Cædmon and Muḥammad: a Comparison

Bede tells Cædmon’s story as a postscript to Hild’s, saying that one of the abbess’ monks was ‘frater quidam diuina gratia specialiter insignis’ (‘a certain brother specially marked by God’s grace’). He says that this monk is known for the ability to compose poetry ‘ex diuinis litteris’ (‘based on the holy scriptures’).Footnote ⁷ Moreover, Cædmon does not compose his poems in Latin. Against the expectations of Northumbrian monastic learning, he learns of each desired subject ‘per interpretes’ (‘through interpreters’), through mediators of God’s word to man,Footnote ⁸ before turning out his delightful poems ‘in sua, id est Anglorum’ (‘in his, that is to say, the English, language’). Bede thus saves the most surprising aspect of Cædmon for the end of his opening sentence. Nor does he quote ‘Cædmon’s Hymn’ when the moment arrives, but offers instead a Latin paraphrase of this poem, whose original appears to survive with mostly minor variants usually in the margins of some twenty manuscripts of the Historia ecclesiastica. Although all Old English versions of the poem were copied after Bede’s death in 735, two of them, as we shall see, may have had his sanction, being added within the following decade in his own Northumbrian dialect to Latin manuscripts of this work in or near his scriptorium in Wearmouth-Jarrow.Footnote ⁹

In his second sentence Bede says that Cædmon’s poems inspired many people to renounce the world and to think of the heavenly life. Other Englishmen, he says, composed religious poems after Cædmon, but their poems could never compare with his. Comparing Cædmon with St Paul on the road to Damascus, Bede says that ‘ipse “non ab hominibus neque per hominem” institutus canendi artem didicit, sed diuinitus adiutus gratis canendi donum accepit’.Footnote ¹⁰ The incident to which Bede refers is St Paul’s revelation before his Damascene conversion, when, as Saul, he was persecuting the early Christians. After a blinding light envelops him and a voice asks ‘Quid me persequeris?’ (‘Why persecutest thou me?’, Acts IX.4), Saul accepts the role of apostle and becomes ‘Paul’. The quotation in Bede’s sentence comes from the Letter to the Galatians I.1, in which St Paul declares himself commissioned by Jesus Christ and God the Father.Footnote ¹¹ Again, in verses 11–16 of the same chapter, Paul tells the Galatians that the message they have heard him preach is the product of divine intervention: ‘neque enim ego ab homine accepi illud, neque didici, sed per revelationem Iesu Christi’. The story is most fully told in Acts of the Apostles IX.3–9, with a variant account in Acts XXII.9. In this way, the Pauline words ‘non ab hominibus neque per hominem’ prepare us for a revelation in Whitby.

Cædmon’s poems, which were consequently never frivolous but always elevated, says Bede, came from a man who had worked on Hild’s estates as a labourer ‘usque ad tempora prouectioris’ (‘to an age beyond man’s middle years’). He never learned singing and in festive gatherings would get up and go back home when he saw the lyre being handed towards him. Leaving on one such occasion, says Bede, Cædmon goes to the cowbyre to take charge of the cattle, and falls asleep:

Bede gives the sense of this poem through a Latin version in prose:

Having announced that this sense is unfaithful to the order of Cædmon’s words, Bede regrets the impossibility of translating ‘carmina, quamuis optime conposita’ (‘songs, even if excellently put together’) from one language to another without some loss of ornament and weight.

When Cædmon awakes, he remembers all he has sung in his dream and soon adds ‘plura in eundem modum’ (‘more in the same style’), all in praise of God. Taking this to his overseer, who brings him before the court of Abbess Hild, Cædmon ‘iussus est, multis doctioribus uiris praesentibus, indicare somnium et dicere carmen, ut uniuersorum iudicio quid uel unde esset quod referebat probaretur’.Footnote ¹⁴ The scholars must be there to align his work with the scriptures, although their agreement that Cædmon’s gift is divinely inspired also hints at an initial fear, in the words ‘quid uel unde esset’, of a demonic intervention. Either motive works against Daniel P. O’Donnell’s conclusion that both Bede and the scholars who vetted Cædmon were ‘more interested in Cædmon’s ability as a poet than in the origins of his gift’.Footnote ¹⁵ Then the ‘doctores’ test him further by reading him ‘quendam sacrae historiae siue doctrinae sermonem’ (‘a passage of sacred history or doctrine’) which he must put into verse. Cædmon accepts the task and returns the following morning with ‘optimo carmine quod iubebatur conpositum’ (‘the requested passage composed in excellent verse’). Abbess Hild orders him to take monastic vows, to be received into the community of monks, and ‘seriem sacrae historiae doceri’ (‘to be taught the whole course of sacred history’). Cædmon listens, learns and memorises; ‘quasi mundum animal ruminando’ (‘like a clean animal chewing the cud’), he turns what he hears into the most melodious poetry; hearing this in turn, his teachers become his audience. In due course, Cædmon turns all sacred history into verse: effectively Genesis, Exodus, Joshua, as well as singing ‘de aliis plurimis sacrae scripturae historiis’ (‘on many other stories from holy writ’); then Christ in the Gospels, Acts of the Apostles, and the Apocalypse; as well as lyrics ‘de beneficiis et iudiciis’ (‘about divine mercies and judgements’) with which to turn his audience from sin towards the love of good works.

Bede’s sources for this story plausibly start with a rumour or legend from Whitby, to which he has added the allusion to St Paul from Galatians I.1. One other source that has been found is for Bede’s metaphor of Cædmon’s ruminating like a ritually clean beast. Initially this was traced to Leviticus XI.3 and Deuteronomy XIV.6.Footnote ¹⁶ However, the line appears to come from a description of the blind scholar Didymus of Alexandria, written by Rufinus of Aquileia as an addition to his translation of Eusebius’ Ecclesiastical History (xi.7), completed before 395.Footnote ¹⁷ Rufinus, having been taught by Didymus between 371 and 375, says that his teacher perfectly memorised what he heard read to him and then, ‘tanquam mundum animal ruminans cibum quem ceperat ex integro revocabat’ (‘like a clean animal chewing the cud, would bring back all the nourishment he had received’).Footnote ¹⁸ This loan is made more plausible by the likelihood that Bede adopted Eusebius’ ecclesiastical mode of history from reading Rufinus’ translation.Footnote ¹⁹ The typology of Didymus also fits with St Paul, whose revelation on the road to Damascus transforms him while he is blind for three days (Acts IX.9).

Although no other source for Cædmon’s story has been accepted, there are several Anglo-Saxon analogues, of which the most prominent is the eleventh-century story of St Dunstan being taught to sing an antiphon by an angel in a dream.Footnote ²⁰ O’Donnell lists this and forty-four others from around the world.Footnote ²¹ One analogue to which he devotes particular attention is the story of Muḥammad’s first revelation outside Mecca in 610.Footnote ²² The surviving written evidence for this story is dated to more than a century after the Prophet’s lifetime (c. 570–632). Details of his life may be inferred from the Qur’ān, whose text did not reach canonical form until the reign of Caliph ‛Abd al-Malik (685–705).Footnote ²³ Otherwise the Prophet’s life is known from hadiths (aḥadīth), oral narratives alleging chains of informants. Upon these uncertain sources, biographies were written which survived in other men’s redactions from the mid-eighth century onwards.Footnote ²⁴ The leading extant biography is contained in The Life of the Prophet (as-sīrat an-nabawiyyah السيرة النبوية), a work from the beginning of the ninth century, written by ‛Abd al-Malik ibn Hishām (d. 833). This work incorporated an otherwise lost biography of the mid-eighth century by the scholar Muḥammad ibn Isḥāq (d. 767/8), The Life of God’s Messenger (sīrat rasūl allāh سيرة رسول الله), also known as The Book of the Campaigns (kitāb al-maghāzī كتاب المغازي.)Footnote ²⁵These and some later works on the Prophet organise his biography according to biblical typology and the Islamic theology of their day. Presently there is disagreement about how to read these later sources from the eighth to tenth centuries: between ‘sanguine’ scholars who accept them as true for the seventh century, ‘traditionalists’ who treat them as the basis for reconstructing Muḥammad’s history in the absence of contemporary Arabic sources, and ‘sceptics’ who prefer the contemporary evidence of mostly non-Arabic origin which might throw light on how Muḥammad’s new religion developed.Footnote ²⁶ With these caveats in mind, let us continue with the story of the Prophet’s first revelation, as told by Ibn Isḥāq, through the redaction of Ibn Hishām.

