[Robartes:] An attempt to go to Mecca ended in my finding myself in a remote Arab town without money or friends without money in a remote Arab town. Thanks to a small medicine chest, which I always carried I became first doctor, & then a kind of steward to an Arab chief, or petty king. One day whenwe were our town, or rather village was visited by a tribe of Arabs of the obscure company of B________ an Arab sect well known at Fez in the time of Leo Africanus, who gave a very unfavourable description of them. One finds amongst them the greatest extremes holy men & others exceedingly licentious, or even cruel living in the same little wandering together in tolerable amity. Indeed they lacked moral indignation beyond any people I have ever met. One of them an old man celebrated for his wisdom whom I was treating for some complaint cried out in wonder at see the title page of my book [Giraldus’s Speculum]. It contained he said certain doctrines of the (?Bacleones) & on learning that in their hands was He possessed an ancient Arab Ms, called ‘The Camels back’ containing their doctrines I resigned my post & wandered with him and his people for some years.
The Arab element in the fictions of A Vision predates the entry of Kusta ben Luka, and in fact brought it about, but it appears to have balanced the European Giraldus from the first. The initial impetus behind Yeats’s choosing an Arabian background for part of the scenario in which Robartes would expound the ideas of the Automatic Script is not entirely clear, but two elements in particular seem to have suggested this context. The first is the appearance of the twenty-eight phases of the Moon, since ‘Their number is that of the Arabic Mansions of the Moon’ (AV A 12), and the second is the link with Per Amica Silentia Lunae and the Daimon.
The schema of the twenty-eight phases of the Moon first began to emerge at the end of November 1917, and the Yeatses must have been intrigued by the possibility that they bore some relation to the twenty-eight Mansions of the Moon from Arabic astrology. Though there are systems of lunar Mansions in Chinese and Indian practice, the Chinese sieu are not really lunar, and the Indian nakshatras are usually reduced to twenty-seven. It was also the Arabian tradition that had entered into mediaeval Europe through Spain, and influenced the astrology of Western tradition, with classical and Hellenistic ideas often coming via Arabian writers, so that, for instance, the ‘Lots’ of Hellenistic practice are still known as the ‘Arabic Parts’ (Part of Fortune, Part of Spirit etc.) and the Mansions of the Moon are more commonly called by Arabic names (or versions of them) than by Latin ones.
One of the earliest drafts, the ‘Untitled Manuscript’ quoted above, refers to ‘a company of B________ an Arab sect well know at Fez in the time of Leo Africanus’ (YVP 4 122). The reference to Leo is significant, as Yeats had researched Leo Africanus prior to 1915, when he had considered that this Arab geographer was his Daimon (see ‘The Manuscript of 'Leo Africanus'’), and had alluded to him in Per Amica Silentia Lunae when he comments that ‘a strange living man may win for Daimon an illustrious dead man’ (Myth 335). It is possible that this Daimonic association provided another initial impulse towards an Arabian context. Douglas Goldring remembered hearing Yeats discoursing on ‘the influence on character of the phases of the moon’ and claiming that ‘the teaching he was about to expound . . . had been revealed to him 'in a dream' by a Moorish initiate with whom he had made contact on the astral plane’ (The Nineteen Twenties, 118). The Moorish initiate sounds more like Leo Africanus (al-Hassan ibn Mohammed al-Wazzan, c.1494-c.1552) than any of the actual Instructors, though the spirit known as Leo came to be regarded as one of the chief ‘Frustrators’ who were trying to impede the development of the System. Goldring’s account, though it may not be entirely accurate, is also evidence of the way in which Yeats developed his fictions through rehearsing them as oral inventions as well as in written form, no doubt modifying and trying new aspects with each retelling. (Goldring dismissed the story as ‘a device for riveting the attention of an audience he was eager to impress’, since he had sub-let Yeats’s rooms in Woburn Buildings and recalled ‘that one of the books in Yeats’s kitchen was a treatise on the mystical significance of the phases of the moon, which had been published in America’, which raises another fascinating possibility, which depends upon ascertaining what book he might be referring to.) The communication with the Moorish initiate on the astral plane also recalls the contact made in 1908 by R. W. Felkin, his wife and daughter (from a previous marriage), with Ara ben Shemesh ‘a discarnate Arab who claimed affiliation with the desert-temple visited by Father Christian Rosenkreutz on his Middle Eastern pilgrimage’ (Colquhoun, Sword of Wisdom). Felkin, who was leader of the Stella Matutina, one of the splinters of the original Order of the Golden Dawn and the one with which the Yeatses continued involvement (see Harper, Yeats’s Golden Dawn, Chapter 9), ‘accepted Ara ben Shemesh as his teacher and the latter's 'Sun-Masters' as his Secret Chiefs’ (Colquhoun). Ithell Colquhoun, incidentally, suggests that Yeats’s first contacts with Leo in 1909 might have been a form of emulation, though this was only fleeting, and regular contact at séances did not start until 1912.
