by Scott McGill
As the use of other literary languages contracted, the role of Arabic as a medium of culture and learning was magnified through the very extensive Arabic reception, via selective translation, of the literatures surviving late antiquity, culled as needed and desired by patrons and professionals among the Arabic‐speaking readership. Classical Arabic literature grew as a remarkable blend of several hitherto largely separate ingredients, creating a new, cosmopolitan culture of scholarship and letters in the process. The large‐scale translation of texts from Iranian languages (especially Middle Persian), Sanskrit, Aramaic (especially the Syriac dialect), and Greek had begun already in the second half of the eighth century. Scholars and patrons of scholarship demanded old learning in the now‐current medium. Old traditions from different countries were synthesized into new Arabic forms and put to new purposes. Translations from Greek, in particular, became especially numerous and prestigious in the ninth and tenth centuries.
The Arabic reception of late ancient literature is sometimes characterized as having been effected by a “translation movement.” The expression was put into wide circulation by Dimitri Gutas (1998) to describe concisely the phenomenon of sustained widespread patronage of translations from earlier literary traditions, especially ancient Greek, into Arabic. While this certainly can be characterized fairly as a movement, it was not motivated by a uniform interest evenly distributed over places and times. Gutas simultaneously demonstrated generational variation in the reception. Interests changed with political and social circumstances. Translations of ancient works themselves sometimes stimulated interest in further translations. The range of topics for which translations were in demand grew. I discuss examples of these variations in what follows.
In the first 75 years of the secular and scientific translations, interest in which was centered in Baghdad (ca. 762–850), it was mostly only physicians, astrologers, and specialists in religion who used these ancient books. They needed to conduct their professions using the state of the art available in the preexisting literary traditions, as it was written in the languages of their communities of origin. Works like Aristotle’s Physics and his logical treatises came quickly to play an important role in the interreligious and sectarian theological debates fostered by the Abbasid caliphs’ courtiers and those in their orbit. Later, the audience of these translations broadened with the growth of Arabic intellectual life and the desire of intellectually inclined Muslims, especially the secretarial class (kuttāb), to participate in all varieties of learning. By the turn of the millennium, most educated, cosmopolitan Arabic speakers had encountered ancient learning in Arabic translations and abridgments. They could be expected to understand the basic elements of ancient Greek philosophy, medicine, science, and wisdom, even when they were not themselves professional practitioners of the ancient arts and sciences. They could name the humors of the Galenic body, describe the spherical Ptolemaic cosmos, outline the fundamental doctrines of the philosophers, and recite a fair bit of memorized gnomological lore in the name of Greek figures such as Socrates and Diogenes the Cynic or of Persians such as Buzurgmihr, legendary sage of the court of the Persian king Khusrō I (r. 531–579). They could spar with each other about first principles and God’s attributes, the dualism of Zoroastrians, and the trinity of the Christians, citing late antique commentaries on Aristotelian logic while sometimes unknowingly echoing early Christian apologetic and polemic. By the mid‐tenth century, such topics had become not so much a matter of the deliberate recovery of ancient thought by individuals as they were merely a part of the fabric of Arabic learned culture. In this respect, the reception had been accomplished and the material was digested. The theories, systems, and ideas translated would henceforth be elaborated and innovated upon in new ways, leading to the gradual abandonment of the direct use of ancient classics in Arabic translation in favor of recourse to the reconfigured and adapted syntheses of ancient scholarship produced from this time onward by Muslim scholars such as Ibn Sīnā (d. 1037).
