1 Introduction: Toward a New Definition of Translation

The Arabic translation that occurred during the Middle Ages has played a more effective role in transforming the Arabic culture than contemporary translation theory and practice, and is entirely worthy of careful and detailed description. The Arabic translators of the Middle Ages did not simply render the meanings of source texts, but they critically and creatively engaged with them. That is to say, translation was not a means of knowledge transfer where equivalence was the major priority, but rather was a tool for knowledge production whose main purpose was applying knowledge and adding to it. As Uwe Vagelpohl states, “The texts produced during the Greek-Arabic translation movement are independent literary facts: they are based on Greek and/or Syriac source texts but they often enough put forward arguments and make points which differ from what we would expect on the basis of our carefully collated and thoroughly annotated and researched editions.”Footnote ¹

The translators began a tradition of engaging critically with their source texts, adding new thought to them, which the Muslim scholars developed further in original composition. The most fascinating outcome of this approach was the transformation of the Arabic language, turning it into an ideal medium for the advancement of philosophy and science. In fact, the precision of some of the translated terms “even surpassed that of the original Greek texts,” so they became the model for the western languages when these texts were later translated into Latin. In addition to the transformation of the Arabic language, the Arabic approach to translation helped create whole new disciplines such as Islamic philosophy. Using the philological outlook, which emerged only recently in modern Europe, as a paradigm to analyze and assess Arabic translation would reduce the Graeco-Arabic translation movement to the preservation of Greek knowledge until it was rediscovered in Europe, which is not the full story of Arabic translation. Thus, a proper description of the Graeco-Arabic translation movement needs to begin with a redefinition of translation, since traditional definitions are not comprehensive enough to encompass the history of Arabic translation.

Surveying the history of science translation in the Roman, Syriac, Indian, Pahlavi, and Arabic traditions in the premodern world, Scott Montgomery defines translation as “the process of transforming a specific piece of one language (commonly a text of some sort) into another language.”Footnote ² Two words are significant in this definition: process and transformation. The word “process” implies an ongoing activity, so the target text is tentative, fluid, and in constant movement. The word “transformation” implies a complex process that Homi Bhabha calls “cultural hybridity.”Footnote ³ There are two features of this cultural hybridity. First, the target text is not a copy of the source text; it combines knowledge of the source text with knowledge from the target culture. Second, the source text is different from the target text “without assumed or imposed hierarchy.”Footnote ⁴ This applies to Arabic translation in the Middle Ages. Arabic translations were a cultural hybridity because the translators added new thought to their target texts, and because, unlike the Roman translators who saw the Greek texts as masterpieces to imitate, they saw their language as equal to the Greek language and capable of expressing scientific ideas. They did not imitate and displace the Greek texts. They acknowledged them, but also corrected and improved them.

In order to account for the history of the practice of Arabic translation, I define translation as combination, that is, a process of remaking meaning that synthesizes multiple realities. The Arabic translators of the Middle Ages did not simply find an equivalent to the source text but combined its meaning with the target language, culture, and sociopolitical context. In addition, Arabic translators combined knowledge from the source text with their own knowledge and experience. Thus, part of translating a text was to add new thought to it. A translation did not simply communicate the original meaning but improved it. To distinguish the source text from the added parts, these new parts were written in different or thicker ink, introduced in separate sections, or marked with phrasal verbs that identified the author. The addition of original composition and thought ensured the utmost critical and creative engagement with the source text and set translation on a par with writing.

However, translation as combination does not fully explain the success of the Graeco-Arabic translation movement. Accompanying this feature is another that I call the “spiral of influence.”Footnote ⁵ The spiral of influence explains how translation enriched the Arabic language and turned it into the language of science. For example, when a term was introduced through translation, it moved up the spiral. That is, once the term was introduced, more scholars, then more people from the masses, began to use it. Therefore, it became part of Arabic language, culture, and daily life. Such an upward movement was contingent upon an environment in which more scholars and more people from the masses could engage in science. If that was not the case, it would be difficult for the term to enter common usage. It would probably confine itself to the discourse of the elite or fade away altogether. It was also possible that a term could move down the spiral after its introduction because it was not optimal enough and, in this case, a new term was usually introduced to replace it. The spiral of influence also explains the failure of translation in the contemporary Arab world, since its upward movement is too slow, given the high level of illiteracy.

My investigation relies on three sources. First, the statements put out by the Muslim scholars in the Middle Ages about translation are useful, even though they are too sporadic, brief, and general to lay the foundation for a theory. Second, more recently Graeco-Arabists have edited and published many manuscripts, which they have compared with the original Greek sources. This work produced significant results, which served as the main source in my investigation. Third, I was able to collect some manuscripts and their English translations and analyze them myself.

My purpose is straightforward: to (a) describe the history of the practice of Arabic translation, (b) redefine translation studies to encompass the history of Arabic translation practice, and (c) identify the knowledge sets and skills that translators needed to acquire in the Middle Ages and compare them with the knowledge sets and skills required in other periods of time. The results of such research can be expected to advance translation studies in various ways.

To begin with, taking the Arab world and Arabic as its points of departure, part of the significance of the study is that it discusses translation in areas whose contributions to translation studies have not been large. Many scholars have recently proposed that translation be studied in different parts of the world to establish translation theory on more comprehensive and solid grounds and to limit ethnocentrism in translation studies.Footnote ⁶ Indeed, as Luc van Doorslaer says, “translation studies can only be strengthened if there is more discourse from and about China, other east Asian nations, southeast Asia, India, Turkic cultures, the Arab world, and so forth.”Footnote ⁷ In his article “Continentalism and the Invention of Tradition in Translation Studies,” Dirk Delabastita raises a very legitimate question: “To what extent are the well-known translation models – say, those discussed in Anthony Pym’s Exploring Translation Theories (2010) – tailored to fit translational practices existing in the West only? To what extent is there an ethnocentrism at work in them which can be contested by considering the practices and theories from different parts of the world?”Footnote ⁸ To these suggestions, I add that studying translation in different periods of time can also be a good test for our contemporary translation theory and practice. Although the history of translation in different parts of the world has recently been the center of attention for some scholars, there is still much to be done. Examples include Jean Delisle and Judith Woodsworth, who review, in their book Translators through History,Footnote ⁹ the history of translation in different parts of the world focusing on the agent, translators themselves, rather than the product of the process.Footnote ¹⁰ In his book, Translation and Identity in the Americas, Edwin Gentzler describes an emerging American theory of translation that, according to Delabastita, “allows for creativity, freedom and change, and has a living relationship with translation practice. It asserts itself against traditional translation theory, which is essentially European, preoccupied with national languages and national literary canons, and entertaining naïve ideas about universality and the reproductivity of identical meanings.”Footnote ¹¹

Further, translation theory suffers from a disproportion between studies of literary translation, on the one hand, and scientific translation, on the other hand. Even though there has recently been more attention paid to scientific translation, translation theory is still largely dominated by discussions of literary and religious texts.Footnote ¹² This study focuses mainly on Arabic translation of science in the Middle Ages and the early nineteenth century. Scientific translation, as described in this study, is substantially different from contemporary translation theory and practice, since translation was not simply a means of communicating meaning but was also a tool of knowledge production.

