and confused mêlée of late twentieth century Anglo-American ‘Theory’, and in that context, the ‘Death if the Author’, offering ‘empowerment’ to all those committed to the beholder’s share in critical thinking, was, for a season, well-nigh irresistible.

60A good example would be found in Michel Foucault’s intervention which deliberately shifted attention from ‘livre’ and ‘oeuvre’ to ‘texte’ and ‘textualité’. It suddenly became fashionable for students to pick up the idiolect. Theory-speak insisted on ‘texts’ when such terms as ‘poem’, ‘story’, ‘novel’, ‘play’, ‘essay’ or ‘book’ were demanded by the context of their remarks: ‘text’ replaced ‘work’ with little sense that to do thus was as much an evasion as a deliberate choice.

61In the same way, the Anglo-American editorial flirtation with the ‘Death of the Author’ emerged in a partial eclipse of authorial intention, even though neither Barthes nor Foucault seems to have envisaged or encouraged an editorial (rather than a readerly/critical) outcome for their theories. After Foucault, the emerging Franco-American critical tradition chose to forget that both theorists had claimed Mallarmé as their chief precursor, ignoring that the same Mallarmé had idealized the Book, Le livre: instrument spirituel. And if books were an early casualty of the Barthesian-Foucauldian critical tradition, it is fair to ask how Historical Bibliography and ‘Theory’ —that critical tradition in which I include the editorial thinking and practice of textual theorists who invoke the names of Barthes and Foucault—now stand in relation to each other, especially regarding the case of Yeats?

62The rethinking of Anglo-American editorial theory had gone in parallel with the dominance of Foucault and Barthes in French critical theory. Within Anglo-American editorial practice, the eclectic tradition of W. W. Greg and Fredson Bowers had had an iron grip, but ‘things thought too long’ could be ‘no longer thought’ (VP 564). Editors and bibliographers such as G. Thomas Tanselle, James Thorpe, Philip Gaskell, D. F. McKenzie and Jerome J. McGann formulated editorial approaches as Historical Bibliography was being restructured as the History of the Book. Some of their most profound thought was given over to the problem of authorial intention and the imperatives it raises for practising editors working with modern authors, for whom extensive author/publisher archives are extant and the evidence points to the ‘social text’.88 In such cases, verifiable evidence of the interconnexions between writers, literary agents, publishers (and above all, publishers’ readers), printers, is recoverable in the form of contracts, costings, letters between the various parties, proofs, book designs and samples of book production, all of which demonstrate how intention is ‘socialized’ in the publishing process. Intention as authorial delegation and authorial expectation89 offers trails of supporting evidence to show the academic editor just how the texts in printed editions and the forms of the books in which they appear had been arrived at. Above all, the pragmatism of the new Anglo-American editorial theorists appears unchallengeable.

[T]he editor should not base his work on any predetermined rule or theory. In general he will try to produce an edited text that is free from accidental error and from unauthorized alteration, and is presented in a way that is convenient for its intended readers. Beyond this every case is unique and must be approached with an open mind.90

G. Thomas Tanselle widens this pragmatism:

Failure to recognize that different kinds of edition may be required to satisfy different historical interests has weakened many of the editorial debates of the past half-century.91

63It is here that the case of Yeats enters into the theoretical debate. The case is that of Yeats’s The Poems: A New Edition, commissioned in 1976 and published in 1983–1984. Richard J. Finneran had edited it on an avowedly intentionalist set of principles as the first volume of a new ‘final text’, annotated, Collected Edition of the Works of W. B. Yeats. In deciding on the overall arrangement and major copytext for that volume, and setting a textual blue-print for the series, Finneran argued that the two competing arrangements of the poems sanctioned by Yeats in the Thirties offered

not merely two different ‘arrangements’, but two different incarnations of the archetypal ‘Sacred Book’ of the poems, thus two different experiences of reading Yeats and of attempting to come to terms with his massive achievement. Which plan is in fact Yeats’s ‘final intention’?92

64We will return below to that so-called ‘archetypal “Sacred Book”’ below.93 Finneran’s reading of the Archive was challenged by my argument that he had not fully considered its evidences and had made the wrong choice of copytext for the bulk of the task, a challenge which set off a ten-year controversy.94 That controversy was cited rather more often than discussed with reference to the relevant literature. In A Critique of Textual Criticism (1983) Jerome McGann had offered various qualifications to the ‘ideology’ of an ‘author’ s final intentions’in the general business of establishing copytext. Gesturing to ‘a storm of scholarly dispute’ after the appearance of The Poems: a New Edition (1983), McGann devoted a couple of pages of The Textual Condition (1991) to the ‘notorious case of Yeats’ as an example to ‘indicate how and why to concept “author’ s (final) intentions)” cannot be used to determine copytext’, concluding ‘Suffice it to say that no one editing Yeats, given the state of the archive as it is presently known, could use “author’ s (final) intentions” as the determinative criterion for deciding on copy-text for many, perhaps most, of the poems, and especially the so-called Last Poems’, says McGann, warily gesturing to Richard Finneran’s accompanying volume, Editing Yeats’s Poems.95

65‘Final intention’ —rightly in my view—had been Finneran’s criterion in his choice of principal copy-text, even if—disastrously—he had failed to see that the Collected Poems was a stop-gap, a ‘radial development’ (to use McGann’s term96) of Yeats’s larger publishing programme during the Depression. Mrs Yeats and Thomas Mark had left plenty of evidence that the decision to end the Last Poems volume-unit with ‘Under Ben Bulben’ was her decision on his suggestion, despite Yeats’s placement of the poem at the beginning of Last Poems and Two Plays. Mrs Yeats and the publisher’s reader were following the convention of placing a funerary poem or an epitaph at the end of a collected edition. I return to this matter below, and mark here only her candour.

