The Greek death of Sībawayhi and the origins of Arabic grammar

Abstract Sībawayhi, the founder of the Arabic grammatical tradition, was said to have died in Persia of sorrow after losing to Kufan rivals in a competition in Baghdad. The first part of this article demonstrates the artifice of Sībawayhi's biography, his death tradition in particular, and the stakes involved in its elaboration in early Islamic culture. The second part argues that the tradition of his death was based on the model of Homer's death, which can be shown to have circulated and been creatively adapted in contemporary Syriac historiography. The third part considers the consequences of Sībawayhi's Greek death for the old question of the influence of Greek on early Arabic grammar.

whose practitioners wielded considerable authority, were high. 4 According to modern andmore importantlyancient narratives, the raison d'être of the science of grammar was religious. 5 In the telling of Ibn al-Nadīm (d. 385/995), Abū l-Aswad al-Duʾalī (d. 69/688)whose knowledge of grammar was due to none other than ʿAlī b. Abī Ṭālibwas asked by Ziyād b. Abīhi (d. 53/673), who was governor of Basra and the eastern provinces and well-known for his oratory, to compose a guide to aid believers in accurately reciting the Quran. Abū l-Aswad demurred, but when he overheard someone mangling Q 9:3, disastrously declaring that God was "free of his messenger", he realized the threat posed to Arabic and, by extension, Islamby the people they had conquered. 6 The belief in their responsibility for the corruption of the language of the Quran echoes throughout the Islamic tradition, as well as in modern scholarship. Sībawayhi would thus appear an unlikely founder for the Arabic grammatical tradition. In another version of the foundation narrative reported by Ibn al-Nadīm, it was a Persian's malapropism, proclaiming a limping (ẓāliʿ) horse "strong" (ḍāliʿ), which prompted Abū l-Aswad to exclaim: "these mawālī . . . entered Islam and became our brothers; if only we had taught them how to speak". 8 Such a sentiment would have been out of place, however, in the time of Sībawayhi and the reception of his work in Abbasid Baghdad, which was marked by Arab-Persian tension, reflecting Arab anxieties as well as increasing attention to Persian identity. 9 One of the targets of the strident anti-Arab polemic, associated with a Persian-led Shuʿūbiyya movement, 10 was Arabic oratory, and the Arabic language more generally. 11 It was therefore easily perceived as a threat to Islam, for instance by al-Jāḥiẓ (d. 255/868). 12  al-Qurʾān) in linguistic terms, a thesis that gained in strength during the third/ ninth century, partially in response to Shuʿūbī claims. 13 The status of Arabic on the one hand, and of Persians on the other, were then both matters of some sensitivity in Islamic society in this period, and they merged in the figure of Sībawayhi. We find Sībawayhi's work being used alongside that of Homer, Euclid and Ptolemy, as well as icons of Persian identity, in contemporary polemic against the inimitability of the Quran. 14 One participant in this polemic was Ibn al-Rāwandī of Khurasan (d. 298/910?), who was considered a heretic and either died in hiding in Baghdad or returned to Persia. 15 The pressure on Persians (or those of Persian descent) who by contrast sought to fit in is apparent in the virulent anti-Persian polemic for which they were responsible, as in Ibn Qutayba's (213-70/828-89) The Excellence of the Arabs; among grammarians, Ibn Fāris (d. 395/1004) went to greater lengths than his Arab colleagues in his insistence upon the superiority of Arabic. 16 Sībawayhi was hardly unusual in being an Arabic grammarian of non-Arab descent, 17 but the large number of mawālī would have placed pressure on grammarians to establish the standing of their nascent discipline. Furthermore, as the dichotomy between "Islamic" and "foreign sciences" hardened, it would have been in their interest to align their tradition with Islam and dissociate it from foreign influence. 18 It has indeed been suggested that the substantial effort involved It is particularly interesting that Sībawayhi's grammar is being used against the language of the Quran, when his discipline was ostensibly entrusted with its preservation (as well as entirely sensible that the rules he deduced undermine the claim for its inimitability). 15 J. van Ess, "Ebn Rāvandī", in E. Yarshater (ed.), Encyclopaedia Iranica (London: Paul Kegan and Routledge, 1982-present) (online). Ibn Rāwandī was a student of al-Mubarrad, mentioned below. 16 See Versteegh, "What's it like", 213 for Ibn Fāris, who belonged to the Kufan school.
