Peer Reviewed

Exclusive to M/m Print Plus

Crediting Ghosts: W. B. Yeats and George Yeats Collaborate with the Dead

“How on earth, we wonder, could a man of Yeats’s gifts take such nonsense seriously?” exclaimed W. H. Auden. “How could Yeats . . . take up something so essentially lower-middle class—or should I say Southern Californian?”[1] Auden’s incredulous geography was more accurate than he knew. Perhaps the peak of Yeats’s lifelong interest in what Auden dismissively called the “mumbo jumbo of magic” (“mediums, spells, the Mysterious Orient—how embarrassing”) were the “sleeps,” joint experiments by W. B. Yeats and his young wife George Yeats (née Hyde-Lees), begun in earnest in a railway carriage somewhere near Pasadena in Southern California. During an American lecture tour, in “one of those little sleeping compartments in a train,” as Yeats remembered, “my wife, who had been asleep for some minutes, began to talk in her sleep, and from that on almost all communications came in that way.”[2]

So emerged into hearing one of the most extraordinary of modernist collaborations. These “communications” involved more than just the two Yeatses. They were, it seemed, in collaboration with the dead. Dipping (it might appear) into the dream language of a Jungian collective unconscious, came “teachers” who “did not seem to speak out of her sleep but as if from above it, as though it were a tide upon which they floated” (Yeats, A Vision (1937), 8). These teachers manifested themselves through all kinds of sensory phenomena: juddering pictures, flashes of light, drops of water or warm breath; the smells of incense, snuffed candles, violets, old carpets, and cats, who miaowed and made the sleeping George try to lap milk (after a period of “cat trouble” W. B. in response barked like a dog, so scaring his wife he “never dared repeat it”) (8). Sounds were especially prominent, of bells, “ghost whistles,” and snatches of music; occasionally a “direct voice” was heard by one or both parties while awake. Most important was the indirect speech W. B. Yeats describes, as George vocalized utterances and responses to questions while in half-trance. As Yeats’s poem “Fragments” has it, truths poured from “out of a medium’s mouth.”[3]

If this sounds silly, a throwback to the table-rapping of Victorian parlors as much as an anticipation of Californian New Age spiritualism, the conditions, language, and precise texture of these collaborations could have happened at no other time. They enacted a prototypical modernist collision of old and new, tradition and technology, mind and matter, fantasy and feminism, science and psychology. It was timely that, outside a Chicago bookshop, a voice directed George Yeats to Sigmund Freud’s Totem and Taboo, whose essay “Animism, Magic and the Omnipotency of Thoughts” lauded the systemic nature of such beliefs as creating “the first complete theory of the universe.” Freud’s secular explanation (“It might be said that in the last analysis the ‘spirit’ of persons or things comes down to their capacity to be remembered and imagined after perception of them has ceased”) was in tune with the earliest of the sleeps’ gnomic revelations: “They only hear & see, when living think of them, as dead. They are dreaming.”[4] That recalling the dead in memory is close to mediating their thoughts (and dreams) through a living mind is axiomatic to contemporary poems like “All Souls’ Night.” There W. B. Yeats’s dead friends, fellow occult researchers, are called up in witness to the “mummy truths” the poem scrupulously does not tell us, as “none but the living mock.”[5]

Answering Auden’s half-mocking challenges (or probing the explanations of Freud or the Yeatses themselves) turns on questions of credit. How is a properly skeptical scholar to credit such phenomena—and how are they to ascribe credit to those involved? Even if an (un)willing suspension of disbelief is indulged, crediting the various contributions is difficult: who was involved how, and what did they think they were doing? Jacques Derrida’s Specters of Marx proclaims: “There has never been a scholar who really, and as a scholar, deals with ghosts. A traditional scholar does not believe in ghosts—nor in all that could be called the virtual space of spectrality.”[6] Our response, as Catherine Paul wisely notes, can depend “whether we take that word ‘deal’ to refer to horse-trading, negotiations, card-playing, doing battle, or simply associating ourselves with something.”[7] Derrida’s demands for ghostly dealings multiply “there is then some spirit. Spirits. And one must reckon with them,” as he tries to conceive “someone mad enough to hope to unlock the possibility of such an address” (Derrida, Specters of Marx, xx, 12). Derrida is thinking of Marcellus’s terrified plea before the ghost in Hamlet: “Thou art a scholar, speak to it, Horatio.”[8] Derrida’s urging has spawned a whole host of scholarly hauntologies, a “spectral turn” that has inspired considerable productivity and skepticism.[9]