Besides, Graeco-Arabic translation is a special case that requires new models of translation to account for its unique challenges, whether in the past in the way that translators performed their work or in the present in the designs that researchers use to study these translations. As far as the process of translation is concerned, a major problem that translators faced during the Graeco-Arabic translation movement was the scarcity of manuscripts. More often than not, translators worked from a single manuscript that may have been damaged and they needed to travel long distances from one learning center to another to obtain a certain manuscript. The translators were committed, under these working conditions, to engaging critically with their source texts, correcting them and filling their gaps. Even when translators had several manuscripts at their disposal, they had to do additional manuscript work, collating a single manuscript or selecting one variant and including all of the others in the margin or interlineally. That process of selecting a variant was not random and required a great deal of background knowledge and critical thinking.

These finished products then went through a long journey of hand copying, copy editing, and revisions before they reached us. In these cases, a translation is defined as a manuscript that is known as a translation and encompasses the original work of the translator as well as the work of subsequent revisers and editors. The long process of hand copying, in addition to rigorous revisions of translations, requires extra caution from the researcher when examining translation strategies. In addition, data on the original translator and the list of editors and revisers are always unavailable or, at best, doubtful, which is an obstacle for studying a certain translator or a certain period of time. That is why the Risāla Ḥunayn ibn Isḥāq is substantial for the history of Arabic translation.Footnote ¹³ It informs us about the translations produced by Ḥunayn, as well as the original translators of some texts that Ḥunayn later revised. It also enlightens us on the practice of translation at that time. For example, it makes us aware that the revision of old translations was a common practice and highlights the difficulty of assigning a translation to a certain translator. Although this was an obstacle for medieval Arabic translation research, it does not pose a serious problem for this study, since it examines the common characteristics of Arabic translation, regardless of the identity of the translator or the period of time.

In addition to a lack of secure data and continuous corruption of manuscripts, many source texts are no longer extant today, which hinders the comparison between source and target texts, particularly those translated out of intermediary Syriac. Further, many of the manuscripts that are extant today are still unedited, and much evidence is still buried in them. In spite of that, the available manuscripts are enough to demonstrate a unique translation practice that, unless clouded by our understanding of translation theory today, yields a rich source for understanding the process of translation and its impact on culture.

2 The Graeco-Arabic Translation Movement: Historical Background

Although the phrase “the Graeco-Arabic translation movement” is usually used to refer to the translation activities during the Abbasid Caliphate, its early signs and roots can be traced back to the Umayyad period. During the reign of the Umayyads (661–750 CE),Footnote ¹⁴ Muslim rule expanded over vast areas and the Muslim Empire became one of the largest in history. At their peak, the Umayyads ruled the Middle East, North Africa, the Iberian Peninsula, Persia, and Central Asia. Despite that power, the Umayyads’ rule was short lived, spanning a period of only ninety years. The decline began with the military defeat by the Byzantine Empire in 717 CE, and it continued throughout the later Marwanid period (720–50 CE). During that period, dissatisfaction with the Umayyads grew, revolts against them increased, and the Hāshimayya, a movement that denied the legitimacy of Umayyad rule, gained more supporters. Eventually, the weaknesses of the Umayyads encouraged the Abbasids to declare a revolt in 747 CE. Supported by the Persians, Iraqis, and Shiʿites, the Abbasids won a victory against the Umayyads in the Battle of the Great Zab River in 750 CE and put an end to the Umayyad Dynasty.

Nevertheless, the Umayyad Dynasty enjoyed many years of luxury and prosperity, and it was not unlikely that some Umayyad figures such as Khālid ibn Yazīd (668–704 CE) or ʿUmar ibn Abd al-ʿAzīz (682–720 CE) commissioned some scientific translations into Arabic. Khālid ibn Yazīd was the son of the second Umayyad caliph, Yazīd I, and brother of the third caliph, Muʿāwiyya II. He was supposed to succeed Maʿāwiyya II, but the throne was taken over by his second cousin, Marwan ibn al-Ḥakam. As a result, it was said that Khālid turned his attention to science. In his book Al-Fahrist, Ibn al-Nadīm mentions that Khālid ibn Yazīd ibn Muʿāwiyya, who loved science and paid great attention to medicine, commissioned some Greek physicians who used to visit Egypt and who were good at Arabic to translate books from Greek into Arabic.Footnote ¹⁵ Even though these reports were contested,Footnote ¹⁶ they indicate that scientific translation was not totally extrinsic to the Umayyad culture. According to Franz Rosenthal, “practical usefulness” was the purpose of translation in the Umayyad period.Footnote ¹⁷ The Arabs needed information, whether medical, chemical, or even philosophical, and they were happy to get that information from an Arabic translator or directly from a Greek scholar. Although these activities were individual projects, they indicated a rich intellectual life and a good foundation for the Abbasid translation movement.

Another significant event that took place during the Umayyad Dynasty and paved the way for the Abbasid translation movement was the translation of the dīwān (administrative apparatus) into Arabic. The dīwān was not simply an official record of salaries, lands, and taxes; it included far more complex operations. As George Saliba convincingly argues,

Although we do not have the manuals translated for administrative purposes, it is reasonable to say that the translation of the dīwān required a high-level effort to translate scientific works and train new staff.

Further, although the number of works translated during the Umayyad period was few, they were significant, since they set the framework for later work. The early translations from Syriac and Pahlavi introduced the Arabic translators to the techniques of translation and laid the foundation for critical and creative engagement with the source texts.Footnote ¹⁹ There are then two immediate traditions that we need to examine in order to understand medieval Arabic translation.