Mottahedeh, "The Shuʿûbîyah controversy", 179: "it is remarkable how many of the anti-shuʿûbîs like Ibn Qutaibah and az-Zamakhshari were non-Arabs". 17 Bernards, "Contribution of mawālī". See Fück, ʿArabīya, 27, on the attraction of mawālī to grammar as a means of assimilation, also reflected in Sībawayhi's motivation to learn grammar. in establishing Abū l-Aswad as the founder of Arabic grammar was motivated by the need for an Arab founder, perhaps in response to Shuʿūbī criticism. 19 Abū l-Aswad was especially attractive as a founder for the "school" of Basra, because Basran grammarians were traditionally depicted as radical and as departing from the religious tradition that the Kufans were reputed to uphold. 20 Conversely, the Kufans claimed that it was Abū l-Aswad who had made the crucial grammatical mistake that launched Arabic grammar. 21 The figure of Sībawayhi, author of the "Quran of grammar", 22 could not but play a central role in the struggles over disciplinary history. His book was far from being an immediate success, even among Basrans, but by the fourth/tenth century a chain of transmission had been constructed to lead from Sībawayhi all the way back to Abū l-Aswad. 23 The Kufans for their part claimed that Sībawayhi had a speech defect and didn't know how to speak Arabic. 24 The derision directed at him focused on his Persian descent, 25 which notably was also a source of controversy in the case of both of his Basran teachers, al-Khalīl (d. 160-175/776-791) and Yūnus b. Ḥabīb (d. 182/798). 26 Sībawayhi's nemesis al-Kisā'ī was of Persian ancestry as well, and like Sībawayhi is said to have sought out grammatical instruction in Basra following some difficulties in Arabic; but he rejected the Basran doctrine, and his reading of the Quran came to be recognized as one of the seven canonical readings. 27 Whereas the Persian identity of al-Kisā'ī and others was elided, Sībawayhi's was emphasized, as the predominance of his distinctly foreign namein place of an Arabic patronymicdemonstrates. 28 Sībawayhi's biography, as we have seen, similarly stresses his foreignness to Iraq and to Islamic scholarship and his rejection by its foremost authorities. 29 Within the contexts of his work in the second/eighth century and its reception in the third/fourth and ninth/tenth centuries, these biographical traditions clearly did not form randomly and were not shaped by purely aesthetic considerations. They were rather a resource which in these charged circumstances was manipulated to inflect the history of Arabic grammar in the service of competing interests. In an attempt to partially recover the process of these traditions' formation, we now turn our attention to the tradition of his death.

The death of Sībawayhi
Sībawayhi's death, variously dated somewhere between 161/778 and 194/810, was already a matter of controversy in the third/fourth and ninth/tenth centuries. The earliest extant account of Sībawayhi's life and death is by the aforementioned Ibn Qutayba in his al-Maʿārif: He is ʿAmr b. ʿUthmān. He was primarily a grammarian. He came to Baghdād, and a meeting was arranged between him and the grammarians.   37 This tradition dominates later biographies, which frequently feature the phrase ghamman bi-l-dharab. 38 The death of a young man from heartbreak appears suspect, to say the least; and if such an account was current already at the beginning of the third/ninth 31 Carter, Sībawayhi, 8, reads fa-rajaʿa wa-maḍā as signifying one action; I think it is better read as two separate journeys in the light of Ibn al-Nadīm, al-Fihrist, 1:143, and the tradition from al-Akhfash discussed below. 32 Cf. Humbert, Les voies, 3. century, it was apparently already regarded with suspicion by Ibn Qutayba and the fourth/tenth-century sources who did not include it in their accounts of Sībawayhi's life. Where, then, did the story come from and, more importantly, how did it gain widespread acceptance? While some of his biographers were content to leave his premature death unexplained, others will have had a stake in his possessing a more robust biography. If Sībawayhi was to serve as a founder for the "school" of Basra, at least, it was a problem that there was so little to say about him, and especially about his death. 39 The most economical way to construct a death tradition out of the spare available biographical material was offered by emplotment, that is, the causal connection of two eventshis humiliation and his deathwhich in Ibn Qutayba and some of the later sources were kept separate. 40 This move evidently struck the fancy of the audiences of Sībawayhi's biography. A tragic air pervades later versions of his biography, which frequently describe him as handsome and narrate in pathetic detail the tears of his brother, in whose bosom he perished. Neither the economy of the emplotment nor its aesthetic qualities appear entirely sufficient, however, to explain the success of this incredible death in driving competing traditions (e.g. al-Marzubānī's) out of circulation and establishing its own authority. Its force, I submit, resided in its resonance.