Both Yeatses spoke to ghosts; George Yeats even spoke for them. Yet even if we conceive them as Derridean scholars avant-la-lettre, their ghostly dealings present a challenge. Derrida’s emphasis on time and the voice when addressing revenants, those who come again, is instructive. He finds the twentieth century characterized by the hauntings of technology, the living-on of voice and image in ghostly traces and after-effects. These returning anachronisms haunt modernism productively. So the Yeatses’ experiments engage with contemporary theories of mind, particularly changing understandings of time and perception, and respond to the intrusions of technology, whether railways or the “dead voices, lost sounds, forgotten noises” of past musical instruments or wireless radio, and the aftermath of technologies that killed so many.[10] Paul Valéry’s contemporary essay “La crise d’esprit” (1919) conjures a bleak postwar landscape in which “our European Hamlet is watching millions of ghosts,” a crisis of intellect linked to the looming presence of the war dead.[11] As he tramped the stairs of Soho or Holloway, Yeats’s essays register a renewed wartime vogue for contacting the dead by friends and family members. And, as he notes, “there are more women than men mediums today; and there have been or seem to have been more witches than wizards.”[12] This requires in turn examination of the role of women in such communications. If technology was typically gendered male, and men often sought the dead through machine means, culturally speaking mediumship was gendered as female. The sensitivity required allowed an unexpected but powerful space for female artistry.[13]

Such ghostly addresses might therefore prompt a rethinking of modernism’s collaborations. Scholars of material culture have traditionally neglected attention to the spoken word, to music, to a world of sound technology that includes radio broadcasting and film. With increasing focus on media technologies, multi-modality, and inter-art animation, that emphasis is changing.[14] It is admittedly another thing to direct attention to spirit communicators who are by their own self-definition immaterial, but any “material turn” of modernism must acknowledge the utter saturation of the period with an immaterial emphasis from radio waves and music to acoustic phenomena and ethereal spirits. David Trotter suggests modernism’s media theory “derived from a late nineteenth-century electromagnetic vernacular.”[15] If so, it must include a lexicon of supernatural phenomena, as well as paying more attention to the incorporation of acoustic science in ideas of hearing and embodiment.[16] With the language of supernatural discourse conspicuously flavoring technology (and vice versa), the role of women in furthering esoteric study overall, as well as developing the particular auditory and vocal skills required for mediumship (and thus forming modernist theories of media) might still be more widely acknowledged.[17]

In this context the Yeatses’ railway-carriage revelations represent a unique but widely significant collaboration. Central to them was an indivisible partnership between George Yeats and W. B. Yeats, both of whom had spent their lives involved in occult research. Previous experiments had begun during their depressing Sussex honeymoon, when within earshot of the Western Front’s booming guns George Hyde-Lees “attempt[ed] automatic writing.” If at first it seems she was faking, “after some vague sentences it was as though her hand was grasped by another hand” (Yeats, A Vision (1937), 10). These physical manifestations of the power of the world beyond produced a wealth of trance-state writing, diagrams, and other phenomena, generating altogether a vast body of material over six years of intense labor, two editions of Yeats’s philosophical book A Vision, and what he called the “self-possession” of his later work, evidenced by two monumental books of poetry, The Tower and The Winding Stair (Yeats, A Vision (1937), 7). Documentary evidence of the collaboration survives in the Automatic Script, a record consisting of dozens of notebooks and a cardfile, hundreds of sittings, thousands of pages, and many more questions and answers. For Janis Haswell the “script remains one of the most important, long-term, and provocative examples of literary collaboration in the English language.”[18]

This piece is too brief to describe all the intricacies of the script or system, or its extensive results in prose and poetry. But by concentrating on one method, the sleeps, it hopes to sketch the context, texture, and significance of their production as means of communication and understand how an interest in its own process contributed to a self-consciousness characteristic of modernism.

Nobody working in this area can fail to acknowledge the work of George Mills Harper, the Automatic Script’s general editor.[19] Margaret Mills Harper in the aptly titled Wisdom of Two has rightly insisted on the Script’s artistic and collaborative quality, and recent work has left no doubt that George Yeats’s reading and handcrafted creativity (as well as spiritual sensitivity) was vital to the project.[20] So too was her voice. Not surprisingly scholars have concentrated on the material evidence of automatic writing, especially the script of the first few years which is marked by quickfire question and answer on paper, a kind of semi-lucid written Socratic dialogue. What remains of the sleeps, in particular, is perhaps more challenging. The influence of the indirect contact of the spirits is ever-present but harder to differentiate; on the other hand, the evidence of an intense collaborative effort is if anything stronger. With several handwritings, corrections, combined entries, after-the-fact recollections of sleep conversations and of visionary dreams recalled, it is harder to unpick each party’s contribution, including the several spirits and guides, whose voices become audible through at least three acts of mediumship—or mediation.