According to The Life of God’s Messenger, Muḥammad was a herdsman around Mecca, about forty years old and married to his cousin the merchant Khadījah, also his employer. One evening in the month of Ramaḍān (around December) he retired to meditate on the slopes of Mt Hirā’ just east of Mecca. In a cave near the summit, Muḥammad had a vision, which is reported in his alleged words:

Then the archangel declaimed lines which the illiterate Muḥammad ‘read’. Having awakened and finding these lines, his first divine message, ‘written on his heart’, Muḥammad recites them the next day to Khadījah, who tells his story and the angel’s words to her cousin Waraqah ibn Nawfal, a Christian. Waraqah approves them as divinely given and proclaims Muḥammad to be the prophet of his people. After Muḥammad’s retreat is ended, Waraqah proclaims this to him also as he walks around the Kaaba.Footnote ²⁸ The lines in question, quoted by Ibn Isḥāq, are preserved as the opening five verses of Sūrah 96 in the Qur’ān, where they go as follows:

This story can be called the ‘iqra’ narrative’, after the angel’s imperative in the first and third verses.Footnote ³⁰ The word iqra’ is from the verb qara’a, ‘to read, to recite’, which shares a root with the noun qur’ān, for ‘the act of reciting, or reading’, whence the name for the holy book of Islam.Footnote ³¹ The miracle of Muḥammad’s first revelation is that he reads, despite being illiterate. It is worth recalling that blind Didymus, gifted with the ability to memorise despite being functionally illiterate in the 370s, provides the basis, through Rufinus, for Bede’s image of Cædmon composing poems.Footnote ³²

This story of Muḥammad resembles Isaiah XXIX.11–12, in which Isaiah says that the Lord has stupefied His prophets, as if with wine:

It has been wondered if these verses, with the illiterate’s refusal to read, might not have separately inspired the narratives of Muḥammad and Cædmon.Footnote ³⁴ Isaiah’s emphasis on reading as the recognition of signs lends credibility to the idea of its influence on Muḥammad, but less so where Bede is concerned, given that his speaker emphasises singing rather than recognising signs.

Nor do Isaiah’s verses contain the narrative which is common to Muḥammad and Cædmon and which has first a refusal to comply, second a question about the subject, and third a recitation about the Creator. This common structure caught the eye of Francis P. Magoun Jr. more than seventy years ago, when he referred readers of Cædmon ‘to the extraordinarily parallel but totally unrelated story of Mohammed’s first revelation’.Footnote ³⁵ And yet Magoun, despite outlining this story in a footnote with more detail than the space allows, seems reluctant to let the matter be pursued. Two decades later a stronger case for influence appeared in which G. A. Lester compared Cædmon’s with eleven other stories of dream-vision inspiration, including the legends of Homer, Aeschylus, the Heliand poet, and St Dunstan.Footnote ³⁶ Lester’s comparative survey grades the Muḥammad resemblance with an overlap of 63 %, with fifteen narrative ‘essentials’ in common out of twenty-four: each story is (1) set within a religious environment, with a protagonist who (2) tends animals, (3) is previously untrained in poetry, and (4) is asleep when (5) he receives a visit from a certain person, who (6) instructs him to sing or recite; when (7) the protagonist replies that he cannot sing or recite, (8) the instruction is repeated; when (9) the protagonist requests guidance as to what to sing, (10) he is told to celebrate God’s Creation; having done so, (11) he recognises his gift as divine; (12) others recognise his gift as divine, whereupon (13) he begins to compose more pious and religious verse which is both (14) in his own language and (15) of the highest quality. In these last three respects, some continuity may also be observed whereby Cædmon’s production of new biblical poems resembles Muḥammad’s reception of a further 6,235 verses from the same source, Archangel Gabriel.

Lester also lists points of difference: (1) Cædmon is said to be advanced in years, whereas Muḥammad is around forty years old; (2) unlike Muḥammad, Cædmon is an outsider in his community, (3) is greeted by name, and (4) sings praise of God still in his sleep, (5) adding more verses on awakening; (6) unlike Muḥammad, Cædmon is tested the next day by scholars who wish to ascertain that his vision was not demonic; unlike Muḥammad, (7) Cædmon is then provided with new subject matter on which to compose, which (8) he produces after a day; and (9) Cædmon becomes a celebrated poet, whereas Muḥammad the next day is keen not to be called a poet.Footnote ³⁷ Some of these differences are less apparent, however: as for (1), neither man is young when he receives his vision; as for (2), Cædmon, insofar as he takes part in a feast, belongs to his community; as for (6), although Muḥammad’s vision is not tested, Khadījah’s visit to her Christian cousin Waraqah the next day parallels Hild’s recourse to the scholars of Whitby, as Sean W. Anthony has said.Footnote ³⁸ If these revisions were added to the list of existing parallels, there would be an overlap of 75 %. In short, although neither percentage is set in stone, the visions of Cædmon and Muḥammad appear to be remarkably similar. Contextually, it is worth adding an implication of Muḥammad’s homeland in the context of Bede’s opening quotation of Galatians I.1, for St Paul, whose Damascene conversion prepares us for his story of Cædmon’s dream, caps his own story of this experience in Galatians I.11–16 with the statement that ‘neque veni Ierosolymam ad antecessores meos Apostolos: sed abii in Arabiam, et iterum reversus sum Damascum’ (Galatians I.17).Footnote ³⁹

Since Lester, the idea of a loan is accepted by Katherine Scarfe Beckett and praised by Joaquín Martίnez Pizarro, who calls Muḥammad’s story the ‘closest fit by far’, but neither goes in for a discussion.Footnote ⁴⁰ O’Donnell, who does, argues against a connection by dwelling on the differences.Footnote ⁴¹ Outside Anglo-Saxon studies, Magoun’s note on Muḥammad has been treated more favourably by Arabists and other specialists aligned with them. Nine years after Lester, but without appearing to have read him, Klaus von See responded to Magoun’s misgivings with his own case for the influence of Muḥammad on Cædmon. Treating the latter’s story as a fiction with which Bede wished to create an ‘Initiations-Szene’ for all northern vernacular biblical verse, von See hypothesised an oral transmission of the iqra’ narrative to Northumbria through a meeting between Duke Eudo (or Odo) of Toulouse and his son-in-law, Governor ‛Uthman ibn Naissa (Munuza) of Septimania, in a time of peace after the Arabs’ conquest of Hispania in 711–20.Footnote ⁴² If true, this would mean that Bede heard a version of this narrative older than that which featured in Ibn Isḥāq’s biography of the mid-eighth century. Von See disclosed his idea to a traditionalist on early written sources for Islam, Rudolf Sellheim, who responded by taking Cædmon as evidence that this narrative was known in 720s Toulouse.Footnote ⁴³ A belief in this influence had in any case spread. Citing Lester fifteen years after his case for influence, but not von See, Zacharias Thundy wrote his own endorsement of this narrative as a source for the story of Cædmon’s vision. Thundy theorises that Bede, treating Muḥammad’s revelation as a Christian story, also ‘gathered as much information about Islam as he could’.Footnote ⁴⁴ Following von See and Sellheim, but without citing Lester or Thundy, the Arabist Gregor Schoeler used Cædmon again to claim that Ibn Isḥāq’s iqra’ narrative was decades older than his book.Footnote ⁴⁵

More recently, the Arabist Connell Monette, following Lester but without citing von See or Thundy, has carried out a more cautious comparison which acknowledges the high number of parallels, but which finds no explanation for the resemblance, despite the probability that these stories are ‘somehow connected in the medieval Abrahamic tradition’.Footnote ⁴⁶ However, not everyone is cautious. Without going into detail or citing any of these scholars, Abdul-Jabbar Jassim Mohammad treats Bede’s use of Islam as ‘undeniable’, concluding simply that since Bede never admits his debt to the Qur’ān, ‘it is possible to accuse him of plagiarism’.Footnote ⁴⁷ Sean W. Anthony has considered this idea more carefully. Describing the resemblance between the stories as ‘certainly remarkable, and even uncanny, not merely because of their chronological proximity’, he adduces every possible analogue from European sources, as well as the aforesaid Isaiah XXIX.10–12.Footnote ⁴⁸ Weighing up these, Anthony twice warms to an idea of an early English loan from Muḥammad, whereby Bede’s story would have been a ‘prestigious imitation’, an early version of King Offa’s imitation dinar.Footnote ⁴⁹ Finally, however, he returns for his answer to ‘a common well of narrative tropes, motifs, and archetypes that go far beyond biblical material’.Footnote ⁵⁰ Like O’Donnell, he objects to any but an indirect possibility of influence, because the resemblance is imperfect and because there is no evidence of transmission.

That no one has hypothesised a route for this is probably because of the distance between Mecca and Northumbria and the lack of evidence for (the knowledge of) Islamic texts in or near Anglo-Saxon England in the century of Arab expansion after the Prophet’s death in 632.Footnote ⁵¹ Nonetheless, the high number of parallels continues to invite further consideration of possible links in support of a loan from one story to the other, with the object of finding the most plausible route of transmission. Only three generations divide the two incidents on which these stories of Cædmon and Muḥammad are based, while the conversion histories of England and Arabia are surprisingly similar.Footnote ⁵² Had cowbyre and cave been somewhat closer in location, the high number of matching narrative motifs might have encouraged a more positive approach to the idea that Bede tells the story of Cædmon on the basis of an illiterate Arab herdsman reciting God’s words on Creation for the benefit of a society keeping paganism at bay. This approach will be followed here. Before we go further, however, there are two theological problems to consider that constitute the greatest objections. One is a textual implication that Cædmon’s Hymn represents the revealed word of God; the other is Bede’s condemnation of the Arabs as heretics.