To begin with, the Syriac tradition of scientific translation provides a good source for understanding the practice of Arabic translation. During the fifth and sixth centuries, translation activities moved eastward into Persia (Syria and Iraq). This resulted in waves of immigrations of Christian dissidents who suffered ill treatment under the Orthodox Byzantine Church. Under prejudice against pro-Nestorian, and to a lesser degree pro-Monophysite, teachings, teachers and intellectuals of these communities had to immigrate to Persia and rebuild their schools, where they resumed their intellectual activities of studying, commenting on, and translating the Hellenistic knowledge. These new learning centers, an example of which was Jundīshapūr, were a melting pot for two traditions: the Roman and the Perso-Indian traditions.Footnote ²⁰ These Syriac translators brought with them the long tradition of Roman translation, but, in their new context, they were able to substantially develop it, whether with the quantity or content of translation. Therefore, beginning from the early fifth century, the focus of translation was no longer the Bible and patristic texts as a larger number of secular texts began to be included. By the seventh century, a significant amount of Greek science and philosophy was being translated into Syriac.Footnote ²¹ These translations were totally different from the Roman appropriation of Greek knowledge, as they were literal renditions. As a result, many new Greekisms – loan words as well as syntactic structures – were introduced into Syriac.

This Syriac translation movement was significant for the Arabic translators because it provided “the stimuli for an increased awareness of technique and the honing of translation skills. Indeed, in more than one instance, the work of these earlier translators was used as a focal point for debate on the best methods to render Greek works into Arabic.”Footnote ²² One example is how Sergius of Reshʿayna was said to describe his translation approach: “I have taken great care to remain entirely faithful to what I found in the manuscript, neither adding anything to what the philosopher wrote nor leaving anything out.”Footnote ²³ Although Sergius claimed that his method of translation was word for word, modern criticism confirms that his translations were a mixture of literal and free rendering, the very method that is attested to by analyzing Arabic translations, even though Arabic translators also made claims similar to those of Sergius.

The Perso-Indian tradition of scientific translation is another important source for understanding the Arabic approach to translation.Footnote ²⁴ Scott Montgomery gives the example of the “zīj” text as a representation of the practice of Pahlavi translation and concludes the following:

To sum up, the Arabic translation movement combined features from the Roman, Syriac, and the Pahlavi translation traditions. Like the Romans, they respected the Greek philosophers, scientists, and astronomers and looked to them as models to imitate, thus restricting their originality to epitomes and commentaries on these works. However, they did not displace the Greek texts. Like the Syriac translators, they acknowledged the Greek authors and texts, and mixed word-for-word and sense-for-sense approaches to make their texts accessible to a wider readership. Finally, like the Pahlavi translators, they did not look at the Greek texts as sacred texts; hence, they engaged critically and creatively with them, correcting, reordering, and adding new thought to them.

These features are present in the practice of early translators such as Māsarjawayh. Māsarjawayh, known as the physician of Baṣra, where he lived, was a Jewish physician and translator. He translated the medical Pandects of the archdeacon or presbyter Aaron of Alexandria (c. 610–41 CE) from Syriac into Arabic. This translation, which consisted of thirty chapters – to which Māsarjawayh added two chapters of his own – might be the first scientific translation into Arabic. This translation was preserved in the library of the court until Caliph ʿUmar ibn ʿAbd al-ʿAzīz put it to the use of Muslim scholars. Māsarjawayh also authored a few treatises in Arabic including “The Virtues of Foods: Their Advantage and Their Disadvantage” and “The Virtues of Medical Plants: Their Advantage and Their Disadvantage.” Neither the translations nor the treatises were preserved, and they were known only by references in other writings.Footnote ²⁶

Another important early translator of the Umayyad period, whose translations assist in understanding the Arabic translation practice in general, is ʿAbd Allah ibn al-Muqaffaʿ (d. 759/60 CE). One of Ibn al-Muqaffaʿ’s major contributions is the use of literature to advise kings and princes in what is known as the literary genre of “mirrors for princes.” He also served as a kātib (chancery secretary) in the Umayyad Caliphate before its fall, which gave him experience in the affairs of the state, which, in turn, helped him to perfect the political writings of his time. He is well-known, however, for his translation of Kalīla wa-Dimna from Pahlavi into Arabic.

3 The Golden Age of the Abbasids

The Abbasids took their name from the Prophet Muhammad’s uncle ʿAbbās ibn Abd al-Muṭṭalib (566–653 CE), from whom they descended. They were able to sustain their caliphate from 750 to 1517 CE, although from 861 CE, they began losing different areas of their empire to autonomous dynasties. The first Abbasid caliph was Abu al-Abbās al-Saffāḥ (722–54 CE). The title “al-Saffāḥ” refers to the blood shedding undertaken under his leadership to eliminate the Umayyads. Al-Saffāḥ was succeeded by Abū Jaʿfar al-Manṣūr (714–75 CE), who was the first Abbasid caliph to sponsor the translation movement.

Under al-Manṣūr’s leadership, many prominent translators flourished. Abū Yaḥya ibn al-Baṭrīq (730–815 CE), a Syrian scholar and translator, was the official translator of al-Manṣūr. He translated some medical works of Galen and Hippocrates from Greek into Arabic. He also translated Ptolemy’s Tetrabiblos (Kitāb al-Samāʾ) and Aristotle’s Meteorology (Kitāb al-Āthār al-ʿUlwiyah). When al-Manṣūr sent emissaries to the Byzantine emperor in Constantinople requesting manuscripts, the emperor sent some manuscripts, including Euclid’s most famous work, Elements, which was translated by ibn al-Baṭrīq.Footnote ²⁷