Once he had decided on the Arabian context, Yeats went to Sir Edward Denison Ross to give him more authentic Arabic clothing for his ideas, and wrote to Augusta Gregory in 1918 about ‘a sect of Arabs called the Judwalis (diagrammatists). Ross helped me with the Arabic’ (L 644). Suheil Bushrui, in ‘Yeats’s Arabic Interests’, comments that ‘the name was invented by someone who knew enough Arabic to be able to derive it from the word 'jadwal' (in this sense meaning 'diagram')’ but that ‘Ross was much more of a Persian scholar than an Arabist’, since the form is incorrect (In Excited Reverie, 295-96).
The ‘Untitled Manuscript’ seems to pre-date Ross’s input, since it does not mention the Judwalis, Kusta ben Luka or his lost book, ‘The Way of the Soul between the Sun and the Moon’, referring instead to an existing ‘ancient Arab Ms, called 'The Camels back'’. The ‘Appendix by Michael Robartes’ seems to come from a similar date, referring to ‘The Camel’s Back’, and vague, unspecified ‘Arabs’ (Yeats and the Occult, 210-215). In it Robartes refers to ‘copies of two diagrams from the "Camel’s back"’, which he has ‘given their arabic names, the first "the holy women and the two Kalendars" because it describes the movements of two symbolic suns and moons; and the second "the dance of the Eunuch with the favourite wife"’ (YO 210). These names probably derive from Yeats’s reading of the Arabian Nights, an important imaginative influence, which he used for flavour and romance. In the Introduction to A Vision A Yeats comments that: ‘Doubtless I must someday complete what I have begun, but for the moment my imagination dwells upon a copy of Powys Mather’s "Arabian Nights" that awaits my return home. I would forget the wisdom of the East and remember its grossness and its romance’ (AV A xiii), referring to a relatively recent purchase, The Book of the Thousand Nights and One Night (1923), but the Arabian Nights already pervade much of the fiction surrounding the exposition.
The ‘Untitled Manuscript’ also contains a note to himself concerning the desert animals: ‘(?what beasts. look up Doughty)’ (YVP 4 123), indicating one of his other sources, this time for verisimilitude, Charles M. Doughty’s Travels in Arabia Deserta. Like the Mathers translation of J. C. Mardrus’s Les mille et une nuits, Yeats’s complete copy of Travels in Arabia Deserta was only published in 1923, but he already had a two-volume abridged version (Wanderings in Arabia [1908; YL 539]. The fact that he was buying new and fuller versions of works that he already partly knew shows their importance to him, and Suheil Bushrui concludes that Arabia Deserta and the Arabian Nights ‘stand out singularly as the most important sources of his inspiration and the most powerful factors in influencing his imagination so far as his Arabic researches are concerned’ (In Excited Reverie, 314).