‘Uersus, quos numquam audierat’

Since Bede initially refers to Paul’s divine revelation, later he might imply the revealed word of God for Cædmon, who ‘coepit cantare in laudem Dei conditoris uersus, quos numquam audierat’ (‘began to sing in praise of God the Founder verses which he had never heard’).Footnote ⁵³ The outcome of Cædmon’s revelation, however, is to create Northumbrian poems according to a divine template.Footnote ⁵⁴ It is likely, therefore, that Bede means Cædmon’s words to be ‘of God’, rather than ‘from God’. Despite the speaking angel in both stories, Gabriel on Mt Hīra’ and apparently in Whitby, God’s gift to the Northumbrian cowherd is confined to his new inspiration to compose vernacular poetry in Hild’s double monastery apparently before her death in 679/80. Whereas the Arab herdsman recites the revealed word of God, as he does here and in all other Sūrahs of the Qur’ān, any notion that Cædmon did on one occasion likewise would entail a readjustment in Bede’s salvation history sufficient to turn Cædmon’s later programme into a movement of its own. Bede claims nothing of this sort or scale for the Northumbrian cowherd, so his words must be read differently.

The content of Cædmon’s Hymn is learned and close to liturgy. Written in prose, its nine verse-lines were passed down over as many as four centuries in a manuscript transmission separate from the Latin text of Bede’s Historia ecclesiastica. The Hymn has thus survived in twenty-two manuscripts of which twenty have reached the present intact, both in Bede’s Latin and in later West Saxon translations.Footnote ⁵⁵ In the Latin context the poem was copied as a gloss in a margin or elsewhere in the manuscript; only in the Old English translations of the Historia was the poem part of the text, naturally without a version of Bede’s paraphrase and without his apology for translating the words.Footnote ⁵⁶ The poem’s oldest versions survive near the Latin text of Cædmon’s story in two manuscripts: in St Petersburg, National Library of Russia,Footnote ⁵⁷ lat. Q.v.I.18, 107r; and in 128v of Cambridge, University Library, MS. Kk.5.16, otherwise known as the ‘Moore Bede’. Both manuscripts are datable to 736 × 746 and appear to be close to Bede’s scriptorium in Jarrow, but they differ in their placing of Cædmon’s Hymn. The St Petersburg manuscript was copied by four scribes (ABCD) before 746 and appears to be close to Bede’s finished draft, having only six errors in the Latin text.Footnote ⁵⁸ Scribe D copied both the Latin of Bede’s chapter in the regular format, two columns of twenty-seven lines per page, and the Northumbrian English text in a smaller hand (insular minuscule) in three prose lines spanning the foot of the page. With his word division being ‘as scrupulous as in the Latin text’, it seems likely that Scribe D supplied this poem as the original for Bede’s Latin paraphrase.Footnote ⁵⁹ The Moore Bede, copied at around the same time, differs from the St Petersburg Historia in having one scribe who copied the Hymn in prose lines near the end as a hurried addendum.Footnote ⁶⁰ The St Petersburg Bede has the more careful text:

This addendum may be ordered into verse lines as follows, with a literal translation:

The opening statement that the works must praise their Creator differs from Bede’s translation (Nunc laudare debemus) in which Cædmon exhorts himself and his audience to do the same. For this reason, the poem is less likely to be a back-translation from the Latin, as some have supposed.Footnote ⁶³ Moreover, Bede’s apology, ‘Hic est sensus, non autem ordo ipse uerborum’ (‘This is the sense, but not itself the order of words’), probably shows his connection with the underlying original. This connection, in turn, might be read as evidence that he sanctioned the poems that were added so differently to two manuscripts of the Historia in the decade following his death.

As an early text, Cædmon’s Hymn has been compared with the pre-Christian Germanic creation myths reflected in poems such as the Old Icelandic Vǫluspá and the Old High German Wessobrunner Gebet, with which it has vocabulary in common.Footnote ⁶⁴ In light of these likenesses, the Hymn may be called traditional in its replication of formulas from preliterate Germanic verse, and yet novel in the way it uses them. André Crépin put this rather strongly when he attributed ‘the miracle in Caedmon’s story’ to his success in ‘adapting ancient traditional themes and formulas to new revolutionary ideas’.Footnote ⁶⁵ Eric Stanley supplied an illustration with ēci dryctin (‘everlasting Warlord’), an apparently new formula used twice in relatively quick succession.Footnote ⁶⁶ As ēce dryhten (‘eternal Lord’) in the corpus of later poetry, this formula became a cliché, or, as Stanley says, ‘so common in Old English verse that we have almost lost our understanding of its full significance’.Footnote ⁶⁷ Read literally, however, the formula reminds us, by combining eternity with the word for a military leader of low life-expectancy, that only the heavenly Lord lives forever. There is a certain freshness here with which the poem appears to vindicate Bede’s implicit claim that Cædmon was the first to compose sacred verse in English.Footnote ⁶⁸

Nonetheless, Cædmon’s Hymn is unlikely to be the first work that he composed, nor is it plausible that he composed his other English works in the order given by Bede, that of a pandect.Footnote ⁶⁹ The metre of the Nū scilun herga opening of the Hymn has been judged as consistent with single-stressed opening half-lines in later Old English poems.Footnote ⁷⁰ Bede’s translation also reflects an ornamental sophistication in the Northumbrian, which, as he indicates in his apology, cannot be rendered in his Latin. Moreover, as Kevin Kiernan observes, Bede’s translation leaves out three half-lines (6b, 8a, 9a).Footnote ⁷¹ Andy Orchard adds that Bede’s translation falls into hexameters just here, in a poetic move which is interpreted by O’Donnell as Bede’s way of compensating for the variation missing in his Latin.Footnote ⁷² At a deeper level in Cædmon’s inherited poetics is the formula, which has been defined by one scholar as ‘a group of words, one half-line in length, which shows evidence of being the direct product of a formulaic system’; and by another as ‘one of a set of verses (or formulaic system) of a similar metrical type in which one main verbal element is constant’.Footnote ⁷³ In the context of Cædmon’s dream, the ability of poets to generate entirely new verse lines with each use of this metrical system may explain Bede’s statement that he ‘coepit cantare in laudem Dei conditoris uersus, quos numquam audierat’.Footnote ⁷⁴ Treating these words as Bede’s allusion to spontaneous oral formulaic composition is one way in which to dispel the theologically disturbing idea that Cædmon recites the revealed word of God. Indeed, it appears that Bede is himself aware of this risk, insofar as his opening is a way of removing Cædmon from it. With Nunc laudare debemus the cowherd exhorts himself and his audience to give praise, in contrast to the Hymn’s older opening Nū scilun herga, which is more like an angel’s call for creatures to praise their Creator.Footnote ⁷⁵ The latter construction is found in the three oldest English texts: in the St Petersburg Bede, in the Moore Bede, and in the earliest extant West Saxon translation of the Historia from the first third of the tenth century, in Oxford, Bodleian MS. Tanner 10.Footnote ⁷⁶ It resembles the wording of Psalm CXIV:10, in which the Psalmist says ‘Confiteantur tibi, Domine, omnia opera tua’ (‘Let all Thy works acknowledge Thee, O Lord’).Footnote ⁷⁷ This likeness is unambiguous, contrary to what O’Donnell implies in his objection to the parallels for the construction.Footnote ⁷⁸ As Stanley shows, Cædmon’s line comes close to St Augustine’s gloss of the above psalm-verse as ‘Laudent te opera omnia tua’ (‘Let all Thy works praise Thee’).Footnote ⁷⁹ There is yet another parallel in the Benedicite Canticle for Lauds on Sundays.Footnote ⁸⁰ These are the most credible analogues for the Northumbrian composition in the St Petersburg and Moore Bedes, as well as for the West Saxon version in Bodleian MS. Tanner 10; and especially so, if the programme of Cædmon’s alleged later compositions, as Bede represents it, is ‘almost exactly in keeping with what was prescribed by Augustine for religious scholars and Latin poets’.Footnote ⁸¹ The learning in Cædmon’s poem shows that it is unlikely to be the first work of an unlettered poet.Footnote ⁸² If Bede’s aim in changing the opening construction was to hide the poet’s learning, it must be said that thereby he would no longer be illustrating the miraculous aspect of this tale. A better explanation for Bede’s Nunc laudare debemus, besides its avoidance of angelic command, is that it anticipates Cædmon’s audience the next day, the abbess and her scholars. By giving thus Cædmon, rather than the angel, control over what he recites, Bede distracts from the implication that his ‘uersus’ are God’s words.