Another famous Greek-into-Arabic translator in the court of al-Manṣūr was Jūrjis ibn Jibrīl (d. 771 CE), who came to the court to treat the caliph and, having succeeded in his mission, he and his pupils after him were hired as court physicians. According to the report, al-Manṣūr had an upset stomach in the year he built Baghdad, and the more his physicians treated him, the worse he felt. Finally, he consulted with his physicians, who all agreed that Jūrjis ibn Jibrīl, at that time the head physician in Jundayshāpūr, the intellectual center of the Sassanid Empire and home to a teaching hospital, was the most acknowledged physician. Listening to the caliph and examining his condition, Jūrjis successfully treated al-Manṣūr, who was soon his old self again. This action gained Jūrjis the favor of the caliph. Jūrjis continued to serve al-Manṣūr until 768 CE, when Jūrjis fell seriously ill and asked the caliph for permission to go back to Jundayshāpūr. Later, al-Manṣūr asked Jūrjis to come back to Baghdad, but Jūrjis was not well and sent one of his students to the court. Other members of the Bakhtishūʿ family of physicians included Jūrjis’ son, Bakhtishūʿ, who was no less skillful than his father and who served as the personal physician of Hārūn al-Rashīd. The grandson of Jūrjis, Jibrīl ibn Bakhtishūʿ (d. 824/5 CE), was also a physician, and he was the personal physician of Hārūn al-Rashīd and al-Maʾmūn. Among the eminent pupils of Jibrīl ibn Bakhtishūʿ was Yūḥannā ibn Masawayh (777–857 CE), a physician from Jundayshāpūr who translated medical works from Greek into Syriac.

Al-Manṣūr was succeeded by his son Abū ʿAbd Allah Muhammad (745–85 CE), and since it had become the tradition of the Abbasid caliphs to take exalted titles that gave them divine authority, al-Manṣūr – himself having a title that implied divine victory and assistance – gave his son the title “al-Mahdi,” that is, “the one guided by God.” It was a prominent name because, according to Sunni Muslims, it refers to the eschatological figure who would save the world at the end of times. Al-Mahdi, who reigned from 775 CE to his death in 785 CE, was succeeded by al-Hādi, who in turn was succeeded by his son, Hārūn al-Rashīd (766–809 CE). Hārūn al-Rashīd decided to divide the empire, upon his death, between his two sons, al-Amīn and al-Maʾmūn, whose titles meant “the trustworthy” and “the trusted one,” respectively. That division resulted in a civil war that lasted for over two years (811–13 CE) and ended with the victory of al-Maʾmūn, who reigned from 813 CE to his death in 833 CE.

Among the well-known translators in the courts of Hārūn al-Rashīd and al-Maʾmūn were al-Ḥajjāj ibn Yūsuf ibn Maṭar (786–833 CE), Ḥunayn ibn Isḥāq (809–73 CE), and his son Isḥāq (830–910 CE). Ibn Maṭar was a mathematician and translator. He produced two translations of Euclid’s Elements. The first, apparently based upon one manuscript, was translated in the early ninth century for Yaḥyā ibn Khālid (d. 805 CE), the vizier of Caliph Hārūn al-Rashīd. In the 820s CE, Ḥajjāj produced his second revised translation for Caliph al-Maʾmūn. Ḥajjāj also translated Ptolemy’s Almagest. Two versions of that translation are extant today. One of these two translations is complete, whereas the second only contains Books I–IV.

Ḥunayn ibn Isḥāq and his son Isḥāq were among the most influential translators and scientists in the Graeco-Arabic translation movement. Ḥunayn ibn Isḥāq was born in the ancient city of al-Ḥīra (south of present-day Kufa in Iraq). He traveled to Baghdad to seek a career in medicine. In Baghdad, he studied under Yūḥannā ibn Māsawayh, the court physician at the time. It seemed that Ḥunayn was too curious, and he asked many questions. His teacher, already short tempered, was disappointed, particularly in the fact that Ḥunayn was from ʿIbād of Ḥīra, whose members were mostly merchants making a living from exchanging money. The teacher, not unlike most of the physicians of Jundīshāpūr, was unsympathetic toward the people of Ḥīra, and he was not willing to teach his profession to the merchants’ sons. Thus, one day when Ḥunayn asked his teacher to clarify a point, the teacher was so upset that he dismissed Ḥunayn from his classes, reprimanding him and advising him to work as a merchant like his people instead of learning medicine. The humiliation was so immense that Ḥunayn left the class crying and disappeared for almost two years.Footnote ²⁸

Ḥunayn did not give up the study of medicine but turned his attention to the study of Greek for a while. At that time, the Abbasid caliph, al-Maʾmūn, had at his disposal medical Greek manuscripts that the court physician, Jibrīl ibn Bakhtishūʿ, commissioned Ḥunayn to translate. That was the beginning of a glorious career for Ḥunayn. He became the most productive translator in the translation movement and authored important medical works in Arabic. He was also the teacher of important translators in the movement such as Ḥubaysh ibn al-Ḥasan (c. the second half of the ninth century).

After al-Maʾmūn, the translation movement continued to flourish, and Baghdad continued to attract prominent translators and scholars. Examples include Qusṭā ibn Lūqā (830/40–912/22 CE) and Thābit ibn Qurra (826/36–901 CE). Qusṭā ibn Lūqā was born in the city of Baʿalbak (present-day Lebanon) and moved to Baghdad, probably during the reign of al-Mutawakkil, who reigned from 847 to 861 CE. Lūqā, of Greek Christian origin, was a physician, philosopher, mathematician, and translator. He knew Greek, Syriac, and Arabic, and he translated, revised, and wrote treatises on mathematics, medicine, astronomy, and philosophy.Footnote ²⁹ He was among the most prominent figures in the Graeco-Arabic translation movement, whose authority in medicine, according to Ibn al-Nadīm, surpassed that of Ḥunayn ibn Isḥāq.Footnote ³⁰

Thābit ibn Qurra was known as al-Ṣābiʾ Thābit ibn Qurra Al-Ḥarrānī, where “al-Ṣābiʾ” refers to his religion as a worshiper of the stars, and Ḥarrān (in present-day Turkey) refers to the city where he was born and grew up. As a member of the Sabians, who were a Hellenized Semitic astral cult and worshipers of stars, Thābit ibn Qurra mastered astronomy and astrology. He moved to Baghdad as a young man, on the suggestion of Muḥammad ibn Mūsa ibn Shākir, who was impressed by Thābit ibn Qurra’s knowledge of Syriac, Greek, and Arabic. In Baghdad, he received training in mathematics and medicine. He returned to Ḥarrān where he was accused of heresy and was forced to escape back to Baghdad. He was appointed the court astronomer for Caliph al-Muʿtaḍid (reigned 892–902 CE). He became one of the most influential mathematicians, physicians, astronomers, and translators in the Abbasid period. His translations include works by the Greek mathematicians Euclid, Archimedes, Apollonius of Perga, and Ptolemy. He revised Ḥunayn ibn Isḥāq’s translation of Euclid’s Elements. He was also the author of important treatises on mathematics and astronomy. By the time of Thābit ibn Qurra and his fellow translators, the Arabization of Greek knowledge had reached a peak. Like the circles of Ḥunayn ibn Isḥāq and al-Kindī, Thābit contributed a great many neologisms to the Arabic language, which facilitated the comprehension and utilization of science and, eventually, of original Arabic science.Footnote ³¹