Dying as a penalty for failing to solve a riddle is a well-attested motif of folk literature, but we can be more specific here, for dying of sorrow over such a failure is a variation that is apparently unattested outside of Hellenistic culture. 41 The sphinx famously committed suicide after Oedipus solved her riddle, and while it is not clear that she did so out of sorrow, some Greek sources had the seer Calchas perish, heartbroken after his colleague Mopsus, Tiresias' grandson, correctly answered a riddle about figs (or piglets). 42 In these cases it is the 39 Humbert, Les voies, 3: "comment la date de la mort d'un si grand personnage a-t-elle pu passer inaperçue aux yeux de ses contemporains?" More broadly, the importance of death traditions is a function of the role of death as the end-point of biography, which fundamentally is narrative; see P. Brooks, Reading for the Plot: Design and Intention in Narrative (New York: Knopf, 1984), ch. 4 (p. 103: ". . . the narrative must tend toward its end, seek illumination in its own death. Yet this must be the right death, the correct end"). On the contemporary Islamic practice of filling out lacunose narratives, cf. L.I. Homer's rich biographic traditions, which circulated widely in late antiquity and beyond, prominently included a public competition with his most important rival, Hesiod. 43 Like Sībawayhi, Homer surprisinglyperhaps unfairlylost. 44 The loss had little impact on him, but he later found himself on the losing end of a rather more fateful wisdom competition, having forgotten that the oracle at Delphifamiliar also from the deaths of Calchas and the sphinxhad warned him of a children's riddle on the island of Ios. 45 As he sat there by the sea, he noticed some boys returning from fishing; addressing them in riddling form as hailing from (landlocked) Arcadia, he asked what they caught. 46 The boys responded with a riddle of their own: "what we caught, we left behind; what we didn't catch, we carry with us". 47 Homer was at a loss, and when the boys explained that they caught no fish but partially deloused themselves, he recalled the oracle and realized his days were numbered. According to one tradition, on his way back from the sea he slipped in the mud and subsequently died. 48 Another version, however, which pointedly claims that Homer was already ill when he arrived on the beach, 49 shows the tradition of his fall to be a rationalizing attempt to indirectly connect the boys' riddle with Homer's death in order to avoid directly associating them. 50 The tradition to which such rationalizing versions are responding is found in numerous sources, which present his death as a result of his sorrow over the failure to solve the riddle. 51 We thus find both Homer and Sībawayhi to have died of heartbreak stemming from defeat in a wisdom competition decided by a question concerned with animals who were themselves (traditionally) agents of death. 52 But did Arabic-speakers know Homer and the story of his death? These are two separate It is significant that the louse sucks human blood, and that phtheir (louse) was associated in antiquity with phtheirein, "to die" (Kahane, Diachronic Dialogues, 21; Porter, "New readings", 4 n. 11).

T H E G R E E K D E A T H O F S Ī B A W A Y H I A N D T H E O R I G I N S O F A R A B I C G R A M M A R
questions; the answer to both is affirmative. Graeco-Arab contact had a long history going back well over a millennium before the time of Sībawayhi, and Hellenistic culture persisted under Islamic rule and in Baghdad, in particular, was buttressed by Syriac scholars immigrating from Edessa. 53 Theophilus of Edessa (695-785), who served as an astrologer in the Abbasid court in Baghdad around the time of Sībawayhi, was said to have translated into Syriac "two books on the conquest of the city of Ilion [Troy] . . . by the poet Homer", and while he was most likely not translating from the Iliad, Syriac excerpts of that work were circulating in Mesopotamia in the third/ninth century. 54 The famous translator Ḥunayn ibn Isḥāq was known not only to recite Homeric poetry in the court of Hārūn al-Rashīd in Baghdad only a few decades after Sībawayhi's death, but was even conversant with Homeric exegesis. 55 Though the Arabic translation movement focused on philosophical and scientific texts, the texts that were translated not infrequently mentioned Homer, and even quoted him. Through such quotations Homeric verses made their way into Arabic literature, at which point some began to circulate independently, appearing in original Arabic works such as al-Bīrūnī's Description of India. 56 It matters little for our purposes that not all such quotations attributed to Homer were actually from the Iliad or the Odyssey; as with the Syriac Homer, so too the Arabic Homer was not merely the author of the two epics. When al-Sijistānī (d. c. 379/985) noted that some Homeric poetry had been translated into Arabic, he was in fact referring to a gnomological collection which the Greeks attributed to the comic playwright Menander. 