Bede on the ‘Saracens’ as Heretics

Secondly, it is highly unlikely that Bede would have modelled Cædmon on Muḥammad, if his idea of the latter conformed with Islam as we know it from the Qur’ān. Bede’s understanding of Islam is hard to ascertain, although he ends by accusing the Arabs of heresy. Near the end of Historia ecclesiastica is a comment on the appearance of two comets (i.e. one comet with two tails) in 729 (for 728):

Bede has already used the word perfidia (‘treachery, heresy’) on the Arian and Pelagian Christians in the early chapters of Historia ecclesiastica (i.8 and i.17).Footnote ⁸⁴ There can be no doubt that he ended by condemning Islam as a sect in a similar way.

In the above passage, neither the name nor the date of the defeat which rewards the ‘Saracens’ for their heresy are specified, as if to a reader who knows current events. If, as seems unlikely, the above phrase ‘quo tempore’ covered the decade before 728, the passage might allude to the decisive Christian victory in which Duke Eudo of Aquitaine defeated and killed the wālī (‘governor’) of Arabic Spain (al-Andalus), al-Samḥ ibn Malik al-Khawlāni (719–21), in the battle of Toulouse in 721.Footnote ⁸⁵ However, contemporaries believed that the Christian victory over the Arabs in Tours or Poitiers in 732 was more significant. In this year ‛Abd al-Raḥmān ibn ‛Abd Allāh al-Ghafīqi, appointed governor of al-Andalus in 730, was killed in a skirmish with Charles Martel, the Austrasian palace mayor. As this setback for the Septimanian Arabs came only four years after the comet, Bede could have added this sentence to his Historia a year or two after he finished it in 731.Footnote ⁸⁶ Apparently with the later defeat in mind, the passage plays rhetorically on ‘lues’ (‘plague’) and ‘luebant’ (‘they received’) in a way that points to some rejoicing by Bede.

Scholars differ over whether Bede was hostile to the Arabs from the beginning of his career. Katherine Scarfe Beckett sees little change over time, whereas J. M. Wallace-Hadrill and Calvin B. Kendall conclude that Bede’s view of the Arabs, neutral at first, darkened as the impact of their armies on Christian communities became known to him.Footnote ⁸⁷ Above, in what seems to be his final reference to the Arabs, Bede refers to them with the negative non-biblical name that was common to him and most other western Christians. The ‘Saracen’ name had passed from Arabic شرقيين sharqiyyīn (‘easterners’) or سارقين sāriqīn (‘thieves’, ‘raiders’) into Greek Σαρακηνοί and Latin Sar(r)aceni as a pejorative term for the Arabs.Footnote ⁸⁸ Over time, Sar(r)acenus was etymologised according to the story of Ishmael in Genesis, in which Abram (later Abraham) begets a child on Hagar, handmaid of his childless wife Sarai (later Sarah). Sarai drives Hagar into the desert, but there an angel tells Hagar that her son will be named Ishmael, a wild man at odds with every man; and with all men, including his kinsmen, at odds with him (Gen. XVI.11–12). Hagar returns to give birth to Ishmael. When the boy is sixteen, Sarah gives birth to Isaac; on the day of his weaning, fearing for her son’s inheritance, Sarah has Abraham drive Hagar and Ishmael into the desert (Gen. XXI.10–11). Muḥammad’s genealogy is traced from Adam via Ishmael’s firstborn son Nabīt in The Life of God’s Messenger. Footnote ⁸⁹ The descendants of Ishmael’s twelve sons were the Ishmaelites, also known as ‘Hagarenes’. Later the Ishmaelites or Hagarenes were identified with the ‘Saracens’, then all three names with the Arabs.Footnote ⁹⁰

Bede, though he lists ‘Sarracenos’ as the fourth of four named tribes in Arabia, appears to avoid the negative etymology of Church Fathers before him.Footnote ⁹¹ Their explanation, that the Ishmaelites called themselves Saraceni so as to claim descent from Sarah, may have begun with St Jerome (347–420); Archbishop Isidore of Seville (c. 560–636) spread this notion further in his Etymologiae: ‘nunc corrupto nomine Saraceni, quasi a Sarra’ (‘now corruptly named “Saracens”, as if from “Sarah”’).Footnote ⁹² The ‘would-be Sarah’ etymology implies that the Arabs, disputing the story of Ishmael’s illegitimate birth, claim supremacy over Isaac and his offspring. This etymology was almost certainly known to Bede, who read Isidore’s Etymologiae, but he nowhere uses it.Footnote ⁹³

Bede’s first impression of the Muslim religion was probably derived from a description of Jerusalem and Damascus in the reign of Mu‛āwiya I (661–80), first caliph of the Umayyads. This was contained in De locis sanctis (‘On the Holy Places’), an alleged pilgrim’s diary written by Abbot Adomnán of Iona (679–704), probably in 683–86.Footnote ⁹⁴ In three books, Adomnán’s work purports to be based on the recollections of one Arculf, ‘episcopus, gente Gallus’ (‘a bishop of Gaulish race’), concerning his pilgrimage to the Holy Land, Alexandria, Constantinople and Rome. Contextually, Arculf’s visit is dated very approximately to 680.Footnote ⁹⁵ Adomnán brought this work with him to King Aldfrith of Northumbria on one of two visits in 686 and 688; on the second visit he also stayed with Abbot Ceolfrith in Jarrow.Footnote ⁹⁶ Young Bede, probably having met the Gaelic abbot there and reading his work, abridged it in his own De locis sanctis in 702 × 709.Footnote ⁹⁷ Bede gave his version the sheen of greater objectivity, removing reference to Arculf in all but two of Adomnán’s eighty-nine instances of this name.Footnote ⁹⁸ Many years later, perhaps in the late 720s, Bede quoted passages from his own adaptation, though calling it Adomnán’s work, in the two chapters that follow his account of the abbot in Historia ecclesiastica (v.15–17).Footnote ⁹⁹ According both to Adomnán in his preface and to Bede in the passage in his Historia, Arculf was blown off course to Britain on his way back to Gaul; after an indeterminate period in western Britain, he was taken in by the abbot in Ireland or Iona, to whom he told his story; most scholars assume that this was in the abbey on Iona. The reality of Arculf as a source will be discussed below.Footnote ¹⁰⁰

There are three places in the De locis sanctis of Adomnán and Bede where Muslims of the Holy Land are described. In one, Adomnán describes a mosque big enough for three thousand people in the lower city of Jerusalem, over the ruins of the Great Temple by the eastern city wall:

This description appears to have become confused in transmission. David Woods, taking Adomnán’s account to be based on an Irish copy of a document of an ‘Arnulf’ who was now no longer there, suggests that the word tabula was mistranslated from Greek στήλη (stēlē) as ‘column’, and that the structure was supported by stonework.Footnote ¹⁰² Robert G. Hoyland and Sarah Waidler, though they take Arculf for the abbot’s living informant, believe likewise that ‘marble’ was originally intended by tabula, which Adomnán, up on Iona, acclimatised as ‘wood’.Footnote ¹⁰³ As for the reference to Muslim construction, Lawrence Nees regards Adomnán’s phrase uili opere (‘of poor workmanship’) as a slighting reference to Islam in a lost written source.Footnote ¹⁰⁴ Bede, repeating the number of worshippers, follows Adomnán closely: ‘Nunc ibi Saraceni quadratam domum subrectis tabulis et magnis trabibus super quasdam ruinarum reliquias uili opere construentes oratione frequentant’.Footnote ¹⁰⁵ His re-use of uili opere has likewise been considered to refer to Islam as something ‘in some manner deviant’,Footnote ¹⁰⁶ although there is no other evidence of hostility to Muslims in either account. If this phrase was Arculf’s, it may indicate more simply that the structure, once standing in the compound of today’s al-Aqṣā Mosque in Jerusalem, was unfinished or under repairs when Arculf saw it.

The construction of this mosque has been dated to the caliphate of Mu‛āwiya I (661–80).Footnote ¹⁰⁷ Both Adomnán and Bede introduce his reflex, Mauuias rex, in the context of a feud between two Jewish communities in the city which he finally resolves through a prayer to Jesus and then through a miracle. According to Adomnán, the shroud which had covered Jesus’ head before his Resurrection was stolen from a church by a pious Jewish Christian who on his deathbed bequeathed it to one of two sons, asking them to choose between his wealth or the shroud (i.11). The brother who inherits his father’s wealth soon spends it all, while the brother with the shroud becomes as rich as his father. From him it passes down five generations into the hands of Jewish unbelievers. Having learned of its whereabouts after an undisclosed interval of time, the Christian Jews of Jerusalem bring a case to get it back, appealing to King Mavias, who judges through a prayer to Jesus. Taking the shroud ‘cum magna reuerentia’ (‘with great reverence’), the king orders a bonfire to be lit in a courtyard, goes up to this with the shroud, and addresses both parties as follows:

Then the king throws the shroud on the flames. The air currents waft it up and over to the Jewish Christians. One chapter later, Adomnán makes it clear that this cloth is kept in the church of St Mary built by the apostles on Mt Sion, along with another linen shroud reputedly woven by Mary herself (i.12).