66There is, by contrast, absolutely no evidence that Yeats’s views were not followed in the preferral of the lifetime array of volume units for the postponed Edition de Luxe seven-volume project over the two-part (‘Lyrical’ and ‘Narrative and Dramatic’) arrangement of 1933. In the court-room drama conjured by Richard Finneran, the ‘onus of proof’ must lie with those who prefer the arrangement of the two part ‘radial development’ and thereby imply that Mrs Yeats and Thomas Mark did not carry through Yeats’s wishes.97

67The sub-title of the ‘Prologomena’ to both Editing Yeats’s Poems and Editing Yeats’s Poems: a Reconsideration is ‘The Myth of the Definitive Edition’. The argument there is that the so-called ‘Definitive Edition’ of 1949 was not definitive because of the posthumous work on text and poem order by Yeats’s delegates, George Yeats and Thomas Mark. Both Editing Yeats’s Poems and Editing Yeats’s Poems: a Reconsideration warm to this theme.

Yeats died on January 28, 1939. He had not been long in his temporary resting-place at Roquebrune before the process began of—not to put too fine a point on it—corrupting the texts which he had worked so hard to perfect.98

68Avoiding the names of the agents who set such a process in motion is a sign not of delicacy but of the ‘othering’ of Yeats’s wife, executrix and publisher’s reader in this Gothic sentence. Elsewhere99 I have suggested that there are only two possible scenarii, a comic one in which the incompetent executrix and publisher get everything wrong, and a gothic one in which they work to undermine Yeats’s intentions, but, from the above, the latter would seem the stuff of Professor Finneran’s conclusions. There can be little point in defending the term ‘Definitive Edition’, but once one begins to investigate why its dislodgement had seemed to symbolise for Richard Finneran something larger, one is filled with dismay at his isolationist lack of empathy with the world of European publishing before, during, and after the War, and the predicament in which Mrs Yeats found herself.

69By the time the controversy had been running for six years, Editing Yeats’s Poems: A Reconsideration (1990) had erased the passage, substituting a new angry chapter (7), ‘The Order of the Poems’, which descends to a coarsely-drafted letter as if from Yeats himself, to represent the ‘quasi-chronologists’,100 in an attempt to obliterate the criterion of ‘final intention’.101 Looking no further than the American theorists McGann and Hershel Parker, and finding that they ‘agree that the concepts of “final intentions” and “definitive edition” are no longer telling’, Finneran proceeds to undercut the entire reasoning behind his own choice of principal copy-text. No published or proposed edition of the Thirties he avers,

can be said to be ‘definitive’ or to represent ‘[Yeats’ s] final intentions’. Even were we to replace ‘final intentions’ by ‘last known intentions’, and even if we somehow could resolve the question of documentary versus non-documentary evidence, those intentions are not necessarily identical with whatever Yeats might have desired in a Platonic world where the Sacred Book of his poems could be inscrited [sic] without mortal interference.102

70The full (and ‘vexatious’) history of the use of the word ‘definitive’ in relation to projects and published volumes of Yeats’s works could have been traced in ‘Appendix Six’ of the Jeffares edition, Yeats’s Poems (1989), though to have done so would have been to recognize that there was a powerful countervailing argument.103 The term ‘complete and definitive’ was first used in a letter from Harold Macmillan to Mrs Yeats of 8 February 1939, and the words were also used in a ‘Preliminary Notice’104 headed ‘The Collected Edition of the Works of W. B. Yeats’ which was released to the trade press before April 4, 1939.105 Thereafter ‘a definitive edition’ was used in the published prospectus for The Poems (1949) but not in the volumes themselves. It would seem that the application of the phrase ‘The Definitive Edition’ to those volumes did not begin until they were thus described in the 1951 first edition of Allan Wade’s Bibliography.106

SACRED WRIT

71What was, however, disturbing, is that this second reference to the ‘Sacred Book’ apparently promotes it from loosely ‘archetypal’ (in the casual, modern sense) into the ‘Platonic’ world of pure forms. This would be simply rather silly (no one who queried Finneran’s choices of copy texts based any of their arguments upon any such ‘Platonic world’ or upon immortal inscription) were it not inbred and confused. For the archetypal “Sacred Book” invoked here does not refer to the Mallarméan thinking of Yeats, which—I reiterate—he never applied to his own poems, but reserved for ‘the new sacred book, of which all the arts are beginning to dream’.107 Instead, it is the opportunistic title of an essay (1955 and much reprinted) by Hugh Kenner. Without bothering to cite Yeats’s prose works and thus recover the contexts wherein Yeats used the phrase, Kenner simply took the phrase on trust from the table-talk of Ezra Pound.

Against the poet as force of nature [Yeats] placed of course the poet as deliberate personality, and correspondingly against the usual ‘Collected Poems’ (arranged in the order of composition) he placed the oeuvre, the deliberated artistic Testament, a division of that new Sacred Book of the Arts of which, Mr. Pound has recalled, he used to talk.108

72Further to my discussion of this essay, its premiss and its self-subversion in my ‘Resurrection’ essay,109 I reread it here from Kenner’s opening ‘Catechism’, a faux-Socratic dialogue between Q, a patronizing professor, and A, a student of some promise. The student is sent away, peremptorily, to write an essay (‘And please refrain from putting in many footnotes that tire the eyes’, 577), which duly follows in two further parts. Either the student wishes merely to tickle the professor’s vanity, or Kenner is unable to maintain the student’s independence. Critics write ‘bastard biography’ and a ‘Pécuchet’s industry’ is the command economy of ‘exegetes’ who copy ‘parallel passages from A Vision (first and second versions), from letters and diaries, from unpublished drafts, and occasionally from other poems’ (577), Kenner’s contempt is for biography (especially the work of Jeffares) and exegesis (Jeffares’s again) other than his own.