57 Just as Homer usurped the latter's sayings, so his biography in al-Mubashshir b. Fātik's collection of sages' lives takes over elements from Aesop's. 58 Whatever transformations he underwent, he nevertheless remained a figure of venerable authority, and only a vague awareness of Hellenistic culture would have been required in order to be familiar with his name: he was referred to as "the Greek Imruʾ al-Qays". 59 The story of Homer's death travelled east, too. A version of his death, related in Pseudo-Nonnus' scholia to Gregory of Nazianzus' First Invective Against Julian, was translated into Syriac, probably during the sixth century CE. 60 Some three or four centuries later, Homer's exchange with the boys is quoted in Antony of Tagrit's Rhetoric. 61 But Homer's death tradition not only circulated in early Islamic Mesopotamia, it was also appropriated and creatively adapted in Syriac historiography. 62 Three Syriac chronicles, with some differences in detail, relate the events of the battle of al-Qādisiyya (c. 14-17/635-38), the turning point in the Islamic conquest of Iraq, as follows. As the Arabs and Sasanians set up camp on the banks of the Euphrates, near Kufa, the Sasanian king sends a local Christian spy from al-Ḥīra (or Ḥirtā) to the enemy camp. When he sees a Maʿaddī tribesman bent over, urinating or defecating, eating bread and delousing himself, he asks him what he is doing. The tribesman responds: "as you can see, I am bringing in the new, removing the old, and killing my enemies". The spy immediately understands: a new people is about to enter, the old people will depart, and the Sasanians will be killed. Inevitably, they are routed and chased all the way to Ctesiphon, the Sasanian capital. 63 The anecdote of the spy and the tribesman repays close reading. The Christian spy, caught between the Muslim and Persian armies, stands for the Syriac reader, who is granted privileged access behind the scenes of history. As the spy decodes the tribesman's message, so the reader is called to interpret the narrative of the battle of al-Qādisiyya as bearing more general significance, which was a convention of late antique historiography. 64 Alongside the Syriac perspective, the tribesman's speech gives expression to a Muslim point of view. He in fact embodies a variation on the Islamic topos of the pre-battle meeting between a representative of the imperial -Byzantine or Persianpower with the "poor and pious" Muslim warrior, which subverted the pre-Islamic dynamic whereby the Romans would dispose of the Arab by disdainfully bribing him. 65 The tribesman's oracular response fittingly signals that agency is now his, that it is his speech which requires interpretation. The one viewpoint that is missing is appropriately that of the imperial representative, who is removed from the scene while the tribesman appropriates and rewrites a piece of Hellenistic lore.
The Syriac chronicle writers' source for this incident will have been Dionysius of Tellmahre (d. 845) 66 who, for the events of the seventh century, may well have relied on the aforementioned Theophilus, the "Hellenophile" translator of Homer whose chronicle has been described as "classicizing". 67 In spite of differences in detail, the Syriac anecdote, as Sebastian Brock has  , 1992), xv, notes that these encounters "echo" Shuʿūbī arguments. Yazdagird III, in particular, like the spy, recognizes that the Arabs are to win by interpreting Muslim answers to his questions regarding their attire as oracles (I:2239). Rustam, who also recognizes the inevitability of Muslim victory, with the help of a Ḥirtan translator expounds parables referring to the Arab squalor, and the Muslims respond with a parable of their own. In addition, there is some discussion of spies in al-Ḥīra, though on behalf of the Muslims (I:2249, 2255-6). It is also tempting to see the Syriac version as an adaptation of the report of the spy who observes the Muslims cleaning their teeth before prayer, which would be meaningful only to a Muslim audience (I:2291, with Friedmann, The History, 86 n. 293). Curiously, the encounter with Yazdagird is followed by the capture of fishermen, leading to the "Day  Sībawayhi's death, which I am arguing to be an Arabic iteration of Homer's death tradition, can be found to function similarly to the Syriac iteration, and is interestingly related to it. As the approximate starting point of the battle of al-Qādisiyya, Kufa was entwined with its end-point, Ctesiphon; after the latter's fall, Saʿd b. Abī Waqqāṣ, the commander of the Arab army at al-Qādisiyya, founded Kufa and was said to have built his own palace there on the model of Ctesiphon's White Palace. 