Bede, retelling Adomnán’s version of the story, compresses it in line with biblical typology, with the elder brother choosing the wealth and the younger the shroud: this is like Ishmael when he loses out to Isaac, who is the type for Israel and its Christian successors.Footnote ¹⁰⁹ In the narrative present, the shroud is sought by Jewish Christians claiming to be heirs of Jesus, as opposed to Jews who retain their original faith. After long strife in court, Bede says:

Calvin B. Kendall, to whom this version of Adomnán’s story is ‘indicative of Bede’s caution’, translates salus in ‘pro suorum salute’ as ‘benefit’, as if Mavias considers the shroud to have no spiritual value.Footnote ¹¹¹ Scarfe Beckett likewise assumes some bias on the part of Bede, acknowledging that ‘benefit’ is keeping with the tangible wealth which the Jewish Christians hope to receive.Footnote ¹¹² However, the subjunctive in Bede’s ‘dignaretur’ presents the clause with ‘pro suorum salute’, which acknowledges Jesus’ gift of salvation to Christians, as an expression of King Mavias’ words. This feature aligns Bede’s meaning with Adomnán’s, that the Muslim king calls Jesus ‘mundi Saluator’ (‘Saviour of the world’). Bede seems merely to have condensed Mavias’ given prayer to Jesus in order to put it into indirect speech. With Mavias presented in this way, Tolan is right to say that ‘rien ne peut nous laisser soupçonner qui’il n’était pas chrétien’.Footnote ¹¹³

The third reference to Muslims in both versions of De locis sanctis concerns the seat of Mu‛āwiya’s caliphate in Damascus. Presumably following Arculf, who may have been there before the caliph’s death in 680, Adomnán says that this city is where

Adomnán’s account contains no denunciation of Islam, nor does Bede add one of his own. Indeed he removes Adomnán’s word ‘incredulorum’ (‘unbelievers’) and so represents Christians and Muslims as worshippers in parallel worlds, with each site of worship an ecclesia; moreover, Bede gives the caliph a greater agency than in Adomnán’s account: ‘Vbi dum christiani sancti baptistae Iohannis ecclesiam frequentant, Saracenorum rex cum sua sibi gente aliam instituit atque sacrauit’.Footnote ¹¹⁵ Mavias’ faith is otherwise unspecified by Adomnán or Bede, but in connection with the caliph on whom he is based, John Tolan shares Kendall’s belief that Bede, though ignorant of Islam, updated his biblical view of the ‘Saracens’ by showing in his version of De locis sanctis that some had once been converted to Christianity, that others were now spiritually close to it, and that Mu‛āwiya and his Syrian Arabs, in particular, ‘were Christian’ or ‘a people most favourably inclined towards Christianity’.Footnote ¹¹⁶

In another early work, Bede continues to represent the caliph and his religion without bias. In his Nomina locorum de actibus apostolorum (‘Names of places in the Acts of the Apostles’), written in 709 × 715, Bede says that Damascus ‘nunc Sarracenorum metropolis esse perhibetur, unde et rex eorum Mauuias famosam in ea sibi suaeque genti basilicam dicauit’, whereupon he cites the shrine of St John the Baptist as an ‘ecclesia’ (ix.2).Footnote ¹¹⁷ The difference between this and his earlier account is that Bede’s new term basilica denotes a church or public building of high status.Footnote ¹¹⁸ This Byzantine term appears to acknowledge the caliph as a king in the Umayyad mosque in Damascus.

In his Expositio in actus apostolorum (‘Commentary on Acts of the Apostles’), completed a little after 709,Footnote ¹¹⁹ Bede seems to know that Muslims have no idols and that they congregate for worship on Fridays. The first detail emerges in his commentary on Stephen’s defence to the priesthood in Jerusalem in Acts VII.42–43. Stephen, quoting from Amos V.25–27, which refers to the lapse below Mt Sinai, tells the priests in Acts VII.43 how the Israelites, forsaking the true God, carried the shrine of Moloch ‘Et sidus dei vestri Rempham, | Figuras quas fecistis adorare eas’.Footnote ¹²⁰ Glossing the star Rempham, Bede says ‘Significat autem Luciferum cuius cultui Sarracenorum gens ob honorem Ueneris erat mancipata’.Footnote ¹²¹ For this note Bede used two works of Jerome. One was Jerome’s commentary on Amos, which here says: ‘id est luciferi, quem sarraceni hucusque uenerantur’.Footnote ¹²² The other was Jerome’s Vita s. Hilarionis (‘Life of St Hilarion’, ch. 25), in which Lucifer is identified with the goddess Venus. Jerome says that Hilarion (a fourth-century anchorite) and his monks convert the Saracens of Elusa after arriving in this city ‘eo forte die, quo universaria solemnitas omnem oppidi populum in templum Veneris congregaverat’.Footnote ¹²³ Bede’s commentary on Acts VII.43 modernises this story by moving Jerome’s imputation of present-day ‘Saracen’ idolatry into the past.Footnote ¹²⁴ Secondly, he alters Jerome’s implication that today’s Arabs worship the devil in the form of Venus: whereas Jerome says that the ‘Saracens’ ‘colunt illam ob Luciferum’ (‘worship her on account of Lucifer’), Bede says that they worshipped ‘Luciferum […] ob honorem Ueneris’ (‘Lucifer in honour of Venus’). By this inversion, Bede remains focused on Rempham as a sidus and defines Lucifer as the morning star which at certain times is the planet Venus. In his De temporum ratione (‘On the reckoning of time’), which he wrote in around 725 to enlarge on his De temporibus (‘On times’) of 703, Bede says that the gentiles named the weekdays after gods to whom they consecrated planets (ch. 8).Footnote ¹²⁵ Since Venus’ day is Friday, Bede’s allusion to the ancient Arab cult of her planet in his commentary on Acts VII.43 may align this day with the Muslim day of congregation.

In his earlier works, in these ways, Bede alludes to the ‘Saracen’ religion with a neutrality bordering on tolerance. Nowhere does he use Isidore’s ‘would-be Sarah’ etymology for Saracenus, which he would almost certainly have seen and which imputes to the Arabs an attempt to purloin the rights of a first-born son. In his De locis sanctis, written in c. 702 × 709, Bede reiterates the leniency towards Caliph Mu‛āwiya that Adomnán appears to have passed on from Arculf, his source, as well as rendering their image of near-equivalence in the words with which he describes the Christian and Arab sites of worship in Damascus. He modernises Jerome’s account of ‘Saracen’ religion, showing that the Arabs are no longer idolaters nor devil-worshippers, as well as implying that they have inherited Friday as their day of congregation from the pagan time. He is faithful to the Pauline doctrine that the Ishmaelites will be converted like all other gentiles.Footnote ¹²⁶ Modern historians have attributed this hope to the Christian subjects of Muslim rulers in North Africa and the Near East in the later seventh century.Footnote ¹²⁷

From around 715 onwards, however, Bede’s neutrality towards the Arabs appears to change. In two works of around this time Bede refers to the ‘Saracens’ as universally hated enemies of the church. First in his commentary In Cantica Canticorum (‘On the Song of Songs’), finished about a year later,Footnote ¹²⁸ Bede refers in this way to the descendants of Kedar, second son of Ishmael. First, he makes David into a type for Jesus. For the verse in which a bride calls herself as dark and beautiful as the tents of ‘Cedar’, or as the curtains of Solomon (I.4), Bede refers to David in his concealment from Saul in the wilderness of Paran in I Samuel (XXV.1) as a type for Jesus during the forty days in the desert before his mission. Bede then glosses the Cedar name with the story of Ishmael, Kedar’s father, according to the angel’s prophecy about Ishmael in Genesis XVI.12, that he will be a wild man, with his hand against all, and the hands of all against him. Of Ishmael here, Bede says:

Bede’s next such reference to the ‘Saracens’ may be found in his commentary on I Samuel, In primam partem Samuhelis, which he finished by the summer of 716.Footnote ¹³⁰ In this work he names the desert of Ishmael’s exile ‘Faran’ (‘Paran’), quoting the verses from Psalm CXIX: 5–7, in which the ‘Psalmist’, David, refers to his refuge with the Ishmaelites (in Meshech). David’s words are said to express the complaint of Sarah’s son Isaac, in a way which makes him a type for Israel, and consequently for Christendom itself:

Bede ends with a use of Galatians IV.22–31, in which St Paul redefines Ishmael and Isaac as an allegory through their mothers: Hagar represents the covenant of today’s Jerusalem, a city of the enslaved, whereas Sarah represents the covenant of heavenly Jerusalem, for the children of God’s promise.Footnote ¹³²

Two more references confirm a change in Bede’s attitude towards the Arabs. In his Interpretatio nominum Hebraeorum (‘A reading of Hebrew names’, possibly of 722 × 725), Bede offers an appropriate etymology, apparently based on Hebrew גר ger (‘stranger’), for the Hagarene name: ‘Arabum Agareni aduenae [sc. gentes]’ (‘of the Arabic Hagarenes, hostile [people]’).Footnote ¹³³ Probably after 716, Bede’s quotation of the angel’s prophecy about Ishmael in Genesis XVI.12 has been read as an acknowledgement of the present-day endurance of Christians under Muslim rule.Footnote ¹³⁴ In this commentary on Genesis up to Isaac, In principium Genesis usque ad natiuitatem Isaac, Bede cites the angel’s prophecy about Ishmael, that he will be a wild man, with the hand of all against him, pitching his tent in exile from all his brothers (Genesis XVI.12). His commentary first reflects antiquity:

Then he says:

Here he appears to update his older commentary with current events. Kendall refrains from connecting this addition with the Arab invasion of Hispania, which ran from 711 to 720, but it seems unlikely that Bede’s animus is based on the Bible or on a memory of the Arab expansion a generation before his birth.Footnote ¹³⁶

The Arabs’ early Mediterranean conquests (Damascus in 635, Jerusalem in 638, Alexandria in 642, and Cyrenaica in 642–45) were a memory transmitted to Bede through his connections with Canterbury, ultimately from Archbishop Theodore (c. 602–90) and Abbot Hadrian (c. 630–709). These men had once been refugees, from Antioch and North Africa respectively, and yet they had escaped at a relatively early age. Theodore had lived in Constantinople and Rome for more than thirty years before arriving in Canterbury in 669, whereas Hadrian, arriving there a year later, had lived in Italy for more than twenty years previously, having left Cyrenaica around a decade after the Arab invasion there.Footnote ¹³⁷ The Canterbury biblical commentaries which derive from their lectures contain a note on Ishmael that may be personal: ‘Sic fuit genus eius Saracenis, numquam cum omnibus pacem habentes sed semper contra aliquos certantes’.Footnote ¹³⁸

For those not affected, on the other hand, these conquests lay in the distant past.Footnote ¹³⁹ Bede does not cite them in his historical references to the Arabs. In his account of universal history at the end of De temporum ratione, completed in 725, which covers the best part of six ages from the creation of the world, he (mis)dates the Arabs first for the year 4,639 (CE 687), when the ‘Sarraceni’ raid Sicily, returning with their spoils to Alexandria; then for 4,649 (CE 697), when Emperor Justinian II signs a treaty of peace with them for ten years, the ‘Saracens’ destroy Carthage, already having occupied the Roman province of Africa; and finally for 4,680 (CE 728), when they lay siege Constantinople for three years, ‘donec ciuibus multa instantia ad Deum clamantibus plurimi eorum fame frigore pestilentia perirent ac sic pertaesi obsidionis abscederent’.Footnote ¹⁴⁰ These Eastern Mediterranean events postdate the Arabs’ earlier conquests as well as Mu‛āwiya’s death in Damascus in 680, but they do not accord with Bede’s sweeping denunciation in his commentary on Genesis XVI.12.

Kendall rules out Hispania as the site of the new Ishmaelite occupation to which Bede refers in this commentary with ‘Europae nonnullam [partem]’, because Bede says nothing of its conquest by the Arabs either here in other commentaries, or in his universal history, or in the Historia ecclesiastica. Footnote ¹⁴¹ However, as the Arabs attacked but did not occupy Bulgaria, Sicily, or Sardinia in Bede’s lifetime, regions which might be called European, Iberia becomes the referent by default. The twofold invasion of Hispania by Berbers and then Arabs in 711 would explain why Bede’s commentaries from around 715 onwards represent the Arabs as hated enemies of the church. Although contemporary Arabic sources are lacking, 714 was the year in which ‛Abd al-‛Azīz ibn Mūsā was appointed governor of the former Catholic Visigothic kingdom now known as al-Andalus, with his capital in Seville.Footnote ¹⁴² These upheavals were either too well-known for inscription, or Bede is too much in denial to record what for him was a disaster for the church.Footnote ¹⁴³ The fact that Archbishop Isidore had superintended Seville, capital of Ibn Mūsā’s new province, makes it unlikely that interested parties failed to carry the news to the libraries of Wearmouth-Jarrow. As we have seen, a familiarity with the Arab conquest of Hispania seems presumed and invoked in Historia ecclesiastica (v.23) by the reference to a defeat of Septimanian Arabs probably at Tours in 732. It seems no coincidence that, from a watershed contemporary with the founding of the first Muslim province in Hispania in 715–16, Bede’s works begin to include asides on the Arabs as modern warmongers, persecutors of the church, and heretics.Footnote ¹⁴⁴

The upshot of this discussion is that in his works before the first political form of al-Andalus, Bede had little reason to be other than neutral in his attitude towards the ‘Saracens’, in their places of worship in Jerusalem and Damascus and especially in the account of ‘King Mavias’ that he adapted from Adomnán’s De locis sanctis. Far from imputing perfidia (‘heresy’) to this near-contemporary reflex of Mu‛āwiya, Bede preserves the aura of Christianity that was created by Adomnán in reminiscence of Arculf’s impression. And as Bede would have met Adomnán personally in Wearmouth-Jarrow in 688, it is plausible that he learned more about Mu‛āwiya and his religion, including Muḥammad, than he saw fit to add to this account.

Transmission of the Iqra’ Narrative to Bede

An unwritten oral route of transmission cannot be proved. Nonetheless, it seems unlikely that Adomnán deposited his De locis sanctis in Jarrow in 688 without giving an accompanying verbal account of who Arculf was and what he said about the religion of Jerusalem’s rulers. My argument is that Arculf’s Christian guides or hosts in Jerusalem told him the story about Muḥammad, which Arculf would have told to Adomnán. Later, again in person rather than in writing, and within a decade of Arculf’s visit to the Holy Land, Adomnán would have passed it on to Bede.

To begin the case for this transmission, it will be necessary to look more closely at Arculf. As no bishop by that name is known from other sources, some doubt has arisen about his reality in Iona as a pilgrim back from the Holy Land.Footnote ¹⁴⁵ Thomas O’Loughlin, finding Arculf’s word to be unreliable, suggests that he was named by Adomnán merely to add standing to an account based mostly on books in Iona’s library, from a total of seventeen written sources, including some ten by Jerome, together with a further five other works.Footnote ¹⁴⁶ O’Loughlin’s comprehensive survey of sources has led him to define De locis sanctis as an exegesis.Footnote ¹⁴⁷ David Woods suggests that Adomnán took Arculf’s name from an Irish copy of a manuscript containing antisemitic miracle stories which ‘Arnulf’ had acquired with relics in Constantinople and which he had then saved from a rough voyage in the mid-seventh century. In Woods’ view, this Gallic bishop was never in the Holy Land and the word ‘rex’ for ‘Mavias’ in Adomnán’s story is a relic of Mu‛āwiya’s time as the governor of Syria (639–60).Footnote ¹⁴⁸ Lawrence Nees, likewise reading Adomnán’s work as an exegesis, follows O’Loughlin’s theory to the extent of taking the abbot’s description of the new mosque in Jerusalem to be based on a written source, even when none has been found. Nees will not attribute this or any other passage to Arculf as a living witness on the grounds that Adomnán’s praise of Arculf’s reliability must show that he is a fiction.Footnote ¹⁴⁹

Other scholars attribute a reality to Arculf and his pilgrimage in De locis sanctis. Ora Limor has provided a literary-historical context for accounts of pilgrims’ travels, comparing this work of Adomnán’s with the Hodoeporicon (‘road diary’) of Bishop Willibald (c. 700–787), which was written in the 780s by Willibald’s young kinswoman, the nun Hugeburc, on the basis of his recollections of journeys to Palestine and Syria half a century earlier.Footnote ¹⁵⁰ Willibald visited Jerusalem four times in 723–7, before settling in Monte Cassino in 729 and then Eichstätt in 741.Footnote ¹⁵¹ In comparison, it appears that Adomnán was senior to Arculf and would have used him chiefly to corroborate details which he had gleaned from many written sources. As Limor says, ‘Arculf’s voyage and his experiences in the holy places were of less interest to Adomnán than the places themselves’.Footnote ¹⁵² Hoyland and Waidler, similarly finding a space between Adomnán as Arculf’s amanuensis and Arculf as Adomnán’s fiction, argue that ‘Adomnán blends what he can learn about these sites from his written sources with a more current description of the city and its sites obtained from someone who had recently visited it’.Footnote ¹⁵³ Illuminating the context of Muslim rule in Jerusalem in around 680, when Arculf would have been there, they point out that the subjects of all the new Arab territories may have seen their Muslim lords as temporary, whose rule, even in the Near East, ‘would not have been much in evidence’.Footnote ¹⁵⁴

No written source for the story of ‘King Mavias’ has been found.Footnote ¹⁵⁵ According to Adomnán in his De locis sanctis (i, ch. 9), Arculf claims that he heard this from the Christian citizens of Jerusalem:

Arculf’s authorities will have included the hosts who put him up, caretakers in the churches, perhaps a guide through the city. With ‘protestatur’, these fideles cives are said to bear witness to the story of the shroud, as if the miracle were contested. Moreover, this word may also justify the claim that Mavias calls Jesus the ‘mundi Saluator’.