The Arabic reception of late antique literature created a different canon from the literary canon of classical scholarship of modern Europe. A comparison of the two receptions is instructive. The classical Latin tradition of Europe, on which the modern discipline of classical studies is based, was fundamentally shaped by the interests of medieval Italian humanisti, rhetoricians who valued poetry and orations according to an ancient standard of language that they deemed excellent and that they sought to revive. The ancient masters whom they followed, such as Cicero and Ovid, were manifestly dependent on Greek rhetoric and poetry, and their interest is part of what led western European scholars to study ancient Greek directly. “Literature” in these forms became central to the European tradition of learning that would develop into classical scholarship. The reception of ancient learning in Arabic had a distinctly different character due to its different conditions. It stemmed not from the needs of Arabic rhetors, who already had adequate and socially appropriate models from ancient Arabic poetry and from Islamic scripture and sermon, models already esteemed by their patrons. In Baghdad, the Arabic reception was generated rather by the needs of other sorts of court professionals: astrologers, physicians, and experts in interreligious disputation. The Arabic reception of ancient texts in Baghdad did require some searching out and rediscovering of ancient works, but it was based at first more on the transfer of the specialized learning then current in non‐Arabic languages into Arabic than on the revival of lost texts. There was no intention of pursuing that learning in the original languages, because this was not a revival of literary forms expressed idiosyncratically in those ancient and defunct tongues. It was, instead, a reception of disciplines, techniques, and methods of learning and practice. Nor was the learning particular to one language, but it was derived from any available source. It was a rather ecumenical reception, as expressed by Abū Ḥātim al‐Rāzī at the beginning of this chapter. The shape of the Arabic reception of ancient literatures thus does not just throw into relief the conditioned character of western European notions of “literature.” It also demonstrates that an entirely different selection process in the reception of ancient Greek literature was both possible and actual, and it entailed a mixture with materials translated from other literary languages employed to the east of the Roman Empire, not with materials from Latin. It also includes texts that do not survive in their original, but only in Arabic translation. Classical scholars have occasionally assumed that what was preserved of ancient Greek and Latin texts was simply “the best” of antiquity, destined to survive basically because of its excellence. Close historical comparison of the multiple receptions of the same body of material should help to improve this view.
There was interest in the learning of all prior nations, whenever anything of it was deemed useful. The ninth‐century Arab philosopher al‐Kindī wrote,
Aristotle, the most distinguished of the Greeks in philosophy, said: “We ought to be grateful to the fathers of those who have contributed any truth, since they were the cause of their existence; let alone being grateful to the sons; for the fathers are their cause, while they are the cause of our attaining the truth.” How beautiful is that which he said in this matter! We ought not to be ashamed of appreciating the truth and of acquiring it wherever it comes from, even if it comes from races distant and nations different from us.
(Ivry 1974, p. 58)
The need felt by al‐Kindī for the comment shows that not all of his contemporaries agreed with the notion he expressed, but it certainly describes an attitude prevalent among urbane intellectuals of the age of translations into Arabic.
When ancient works were not currently in use, they had to be sought. A substantial number of reports survive of Arabic scholars seeking copies of one ancient work or another. One way to find them was simply by hunting down ancient manuscripts surviving in nooks and crannies of old churches and monasteries, family libraries, and neglected archives (van Koningsveld 1998). Another was the commissioning of new copies of a
ncient scientific and mathematical works, which were now being copied anew in in Byzantium. The new copying activity has plausibly been connected to a market demand for such books in Baghdad; at least the two phenomena are fairly synchronous (Gutas 1998, pp. 175–186). These sciences had long coexisted with Christian learning and in some respects were integrated into it. For example, the Greek works of John Philoponos (fl. sixth century), a Christian philosopher of Egypt, came to be well known to Arabic philosophers in Arabic translation. His treatises addressed Greek philosophy as related to the theological needs of a monotheism of a type similar to that enjoined by Islam. For another example, one of the earliest encyclopedic cosmological works composed in Arabic, the Sirr al‐khalīqa, or Mystery of Creation – written in the time of the caliph al‐Maʾmūn (r. 813–833) and attributed spuriously to the ancient wonder‐working sage Apollonius of Tyana – incorporates in its major recension large sections of Nemesius of Emesa’s De natura hominis, a work of Christian anthropology written circa 390 in the ancient Greek philosophical tradition. The latter had been translated into Syriac, and from Syriac it appears to have entered Arabic in this form. The Sirr al‐khalīqa would go on to be influential in the formation of medieval alchemy, without any explicit reference to Nemesius in it, although it incorporated some of its contents. Nemesius himself relied on Middle Platonists and Neoplatonists.