73With admirable economy, the student then suggests that ‘the place to look for light on any poem is in the adjacent poems, which Yeats placed adjacent to it because they belonged there’, The Tower offering the evidence that ‘Yeats was an architect, not a decorator; he didn’t accumulate poems, he wrote books’(578). Kenner—his pretence of a student cannot sustain itself—then claims that, at least from Responsibilities to A Full Moon in March the individual volume rather than the poem or the sequence is the unit ‘in which to inspect and discuss his development’ (578). Through brief discussions of poem order in other collections, Kenner extrapolated to the overall ‘life-time arrangement’ of the 1949 ‘definitive edition’ which begins with The Wanderings of Oisin of 1889 and ends with Yeats’s epitaph, seeing it as a biblical structure where the various books offer a

dramatic revelation… Each volume of his verse, in fact, is a large-scale work, like a book of the Bible. And as the Bible was once treated by exegetists as the self-sufficient divine book mirroring the other divine book, Nature, but possessing vitality independent of natural experience, so Yeats considered his Sacred Book as similar to ‘life’ but radically separated from it, ‘mirror on mirror mirroring all the show’ (588).

74What we might call the volume’s ‘closure’ Kenner inflated into ‘an apocalypse’:

The last division of his Sacred Book closes with an apocalypse, superhuman forms riding the wintry dawn, Michaelangelo electrifying travellers with his Creation of Adam, painters revealing heavens that opened. (587)

75In the Irish Writing and Sewannee Review versions, Kenner’s sentence quoted above about the volumes ‘at least from Responsibilities to A Full Moon in March’ had subtended to it a bashful footnote reading ‘It isn’t clear how much, if any, of Last Poems was arranged by Yeats himself, except that he wanted “Under Ben Bulben” to go at the end’.110 Nobody seemed to notice that the note had begun to undermine the epic Hollywood ‘apocalyptic’ argument of the essay. Worse was to follow, when in, complimenting Kenner on his essay in 1956, George Yeats told Kenner that it was not in fact the case that Yeats had wanted ‘Under Ben Bulben’ to appear as the final poem in his collected Poems. Kenner remained troubled enough to modify that note in reprintings. In the Gnomon (1958) and the Unterecker reprintings it is shortened to ‘It isn’t clear how much, if any, of Last Poems was arranged by Yeats himself’.111 If this had not been enough to startle even tired eyes, that footnote itself had a further nervous history. In Finneran’s Critical Essays on w. b. Yeats the note had praised the volume’s editor:

In Editing Yeats’s Poems … 65, Richard J. Finneran has offered conclusive evidence supporting Curtis Bradford’s conjecture that Yeats was not responsible for the contents or order of Last Poems & Plays (1940).112

76By 1990, in Editing Yeats’s Poems: A Reconsideration, a letter of Kenner’s of 1984 is quoted, offering his memories of what Mrs Yeats had said in 1956:

But I had incautiously said that WBY wanted ‘Under Ben Bulben’ to come at the end, and that, she stated, was not so… [S] he gave no further details, and I was insufficiently informed to know what to ask’.113

77In perhaps the most reductive document in the entire controversy, Hugh Kenner’s ‘Whose Yeats is it, anyway?’, it becomes apparent that Kenner had indeed consulted the Yeats’s Poems appendix from December 1989, before concluding that the ‘Jeffares edition is chronological, the Finneran is not. I find it’s not something I can get agitated about’, a remark which deflates the rapture of ‘The Sacred Book of the Arts’.114 By 2000, the note to the essay had been expanded again; its second sentence an ill-fit with the first:

It isn’t clear how much, if any, of Last Poems was arranged by Yeats himself except that he wanted ‘Under Ben Bulben’ to go at the end. I learned many years ago that this was untrue. Yeats wanted ‘Under Ben Bulben’ at the beginning of Last Poems, where the Finneran edition has placed it. HK, 1999]. [sic].115

78Neither the Jeffares edition nor the Albright edition is cited; all the more regrettable in that the former of these includes Mrs. Yeats’s reply of 14 June 1939 to a suggestion of Thomas Mark’s is given:

Certainly put ‘Under Ben Bulben’ at the end of the volume. Its present position was WBY’s, but I t [h] ink now it should undoubtedly be at the end as you suggest.116

79Above all, the misapplied ‘Sacred Book’ in Kenner’s title seems designed to confer the authority of a fin-de-siècle concept of Yeats’s upon Kenner’s own thinking. The resulting implication is false coin: to Yeats, his own writing and shaping of his assembled poems were never intended to realise a new ‘Sacred Book’ of his own. But Kenner’s misapplication of a leading fin de siècle concept of Yeats’s own has been influential: indeed, for Finneran, Kenner’s essay itself is ‘sacred writ’ for some inner order whom Finneran calls ‘Yeatsians’, at a fourth remove from the poet’s words in context.117 The resultant rhetorical appeal to ‘Yeatsians’, couched in ‘Yeatsese’ is an appeal not to Holy Writ, but to the Dictionary of Received Ideas.