Whether or not Mu‛āwiya said this of Jesus in public is less important than an assumption in the citizens’ report to Arculf that he was a Christian, or close to being one. As governor and caliph of Palestine and Syria for more than forty years, Mu‛āwiya was a well-liked ruler among a predominantly Christian population. Most of Mu‛āwiya’s subjects were Greek-speaking Arab Christians, some of whom had served the Byzantine emperor and now served him.Footnote ¹⁵⁷ There was also a remigration of Jews in Mu‛āwiya’s reign,Footnote ¹⁵⁸ whom Arculf’s informants seem to have resented. Adomnán’s story in De locis sanctis about the Jews who remain Jewish reflects an attitude that presents them, not the Muslims, as a menace to Christianity.Footnote ¹⁵⁹As governor, Mu‛āwiya had married a Christian princess, the daughter of the chief of Kalb, and their son Yazīd married a Christian also, the daughter of the last Ghassanid Christian king.Footnote ¹⁶⁰ When Mu‛āwiya became caliph in 661, he prayed to Jesus at Golgotha and to Mary at her tomb in Gethsemane.Footnote ¹⁶¹ This was probably to please the Christian crowds who witnessed the homage of his Arab lords in this city, although to early Muslims Jerusalem was as holy as Mecca; prayers to Jesus and Mary are also Islamic devotions.Footnote ¹⁶² Some material evidence for Mu‛āwiya’s tolerance has been found with a coin dated to 663 and found at Hamat Gader near the Sea of Galilee, in which a Greek inscription to Mu‛āwiya āmir al-mū’minīn (‘Commander of the Believers’) is preceded by a cross.Footnote ¹⁶³ Adomnán’s account of an ‘eclesia incredulorum’ in Mavias’ Damascus is supported by an Arabic record that Mu‛āwiya asked the Damascene Christians for permission to use the eastern half of St John’s church (presumably the courtyard) as a site for his new mosque; later, when he asked to extend this into the church, and they refused, he refrained from further action.Footnote ¹⁶⁴ In 679, Mu‛āwiya is also said to have restored the church of Hagia Sophia in Edessa, damaged by an earthquake.Footnote ¹⁶⁵ Although he was not himself a Christian, he allowed Christians and Jews to worship as they pleased.Footnote ¹⁶⁶

Related to the caliph’s laissez faire is the question of his interest in Muḥammad. This is relevant to the transmission of the iqra’ narrative, if it was the Christians of Jerusalem who told Arculf the story about Mu‛āwiya with an unrepressed positive spin. Contemporary Muslims were divided about the strength of Mu‛āwiya’s belief. Because neither he nor his son Yazīd I (680–3) was related by blood to Muḥammad, his right to rule was questioned by Muslims outside Syria. From the Prophet’s clan had come all Mu‛āwiya’s predecessors, Caliphs Abu Bakr (632–4), ‛Umar (634–44) and ‛Uthman (644–56), whereas Mu‛āwiya was related to Muḥammad by marriage. His sister, Umm Ḥabība, had been married to Muḥammad to seal the peace in 630, after six years of intermittent fighting in which the Prophet succeeded in taking Mecca from their idolatrous father, Abu Sufyān ibn Ḥarb of the Quraysh. After Mu‛āwiya’s clan converted, the Prophet made him a scribe of his revelations (one of eighteen scribes).Footnote ¹⁶⁷ When Muḥammad died in 632, his followers rode northwards and Mu‛āwiya rose through the ranks to become a commander in the Levantine campaign of 633–4 that resulted in Theodore’s flight from around Antioch.Footnote ¹⁶⁸

In around 639, Caliph ‛Umar made Mu‛āwiya the governor of Syria. When Caliph ‛Uthman was assassinated by rebels in 656, Mu‛āwiya sparked off the first fitna or Arab civil war by contesting the succession with the next in line, ‛Alī ibn abi Ṭalib, who was the Prophet’s cousin and son-in-law. Five years later, Caliph ‛Alī was outmanoeuvred by Mu‛āwiya and died at the hands of former allies, whereupon Mu‛āwiya bribed Ḥasan, son of ‛Alī, to abdicate, so becoming caliph himself. Mu‛āwiya then moved his capital away from Medina, which had been favoured since the time of the Prophet.Footnote ¹⁶⁹ A contemporary Maronite (eastern Catholic) chronicle reports that ‘he placed his throne in Damascus and refused to go to Muḥammad’s throne’.Footnote ¹⁷⁰ In 676 Mu‛āwiya broke his agreement with Ḥasan by nominating his own son Yazīd to succeed him. Upon the death of Mu‛āwiya four years later, Ḥasan’s younger brother, Ḥusayn ibn ‛Alī, came north from Mecca to challenge Yazīd for the succession, but was ambushed and killed at the battle of Karbalā’ on 10 October 680. By Ḥusayn’s party, who became the Shi’as, and by the pietist Kharijites, as well as by the Abbasids nearly a century later, Mu‛āwiya was accused of turning Muḥammad’s covenant into an hereditary monarchy.Footnote ¹⁷¹ His son Yazīd broke with Muḥammad further by dating his coins from Year One of his reign, so abandoning the hijra calendar which starts with the Prophet’s migration to Medina in 622.Footnote ¹⁷²

Mu‛āwiya defined his religion in a public letter when he was governor of Syria. Besieging Constantinople in 651–2, he wrote to Emperor Constans II (641–68) with a challenge ‘to deny that Jesus and turn to the Great God whom I worship, the God of our father Abraham’.Footnote ¹⁷³ This pious injunction contains no reference to Muḥammad. When he became caliph, Mu‛āwiya called himself ‘Deputy of God’ (khalīfat allāh خليفة الله), not ‘Deputy of the Messenger of God’ (khalīfat rasūl allāh خليفة رسول الله), nor did he refer to Muḥammad or Islam in his public proclamations.Footnote ¹⁷⁴ Some inscriptions confirm this independence from the Prophet. Not only does the coin with cross from Hamat Gader in 663 fail to mention Muḥammad alongside Mu‛āwiya ‘Commander of the Believers’, but he is absent in an Arabic inscription to the caliph by a dam near Ta’if, not far from Mecca, dated to AH 58 (CE 677/8). This is probably the oldest Islamic inscription surviving. It attributes the dam to ‛Abdullāh Mu‛āwiya, ‘Commander of the Believers’ (āmīr al-mū’minīn أمير المؤمنين), asking God to pardon him and strengthen him with victory.Footnote ¹⁷⁵ Assigning this title to Mu‛āwiya three times, the inscription makes no reference to the Prophet at all, despite the site’s proximity to Mecca. In these ways, it appears that Muḥammad was no longer at the forefront of the caliph’s religion when Arculf visited Jerusalem and Damascus in around 680, nor in any of the other years suggested.

The period, therefore, in which Arculf visited Jerusalem is one in which the story of Muḥammad on Mt Hirā’ could circulate between Muslims and Christians without sensitivity or prohibition. Hoyland believes ‘that Muslims did not initially see their faith as totally distinct from the other monotheistic confessions’.Footnote ¹⁷⁶ Islam was not yet an institutionalised faith. Reading non-Muslim and Muslim sources, historians have interpreted a development in Muḥammad’s religion from the seventh-century period of ‘the believers’ (al-mū’minīn) or ‘the submitters’ (al-muslimūn), warlords and pietists whose tenets variously overlapped with those of Jews and Christians, to a more uniform eighth-century confession in which all Muslims began to submit to God through the Qur’ān of His Messenger. Although there has been some argument about the role of faith as opposed to booty and prestige in the Arabs’ motives for their rapid territorial gains after Muḥammad’s death in 632, a broad consensus has emerged about a period of ‘proto-Islam’ from the 630s into the last two decades of the seventh century.Footnote ¹⁷⁷ Not until after the second Arab civil war (680–92), late in the caliphate of ‛Abd al-Malik (685–705), were the laws of Islam codified, the Qur’ān standardised, non-Muslims taxed as subordinate citizens and a double shahādat (‘creed’) instituted as an oath to one God and to Muḥammad as His Messenger.Footnote ¹⁷⁸ This is not the version of Islam that Arculf would have encountered or that Bede or anyone else in England could have heard of in the late 680s.