It is common to divide the Abbasid Caliphate into three overlapping periods as far as translation is concerned:Footnote ³²

• – The first period is 753–809 CE, the time of the two caliphs Abū Jaʿfar al-Manṣūr and Hārūn al-Rashīd. During that period, al-Manṣūr established Baghdad (in 762 CE), and Hārūn al-Rashīd was said to establish the bayt al-ḥikma (house of wisdom) to be a public library after the palace library became too small for the manuscripts collected for the Abbasid court and their translations. With the exception of Ibn al-Muqaffaʿ, the celebrated Persian translator and thinker, the translators from this period are little known. They were mostly from Christian Aramaic clans from southern Iran and Iraq, and their translations were mostly of astronomy and astrology. This group of translators is different from the translators of the second and third periods, who were Hellenized.

• – The second period covered two centuries and started during the time of ʿAbd Allah al-Maʾmūn. This second period is the golden age of translation, when translation became a state-sponsored activity. It received abundant support, both financially and politically, from the court and the elite. The two major groups of translators, that of al-Kindī and Ḥunayn ibn Isḥāq, operated in that phase. By the end of this period, most of the Hellenistic works had been translated, and the standardization of methodology and terminology was achieved.Footnote ³³

• – The third period started in 912 CE and continued until the fall of the Abbasid and the beginning of the Mamluk Sultanate, and it was generally a downward movement of translation. During that period, there were a few Aristotelian treatises that were never translated into Arabic, such as the Posterior Analytics and the Poetics, which were taken care of by Abū Bishr Mattā, Yaḥya ibn ʿAdī, ʿIsā ibn Isḥāq ibn Zurʿa, and al-Ḥasan ibn Suwār. The reason why these works were not among the early translations is that these parts were removed by Christian authorities from the curriculum in Alexandria because they were considered harmful to Christianity. This restriction was lifted after the Muslim conquest. It is the generation of Abū Bishr Mattā that was credited with introducing these parts of the Organon to the Arabic culture.Footnote ³⁴

Therefore, there is no need, for the purpose of this study, to describe the political situation of the Abbasids beyond al-Maʾmūn, since it was around that time that the Abbasids enjoyed the peak of their golden age and during which translation reached its peak. It was a period of peace and prosperity during which the arts, poetry, science, translation, and learning in general were generously sponsored by the caliphs.

In contrast with the Umayyads, who annexed vast areas to the emerging Muslim empire, the Abbasids focused primarily on internal affairs. Among the significant characteristics of the Abbasid reign was the inclusion of non-Arabs in the community. They relied on the Persians (particularly the Barmakids) in administration, and they adopted many of their etiquettes. The combined accomplishments of the Umayyads and the Abbasids (i.e. the expansion of the Muslim empire over vast areas and the acceptance of multicultural and multilingual norms) were significant for the translation movement for two main reasons.

First, the success of the translation movement relied on men capable of translating from Syriac, Greek, and Persian. Such men existed as part of the Islamic empire after the annexation of huge areas from other empires surrounding Arabia. In addition, the existence of speakers of other languages in the Muslim empire made translation, particularly for business and administrative purposes, a necessity. Further, soon under the Umayyads’ reign, Arabic became the state official language of administration, commerce, and learning. Accordingly, those who spoke Arabic among the administrators, merchants, men of letters, and scientists “found themselves in an advantaged position, not only politically and socially but intellectually.”Footnote ³⁵ In this sense, the history of Arabic translation had its roots in the Umayyad period. Second, the lifting of the political and religious barriers helped equip the Abbasid court with a great number of scholars who were experts in their fields, as well as multilingual, so they were able to transmit their knowledge and translate key texts in their fields.Footnote ³⁶

The newly founded capital of al-Mansūr, Baghdad, became, together with older cities such as Jundayshāpūr and Ḥarrān, a center of learning. In these learning centers, translations of classical works of medicine, mathematics, astronomy, and natural philosophy flourished. That rise of translation activities was the result of a constellation of circumstances. In his book Greek Thought, Arabic Culture, Dimitri Gutas analyzes the Abbasid translation movement as a social phenomenon that was the result of its context. Above all, without the prosperity and luxury of the court and the interest of the caliphs themselves, there was doubt that translation could gain its momentum. Abbasid caliphs such as Hārūn al-Rashīd and al-Maʾmūn showed great interest in making manuscripts available for Arabic translators, and they were ready to endow bags of gold for finished translations. The translators themselves were content to translate for the caliph’s gold, but they also had clear research interest in their work.

Sometimes, the individual propensities of the caliph determined the kind of material to translate. For example, al-Ghassānī compiled a book for al-Mahdī on falconry because the caliph (al-Mahdī) liked falconry.Footnote ³⁷ Al-Ghassānī not only used Arabic sources but also incorporated foreign sources. That book, which became the archetype of Arabic literature on falconry, was mainly possible because of al-Mahdī, who liked falconry, and because of the translation culture that enabled Ghassānī to include both Arabic and foreign sources, since without that culture, he could have compiled only Arabic sources.