From the fully armored ghost in Hamlet Derrida posits a “visor effect”: we do not see who looks at us from the other side. So, he suggests “the one who says ‘I am thy Fathers Spirit’ can only be taken at his word,” and, “since we do not see the one who orders ‘swear,’ we cannot identify it in all certainty, we must fall back on its voice” (Specters of Marx, 7). Despite employing the word “specter,” seeing could never be enough: as a scholar, Horatio must promptly address the ghost’s “sound and use of voice” (Hamlet 1.1.131). So must we. That the Yeatses’ experiments depend on voices of doubtful authenticity does not require us, as does Hamlet’s Wittenberg-educated Protestant skepticism, any more to delay. Examining the precise textures and contradictions of such ghostly voices is thus imperative. Auditing the sleeps’ sonic proficiency and observing the summative capacity of the surviving record, it is plain the partnership involved the meeting and intermingling of conscious as well as unconscious minds.

A new method involving “two consciousnesses dovetaled [sic]” was foretold in January 1919, a year before it became routine (Yeats, Vision Papers, 2:183). But the consciousnesses in question were not necessarily those of the two Yeatses: George Yeats’s connection was with the spirits. The sleep method appeared to be less impeded by the questioning, systematizing, contradicting and even skeptical force of her husband. To “Talk not write in [the] dark” (2:183) made for a less taxing and purer form of communication. Allan Kardec’s classic book on mediumship, well known to both Yeatses, described “hear[ing] the voice of the spirits.” Crucially, “when a person is not himself endowed with this faculty, he can communicate with a spirit by means of an auditive medium, who fills the office of interpreter.” It is this word which emerges in the sleep notebooks as that preferred over medium. “Interpreter” represented a fitting ascription of agency, as Kardec explained:

The part of the mechanical medium is that of a machine, the intuitive medium acts as an interpreter. In fact, to transmit the thought, he should understand it; appropriate it in some sort, in order to translate it faithfully; yet this thought is not his—but it passes through his brain.[21]

Substituting “she” for “he” gets somewhere close to what George Yeats was doing. Her method went beyond mechanical means, requiring scholarship as well as intuition. W. B. Yeats described George as “a scholar in several languages but young and nothing of a pedant.”[22] Employing these skills, her part was not just to mouth or mediate but to translate, to interpret, to appropriate the thoughts of the other side.

Her husband’s part was thus circumscribed: secondary, though undoubtedly required. This was made plain when the interpreter went into too deep a trance, and was found asleep even from the point of view of “Ameritus,” one of the spirits. A hopeful W. B. Yeats intervened, but was crushed: “I asked if he came more through me than her. He said no.” He was rather an assistant at an experiment, playing an accustomed magician’s role, casting on command impressions by thought transference, evoking symbols, colors, sounds, even silences: “When she went to sleep for the night she talked about trumpets that annoyed her & on the advice of Ameritus I suggested that the trumpets made no sound” (Yeats, Vision Papers 3:28). The interpreter’s sensitivity to sounds real and imaginary was a common theme:

[WBY] In train 9 p.m. April 13

A sleap of nearly an hour owing to sounds when train stopped [under?] [GY] being a greater interruption than sounds lost in the noise of the train. Spoke once of train to sleeper but she did not know what I meant. She only knew of the movement. The “movement” had stopped. Was told to imagine purple light after which she could hear no sound, not even my voice. In this state however she began an argument with Ameritus, I of course only hearing her part. She said he was making a fool prophecy. He was telling her that she could have a child with [8 astrological symbols]. [Afterwards in Speaking] I then had to imagine blue light (the train now moving)  In rest of sleep spoke of those who wait, and of Anne Hyde   Those who wait [convey] speak to us only detached sentences   They have no constructive power.  . . .

All facts about Anne Hyde correct. All other names & dates of past incarnations of medium & self & others false, though incarnations psychological truth.

 . . .

True horoscope [symbols of Mars & Venus] more doubtfully [Mars Venus] conjunction. [Mars Venus] in each other’s houses [WBY] ([?or] “some thing like that” – “not mutual reception” (obvious confusion WBY) plainly could not get the word)

On [24] 22 and [24 25] 23 April in train could not write until today 28 April owing to hurried journeys (3:12).