71 Kufa thus links the narrative of the battle of al-Qādisiyya and Sībawayhi's biography: just as Arabs setting out from Kufa drove the Persians out of Iraq, so it was an attack by Kufans that brought about the departure of the Persian Sībawayhi. Kufans indeed appear to have played an instrumental role in both traditions, 72 and it is therefore significant that in both cases we have Arabs making their way from Kufa to the capital city situated on the banks of the Tigris (first Ctesiphon, later Baghdad), achieving the establishment of Kufa (literally, or as the pre-eminent grammatical school), as well as the expulsion of the Persians from Iraq. If in the case of case here; for their accounts of the battle of Qādisiyya, see Hoyland, Theophilus of Edessa, 104-5. 68 Again, Brock, "Antagonism to assimilation", 28-9, and Brock, "Syriac views", 13-4. 69 Elsewhere too it can be seen that where the traditions differ, the differences facilitate adaptation. The sea, a crucial component of the Homeric tradition, is separate from the riddle in the Syriac tradition, but is not actually absent: whereas Homer dies on the shores of the Aegean, the battle of al-Qādisiyya officially begins when the Persians cross a canal flowing from the Euphrates, al-ʿAṭīk ("the old [river]"), which is also where Rustam and thirty thousand Persian soldiers find their death (this may explain the name of the first day of the battle, yawm armāth, the Day of the Rafts). Note also that al-Qādisiyya is described in Islamic historiography as a "moat" (khandaq) and in relation to a system of canals in al-Ṭabarī, Taʾrīkh I:2229-30, 2243, 2336-7. And see below on al-Akhfash's final meeting with Sībawayhi, on the "bank" of Basra. 70 Recall that the Syriac tradition itself authorizesindeed, urgessymbolic reading, which was typical of late antique historiography more broadly (see n. 64 above). If the adaptation of Homer's death indeed goes back to Theophilus, it may in fact fit a pattern of using scenes of Greek ruin in relation to the Islamic conquests that runs through his work. Cf. Conrad, "The Arabs and the Colossus", 185, on Theophilus' fanciful ascription of the demolition of the Colossus of Rhodes, "as a symbol of the achievements of Antiquity and its cultural heritage", to the Arabs. It has also been suggested that in his translation of "the conquest of Ilion", Theophilus presented the sack of Troy as a warning to eastern Christianitysee Conrad, "Mawālī", 388, with Hilkens, "Syriac Ilioupersides", esp. 310, but cf. Niccolai, "Epic to parable". the battle of al-Qādisiyya they were explicitly fighting in the name of Islam, in the case of Sībawayhi, as we have seen, they were doing so implicitly. 73 As in the case of Imruʾ al-Qays, where the exile of the king of Kinda from Arabia to Byzantium laid the ground for the rise of Islam, 74 Homer's death tradition is employed in both the Syriac and Arabic traditions to signify the banishment of Persians who were hostile or at least foreign to Islam. While it is clear why the Homeric death tradition would be of use to Sībawayhi's rivals, it appeared above that the author of the Homeric emplotment of Sībawayhi's death may have been the Basran al-Akhfash. But we can see why a Homeric death would also be valuable to al-Akhfash, in whose account Sībawayhi's sad story of rejection is transformed into a narrative of redemption. In the tradition preserved by al-Zubaydī, Sībawayhi himself told al-Akhfash of his loss when he came to bid him farewell on the "bank" (shāṭiʾ) of Basra before departing for Persia. Al-Akhfash then set out for Baghdad, where he went to al-Kisāʾī's mosque and put to him 100 questions, all of which al-Kisāʾī failed to answer. 75 Al-Akhfash's account thus produces a new layer of ring composition for Sībawayhi's story: as al-Akhfash came to Sībawayhi's aid after he was first humiliated by accepting him into his majlis, so he avenged his second humiliation. He in fact accomplished more than mere revenge, for al-Kisāʾī was so struck by al-Akhfash's performance that he entrusted his children's education to him, which plainly serves to signify the Kufans' acceptance of Basran grammar. It is significant that al-Akhfash concludes his lengthy narrative by abruptly returning to Sībawayhi, reporting that he died of sorrow: if by means of this tradition al-Akhfash glorified himself, 76 he also turned Sībawayhi from an unfortunate victim of tribalism or bribery into a heroic, Homeric victim. In the event that the story of Sībawayhi's humiliation was already established, one way of salvaging his reputationindeed, of cementing his authority and legacywould be to have him die of sorrow à la Homer. 