The translation movement was not supported by the court alone; it was supported by the entire elite Abbasid society. The funding was enormous, and it came from both public and private sources. This can be analyzed in light of the Islamic stress on learning (ʿilm) as one of the duties of all Muslims.Footnote ³⁸ This can also be analyzed in light of the needs of the Muslim community. Thus, according to Dimitri Gutas, the reasons for the rise of the Abbasid translation movement were different from those for the rise of translation activity of Aristotelian and other ancient texts from Greek and Arabic into Latin in western Europe in the twelfth century. In Europe, the rise of the new class of the bourgeoisie necessitated a new kind of knowledge that was different from the traditional church learning of the clergy. In the Abbasid society, the longevity of the translation movement can partially be explained by considering “the demand for applied knowledge in the rapidly evolving social climate of Baghdad and [of] the demand for theoretical knowledge by the scientific and philosophical tradition in the process of formation.”Footnote ³⁹

The movement started with an interest in the applied sciences of astrology, astronomy, and mathematics. This initial interest in the applied sciences eventually created a need for theoretical foundation, so the translation movement was expanded to include philosophical sciences. For example, Aristotle’s Topics was needed to teach Muslims debate, so they could defend their religion against Christians and Jews, and Aristotle’s Physics was needed to provide them with factual information to be used in theological debates. Thus, the translated works had the added value of immediate use and applicability, whether it was to medicine, philosophy, or astrology. Similarly, astrological history was translated and adopted to support the Abbasid revolution. Astrological history described cycles of power, such as emergence, dominance, and downfall, as governed by the stars and planets.Footnote ⁴⁰ The cycles of power and the renewal of knowledge were Zoroastrian beliefs. The Muslim translators and scholars Islamized those ideas so that the decrees of the stars were by the command of God, so any resistance to the Abbasid rise would be futile or against the very will of God.

4 Key Characteristics of Arabic Translation in the Medieval Period

As explained above, the early Arabic translations from Pahlavi and Syriac are significant, since they explain key features of the Arabic translation tradition. From the Pahlavi, Arabic translators took the idea that a translator can add new thought to a translation. Thus, when Ibn al-Muqaffaʿ translated Kalīla wa-Dimna from intermediary Pahlavi, he was inspired by translators before him who added their own original stories to the translation. From the Syriac, the Arabic translators learned to be mindful of faithfulness to source texts while integrating word-for-word and sense-for-sense approaches to translation and Christianizing their pagan sources. Therefore, before describing in depth the key features of the Arabic translation movement, we will examine a prominent Syriac translator before Islam, namely Sergius of Reshʿaynā (d. 536 CE).

Syriac was the first Semitic language into which Greek philosophy was translated. The key figure in the first significant translations from Greek into Syriac was Sergius of Reshʿaynā, a physician, theologian, and philosopher in the fifth and sixth centuries. Sergius studied in the school of Alexandria and became well known for his Syriac translations in the fields of theology, philosophy, and medicine, particularly the corpus of Pseudo-Dionysius the Areopagite. He also had a grand project in mind, as he expressed his intention to write about all parts of Aristotle’s philosophy. The project, however, was too ambitious, and only the first few books of the Organon were translated.Footnote ⁴¹ Sergius’ approach to translation is noteworthy, as he reconciled the “pagan” wisdom of the Greeks with Christianity,Footnote ⁴² and he did not introduce Greek philosophy in Syriac as dry texts but as a living system of problems.

Part of the significance of Sergius is that he established a tradition that was followed by other Christian translators and commentators on Aristotle, and he later played a significant role in the rise of Arabic philosophy.Footnote ⁴³ The school of Alexandria, where Sergius flourished, rather than any of the intellectual centers that existed at that time, such as Antioch, Edessa, Nisibis, Qenneshrīn, Jundayshāpūr, Marw, and Ḥarrān, was the main place for learning philosophy.Footnote ⁴⁴ Translators in Alexandria did not simply convey the meaning of the Greek texts but engaged critically with them. That pattern was adopted in Baghdad, which became an intellectual center where translation developed into “a systematic assimilation of Greek scientific and philosophical learning.”Footnote ⁴⁵

In two separate chapters in Interpreting the Bible and Aristotle in Late Antiquity, Adam McCollum and Emiliano Fiori indicate that Sergius of Reshʿaynā employed two different approaches in his translation of De Mundo and Dionysius.Footnote ⁴⁶ In De Mundo, he made changes and employed additions to clarify matters for his Syriac readers only when absolutely necessary, deleting at the same time nothing of substance. With Dionysius, Emiliano Fiori calls Sergius’ approach “dialectical fidelity,” since he practiced control over the text to prevent any of Dionysius’ Neoplatonic nuances from creeping into the biblical text. This analysis allows Josef Lössl and John Watt to conclude that “the activity of Sergius thus turns out to be considerably more coherent than might appear at first sight.”Footnote ⁴⁷ Consistency of approach becomes clearer when we focus on his purpose rather than his approach to translation. His purpose was to offer his Syriac readers an honest translation of the original text that was as consistent as possible with the Bible because the target text could be true only when it was consistent with the Bible.Footnote ⁴⁸

This kind of critical engagement with source texts reemerged with the Arabic translators, philosophers, and commentators. Sergius, in Alexandria, broke “from the Plato and Proclus of his pagan masters and linked Aristotle in the program with the Bible and Dionysius.”Footnote ⁴⁹ The process was reverted around four centuries later in Baghdad when al-Fārābī (870–950 CE), the renowned early Islamic philosopher, broke from the Bible and Dionysius and reconciled the Greek philosophy with Islam. Averroes, in the twelfth century, as Abid al-Jabri explains, purified Plato’s The Republic from unproven opinions (aqāwīl ghayr burhāniyya), so he ignored any reference to Greek mythology. He also did not include the first book and part of the second book because it was a debate without proof. Throughout the book, there were parts that Averroes ignored or rearranged so that the final product was presented in a completely new form. Jawāmiʿ Siyāsat Aflaṭūn (the Arabic version of Plato’s The Republic) was not a mere summary or explanation of Plato’s work. It was a text reconstructed, so it was void of any unscientific sayings. While Plato writes about war and its laws, Averroes discusses jihad and its ethics in Islam. Another example of critical engagement is how Averroes broke the gender bias against women, stating that they could be philosophers and judges. The idea itself was Plato’s, but Averroes did not ignore it like al-Fārābī had. Averroes carried out two complementary processes. First, he familiarized the reader with Plato’s ideas and, second, he aimed to contextualize these ideas (tabīʾa, to use al-Jabri’s term) to fit into the Arabic culture.Footnote ⁵⁰

What the masters learned from Aristotle was to subject any writing, including Aristotle himself, to a relentless critique. They rejected any authority except the authority of reason. Thus, they distinguished between philosophical and theological discourse, which came to be known as Averroism in the latter half of the thirteenth century. This distinction enabled the break with the clerical commentary tradition (i.e. from exposition to critique and the question form of commentary).Footnote ⁵¹ The distinction between philosophical and theological discourse can be interesting for translation studies. The translation of theological texts, particularly scriptures, focuses on the linguistic analysis of text, is concerned with philological problem-solving, and is stuck between word-for-word and sense-for-sense approaches. This kind of translation focuses solely on communicating the meaning of the source text. By contrast, the history of translation of science and philosophy attests to translation activities that go beyond the communication of meaning and looks at translation as a creative interaction and critical engagement with the text. In this sense, translation is a means of knowledge production rather than simply a means of communicating meaning. The concept of translation as a unified product that cannot be broken up unless one produces inaccuracies narrows our view of translation, turns it into a mechanical process, prioritizes philology over philosophy, and imprisons the translator’s reason behind language. It turns the translator into a conservative clergyman before a sacred text.