As with early wireless, the difficulty and contingency of the method was intrinsic to its nature. Fragmentary and incomplete communications introduced the possibility of mistakes and the need to correct the record. Beyond the collapsing of time when ghosts and historical personages (like George’s supposed ancestor Anne Hyde) appeared, there were contemporary time-lapses or gaps, for instance in the latter stages of George’s pregnancy, prefigured above (the pair were trying to have their second child), or when some interfering spirits told them not to record the sleeps: “Probably as many words had been spoken in sleep as had been written but I could only summarize and much had been lost through frustration” (Yeats, A Vision (1937), 11). Sleep notation was affected by circumstance, although the sleeps themselves had the advantage of portability, with the notebooks notably well traveled, recording experiments in Pasadena, New York, London, Oxford, Galway. The comings and goings of the Yeatses’ revenants were not, therefore (unlike most Victorian hauntings) dependent on place. When writing was difficult it was possible, indeed potentially conducive, to experiment aurally in a train. Thereby the experiments were buffeted by external events, especially sounds: “[On] Before waking second time she thought [?noise] noises of shunting trains were bombs” (Yeats, Vision Papers, 3:13). Such things did not seem outlandish. Across Europe the First World War had failed to end, and the increasing violence of the Irish War of Independence for a time confined the Yeatses to Oxford.[23] George “smelt paraffin” when considering a trip there: “does this mean burning?” Yeats queried (Yeats, Vision Papers, 3:76). The sleeps thus proceeded with the collusion and collision of technology.

Yet against the intrusions of the world, the sleeps provided a space of extraordinary intimacy. The different handwritings on view here illustrate how entangled and interactive the method was. Yeats’s own hand [WBY] displays distinctive spellings (including “sleep” as “sleap,” making, as on the notebook’s front cover, a visual rhyme with “leap”) before giving way to George’s hand [GY], predominant throughout the notebooks, sometimes overinscribed or subject to handwritten correction (as here concerning astrological predictions), and even then sometimes plainly written to the vocal dictation of her husband. However one considers their origin, there appeared several (sometimes competing) instructors, interpreted through George’s voice, sometimes unanswered, sometimes questioned by her, as also by her husband; with the results recorded afterwards, sometimes long afterwards, and amended by a combination of both while fully conscious. It was a multilayered process, conspicuously polyvocal.

Here Bakhtin’s philosophy of language might be instructive. “The dialogic orientation,” he posits, “is the natural orientation of any living discourse.” So too, the Yeatses found, was the discourse of the dead. While transcriptions smooth over some bumps, in manuscript the sleep notebooks retain the sense of a “dialogically agitated and tension-filled environment of alien words, value judgments, and accents.” According to Bakhtin, this is inherently novelistic, whereas “in genres that are poetic in the narrow sense, the natural dialogization of the word is not put to artistic use, the word is sufficient unto itself and does not presume alien utterances beyond its own boundaries.”[24] By contrast the sleep method is predicated on the presence of “alien utterances” returning from what Hamlet calls that “undiscovered country,” the “sleep of death” (Hamlet, 3.1.65, 79, 278–79).[25] Drawing concertedly on Bakhtin, Julia Kristeva’s positing of the “polylogue” might be helpful here, especially as one of her talismanic texts is Louis-Ferdinand Céline’s Mort à crédit (Death on Credit, published as Death on the Instalment Plan), a novel that plays with anachronic narration, ellipses, and collapsed-sentence discourse.[26] Céline’s more cynical notion of credit implies something that can eventually run out; whereas in the Yeatses’ world the dead are always there, endlessly if intermittently available, if only one could credit them enough to bother to listen to them. Better perhaps as Kristeva does to recognize and acknowledge (to credit) the interplay of voices that play a part in forming any textual entity, including the poetic, and all the dynamics of rhythm and sound that play a part even in prose. That way the dialogic or polylogic nature of discourse can be discovered playing out in the conscious and unconscious textures of poetry, too. Yeats claimed to be pursuing the sounds of speech in poetry, “bringing all back to syntax that is for the ear alone,” but this need not be unitary: as a poet given to obsessive revisions Yeats (and his readers) had become used to the “stitching and unstitching” required in reworking shadowy material.[27] Interpreting and channeling the muse’s wilder utterances is, according to T. S. Eliot, a task for the critical faculty of the poet, as well as of the reader.[28] So the triple layers of interpreting, questioning, and synthesizing that survive in the complex records of the sleeps perhaps mark stages recognizably present in more familiar writing and reading processes, however singular or collaborative. Indeed, the requirement for such rewording seemed precisely to supply the “constructive power” the speaking voices of the dead themselves sometimes lacked—or supplying sense when they “plainly could not get the word.”