77 The Homeric death tradition would thus have been useful both to Sībawayhi's rivals and his followers. It is not necessary that this death should have been specifically associated with Homer, 78 only that it be recognized as fitting for a founding figure. What is important is that this narrative, in one form or another, was current in Mesopotamia and accessible to Arabic-speakers, 79  ways attractive as a death tradition for Sībawayhi. There was certainly no contradiction in assigning a Greek death to a Persian, for in late antiquity Hellenism was broadly equivalent to paganism. 80 As in the case of Imruʾ al-Qays, Sībawayhi's death tradition could be of service both within Islamic culture and without: while internally it denied the notion that a non-Arab not sufficiently versed in Islam could found Arabic grammar, thereby legitimizing it, externally it crafted a founding figure worthy of competition with rival traditions. 81

The origins of Arabic grammar
Recognizing the death of the founding father of Arabic grammar as Greek is significant not only for the reception of Hellenism in early Islam, but also because it has long been claimed that Arabic grammar itself was built on Greek foundations. This claim has encountered staunch opposition; critics prominently argue that Islamic law rather served as a model for Arabic grammar. 82 But the two claims are not mutually exclusive. In fact, they are both embedded in Sībawayhi's biography. The role of Islamic law features in it quite explicitly: Sībawayhi's grammatical studies, we recall, were preceded by religious studies and fuelled by his failure at them, echoing the widespread narrative about the religious roots of grammar, and in particular the Abū l-Aswad al-Duʾalī foundation narrative which is similarly concerned with a grammatical mistake with theological implications. 83 This article has argued that the claim concerning the foreign roots of Arabic grammar is also implicit in Sībawayhi's biography. It is notable that he is not the only figure who, after being expelled by an exasperated teacher, proceeded to establish himself as the pre-eminent authority in a different field of study; this evidently was a topos, one that links him with another outsider who attained mastery of Arabic, the Nestorian Ḥunayn. After being banished from his medical studies, a driven Ḥunayn embarkedlike Sībawayhion linguistic studies, primarily Greek but also Arabic, acquiring a profession that consisted in importing Greek material into Islamic culture. 84 According to one tradition of his death he too perished of sorrow (ghamm) following a debate of a linguistic nature. 85 That both claims regarding the origins of Arabic grammar are to be found in Sībawayhi's biography cannot be considered incidental. Its artful and artificial structure, in addition to its affinity with the Abū l-Aswad al-Duʾalī foundation narrative, suggests that the biography of the founder of Arabic grammar was also constructed in part as a biography of the discipline he founded. 86 Sībawayhi's biography thus offers another avenue by which to approach the question of the origins of Arabic grammar, indicating that in discussing the possibility of Greek influence it is not sufficient to look for its tracesor the absence thereofin Sībawayhi's theory. 87 We must also ask what, within the environment in which Arabic grammar developed, was the significance of the stories that were told about its foundation? The fact that both Sībawayhi's biography and the Abū l-Aswad al-Duʾalī foundation narrative are intent on showing that non-Arabs could only contribute to the development of grammar negatively implies that there was some concern about foreign influence, or the appearance of such influence, in the formation of Arabic grammar.
There was good reason for concern. The Arabic grammatical tradition did not develop in a vacuum. 88 It was but one of three contemporary grammatical traditions in Mesopotamia, along with Syriac and Hebrew, driven by the demands of reading scripture. 89 A letter written in 168/785 by Timothy I, Catholicos of the Church of the East, to Sergius, head of the monastic school of Abraham in Mosul, bears witness to a sense of rivalry between the Syriac and the Arabic and Greek grammatical traditions. 90 The three traditions are demonstrably related: the Syriac tradition drew deeply from the Greek, and made no attempt to deny doing so; 91 and numerous features of the Arabic tradition, including the names of the vowels and their signswhose invention was attributed to Abū l-Aswadbetray contact with Syriac grammar. 92 What of Arabic and Greek? While the Syriac tradition was not averse to acknowledging its Greek debt, if Arabic grammar were to safeguard the purity of the language of the Quran, it clearly could not admit to any foreign source, just as early Islamic culture more broadly endeavoured to present itself as having formed in sublime isolation from its neighbours. 93 In the third/ninth century, the