80Given such a clarion, it is puzzling that the editors of the new Collected Works have applied the thinking behind a Mallarméan ‘book’ so patchily in regard to annotational policy. To grapple textually with such a ‘Great Work’ requires interpretation without hostility to the record of acts and decisions of previous editors working in very different circumstances. The recovery of ‘intention as expectation’ demands patience and resists adversarial inquisition. It is unrecognizable in the courtroom drama inflicted upon Yeats’s widow and publisher’s reader in, e. g., Finneran’s ‘The Order of Yeats’ s Poems’.118 The archive considered as a whole is only ‘patient of interpretation’ (to use A. N. Whitehead’s remark as adapted by Frank Kermode119) if its scholars are as patient in acknowledgement of the circumstances of acts done in a publishing house in the 1930s and 40s under circumstances vastly different from those which obtained in academic debate fifty years later. Unless the reality of those circumstances is acknowledged, then the textual theory brought to bear upon those acts and decisions is inadequate to the thinking of which it seeks to take account. When the theory of expectation is itself tolerant (and able to account for occasional silences in the record), interpretation of the case of Yeats’s texts points only in one direction.

81Shrewdly or evasively, McGann found the controversy ‘moot’ while reviewing the two editions in 1990.120 ‘Moot’ or mute, he adroitly avoided a summer’s wrangle in the correspondence columns of the TLS. He conceded that Yeats’s own statements had valid borne out his theories of his works as his ‘permanent self’, but his mind was really running on his own preoccupation, the solution he at the time felt that hypermedia archives offered for editorial dilemmas. For

‘writers who exhibit not merely an extreme interest in finished forms… but who obsessively rework their texts in an effort to arrive at their impossible (and changing) dreams’ an electronic edition would try ‘to co-ordinate and therefore go beyond all the scattered texts that have descended to us, or that might be (but are not yet) constructed.121

82Short-circuiting the debate over final intention and to create access to all textual states, including manuscripts, McGann proposed a ‘reading text’, ‘either the Finneran or the Jeffares/Gould constructions would do’, combined with a disk-borne hypermedia archive of all states of Yeats’s texts including Yeats’s abandoned texts. This would be

one image of Yeats’s desire for what Gould names the definitive edition… [I] t would stand more as an arbitrary point of departure than as a fixed point of reference: a certain place from which the reading and study of the Yeatsian texts might (but need not) begin.122

83Finneran responded that:

In a future electronic edition… no doubt… the reader [will]… toggle between… competing formats. But even then, one version will presumably have to be the default arrangement…. yes, Virginia, short of the discovery of a lost codicil to Yeats’s will or the reappearance of his supernatural Instructors, readers-and editors-must make such a choice… our choice of formats has significant consequence for interpretation (TI 28–29).

84Addressed to the University of Virginia at large with the Christmas cliché, an appeal to the Spirit World or a lost Will—I would have more sympathy with the argument but for this angry fancy. However regrettable his own choice had been, Finneran at least understood the responsibility to choose. McGann did not consider that the canon of poems is in fact a canon-within-an oeuvre within an oeuvre. Some texts have multiple existences at every level. While it is necessary that a variorum or hypermedia edition should include all abandoned versions, it would be hard to argue that this should happen in a Collected Edition of the Works which would elsewhere print lyrics rather pointedly never removed from the plays by Yeats himself, in their rightful contexts.

VAIN REPOSE

85Twenty-five and more years on, it is no longer possible to kick an editorial problem into the digital future. Those—such as Jerome McGann—who have plausibly argued for digital solutions now have a more realistic sense of the present impracticality and inaccessibility of digital platforms, particularly those carried on CDs, for the common reader, and so a renewed understanding of why the reading of literature in print form is not going to wither away, and of the necessity to reach those ‘common readers’ who will increasingly absorb books in e-book form, and who at present struggle with, for example, Kindle texts which for all the ease of reading footnotes and of looking up unfamiliar terms, convenient portability, and lower cost, suffer from very poor bibliographical standards.123 Hugh Kenner had warned in 1955 against what is now the entry-level encounter with Yeats’s poems, ‘out of the context Yeats carefully provided’. Kenner’s ‘Anthology Appointed to be Taught in Colleges’ (577) has now been Nortonized, but even—perhaps especially—if a first reading of a poem has been googled from some stray website, fresh readers can be expected to require access to context and referral to cognate texts in the oeuvre.

86For all the publicity given to it, the ‘notoriety’ of Yeats had surprisingly little impact among textual theorists, Paul Eggert has rightly welcomed the exemplary record of the archive found in James Pethica’s Cornell volume on the manuscripts of the Last Poems, observing how Pethica’s volume thus stands out from the others in that series.124 Beyond Eggert, however, few who cite the case of Yeats seem to have conducted any independent review of the documentary record, nor even of the literature it has generated.

87Among the Yeats scholars, Hazard Adams, the well-known critical theorist, published The Book of Yeats’s Poems in 1990, perhaps the first full-length study to apply the Kennerian paradigm of reading to Yeats’s poems after the appearance of The Poems: A New Edition. A critical study of the various volume units of Yeats’s poems, it offers a brief portentous survey, ‘Critical Constitution of the Book’. It then offers summaries of the opposed positions, based—it would seem—on the Finneran summaries in Editing Yeats’s Poems and a draft of part of Editing Yeats’s Poems: A Reconsideration, not at the time, yet published. Adams’s summaries seriously misapprehend the opposing arguments, even suggesting, for instance that those who supported the chronologically arranged volume-sequence had in fact wished for a chronological arrangement of all poems according to date of composition.125