George Yeats, dream sketch of Bird and Tower (19 August 1920)
Fig. 1. George Yeats, Bird and Tower (19 August 1920), Sleep and Dream Notebook 6, NLI MS 36 257, image courtesy of the National Library of Ireland, © IVARO London and Dublin 2021.

Again, though, credit where it is due. Any polylogic or revisionary aesthetic must not deny the force of spoken words coursing through the interpreter, emerging directly “out of a medium’s mouth.” Such a visceral sonic presence caused renewed self-consciousness about the origins of the phenomenon. Certainly, the sleeps’ vivid orality exerted considerable creative pressure, as a break into automatic drawing confirmed. Beneath a sketch of a tower and large bird with open “big mouth” appeared: “The speaker. The creator. / This is a magnet not a lesson” (Yeats, Vision Papers 3:36). Describing thus a vocal interpreter unusually imbued with sensitivity and nuance in making as well as receiving sound could not but affect the image of the poet. Crucially, hearing and speaking were inextricably combined in the sleeper, and this dynamic combination of aurality and orality spurred a heightening of all the senses, in surprisingly synesthetic sensitivity:

This sight was described in one sleep as having also the quality of touch. In other sleep the Spirit was said to obtain the quality of touch by seeing us touching. Sight is emotional. During the shiftings the spirit can only hear us and use our hearing. Hearing is intellectual. (3:17)

These after-death “shiftings” themselves shift stable ground: reversing the customary history of the senses which, even before the Romantics, associated sound and music with emotion, and sight with rational intellect. In the system “intellect” was not merely a function of the rational brain. Soon “Ameritus” would suggest the substitution of the words “‘creative Genius’ with ‘intellect,’ intellect as the word was used up until C18th,” remembering its Latin root intellegere, to discern, to choose between. Adopting this definition, it is the aural sense which potentially represents a discerning but creative force, leaving sight affective but passive. Though not claimed as universal, this is confirmed (up to a point) by the texture of the sleep scripts. Given the communications resulted in two illustrated publications called A Vision, the visual sense was hardly dismissed. Still, the masks, prefaces, and shifting voices of these published texts draw on philosophical dialogues, as if to recall their dialogic origins.[29] And for W. B. Yeats who admitted he had never seen a ghost (though he had heard and smelt them), the interpreter’s extreme sensitivity to sound could only intensify his own interest in sounded poetry.

In direct response his poems become newly sonically charged. In Oxford the experimenters were disturbed by the welcome cry of an owl, real or imagined, and as the clocks chimed midnight, “When the last strike was finished he said ‘Sounds like that are sometimes a great pleasure to us’” (Yeats, Vision Papers 3:41). Hardly surprising that the bells of midnight should inaugurate Yeats’s 1920 poem “All Soul’s Night”, set in Oxford in tribute to the untold “mummy truths” rendered by the sounds and silences of the sleeps. A decade later the “great cathedral gong” of “Byzantium” (actually referring to the orthodox church’s “semantron,” a rhythmically beaten wooden beam) sounds a fitting answering echo.[30]

Such sounds invoke but also undermine the grandeur of organized religion in favor of private understandings of the afterlife disclosed by sonic communications. “Sooner or later I must . . . announce the poetry of belief,” wrote Yeats in 1924.[31] He never did, and five years later admitted: “I agree with Ezra [Pound] in his dislike of the word ‘belief.’”[32] Catherine Paul has aptly described the “problem of belief” when it comes to addressing Yeats.[33] If “belief” was not quite the word, neither was “doubt.” Bakhtin argues that “in poetry, even discourse about doubts must be cast in a discourse that cannot be doubted” (Dialogic Imagination, 286).  Though they perhaps vocalize belief better than doubt, Yeats’s poems credit us with doubt. They are also increasingly self-conscious about their vocalizations, through which their discourse becomes porous, and thus, unusually, as Bakhtin puts it, “forming itself in an atmosphere of the already spoken” (280).  By collapsing time and conceiving a past and continuing interaction with the dead, Yeats’s poems, always anticipating an answer, turning on their origins, increase in self-possession, reckoning with doubters and believers of ghostly voices. His twelve radio broadcasts and the staging of a séance ventriloquizing Swift in The Words Upon the Window Pane are evidence of his connecting techne and technology, media and mediumship, sound and the supernatural.[34] His doing so, again, should help challenge narrow conceptions of periodicity. All this was impossible without the example of the “sleeps.”