Sergius of Reshʿaynā is an interesting example as an early translator from the East for two main reasons. First, Sergius was not a translator by profession. He was a polymath whose knowledge spanned medicine, theology, and philosophy, and translation was part of his scientific project. This also applies to all of the Arabic translators of the medieval East who were philosophers, scientists, and writers. Second, in his translations, Sergius was not simply communicating meaning, but was critically engaging with the source text. By relating Aristotle to the Bible, for instance, Sergius used translation to carry out the social function of “maintaining” the target culture.Footnote ⁵² This use of translation to solve social problems, fill scientific gaps, or maintain target cultures would later dominate the Graeco-Arabic translation movement.Footnote ⁵³

5 Translation Competence

The previous sections introduced some of the most influential translators, philosophers, and scientists in the Graeco-Arabic translation movement in the Umayyad and Abbasid periods, such as Māsarjawayh, Ibrahim al-Fazārī, Abū Yaḥya ibn al-Baṭrīq, al-Ḥajjāj ibn Yūsuf ibn Maṭar, Ḥunayn ibn Isḥāq, Qusṭā ibn Lūqā, Thābit ibn Qurra, Abū Bishr Mattā, Yaḥya ibn ʿAdī, and ʿIsā ibn Isḥāq ibn Zurʿa. Although this list is by no means complete, it gives us an idea of the nature of translation and the skills of the translators. In the following subsections, I describe what is common among these translators and their approach to translation.

6 Toward a Definition of Arabic Translation in the Medieval Period

Given the huge span of time covered by the Arabic translation movement and the wide area of the Muslim Empire where translation was practiced, some scholars argue that there is no coherent body of theories to explain the whole Graeco-Arabic translation movement. This is because of the considerable variation in the practiceFootnote ¹¹⁶ and because the medieval intellectuals rarely discussed the work of the translators.Footnote ¹¹⁷ In spite of that, there is a common theme among all medieval Arabic translations that is almost absent from the literature, namely the use of Arabic translation as a means of knowledge production. Only when we start looking at medieval Arabic translations from this perspective can a coherent body of theories appear to connect all of them. As a means of knowledge production, the most important feature of translation is the critical and creative engagement with the source text.

Translation as creative engagement with the source text is supported by the original meaning of the Arabic word tarjam (to translate). Morphologically speaking, the term tarjam is either from the root rajm, meaning “to throw stones at”Footnote ¹¹⁸ or “to guess,”Footnote ¹¹⁹ or from the quadrilateral root tarjam, which has a few meanings, two of them relating to translation: “to explain speech in another language” and “to explain one’s own speech in the same language.” These definitions are significant because they indicate that translation is not used in Arabic to refer to saying the same thing in another language. According to these definitions, translation refers to explaining, elaborating, or clarifying. That attitude toward translation, namely as explanation, was the most dominant approach during the Graeco-Arabic translation movement. Thus, various nouns such as naql (transfer) and taʿrīb (Arabization), and verbs such as sharaḥ (to explain) and fassar (to interpret), are used by Arabic writers in the medieval period to refer to translation activities.Footnote ¹²⁰

The development of the meaning of “translation” from explanation, as used in the medieval period, to the rendition of equivalent meaning, as used in modern standard Arabic, points to the evolution of the Arab mind and its intellectual activities, since “in the birth of individual words and the stages of their growth could be discovered the stages of evolution of the human mind.”Footnote ¹²¹ Indeed, medieval Arabic translation involved critical and creative engagement, whereas contemporary Arabic translation has imprisoned the translator in producing an equivalent to the source text.

In the contemporary theory of translation, imitation and originality are restricted to a few problem-solving techniques. For example, translators can choose to imitate the source text, translate it word for word, or focus on the sense and convey the meaning in a different packaging. Similarly, translators can choose to either imitate the source text – thus producing a highly foreignizing translation – or domesticate it by adapting it to the target culture. Such procedures have little or nothing to do with the content, since contemporary translators are not theoretically supported to modify, reorganize, or supplement the content. In contrast, medieval Arabic translators used strategies that are no longer considered part of the process of translation, such as adding new information to the target text or presenting old information in a new form.

The Arabic medieval translation can also be analyzed according to an imitation–innovation model. The Arabic translators highly respected their source texts and took them as models to imitate. That is why epitomes and commentaries were the most common forms of scientific writings. However, the Arabic translators did not blindly imitate those texts. In many instances, they criticized, corrected, and improved their source texts. In general terms, the Arabic translators and the Muslim scholars after them attempted to save the Greek writers from any contradiction, correcting whatever they found inconsistent with the writer’s general ideas or approach. At the same time, their great respect for the Greek writers did not inhibit them from amending gaps and absurdities and adding new thought to the target text.

An important question to consider is whether adding original thought to translation coincides with medieval theory or violates its code. A quick survey of the criticisms that were addressed against translation indicates that translations were criticized for weak language or lack of clarity, but they were never criticized for adding fresh material to the source text. This means that the practice of improving and adding thought to the source text (i.e. creative engagement with the source text) was a common practice, and it was by no means aberrant to medieval Arabic translation standards.

7 A Model of Arabic Translation Analysis

A model for the analysis of medieval Arabic translation includes four processes: externalization, combination, internalization, and actualization. Externalization is the pouring of ideas into the world.Footnote ¹⁵¹ It applies to the production of knowledge that is contained in source texts produced prior to translation in another language and culture. However, it does not apply to translation as it is defined today, since it consists of old ideas expressed in a different language. The Greek texts that were translated into Arabic in the medieval age are instances of externalization, produced a long time before the Arabic translation movement.