88Adams loftily announced that the ‘Book of Yeats’s Poems’ as he conceived it did not exist. He had evidently not looked at the Jeffares edition of 1989 which fulfils his demands for such an edition to the letter. Though careful with those trigger-happiest when reaching for ‘smoking guns’,126 Adams in effect said ‘a plague on both your houses’, before opting for the lifetime chronological arrangement as the basis for his critical endeavour exactly. In acting so fully in character, he delighted those of us who then knew him, including his former Cornell colleague, the late Stephen Parrish, the co-general editor of the Cornell Yeats Manuscripts Series.127

89The Yeats editor George Bornstein has written more widely upon editing and conceptualizing the texts of Yeats’s poems, and is far happier than Finneran when it comes to bringing original contexts and materialities of publication to bear upon the interpretations of textual history.128 In an essay entitled ‘Constructing Literature: Empiricism, Romanticism, and Textual Theory’, Bornstein admirably sought to apply ‘traditional empirical standards’ to ‘different constructions’ of literary traditions (such as Romanticism). His method was to ‘ask how well each view fits the evidence, how much evidence contradicts each view, and how internally consistent is each theory’,129 a procedure of exemplary reaction from such theoretical stances of the time as that of Frank Lentricchia, who ‘conceive[d] of theory as a type of rhetoric’ and sought ‘not to find the foundation and the conditions of truth but to exercise power for the purpose of social change’ claiming ‘there is no such thing as eternally “true” theory’.130

90Inconsistently, Bornstein seems content that what was at issue in the Yeats controversy was adequately summed up in the title of the Prologomena to both Editing Yeats’s Poems and Editing Yeats’s Poems: a Reconsideration, ‘The Myth of the Definitive Edition’,131 and draws back from speculating where such thinking leaves Mrs Yeats and Thomas Mark. His co-edition (with Richard Finneran) of Early Essays takes as its copy-text the relevant part of Essays (1924), and the Editors’introduction claims

As with other volumes in the series, we have honored Yeats’s final intentions as expressed by the written record. We thus use as copy-text the final versions of the essays as contained in Essays (1924).132

91With this as their chief criterion in the choice of copy-text, Bornstein and Finneran fudge the issue by their own choice of words. That ‘written record’ which sounds so impressive is in fact meaningless, the archive, though recorded, is disavowed, so that the ‘written record’ is in fact ‘the as-last-published-by-Yeats-himself record’. The ‘final intentions’ thus ‘honoured’ by their choice of copy-text is Yeats’s ‘last-published intention’, the ‘final versions of the essays as contained in Essays [1924]’.

92Elsewhere Early Essays is said to ‘follow the principle of final and expressed authorial intention common to the collected edition’,133 but even this is not accurate. Key volumes in the series were edited upon entirely different principles. Autobiographies, for instance, takes the 1955 Macmillan edition as its copy-text, and that includes ‘lifetime’ authorial delegation as well as posthumous work by Mrs Yeats and Thomas Mark. The point of departure for William H. O’Donnell’s textual introduction to the 1999 volume is the 1932–1936 proofs of the Edition de Luxe, worked on by Yeats, Mrs Yeats, and Thomas Mark, and the corrected page proofs of Dramatis Personae (1936)

that WBY did not himself see, but which he and Mrs Yeats authorized Thomas Mark to correct on his behalf. … The editorial changes introduced in Autobiographies (1955) thus can be regarded as a continuation of the correction process evident on the 1932/1936 proofs for Volume VI’ of the Edition de Luxe… Autobiographies (1955) thus is the best available basis for a text of Autobiographies. The practical result has been that this edition uses the text of Autobiographies (1955) with only thirteen emendations to the 156,000 words of text and three emendations to WBY’s notes.134

93By contrast, Bornstein and Finneran’s copy-text choice for Early Essays would have us believe that that after 1924 Yeats did no work on the text of his essays and that all work delegated by him to Mrs Yeats, Thomas Mark is to be discounted. The introduction offers a brisk chronological guide to the various archives, ranging warily beyond the necessary history of Yeats’s Essays, but selectively omitting any evidence from parallel editions of the prose works, which does not support their case, including that of Autobiographies and Mythologies (2005, and published outside the Collected Works so as to be free from onerous and inappropriate editorial policies).

94The ‘Note on the Text’ to Early Essays is even more suppressive, and the relevant paragraph concludes ‘We have not accepted the weak argument of posthumous delegated authority’.135 But as we’ve seen, this is not, a matter of ‘posthumous delegated authority’ but of lifetime actions by an author. Yeats delegated increasingly as his trust in Thomas Mark (‘that admirable scholar’ was Yeats term for him in a covering letter to Harold Macmillan and in the now lost letter offered Mark, as Mark claimed when he wrote to thank Yeats for his trust ‘a free hand with the punctuation’. The whole correspondence can be followed in the ‘Editorial Principles’ section of the Introduction to Mythologies (2005) and there is little point in repeating the evidence here.136

95Had the textual history of Early Essays been less selective, the comprehension of Yeats’s working methods might well have been greater and the engagement with the work of delegates a little less cursory. Briefly, Yeats was content to leave to Thomas Mark the proposing of such textual emendations as his work across the full range of his work might suggest if uniformity between volumes were to be a consideration—as it was—to Yeats. The 1931–1932 proofs of Mythologies AND The Irish Dramatic Movement offer scores of examples of Yeats’s marginal replies to Thomas Mark’s suggestions. Here are a couple. Mark had questioned on the half-title of the volume whether before quoted speech the ‘“He said” etc.’ phrase should be followed

by either a comma or a colon in the text of this volume. Is it worth while to aim at any uniformity in this detail, as suggested, in the margins?

96Yeats replied

I leave this to Macmillan’s reader. I have accepted his suggestions where ever he has made the correction but I am a babe in such things. Some printers reader put in those colons W. B. Y.