Silicon valley’s obsession with transhumanism, say, shows (as with the sleeps) the supernatural still coexists with Californian technology: if we are just downloadable information, then immortality seems conceivable and “the future belongs to ghosts,” as Derrida promised.[35] But so did the modernist past: which requires us to address them, somehow. I have, therefore, tried the word credit over belief: scholars can and should express doubts, or they can seem untrustworthy second-hand ghost-dealers. Even George Yeats, addressing “that problem ‘Nobody ever felt certain about Yeats’s faith in the occult. Did he really believe it or was it just a playing with fantastic images, etc.?’ could only conclude: ‘I myself find it difficult to know. Of this side in his life I am writing.’”[36] These written answers, if there were any, never appeared. But her voicings of the dead deserve some kind of credit, even as she herself demurred, furious when her contribution was revealed. Female mediums could typically seem attractive as collaborators due to what Galvan calls an apparent “sympathetic excess,” their alleged openness to feeling and spirituality (The Sympathetic Medium, 15). This might be true but is not sufficient in crediting George Yeats. It is thanks in particular to her acoustic sensitivity, but also all of her auditing, interpretative, and vocal powers—and her scholarship—that Yeats’s work answers Marcellus’s call, and Hamlet’s, who remembers “That skull had a tongue in it, and could sing once” (Hamlet 5.2.77, 309). Thereafter all his poetry, prose, and plays calls up, speaks to, questions and cross-questions the dead; which texts thus become no longer inert inscriptions but living scenes that not just record but vividly recall dead voices: acts of “live,” anachronistically present speech—whether or not we credit ghosts.


[1] W. H. Auden, “Yeats as an Example,” The Kenyon Review 10, no. 2 (1948): 187–95, 188; emphasis in original.

[2] W. B. Yeats, The Collected Works of W. B. Yeats, Vol. 14, A Vision (1937), ed. Margaret Mills Harper and Catherine E. Paul (New York: Scribner, 2015), 8. A Vision and surviving records suggest the train was in Southern California; another notebook suggests it might have been “on way to San Francisco”, (see The Collected Works of W. B. Yeats, Vol. 13, A Vision (1925), ed. Catherine E. Paul and Margaret Mills Harper [New York: Scribner, 2008], 218).

[3] W. B. Yeats, “Fragments,” in Yeats’s Poems, ed. A. Norman Jeffares (London: Macmillan, 1996), 321.

[4] Sigmund Freud, Totem and Taboo, ed. and trans. James Strachey (London: Routledge, 1960), 94. W. B. Yeats, Yeats’s Vision Papers, ed. George Mills Harper, et al. (London: Macmillan, 1992–2002), 3:9.

[5] W. B. Yeats, “All Souls’ Night,” in Yeats’s Poems, 340–44, 341.

[6] Jacques Derrida, Specters of Marx, trans. Peggy Kamuf (London: Routledge, 1994), 11.

[7] Catherine E. Paul, “Listening to the Dead: Yeats and Spirits,” (keynote lecture at The British Modernities Group Conference, University of Illinois, Champaign-Urbana, April 2018). I’m grateful to Professor Paul for a transcript of this brilliantly thought-provoking lecture. See also Catherine E. Paul, “W. B. Yeats and the Problem of Belief,” in Warwick Gould, ed. Yeats’s Legacies: Yeats Annual No. 21, ed. Warwick Gould (Cambridge, UK: Open Book Publishers, 2018), 395–316.

[8] William Shakespeare, The Arden Shakespeare: Hamlet, ed. Harold Jenkins, (London: Routledge, 1982), 1:1.48; 168.

[9] For both see Roger Luckhurst, The Invention of Telepathy 1870-1901 (London: Oxford University Press, 2002); Jeffrey Sconce, Haunted Media: Electronic Presence from Telegraphy to Television (Durham, NC: Duke University Press, 2000); and Steven Connor, Dumbstruck: A Cultural History of Ventriloquism (Oxford: Oxford University Press, 2000).

[10] Villiers de l’Isle-Adam, Tomorrow’s Eve, trans. Robert Martin Adams, (Urbana: University of Illinois Press, 2001), 10.

[11] Paul Valéry, “The Crisis of Spirit,” in The Collected Works of Paul Valéry, Vol. 10, History and Politics, trans D. Folliot and Jackson Mathews, (Princeton, NJ: Princeton University Press, 1971), 23–36, 25.