Combination is the process of converting knowledge from the source language to the target language. This is the process through which groups and individuals combine knowledge, potentially creating new knowledge. I define translation, then, not as a process of knowledge transfer, but as a process of knowledge combination. In its simplest form, translation is a combination of two linguistic systems and two cultures. Linguistically speaking, when the ideas of the source text are expressed in another form (i.e. the target language), the translator does not produce a pure target text that is void of any influence of the language of the source text. There are usually traces of interference. However, even if this pure linguistic form is hypothetically achievable, we still have an instance of combination where ideas produced in another culture combine with a local linguistic system, thus being introduced to a new audience.

Combination may also occur at the level of content when the translator engages creatively with the source text. In the Arabic translation movement, we have more than one form of combination. For example, some Arabic translators, such as Ibrahim al-Fazārī, who translated Zīj al-Sindhind al-Kabīr, combine more than one source to produce their translations. When Ḥunayn ibn Isḥāq added a summary to some of his translations, he combined the full target text with his own perception of the more important information to include in a summary. Nevertheless, combination reaches a peak when translators add new thought to translation. In this case, they combine new knowledge, based on their expertise and environment, to knowledge that already exists in the source text. When Ibn al-Muqaffaʿ added fresh material to Kalīla wa-Dimna, he combined the Indian animal fables with his political project.

Since translation is both a process and a product, combination is observed in both. As a product, we can see new information integrated in the medieval Arabic translations, taking the form of glosses, commentaries, or text segments that were added. As a process, translation involves internalization and actualization. Internalization means the translator comprehends the material of the source text fully enough that he can present it in a different form, such as writing a summary or adding new thoughts to it. Thus, internalization is a process of learning through translating. After internalizing the information of the source text, the translator begins to put his newly acquired knowledge to use, produces new knowledge, and adds it to the target text. In this sense, learning through translation is not only about knowing but also about becoming. Medieval Arabic translation is not repetition, imitation, or even participation in some pursuit totally designed by another culture; it is fusion that has created a new self. It is this new self that makes Arabic translation a prominent tradition. Iṣṭifan ibn Basīl was a superior example in this context. Iṣṭifan ibn Basīl did not sit with pen and paper and translate the source text. Between analyzing the source text and synthesizing the target text, he went through internalization and actualization. He studied the source text, learned about the drugs, administered them, and, after that, Arabized them.

Translation as a form of combination does not fully explain the success of the Abbasid translation movement because it is a process that was practiced only by the elite – the translators and scientists. Another aspect, which I call “the spiral of influence,” explains the effect of translation on culture. This “spiral of influence” can explain how the medieval translation movement enriched the Arabic language. When a new term is introduced through translation, the masses who are actively engaged in learning begin to use it, so the term moves up the spiral until it becomes part of the Arabic language. Without the spiral of influence, the term remains in the narrow circles of the elite or it disappears. That is why translation needs a fertile learning environment to achieve its effect.

We have seen how translation in the Abbasid Golden Age introduced new ways of thinking and contributed to a more rational, critical, and creative frame of mind. Through a spiral of influence, the new way of thinking – along with the new terminology and tools of expression – involved more people from the public. This was facilitated by the fact that some translations were produced for the public, which accelerated the spiral of influence, speeding up the impact of translation on the masses. Once the circle of utilizing the products of translation was widened, new gaps began to appear. To close such gaps, more advanced knowledge in new translations or original compositions became a necessity, and so on and so forth.Footnote ¹⁵²

8 How Arabic Translation Changed

In the medieval period, translation played a pivotal role at least twice in the progress of scientific knowledge. The first was the translation movement of Baghdad, when the works of the Greek masters were translated from Greek and intermediary Syriac into Arabic. The second was when the scholars of Toledo, the intellectual center of the Latin West, translated Arabic works into Latin. The Toledo translation movement featured characteristics from three sources: the Romans, the patristic writings, and the Arabic translation movement of Baghdad. The Arabic influence did not continue for long, and it was noticeable in only some of the early translations. The patristic views constituted the stronger influence on all religious and scientific translations, whereas the Roman approach of the rhetoricians was clear in literary translation.

With the advent of the Reformation in the sixteenth century and the challenges it posed to Catholicism, the Church grew suspicious of the effect of translation as an instrument for reinterpreting sacred texts, thus undermining the Church’s authority as providing the “correct” interpretation. As a result, translation in general was discouraged, and greater formal fidelity and literalness was desired by churchmen to block interpretation, as well as by translators themselves to reduce any personal responsibility.Footnote ¹⁵³ At the same time, the rise of philology and the increased interest in classical Latin confirmed that literalism is the best choice to rediscover the legacy of classical Latin and Greek literature.

In the history of Latin and vernacular European translation, literary translation is generally distinguished from sacred and scientific translation. Literary translation is concerned with the creation of new literary forms through translation, hence the idea of translation as creation, whereas sacred and scientific translations are concerned with the transmission of intellectual information. This was further emphasized in the nineteenth century, when literalism was the more widespread technique for scientific translation. In contrast, literary translation, particularly the translation of English Gothic novels into French, was responding to the entertainment demands of the market and, as a result, translators creatively carried out various changes such as inserting a totally new title, deleting whole passages, or adding fresh material to their translations.Footnote ¹⁵⁴

In late nineteenth-century Europe, philosophical and scientific works were no longer imported to Europe through translation, as Europe had become the source of original writings in these fields. Instead of philosophical and scientific works, “modernist translators were attracted to literary novelties.”Footnote ¹⁵⁵ There was a demand for literary translations to be made available soon after publications in the original language.

A similar market for literary translation emerged in the Middle East in the late nineteenth century; however, in the early nineteenth century through to the second half of the nineteenth century, there was a demand for scientific translation. That demand was represented by a strong translation movement in Egypt that emerged as a part of Muhammad Ali’s modernizing project. As practiced within this translation movement, Arabic translation featured characteristics of both the Middle Age and the modern European translation practices. Like the Abbasid translators, the Arabic translators in Egypt engaged creatively with their source texts, but like the European translators, they gave priority to fluency and rhetorical style. Egypt, Syria, and Lebanon had a literary translation movement in the mid-nineteenth century, during which Arabic translation seemed to sever its ties with the Middle Ages and was fully influenced by the modern European translation theory and practice.