97This is but one of the many evidences of a textually embedded history of working with the publishing process itself in matters of prose punctuation. It stands as a rebuke to the imposed ‘belief’ of the editors of Early Essays (adopted from Curtis Bradford’s Yeats at Work) that ‘Yeats often followed “rhetorical as opposed to grammatical punctuation”’,137 and their consequent practice of an ahistorically ‘light’ punctuation. Under the intense pressure of new work in poor health, Yeats was clear about his reliance on Mark to ‘get his work into the final form he wished’. As Mrs Yeats on 17 April 1939 reminded Mark,

WBY wrote to you in September (or October) 1932 about punctuation and generally asking your help, without which he knew he could never get his work into the final form he wished [emphasis added]. There are, however, a few metrical ‘tricks’ as he called them, and tricks of repetition of words and phrases, deliberately used, which we should, I think, carefully preserve.138

98Again, when Mark sought guidance on the spelling and hyphenization of Irish names such as ‘Knock-na-gur’ (or ‘Knocknagur’) and asked in the margin of The Celtic Twilight ‘One word? Like Knocknarea Knockfefin in Vol. I’. Yeats first replied ‘Yes. W. B. Y’ before adding in a long dependant bubble,

It is difficult to decide on uniform usage. The ^ familiar words above are always written without hyphens. On the other hand the names of woods in Vol I 167 seem to require hyphens to help pronunciation & to mark the words they are compounded from. I would be glad if Macmillan’s reader would decide for me.139

99The ‘Prefatory Note’ to and the ‘Editorial Principles and Note on the Text’ of the 2005 edition of Mythologies summarizes the evidence of Yeats’s delegation both in correspondence and in working practices for Mythologies AND The Irish Dramatic Movement. Calling for a revise of other volumes in the series on 5 July 1932, Yeats wrote to Harold Macmillan,

The volume called ‘Mythologies’ I need not see again. You reader can complete the revision better than I could.140

100By selectively omitting evidence such as this, the chronological overview of the archive in Early Essays not only evades the issue of intention as expectation, but fails to understand Yeats’s working methods with his prose in the creative economy forced upon a poet in unpropitious times, one near the end of his energies, ‘old and ill’ and yet summoning ‘an old man’ s frenzy’.141 In addition to prose requiring to be written (the general introduction and two others), there was prose to be rewritten or augmented, prose to be rationalized to avoid repetition, as we have seen with a passage common to Autobiographies and ‘Dust hath closed Helen’ s Eye’, and there was prose ‘finished’ in the sense only of being ready to be tidied up by delegates. Early Essays and Mythologies are in that category and were to be left safely to the publisher’s reader. An excellent example comes, again on the fly title of the 1931/32 proofs of Mythologies where Mark offers Yeats the opportunity to regularize the usage of the title ‘Saint’ throughout, offering the choices ‘S.’, ‘St.’ or ‘Saint’ and Yeats replies ‘use S.’ How does such an instruction weigh up against Finneran’s gothic charge that the delegates worked to undo Yeats’s final intentions, ‘corrupting the texts which he had worked so hard to perfect’?142 My reply would be that Yeats delegated, and only wish to depart from delegated decisions when he sought to rewrite as distinct from correct.

LE DICTIONNAIRE DES IDÉES REÇUES

101Barthes’s 1967 polemic was a joyous piece of sixties antinomianism, designed to shake up a pedagogical system stifled by Lansonism. Now that iconoclasm which ‘the age demanded’, has dwindled to a drollerie inscribed in the corner of the illuminated page of twentieth century literary history. One catch-cry remains arresting in retrospect. ‘A theory of the work does not exist, and the empirical task of those who naively undertake the editing of works often suffers in the absence of such a theory’, Barthes had warned, thinking of course of oeuvre.143 Lacking any such theory, Yeats’s editors in the 1980s and later had no unified approach to the entire Collected Works, and their ‘empirical task’ certainly suffered from the ‘naivete’ of their undertaking. I well remember the earliest discussions of the project with Richard Finneran in the early 1970s, in London. Volumetrics seemed easy enough, Finneran was cheerfully Bowersite in his kneejerk preference for ‘last-published-in-the-author’s lifetime’ texts (which privileged the Collected Poems and Collected Plays), and told me that he thought little of the ‘Narrative and Dramatic’ poems anyway.

102Very little at the time was known of the Macmillan Archive, and much that was not yet in the British Library was in an apple-barn at Birch Grove House in Sussex, the residence at the time of Harold Macmillan, later Lord Stockton. I was working there amongst the papers in order to finish my work on what became The Secret Rose, Stories by Yeats: a Variorum Edition, and interviewed him about his work and that of his publisher’s reader, Thomas Mark, with Yeats in the thirties, As my perspective on the value of the archive to editors grew, I think Finneran’s resolve to defy it only hardened. In Editing Yeats’s Poems, he projected his own critical doubts about the narrative and dramatic poems into an unconvincing little drama of what he thought Yeats’s own views of them might have been.

Moreover, it may well be that both Yeats and his publishers recognised that—with the arguable exceptions of ‘The Wanderings of Oisin’ and ‘The Shadowy Waters’ —the narrative and dramatic poems were not among Yeats’s best and should be placed so as not to distract from his essentially lyrical achievement. It might also be suggested that the twofold division… accords with Yeats’s sense of antithesis as the dominant characteristic of reality’.144

103Fifteen years later, Finneran was far readier to concede the subjectivity of editorial decision-making,145 but even then, with numerous editors working away in comparative solitude, with never so much as an editorial conference, the enterprise was a world away from ‘The Death of the Author’. The General Editors’ukase against intentionas expectation, their relegation of delegation, one might say, had yet the effect of downgrading the author to an ‘author-function’ without even the esprit de corps that a theory, any theory, might have brought to their work.