[12] W. B. Yeats, “Witches and Wizards and Irish Folklore,” in Visions and Beliefs in the West of Ireland, ed. Lady Gregory (New York: The Knickerbocker Press, 1920), 247–64, 258.

[13] See Jill Galvan, The Sympathetic Medium: Feminine Channeling, the Occult, and Communication Technologies 1859–1919 (Ithaca, NY: Cornell University Press, 2010). Female writers of the period from Helena Blavatsky and Annie Besant to Gertrude Stein and Hilda Doolittle (H.D.) seem especially open to collaboration, not only with each other, but with the dead: see Bette London, Writing Double: Women’s Literary Partnerships (Ithaca, NY: Cornell University Press, 1999), and Holly A. Laird, Women Coauthors (Urbana: University of Illinois Press, 2000). The case of the Anglo-Irish partnership “Somerville and Ross” is particularly pertinent: it continued to operate as a creative entity only because Edith Somerville claimed to keep in touch with her collaborator Violet Martin long after her death. Male pseudonyms like “Michael Field” (denoting aunt and niece Katherine Bradley and Edith Cooper) might hide enviously productive female partnerships; as the nineteenth century declined, to gain access to such a vein of creativity, male authors, such as Yeats’s friend William Sharp, were inclined to conjure into being elaborate but entirely fictional female alter-egos with whom to collaborate, in his case a Scottish poet called Fiona MacLeod (see William F. Halloran, The Life and Letters of William Sharp and “Fiona MacLeod,” Vol. 1, 1855–1894 (Cambridge, UK: Open Book Publishers, 2018).

[14] See as examples Broadcasting Modernism, ed. Debra Rae Cohen, Michael Coyle, and Jane Lewty (Gainesville: University Press of Florida, 2009); Gunther Kress, Multimodality: A Social Semiotic Approach to Contemporary Communication (New York: Routledge, 2010); and Stephen Connor, Beyond Words: Sobs, Hums, Stutters, and Other Vocalizations (London: Reaktion, 2014).

[15] David Trotter, “Modernism’s Media Theory,” Critical Quarterly 58, no. 3 (2016), 5–26, 6.

[16] Much work is still to be done to follow the lead of scholars like Alexandra Hui (in The Psychophysical Ear: Musical Experiments, Experimental Sounds [Cambridge, MA: MIT Press, 2013]) and Robert Michael Brain (in The Pulse of Modernism: Physiological Aesthetics in Fin-de-siècle Europe [Seattle: University of Washington Press, 2015]) in, for example, tracing the acoustic theories of the artistic-minded scientist Hermann von Helmholtz and their influence on modernism.

[17] Friedrich A. Kittler dismisses the “glibness” of automatic writing (in Discourse Networks 1800/1900, trans. Michael Metteer with Chris Cullens [Stanford, CA: Stanford University Press, 1900], 228). He furthermore goes as far as to say that the dissemination of new technologies in the workplace led to the “desexualization of writing” [Gramophone, Film, Typewriter, trans. Geoffrey Winthrop-Young and Michael Wutz (Stanford, CA: Stanford University Press, 1999), 187]. Rightly, Jill Galvan (in The Sympathetic Medium) takes issue with this: indeed, with women increasingly involved in the realm of textual production (the word “typewriter” for a time referred to both the mechanical object and the person—often a woman—who operated it) the association between new technology, communication, and the supernatural, as in the Yeatses’ intimate experiments, stood ready to “resexualize” writing, in both senses of the word. The automatic script briefly questions if the medium’s body is a “kee-board” [sic], but the subtlety of touch implied and George’s skill at the instrument, to my mind suggests a piano rather than a typewriter, reminding us of the importance of music and sound technology to the experiments (Yeats, Vision Papers, 1:349).

[18] Janis Haswell, “Yeats’s Vision and the Feminine,” in W. B. Yeats’s A Vision: Explications and Contexts, ed. Neil Mann, Matthew Gibson, and Claire Nally (Clemson, SC: Clemson University Press, 2012), 291–306, 292.

[19] W. B. Yeats, Yeats’s Vision Papers, ed. George Mills Harper, et al. (London: Macmillan, 1992–2002) 4 vols; transcriptions here derive from this edition, retaining original spelling but corrected by sight of the original MSS in the National Library of Ireland.

[20] See Margaret Mills Harper, The Wisdom of Two (Oxford: Oxford University Press, 2006); Charles I. Armstrong, Reframing Yeats (London: Bloomsbury Academic, 2013); Neil Mann, A Reader’s Guide to Yeats’s A Vision (Clemson, SC: Clemson University Press, 2019). Dana Garvin (2020 PhD, University of Limerick) is unearthing new riches in the playful line-making, symbols, and drawings of the written scripts.