A large part of the earliest strata of translations, however, from the late eighth century, came not from Greek but from Middle Persian. The extent of Middle Persian written literature under Sasanid rule is evident more from Arabic translations than from any other source. Current scholarship sees pre‐Islamic Iran as having a largely “oral” culture, at least one in which literature and scholarship were predominantly unwritten (e.g. Huyse 2008). This is misleading, as numerous Middle Persian works, lost in the original, survive only in Arabic translation, including examples of genres otherwise entirely lost. These were clearly written (not oral) compositions in their Middle Persian originals. The mistaken view exaggerating Sasanian “orality” arises because of the character of what does survive in continuous Middle Persian manuscript tradition. With the important exceptions of numerous but fragmentary Manichaean Middle Persian texts recovered from Central Asian deserts and a small number of inscriptions and other miscellaneous items, Middle Persian texts survive in Zoroastrian manuscript tradition, which preserves only such material as was of continuous interest to Zoroastrian scholars for the purposes of their religion. Zoroastrian priests praised the oral, esteeming their orally preserved liturgy above all else. These same scholars came eventually to use Arabic, New Persian, and other languages for texts with nonreligious purposes and kept Middle Persian works largely only for their religion, with a few exceptions. One has to turn to early Arabic translations of Middle Persian to get a glimpse of what else was there late in the Sasanian period (224–651) and immediately after the Arab conquest: astronomy and astrology, physiognomy, narrative history, lapidary science, agricultural lore, wisdom literature, books of manners and protocol, and more. In other words, studying Middle Persian prose literature of late antiquity today requires recourse to Arabic texts at least as much as to Middle Persian itself. This endeavor has barely begun among modern scholars. The fact of it, however, is due precisely to the large‐scale adoption of ancient learning in Arabic translation between 750 and 1000, which began with Middle Persian texts.
Sanskrit texts also came into Arabic early in the period of translations, mostly during the last quarter of the eighth century. The interest in Sanskrit learning originated among men from Central Asia whose families came from beyond the former Sasanian Persian Empire, which had relatively recently fallen under Arab rule in the first half of the eighth century. These sorts of families participated in the rise of the Abbasid dynasty of caliphs. The courtiers of the Abbasid caliph al‐Mans&c.dotbl;ūr and the officers and viziers of the Barmakid family, from Balkh, a city where the Bactrian language had been spoken and where Buddhist scholarship was alive into the eighth century, were the best‐known patrons of translations from Sanskrit into Arabic (van Bladel 2011; 2014). When the Barmakid family was removed from power in 803, the patronage of new translations from Sanskrit effectively ceased. This clearly illustrates the connection between the background of patrons and scholars and the materials sought for translation. Sanskrit mathematical, astronomical, and medical learning of late antiquity became part of the early strata of science in Arabic. Because they were superseded by later works, many of these early translations are lost, but their traces are numerous as citations in later texts on the same genres, such as medical or astronomical compendia, whenever the competing matter from Greek did not supply much material as an alternative.
In the first half‐century of Abbasid rule (750–800), while translators reworked Middle Iranian languages and Sanskrit texts into Arabic, Greek learning came into Arabic sometimes from Syriac and Middle Persian intermediaries rather than directly from Greek. With the new capital of Baghdad founded in Iraq, when the early supporters of the Abbasids, of Central Asian origin, died off, new secretaries came increasingly from a local, Aramaean background. The geographical reorientation, including regular Abbasid campaigns against Byzantium, coincided with a turning of interests toward the traditions of Hellenic learning cultivated among Syriac scholars and, soon, to increasingly numerous direct translations from Greek into Arabic where they could be had. The ninth century accordingly witnessed the birth of Arabic philosophy as such – under the name falsafa – with al‐Kindī, whose interests followed a course that has been described as Neoplatonic, with a heavy infusion of Euclid. Several reports survive of the painstaking search for Greek manuscripts in the ancient cities of the Near East, the plunder of books from Anatolian towns on the yearly raids, and copies sought in the Byzantine Empire. Eventually hundreds of works by dozens of authors were translated, reworked, commented upon, and absorbed into the intellectual life of the new society. Translators were paid large sums in return for which they mastered ancient Greek to a high degree, as the translations themselves often bear out. As far as can be determined, everything both interesting and available to the new Muslim audience was translated over three centuries. The adoption of paper technology in the Middle East facilitated the creation of large private libraries, collections of thousands of books stocked with the works of Galen, Aristotle, and dozens of other ancients along with all the current Arabic poetry, literature, and scholarship.