104And yet, ‘The Death of the Author’ was so pre-disposed to power over those readers whom it actually impoverished in the name of self-empowerment, that it took no account whatsoever of authors’ testimonies, and testaments. Neither Barthes nor Foucault could be said to have had editorial practice to the forefront of his mind. For Barthes ‘the true locus of writing is reading’ while Foucault did not even glance at international developments.146 Both had played to the gallery of ‘disempowered’ readers in France: la nouvelle critique was to energize readers at the expense of writers. A mere critical constraint, Foucault banished authorial intention from scholarly attention And yet, for all that reader empowerment, the ‘Death of the Author’ itself died during the slow detumescence of ‘Theory’.

105Or so it had seemed. ‘The Death of the Author’ became a slogan in very different schools of thought in very different places. In France, poststructuralists of various allegiances used the term as shorthand for the limitations of the Lansonian view they sought to overturn. In as much as invocation of the names of Barthes and Foucault is still a preliminary rite in the writings of certain editorial theorists, it could be said that editorial theory remains an echo-chamber for increasingly distant signals. The slow tocsin from the funeral of the Author also rang for ‘Authorial Intention’ in Anglo-American editorial theory. Editors working in the tradition of Greg and Bowers clutched at its ramifications when seeking liberation from the obligations intention imposed in that tradition of editorial duty.147 The reduction of the author to ‘la fonction «auteur»’148 enhanced the professional ‘authority’ of the editorial enterprise largely because it seemed to allow for the possibility of eliminating intention as a central problem of textual editing. American editorial thinking, traditionally in search of the professionalization of its business, easily adopted such ideas via the new Society for Textual Scholarship (1979–). Professionalism furthered editors’ authority over texts, and many unhesitatingly unfurled their sails on the Foucauldian wind of change even as it ruffled the very idea of ‘authority’, authoriality’ or ‘authorialism’.149 Richard Finneran, heavily embattled in 1991, reported back that ‘final intentions… a term once widely venerated’ was ‘now at meetings of the Society for Textual Scholarship barely whispered in dark corners’.150 As with Kenner, it was no longer something he could ‘get agitated about’.

106With the Author officially dead, Finneran thus felt free to undermine his own editorial rationale. And yet, Yeats’s avowed intentionalism had indeed been the correct—indeed—the only—principled point of departure for the editing of his works. The General Editor’s abandonment of intention-as-expectation and preference for an older ‘last-in-print equals best copy-text’ approach remains a profound riddle. It was not the result of French theory applied or misapplied. Was it, perhaps, the result of a critical distaste for the narrative and dramatic poems, or of déformation professionelle— an envious or possessive attitude towards authorship, even to the extent—as we have seen—of frank personation of the author, and of an envy of and so hostility to the lifetime editors’ access to an author? Theory-driven, textual annotation might have limbered up, and cross references flourished. Expectation and delegation would have still been blighted. But the constrictions of the Yeats Collected Works policies prompt one to ask what it means for editors to deprive their author of authorial agency against the evidence of the archive and thereby to reduce the author to an ‘author-function’?

107The answers include the following:-

• such a top-down policy reduces the humanity of the author in its rejection of history and context as found in written records of a past reality, and in its rejection of the evidence of the work done by the author’s delegates in the publishing process.

• Under the pretence of clearing rubbish from the mouth of the sibyl’s cave, it accords to an intending editor more elbow-room to ‘find the author-in-the text according to the editor’ s subjectivities, and so

• increases the risk of the editor’s subjectivities coming to dominate the editorial procedures.

• It disempowers readers by rejection the author’s own strategies of selfallusion across the oeuvre.

108Long after ‘The Death of the Author’ had itself apparently expired, faint signals of its lingering after-life have recently been discernible in Anglo-American criticism.151 As Historical Bibliographers and Book Historians have been fully engaged for so many years in addressing the wider implications of the materiality of the text, such approaches seem unlikely to stir beyond ‘theory-nostalgia’.

LA CRITIQUE GÉNÉTIQUE

109So far as editorial theory is concerned, there remains the standing challenge to intentionality associated with genetic editing which has grown from and is therefore shaped by, the study of modern manuscripts. The Institut des textes et manuscrits modernes152 has provided a home for this emergent discipline for many years. Various CNRS-funded ‘equipes’ for intensive single author and comparative study flourish at the intersection of European and international digital developments and provide a powerful theoretical basis for demonstrating the ‘author-function’ in manuscripts and proofs. Genetic techniques have proved at their most rewarding when focused upon ‘avant-texte’ and the stages of composition which precede publication, and can be applied to the puzzles left by the many different hands that have shaped the published (and republished) works, when such evidences are admitted to the field of study.

110More generally, as the move to digital ‘knowledge sites’ has gathered pace, Hans Gabler has sought to theorize digital scholarly editions. In 2010 he moved to concede a strictly limited role to authorial intention, whilst asserting the professional independence of the editor.