Though the Mardrus/Mathers version of the Arabian Nights was new, Yeats had first encountered the stories at the age of seven, when his father had given him Five Favourite Tales from the Arabian Nights in Words of One Syllable (YL 676) and he also evidently knew the Burton translation (see Warwick Gould’s essay, ‘'A Lesson for the Circumspect'’, in The ‘Arabian Nights’ in English Literature, 245ff.; see comparisons of various English versions and texts at another web-site). The diagram called ‘the holy women and the two Kalendars’ in the ‘Appendix by Michael Robartes’ prefigures ‘The Dance of the Four Royal Persons’ (AV A 9), and the tale of ‘The Porter and the Three Ladies of Baghdad’ contains the tales of the three ‘Kalandars’ in Burton’s translation, where previous versions had referred to these dervishes as Calenders (Mathers also uses ‘Kalandar’, for Mardrus’s saâlouk, and the men in question are all actually royal persons disguised). The world of the Arabian Nights, both the framing world of Shahryar and Shahrazad (Scheherazade) and the world of the stories, often the Baghdad of Harun al-Rashid, stands behind much of the fictional apparatus of A Vision A. It carries with it a certain Orientalism, in the sense that it creates an exotic other world which defines the West by opposition, in the early appendix, opposing the Western vision of ‘theological Gyraldus’ with the Arab tribes’ ‘singular tolerance’ (YO 212-13; see the Fictions of A Vision and Edward Said, Orientalism).
The engagement with the Mansions and Leo Africanus may have been one factor in his choice of an Arabian setting, but the scenario also starts from trying to find an acceptable context for mediumship and divination. The Automatic Script may not appear to have much in common with divination, but this aspect was suggested by a ‘curious message from Bessie Radcliffe, “They departed with the rewards of divination in their hands”’(L 643-44), which she had sent shortly after the Yeatses had started their sessions. Mediumship and divination therefore were represented by dreams and geomancy, the latter as patterns in the sand, the practice’s original form alluded to in the Arabian Nights. The geomantic sands had also been evoked in Per Amica Silentia Lunae, where Yeats notes that he will not provide personal proof of his ideas, though he could, and ‘I will but say, like the Arab boy that became Vizier: ‘O brother, I have taken stock in the desert sand and of the sayings of antiquity’ (Myth 343), an elliptical allusion to geomancy in the story of ‘King Wird Khan, his Women and his Wazirs’ (Gould, ‘'A Lesson for the Circumspect'’, 245-46; see the text of Burton’s version ['I know this from the sand wherewith I take compt of night and day and from the saying of the ancients'], and Payne’s version ['I know this from the sand wherewith I tell the tale of night and day and from the saying of the ancients'], at another site).
Upon a moonless night
I sat where I could watch her sleeping form,
And wrote by candle-light; but her form moved,
And fearing that my light disturbed her sleep
I rose that I might screen it with a cloth.
I heard her voice, ‘Turn that I may expound
What’s bowed your shoulder and made pale your cheek’;
And saw her sitting upright on the bed;
Or was it she that spoke or some great Djinn?
I say that a Djinn spoke. A livelong hour
She seemed the learned man and I the child;
Truths without a father came, truths that no book
Of all the uncounted books that I have read,
Nor thought out of her mind or mine begot,
Self-born, high-born, and solitary truths,
Those terrible implacable straight lines
Drawn through the wandering vegetative dream,
Even those truths that when my bones are dust
Must drive the Arabian host.
. . . .
A dozen nights of natural sleep, and then
When the full moon swam to its greatest height
She rose, and with her eyes shut fast in sleep
Walked through the house. Unnoticed and unfelt
I wrapped her in a hooded cloak, and she,
Half-running, dropped at the first ridge of the desert
And there marked out those emblems on the sand
That day by day I study and marvel at,
With her white finger.
Illustration by Frank Brangwyn for Edward William Lane’s translation (London: Cecil Palmer, 1921)
from ‘The Gift of Harun Al-Rashid’ (VP 467-68; cf. AV A 125-26)
The Arabian setting of ‘Desert Geometry or the Gift of Harun Al-Raschid’ is located solidly in the world of the Arabian Nights with references to Shahrazad’s famous Caliph and to the story of Jaffer, but with the introduction of Robartes’ quest comes the need for a more contemporary framework. Although Yeats was still thinking in terms of diagrams in the desert sand, with respect to the Judwalis, therefore, Travels in Arabia Deserta is the more important source and in a note for The Only Jealousy of Emer he uses the phrase to place the sect in Doughty’s terrain, referring to ideas ‘which Robartes found in the Speculum of Gyraldus and in Arabia Deserta among the Judwalis’ (VPl 566). Elements of Doughty’s experiences influence Yeats’s picture of Robartes: Doughty had been forced to join a Haj caravan from Damascus in order to try to reach the Nabatean remains of Medáin Sâlih, and, though a Christian, had gained respect from the Bedouin through his medical skills, which included administering vaccines. The names of places in the Arabian peninsular, such as Hâyel and Aneyza , and reference to ‘the Wahâbies’ and their siege of Aneyza (VPl 790) appear to be derived from Doughty, though he uses the slightly different forms of ‘Hâyil’ and ‘Waháby’ (transcriptions still vary, including Hail and Ha’il, Anaiza and Unayzah, Wahabi and Wahhabi).