This efflorescence of learning in Baghdad was one of the major bases of later medieval science, medicine, and philosophy. Both the Byzantine and Latin European learned traditions depend in part on this Arabic tradition for their form and content. Byzantinists are becoming increasingly aware that Byzantine Greek manuscripts contain numerous translations from Arabic (Mavroudi 1998, pp. 392–430), and medieval Latinists have known for many years that Western scholasticism could not have come about as it did without translations from Arabic (Burnett 2005). The different receptions of ancient learning – Byzantine Greek, European Latin, and African and Asian Arabic – are inextricably intertwined.
Scholarship in the caliphal capital of Baghdad provided a model for other courts, where increasingly independent emirs and sultans would gather scholars to serve and advise them and provide them with prestige. For example, the philosopher al‐Fārābī (d. 950) spent time in the Ḥamdānid court in Aleppo. The Sāmānid emirs of Northeastern Iran and Central Asia were patrons of Arabic scholars and inaugurated the patronage of scholarship in New Persian. Ibn Sīnā (Avicenna), who came from that Central Asian region, found his last patrons in Isfahan, in western Iran, where he died. In al‐Andalus (Spain), emirs sponsored astronomy, philosophy, and other sorts of learning. As sites of patronage for learning proliferated, the heritage of antiquity, synthesized as it had been in Baghdad, also spread.
Many pre‐late antique texts by Greek authors were translated into Arabic, including works of Euclid, Aristotle, Plato, and Ptolemy, so long as they were deemed
useful. The translators themselves had no concept corresponding to late antiquity, but if one wanted a list of just authors of scientific and philosophical works of the period 200–600, roughly our late antiquity, we can name Alexander of Tralles, Alexander of Aphrodisias, Ammonius, Cassianus Bassus Scholasticus, Galen, John Philoponus, Olympiodorus the Younger, Oribasius, Palladius the physician, Pappus of Alexandria, Plotinus, Porphyry, Proclus Diadochus, Simplicius, Themistius, Theon of Alexandria, Timothy of Gaza, Vindanius Anatolius, and Zosimus of Panopolis, with works such as the medical Summaria Alexandrinorum among them. This is only a portion of what became available.
In only one region did the Arab conquerors’ descendants adopt another language as a regular medium for scholarship, under different demographic circumstances too complicated to be explicated here: That was the new form of Persian written in the Arabic script, in Central Asia, a language the use of which would be spread westward and eastward, beginning in the eleventh century, with the Muslim Turk warlords whose secretaries and administrators were drawn from that region. Even with its different conditions of genesis, New Persian literature began with the reception of Arabic texts and modes of expression in Persian translation, and along with it came material drawn from late antiquity through Arabic. For example, one of the earliest extant works composed in New Persian is Maysarī’s medical work Dānish‐nāma, or Book of Knowledge (written ca. 980), summarizing treatments corresponding with their ailments in four and a half thousand verses. It is full of material derived from ancient Greek medicine and its entire framework is Galenic. That is because Galenic medicine in Arabic translation was, in the tenth century, the norm from the Atlantic to Central Asia. Thus, even when substantial Muslim populations eventually came to use languages other than Arabic, their reception of the literatures of late antiquity came by way of Arabic. It was in Arabic that the literary products of the age we call late antiquity were interpreted for later peoples of Asia and northern Africa. Wherever Arabic learning went, the learning of late antiquity was carried along with it in continuity with the ancient past.