… [A]uthority, say (where determinable), or intention (where inferable, or actually evidenced, as may exceptionally be the case) are categories that the editor, in constructing the edition text, will ascertain, register and record. Yet the autonomy postulated for the edition text as the editor’s text will put them in a new perspective. They will cease to be an edition’s a priori determinants. Instead they will function as an edition’s potential regulatives, to be actualized or not according to the editorial rationale. Whether they are chosen or not, that is, to regulate the edition, will be subject to critical reflection and decision. In terms of editorial procedure, this means that it will be deemed acceptable and sound to mark ‘authority’ or ‘intention’ out as criteria or establishing an edition text, as it will be legitimate not to tie the editing to them.153

111Which, one might agree, is a rigorous but entirely accurate statement of the case. However, more recently (and not without the reflex genuflexion to Michel Foucault), Professor Gabler ‘fundamentally questions’ the usefulness for editing of the ‘cluster of notions’ to be found in ‘Authorship—authority—authorisation—the author—the author’s will—the author’s intention’. His argument is conducted from the history of editorial praxis, from stemmatics to present-day ‘author-orientation’. As he himself abstracts his case,

What textual scholarship engages with, directly and tangibly, is not authors but texts (and equally not works but texts), materially inscribed in transmissions. In the materiality and artifice of texts, ‘authoriality’ is accessible conceptually only, in a manner analogous to the Foucauldian ‘author function’. Under such premises, as well, ‘authority’, ‘authorisation’ and ‘authorial intention’ become recognisable as exogenous to texts, not integral to them.154

112This approach evaluates only the documents in a textual progression with cheerful confidence. As such, it is an editorial procedure defined—and perhaps punished—by its refusals. It is reluctant to admit into the editorial horizon another and much wider range of human documents, such as the archive of correspondence between author, agent, publisher, and printer, although it does upon occasion resort to the archive for a critical consideration of definable authorial intention. This having been determined might be allowed in individual instances to establish readings for a critical text. Nevertheless, as Professor Gabler put it in a recent private message to me

to edit a text to authorial intention, let alone to do so in an attitude of fulfilling authorial intention is untenable as an editorial principle.

113And yet, in his view, these editorial procedures, tenable or untenable, have nothing to do with ‘the distinction between the real author in real life and the author function in relation to the text’. Rather, for Professor Gabler, just as one speaks of ‘the implied reader’ as a structural dimension or function of a text, so it makes heuristical sense in terms of text analysis to posit the ‘implied author’. So, when he and other editorial theorists have invoked the Foucauldian distinction, it has been in the attempt to envisage that ‘implied author’ as a ‘function’, and this is ‘what the real author constantly is up against in generating texts through creative acts of writing/reading in the course of composing a text’.155

114Gabler proposes that

If it can be said that Roland Barthes’s ‘death of the author’ (1967) has, as a slogan, generally tended to overshadow Michel Foucault’s significant elucidation of the ‘author function’ (1969), it would probably be also true that textual critics and editors in particular, must be counted among those who still hold both tenets in scorn. (They will insist: ‘The author is real: look, these manuscripts are incontrovertible proof that the author is not dead—or was not when he wrote them!’).

115Thus arraigned, I would offer a different defence. I would say ‘Look at all the penumbral documents which confirm the co-identity of the “real-life author” and the “author-function” of this document, and restores to it its humanity’. Gabler, however, continues:

Seen with a colder eye, however, the proof of the author that manuscripts provide in truth only evidences (alike to footprints in the sand) that an author once (or, as the case may be, repeatedly) traced his or her hand and writing implement over the manuscript page. The real-life author, consequently, cannot honestly be conceded more—though also no less—than an empirical and legal authority over the documents carrying the texts of his works. To concede him or her an overriding authority over those texts, and on top of that to consider those texts, as texts, themselves invested with an innate authority, amounts to performing an argumentative leap akin to what psychology terms a displacement. It is this that constitutes the fallacy suggested. [emphasis added].

116That ‘colder eye’ is unfortunate. A ‘cold eye’ was enough for Yeats’s epitaph; but it wasn’t enough for Hugh Kenner, whose A Colder Eye betrayed in its whole approach an envy of Yeats’s precursive authoriality.156 But perhaps that is just the point. For Gabler’s argument stands only if we can find a way to keep at bay the ‘reallife author’. We can try to reduce his or her ‘works’ until we are left merely with ‘texts’. The ‘textuality’ we have then created however, pretends that we can exclude all that of which the materiality of the text demands we take account. What we sought to repress, returns. The reality of the author is pressingly found in the book, as well as in its attendant documents. Genetic methods have yet to be attuned to the full record of published and revised books, whereas manuscript documents can be deemed—inadequately as it seems to me—to show authorship as ‘écriture’, pure and simple.157

I ADD IN COMMENTARY

117Modern authorship and its archival remains, are neither so pure, nor so simple. Theory-driven editorial techniques attuned to the digital age can be challenged by the sheer range of traces of authors’ activities. Textual restlessness is occasioned by a fresh authorial critique which determines that the written did not fulfil what is now seen ‘to have been’ authorial intention, and the relation between (say) poems and their notes offers a case in point. I pursue this question in a forthcoming essay, but for the moment it is enough to say that the author’s zestful engagement in book-making, including symbolic book-design and its polarity for text-assemblage, authorial compromise with and active participation in the means of publishing and with publishing’s human agents, established and demonstrable patterns of authorial delegation to trusted agents such as publishers’ readers, authorial engagement in promotion, sales and marketing, including interviews, lectures, book-launches with their speeches and signings, the bibliographical opportunism attendant on post-publication revision and republication—all of these leave evidence of intentions and confirm or qualify those found in the ‘texts’; themselves, as well as in other works in an author’s oeuvre, including autobiographical and ‘required’ writing. The wider reader as well as the sceptical thinker remain restlessly aware that the editorial tradition after Barthes and Foucault might itself be too doctrinaire, too ‘top down’. All of which makes biblio-biography urgently necessary,158 to make the theory-driven editor’s job harder, with a pragmatic insistence on proceeding on a case-by-case basis. It will come as no surprise that I continue to work in this field myself.