George Yeats told Birgit Bjersby that her husband read Doughty ‘over and over again’ (The Cuchulain Legend 127 n.1), and Lady Gregory shared Yeats’s enthusiasm, reproving Leonard Woolf for ‘his belittling of Doughty’ (Journals 2 337). Yeats later recalled that, ‘During her last year she kept an unmoved face amid great pain. . . . at her side always Arabia Deserta and the New Testament in Gaelic’ (‘Modern Ireland’; lecture, edited by Curtis Bradford in Irish Renaissance). She also had more personal experience of the Arab world, having lived in Alexandria with her husband, where she had had an affair with Wilfred Scawen Blunt. Blunt, like Doughty, had travelled among the Bedouin and his wife, Lady Anne, had written Bedouin Tribes of the Euphrates (1879) and A Pilgrimage to Nejd, The Cradle of the Arab Race (1881). Lady Gregory may therefore have been another source of information or anecdote and provided a further circuitous link to the tradition of the European fascination with Arabia and the desert Bedouins, as well as Egypt.
It is difficult to say exactly what kind of group Yeats imagined the Judwalis to be; Robartes refers to ‘several tribes of this strange sect’ (AV A xviii), ‘the two or three wandering companies of Judwalis’ and to an old Judwali behaving ‘as became the votary of a small contentious sect’ (VPl 790). Yeats refers to ‘a certain obscure Arab tribe’ in the preface to Michael Robartes and the Dancer (1921; VP 853) and to ‘Kusta ben Luka, a philosopher of Bagdad, and his Beduoin followers’ in the notes for The Tower (1928; VP 830). Yet whether Yeats saw the Judwalis as more an ethnic or religious grouping, there is in the end little difference in his mind between the two and, since many sects of the region have tribal affiliations, this is perhaps a nice distinction. The apparent surprise of Robartes that, ‘One finds amongst them holy men & others exceedingly licentious, or even cruel wandering together in tolerable amity’ (‘Untitled Manuscript’), suggests that he might expect more homogeneity, and implies therefore a set of shared moral values, since otherwise the range seems only natural.
In a draft (‘Version B’) Yeats refers to ‘several tribes of this strange sect which owes its foundation to a christian philosopher but outwardly conforms to Mohammedan usage. . . . The Judwalis get their name which means measurer from the diagrams, which their old men [drew] upon the sand for the instruction of the young & these diagrams are derived they say from a certain book written by one Kusta ben Luki’ (YVP 4 142-43; NLI 30,525). In this, the Judwalis could be modelled on any of various sects mentioned by Doughty such as en-Nuseyrîeh or the Nusayris (‘Alawîya, Alawites), Druses or Druze (Muwahhidoon to themselves) (Arabia Deserta 2 401), or others such as the Yazidis (Dasni), all of which hold forms of belief in reincarnation, though the Druze of the Levant show particular points of similarity, and are concentrated near the route from Jerusalem to Damascus which Robartes takes in A Vision A (AV A xviii). The Druze creed arose at a date shortly after the life of Qusta ibn Luqa, during the rule of al-Hakim (996-1021 CE), and has its origins in Isma‘ili Shi‘ism, blending Islam’s unitarian monotheism with elements of Greek philosophy and Hinduism. The Druze are permitted to camouflage their religion and the contents of the sacred books are not fully revealed to ordinary believers. Even their exoteric doctrines are largely secret to outsiders, who are not proselytised, since they consider that everyone had an incarnation during the open period of the religion and had the chance to convert then.