In the Near East, in Egypt, Syria, and Iraq, there is evidence of widespread communication between the religions at the time Arculf visited Jerusalem and Damascus. Christian monks under Arab rule in these regions described Muḥammad relatively neutrally: Sebeos of Armenia, in the 660s, calls him a trader and reviver of monotheism; he was a lawgiver, to John bar Penkaye of Mesopotamia in 686/7, and to George the archdeacon of St Makarios, in Northern Egypt, in the 690s; to Jacob of Edessa, in the 690s, Muḥammad was a king; a century later he was acknowledged as a prophet to the Arabs by the Christian author of the Syriac Zuqnin Chronicle, written in the 780s.Footnote ¹⁷⁹ Jacob of Edessa (c. 640–708), a Syriac scholar contemporary with Aldhelm (c. 639–709), reveals also some knowledge of Islam and the Qur’anic Jesus in his correspondence.Footnote ¹⁸⁰ Likewise, Abbot Anastasius of St Catherine’s on Mt Sinai in his Hodegos (or Viae Dux ‘Lord of the Way’), a treatise against Monophysite Christians finished in 681, appears to show knowledge of the Qur’ān before it reached its first extant written form under Caliph ‛Abd al-Malik.Footnote ¹⁸¹ Probably through Byzantine condemnation, the Prophet goes unmentioned and unnamed in western Christian records before the eleventh century. This has led Kendall to state that ‘the very existence of Muḥammad was unknown to Western Europeans in the eighth century’.Footnote ¹⁸² But absence of record does not mean ignorance of subject. Many Christians besides Arculf and Willibald returned to Western Europe from the Muslim-governed Holy Land in the eighth to tenth centuries.Footnote ¹⁸³ If Arculf met a similarly informed clergy in Damascus or Jerusalem in around 680, they could have told him about the first revelation of Muḥammad.

Gifts of Different Graces: Arabic and English

So far we have seen how the Prophet’s story may have been used as a model for Bede’s story of the dream-vision visit of an angel who commands Cædmon to sing on Creation. The first objection to this idea, that the poem might be taken for the revealed word of God, has been addressed by reading Bede’s statement on its novelty as his reflection on oral-formulaic verse. The second objection, that Bede saw Islam as a heresy, has been scaled back in keeping with the late seventh century as a period in which Islam did not yet subordinate the older Abrahamic religions and in which Bede, at least until he could have heard about the creation of al-Andalus in 714–15, would not have had the modern examples on which to base complaints about ‘Saracens’ persecuting Christians.

That Bede paraphrases Cædmon’s Hymn in Latin, rather than quoting its Northumbrian original, is probably to be explained as his care for foreign readers who would not have understood the local vernacular. Another suggested reason is that Bede was reluctant to reveal Cædmon’s dependence on traditional diction, whether ludic or heroic.Footnote ¹⁸⁴ The existence of profane English poetry in Cædmon’s day is clear enough in Bede’s statement that he ‘nil unquam friuoli et superuacui poematis facere potuit, sed ea tantummodo, quae ad religionem pertinent, religiosam eius linguam decebant’.Footnote ¹⁸⁵ Some caution about English secular verse appears likely also in the context of Bede’s substitution of pagan with biblical Latin exempla in his rhetorical works.Footnote ¹⁸⁶ This reason for not quoting the Hymn in Northumbrian would align with the larger aim of making the Historia suitable for readers outside England.

At any rate, there is no reason to suppose that Bede was shy of the English language, even if it remains doubtful he would have accepted this as a lingua sacra equal to Hebrew. The tradition of exalting the latter with Greek and Latin was founded on John XIX.20, in which Pilate orders the title ‘King of the Jews’ to be inscribed on the cross in all three languages. While Hebrew held its place among scholars as the world’s first language beyond Noah and the fall of Babel, many books of the Old Testament were translated into Greek koinē in the third century BCE, as were the Gospels in the first century CE. By the time of St Augustine (354–430), Latin, the language of the Western Empire, was also in liturgical use.Footnote ¹⁸⁷ After this, other gentile languages became acceptable. Three centuries later, in Bede’s Retractatio, a technical revision of his Expositio actuum apostolicorum which M. L. W. Laistner dated very roughly to 725, Bede’s advice to take plural Latin tenebras (‘darkness’) as a singular, on the evidence of the Greek word σκότος, leads to an interesting aside:

On this evidence, it appears that some Northumbrian clergy were dictating Old English interlinear glosses to the Bible in the 720s. Bede’s own translations are recorded by his pupil Cuthbert, who writes that before his death on the eve of Ascension Day, on 26 May 735, Bede was working on two tasks:

The period in which Cuthbert says Bede was working on these translations, between Easter and Pentecost, corresponds with the seasonal liturgical use of Bede’s first task, to translate the given section of St John’s Gospel (I.1–6.9), probably through English interlinear glosses.Footnote ¹⁹⁰ While the accompanying book by Isidore serves no liturgical purpose, the place of exceptiones quasdam as the object of conuertit includes it in the same likely interlinear project. The implication for Bede’s final year is that Northumbrian English was beginning to come of age as a literary language.

In all this, however, Bede’s view was pragmatic. Two months before his death, he wrote to Bishop Egbert of York, a former pupil, urging him to appoint more priests to teach liturgy. Some of these, he said, were going to be ‘idiotas’ (‘illiterates’) who knew the meaning of liturgy only through report in their own language. Of such men he writes that ‘haec ipsa sua lingua discere, ac sedulo decantare facito’.Footnote ¹⁹¹ Bede adds that he has already lent his own translation of the Creed and the Lord’s Prayer ‘sacerdotibus idiotis’ (‘to illiterate priests’).Footnote ¹⁹² In this way, it seems that he regarded English as an aid to the learning of Catholic doctrine in a time of declining Latinity.Footnote ¹⁹³

These measures have some bearing on what Bede thought of sacred poetry in the vernacular. Cuthbert says earlier in the same letter that Bede, on his deathbed, recited a five-line Northumbrian poem, the so-called ‘Bede’s Death-Song’, about making a good end (and possibly Bede’s own work). As Cuthbert says that this was ‘In nostra quoque lingua, ut erat doctus in nostris carminibus’,Footnote ¹⁹⁴ it is clear that Bede knew Cædmon’s other works, as well as the imitations of these to which he refers. His comment on the difficulty of translating English verse ‘sine detrimento sui decoris ac dignitatis’ (‘without impairing its ornament and weight’) thus appears to be founded on first-hand experience.Footnote ¹⁹⁵ Knowing the verse of his native tradition, Bede could have found a work of Cædmon’s with which to match the lines on Creation in an eastern narrative that he was using as his model. So would Bede have chosen an Old English verse opening to Genesis which he paraphrased in Latin, altering the initial construction in order to to give the exhortation to Cædmon. Then two of his fellow monks in Wearmouth-Jarrow, knowing the same or a similar poem, would have added it to the St Petersburg and Moore manuscripts of the Historia ecclesiastica. Later scribes, leaving the other works of ‘Cædmon’ uncopied, would have transcribed their versions of this extract in the same or similar position, encouraging their pupils to do the same. This is probably how ‘Cædmon’s Hymn’ got a life of its own.

Bede’s acceptance of English as a medium for the sacred was shaped by Pope Gregory I, for whom the tongues of fire in Pentecost, bestowing a knowledge of all languages, reversed the diffusion of Babel by giving the church a voice to each one of its diverse congregations; he was also influenced by Isidore’s conclusion that God speaks to all men in their languages, as well as by Isidore’s example, which is St Paul as God’s mediator in I Corinthians XIII.1.Footnote ¹⁹⁶ Bede responds to this missionary pragmatism in his commentary on the Pentecost in Acts II.4 in Expositio actuum apostolorum (a little after 709). The post-Babel diffusion of languages was no punishment, says Bede, but rather the means for the Holy Spirit to alight on all mankind:

This doctrine of ‘dona uariarum gratiarum’ agrees with Bede’s words on Cædmon, the man ‘diuina gratia specialiter insignis’ who ‘diuinitus adiutus gratis canendi donum accepit’, all through the English language.Footnote ¹⁹⁸ If Bede based Cædmon’s story on Muḥammad’s, the premise would have been the same, but with Arabic, as the story of Pentecost allows: further into Acts, when the Holy Spirit descends on the initial twelve apostles, it is said that each man in a crowd of Jews from all nations, including ‘Arabes’, heard the servants of God speak in his own language.Footnote ¹⁹⁹ Bede was thus primed to accept Muḥammad’s language as a medium for the word of God.