[21] Allan Kardec, Book on Mediums, trans. Emma Wood (York Beach, ME: Samuel Weiser, 1970), 221.

[22] Yeats to Lady Jean Hamilton, June 19, 1920, Collected Letters, (InteLex Past Masters, 2002), 3740.

[23] See Robert Gerwath, The Vanquished: Why the First World War Failed to End, 1917–1923 (London: Penguin, 2016).

[24] M. M. Bakhtin, The Dialogic Imagination: Four Essays, ed. Michael Holquist, (Austin, TX: University of Texas Press, 1981), 279, 286.

[25] Helen Sword notes the quasi-modernist texture of the Yeatses’ automatic writing communications with their interpolations and interruptions (Ghostwriting Modernism [Ithaca, NY: Cornell University Press, 2002]) while Catherine Paul finds a fragmentary poetics in the script’s question and answers, a free form that only ‘Broken Dreams’ among Yeats’s published poetry even aspires to (see Paul, “Listening to the Dead”). Sword’s analysis is penetrating, though I find myself less convinced that the very difficulty of the communications led to an Eliotian mistrust of language per se: in the script W. B. Yeats rather comes across as enjoyably piqued by the contention of spirit dialogues, and, especially when it came to the sleeps, energized by inconsistencies and the presentness of anachronistic collapses of time. Paul’s interest in the moment when ‘reading become[s] listening’ mirrors Yeats’s own, and might be helpfully applied in particular to the precise methods of the sleeps.

[26] See Julia Kristeva, Polylogue (Paris: Éditions de Seuil, 1978), 166–73.

[27] W. B. Yeats, “An Introduction for My Plays,” Essays and Introductions (London: Macmillan, 1961), 527–29, 529; Yeats, “Adam’s Curse,” in Yeats’s Poems, 132–33, 132. For more on the shadowy nature of revisions, omissions, and allusions, which are termed “apparitions of writing” (and including a chapter that investigates Keats’s haunting of Yeats) see Susan J. Wolfson, Romantic Shades and Shadows (Baltimore, MD: John Hopkins University Press, 2018).

[28] See T. S. Eliot, “The Function of Criticism,” in Selected Essays, (New York: Harcourt Brace, 1923), 12–24. The ghostly pressure of past writers and selves on the present and (anachronistically) of the present on the past is vital to Eliot’s critical theory, and might be adjudged important to his own writing life. Recently revealed comments on his correspondence with Emily Hale suggests a need to consider and confront such past writing selves as revenants: “I came to see that my love for Emily was the love of a ghost for a ghost, and that the letters I had been writing to her were the letters of an hallucinated man,” T. S. Eliot, “Directions to my executors regarding the envelope enclosed herewith” (Houghton Library, Harvard, MS: 1960, 1963).

[29] Indispensable recent editions source the exhaustive philosophical readings (and revisions) undertaken by both collaborators: see W. B. Yeats, A Vision (1925) and A Vision (1937).

[30] W. B. Yeats, “Byzantium,” in Yeats’s Poems, 363.

[31] Yeats to Frank Sturm, July 3, 1924, Letters, 4583.

[32] Yeats’s unpublished manuscript from January 1929, quoted in Richard Ellmann, The Identity of Yeats (London: Faber:1983), 239.

[33] See Paul, “W. B. Yeats and the Problem of Belief.”

[34] See Adrian Paterson, “Music will keep out temporary ideas: W. B. Yeats’s radio performances,” Word and Music Studies 12: Performativity in Words and Music, ed. Walter Bernhart and Michael Halliwell (Amsterdam: Rodopi, 2011), 43–76, and Pierre Longuenesse “Playing with Voices and Doubles in Two of Yeats’s Plays: The Words upon the Window-pane and A Full Moon in March,” in Yeats’s Mask: Yeats Annual No .19, A Special Issue, ed. Warwick Gould, (Cambridge, UK: Open Book Publishers, 2013), 103–20.

[35] Jacques Derrida and Bernard Stiegler, “Spectrographies,” in The Spectralities Reader: Ghosts and Haunting in Contemporary Cultural Theory, ed. Maria del Pilar Blanco and Esther Peeren, (London: Bloomsbury, 2013), 37–52, 38. For transhumanism see Mark O’Connell, To Be A Machine (London: Granta, 2017).

[36] Quoted in Ann Saddlemyer, Becoming George: The Life of Mrs W. B. Yeats (Oxford: Oxford University Press, 2002), 594.