Detail of map of the Middle East, from T. E. Lawrence, The Seven Pillars of Wisdom (private, 1926; public, 1935), highlights added. The map shows some of the tribal and sectarian groupings in the region at the time of the First World War, including the ‘Druses’.
[Robartes:] ‘I joined that tribe, accepted its dress, customs, morality, politics, that I might win its trust and its knowledge. I have fought in its wars and risen to authority. Your young Colonel Lawrence never suspected the nationality of the old Arab fighting at his side.’ (AV B 41)
However, there are other pre-Islamic and even pre-Christian sects that could be possible sources of the Judwalis. As mentioned in the discussion of Kusta ben Luka, Suheil Bushrui considers that the ‘only tribe in Arabia (and in Mesopotamia in particular) of which [Yeats] could have been thinking, and which is identical with his description of the Judwalis, is the tribe or group of small tribes known as the Sabians’, which he describes as ‘a semi-Christian sect of Babylonia, the Elkesaites, closely resembling the Mandaeans or the so-called "Christians of St. John the Baptist", but not identical with them’ (In Excited Reverie, 297; cf. Encyclopaedia Britannica [1911]). Others dispute the Elkesaite link (G. R. S. Mead, in Fragments of a Faith Forgotten [1900], even doubts whether the Elkesaites ever existed) and ‘semi-Christian’ description (according to Wilfred Thesiger, in The Marsh Arabs [1964], this and the connection with St. John comes from a European misunderstanding of the frequent ritual baptism practised by the religion). The Koran specifically enjoins tolerance of the peoples of the book, which includes the Sabians (2:62, 22:17), and for this reason there are also ‘the pseudo-Sabians of Harran (Carrhae), who professed the religion of the true Sabians to escape the wrath of the Caliph al-Ma’mun (A.D. 830)’. The Harranians apparently worshipped planetary or stellar deities, with rituals that may have included human sacrifice, and were resurrected in nineteenth-century Syria by Benjamin Disraeli in his novel Tancred, or the New Crusade (1847). With respect to the subsequent confusion, Bushrui wryly comments that it ‘is not likely that these details bothered Yeats; the more mysterious and obscure the background, the better it served his purposes’ (In Excited Reverie 297-98).
Both Doughty and Blunt mention the Sabians, though neither group is nomadic: Doughty refers to the ‘Sabeans, or 'disciples of St. John,' beside the Persian Gulf’ (Arabia Deserta 2 231) and Blunt refers to ‘numerous small tribes and sections of tribes about Baghdad, but none of them deserve notice except the Sabaeans, now found only in the neighbourhood of Souk esh Shiokh, a village on the Shatt el Arab below Hillah, and numbering about 3,000 souls’. She mentions that they see themselves as ‘the descendants of Shem, and to the present day have preserved the ancient [pre-diluvian] tongue unchanged. In it their 'book' is written and is described as a sort of Syriac’ (Bedouin Tribes of the Euphrates 2 194). Though this sect predates Islam, Yeats places the ideas of Kusta ben Luka as deriving from earlier Syriac sources, so that this need not be a problem.
Another possible perspective on the Judwalis is the ascetic dervish tradition of Islamic Sufism. There are many forms of dervishism, including brotherhoods which lead mendicant, settled, and mixed lives, but each group (tariqat) employs specific techniques for focusing the mind on God through ritual meditation called Dhikr (also Zikr; remembrance of God). Amongst the better-known groups are the Mevlevi (Maulawi; whirling dervishes), Rufai (Rifa’i; howling dervishes), Bektasi (Bektashi; an order particularly open to women), Halveti (Khalwatiyya), Kadiri (Qadiri), and Nakshibandi (Naqshbandi; see below). There is usually some secrecy concerning their teachings and, amongst many other elements, Sufism uses poetry, music, spiritual astronomy and mystical geometry in the approach to the mystery of God. The dervish order best known in the West, the Mevlevi Order, was founded after the death of the poet and mystic Rumi (Jalal al-Din Muhammad Rumi, d. 1273), known as Mevlana (our guide), by his followers, and their ritual cosmological dance (sema) could well lie behind Yeats’s Dance of the Four Royal Persons (AV A 9-10). In a journal entry of 4 June 1909, Yeats had recorded a story told to him by R. W. Felkin:
Felkin told me that he had seen a Dervish dance a horoscope. He went round and round on the sand and then circle to centre. He whirled round at the planets making round whorls in the sand by doing so. He then danced the connecting lines between the planets and fell in trance. This is what I saw in dream or vision years ago.
This dance may well have lain behind the early idea of ‘their arabic names, the first "the holy women and the two Kalendars" because it describes the movements of two symbolic suns and moons; and the second "the dance of the Eunuch with the favourite wife"’ (YO 210). Kalandar, kalender or qalandar is something of a blanket term for mendicant dervishes.
In the end, as with almost all of the investigations in this area, the possible sources of the fictions throw up many fascinating perspectives, but without tying Yeats to any single source. The Judwalis have certain traits of the Bedouin, others of the dervishes, and others of various sects in a highly sectarian region of the world. What is most important is the moral effect of their doctrines, so that Robartes comments that they ‘are known among the Arabs for the violent contrasts of character amongst them, for their licentiousness and their sanctity. Fanatical in matters of doctrine they seem tolerant of human frailty beyond any believing people I have met’ (AV A xviii-xix; cf. YVP 4 16). Just as the Caliph, once he has agreed to learn the teachings of Kusta ben Luka, declares ‘I now understand human nature; I can never be surprised again’ (AV A 10), even to the extent of understanding how his trusted companion could have betrayed him with his favourite slave, so also the implication is that the Judwalis will understand and therefore tolerate all variations of human nature, because of their understanding of the Great Wheel. The sensualist of Phase 13 balances the Saint of Phase 27, and they will understand both equally as expressions of human potential.
The following thread is probably more a tracing of a remarkable coincidence than any direct source of Yeats’s ideas. Both Yeatses knew of G. I. Gurdjieff and, perhaps even more, P. D. Ouspensky who split from him (Ouspensky’s Tertium Organum was in the 1920s library in George’s name). However, they would not have been aware of a story pieced together by J. G. Bennett, one of the main expositors of Gurdjieff’s teachings after his death, in lectures given in 1963. Bennett sees one of the keystones of Gurdjieff’s teaching as lying in the management and transformation of energy, linked with the development of what is traditionally called the Resurrection Body, and sees roots in Rosicrucianism, Platonism and mystical Islam. The Enneagram is a symbol which was introduced by Gurdjieff in this context and has since become popularised even further as a form of personality typing and analysis of interactions. Bennett traces the origins of the Enneagram through Gurdjieff’s writings to the Naqshbandi Order of Dervishes, one of the more widespread Sufi groups. ‘Gurdjieff himself gives a clue . . . he describes how he found his way to a particular brotherhood which he said was in Upper Bokhara, where this particular knowledge was available’ and notes that it is in ‘the section of the chapter which deals with the training of the priestesses who performed the sacred dances, and the apparatus that they used for their training’ that the reader will find ‘an unmistakable reference to this symbol of the Enneagram’ (Gurdjieff: A Very Great Enigma, 48-49).
Bennett goes on to place certain Naqshbandi Dervishes in Bokhara and to derive the name from Naqsh, a seal, symbol or sign, and Naqshband, ‘those who make symbols, those who have the power to create a symbolism’ (54). Thus ‘the knowledge that Gurdjieff afterwards taught as his 'Ideas' came from putting together two halves of a single truth. One half is found in the Western—chiefly Platonic—tradition and the other half is in the Eastern—chiefly Naqshband—tradition’ (54), the two halves having an ancient common origin. In a similar way Robartes, wandering the Middle East within a slightly more limited compass than Gurdjieff, attempts to unite Western Giraldus with Eastern Kusta ben Luka and the Judwalis, or makers of diagrams, whose diagrams in the sand are descended from the Dance of the Four Royal Persons before the Caliph. As Bennett comments in another context, though, ‘This may mean anything or nothing, and I must warn you that for anyone who reads Gurdjieff and tries to reconstruct anything about Gurdjieff’s adventures, almost anything he writes may mean anything or nothing’ (37).
