tarjama

Introduction

أَتَعْلَمُ مَلِكَ الحَبَشِ أَنَّ مُحَمَّدًا نَبِيٌّ كَمُوسَى وَالمَسِيحِ ابْنِ مَرْيَمِ أَتَى بِهُدًى مِثْلَ الَّذِي أَتَيَا بِهِ وَكُلٌّ بِأَمْرِ اللَّهِ يَهْدِي وَيَعْصِمِ وَإِنَّكُمُو تَتْلُونَهُ فِي كِتَابِكُمْ بِصِدْقِ حَدِيثٍ لَا بِصِدْقِ التَّرْجُمِ فَلَا تَجْعَلُوا لِلَّهِ نِدًّا وَأَسْلِمُوا وَإِنَّ طَرِيقَ الحَقِّ لَيْسَ بِمُظْلِمِ^{1أبو طالب، ديوان أبي طالب، ‎90‎.}

Do you know, O King of Abyssinia, that Muḥammad
    Is a prophet like Moses and Jesus son of Mary?
He came with guidance like what they brought
    And each, by God’s command, guides and protects
You recite it in your book
    With literal recitation, not with true interpretation
So do not set up equals to God, and submit
    For the path of truth is not dark.²Abū Ṭālib, Dīwān Abī Ṭālib, 90.

The term tarjama has roots outside Arabic. Contemporary scholarship traces its etymology to the Aramaic targum, meaning “translation” or “interpretation.”³Tageldin, “Negative Translation,” 163. Arab lexicographers debated its etymology: some classified it as a borrowed four‑letter root (t‑r‑j‑m), while others attempted to naturalize it as a three‑letter root (r‑j‑m).⁴For a detailed discussion of the etymology and lexicographic debate, see Tageldin, “Negative Translation”, 190. See also Zabīdī and Majmūʿa min al-Muḥaqqiqīn, Tāj, 367–368. The earliest textual evidence, however, supports a foreign origin. Addressing the King of the Abyssinians, Abū Ṭālib ʿAbd Manāf (d. c. 619 CE) employs the word al‑tarjim/al‑tarjum in the sense of textual elucidation and interpretive understanding of scripture.⁵The reference is to a theological argument, explicitly stated in the Qurʾan (61:6) and endorsed here by Abū Ṭālib, which holds that the Bible foretells the coming of a prophet matching Muḥammad’s description. Abū Ṭālib explicitly contrasts mere recitation, which implies superficial understanding, with true comprehension. This use prefigures the later Abbasid association of tarjama with interlingual mediation and overlaps with another Aramaic shade of meaning: “textual interpretation, explanation.”⁶Kaufman et al., Aramaic Lexicon.

While tarjama predominantly signified interlingual translation throughout Arabic literary history—a connotation retained today—early evidence indicates a trend toward semantic narrowing. The borrowed term moved from describing a general, often oral, process of interlingual and intralingual mediation and interpretive understanding to a material, literary, and intellectual practice with a more focused conceptual scope revolving around translation.⁷A tarjama also signified a “biographical notice,” attested by the 5th/11th century and later (Gutas et al., “Tard̲j̲ama,” n.p). Scholarship notes the relative silence in the “lexical record” despite its abundance in biographical dictionaries (Tageldin, “Negative Translation,” 166).

The development from interpretive understanding to systematic translation practice can be traced through intermediary textual evidence. References to tarjama before Abbasid rule are scarce and, when found, largely pertain to diplomatic and administrative contexts.⁸Vagelpohl and Sánchez, “Why Do We Translate?” 257. In early texts, tarjama also appears alongside other terms such as naql (lit.: transfer/transmission) and tafsīr (lit.: exegesis), which were likewise deployed to signify translation.⁹Gutas et al., “Tard̲j̲ama.” One significant anecdote in Sulaym b. Qays al‑Hilālī’s (d. c. 76/695–96) Kitāb Sulaym, described by Ibn al‑Nadīm (d. 380/990) in al‑Fihrist as “the first book to appear for the Shīʿa,”¹⁰Ibn al-Nadīm, Fihrist, 69. illustrates a transitional moment where tarjama encompasses both interpretive, social, and material dimensions.

ʿAlī b. Abī Tālib (d. 40/661) instructs a companion to secure a turjumān (an interpreter) who understands the other party’s language and to have a prophetic text he relayed in the course of the conversation translated into Arabic, framing the request in material terms: he orders that a copy of the text be made by the turjumān in Arabic—fa-l-yansakhhu laka bi-l-ʿarabiyya—in interpreted form (mufassaran). After the copy is brought, ʿAlī furnishes a written exemplar of the prophecy in his own hand, has his son al‑Ḥasan (d. 50/670) read while another compares the new Arabic copy, and they find “not a single letter differed, with no additions or omissions, as though one man dictated to two.”¹¹Al-Hilālī, Kitāb Sulaym, 255. This early account moves tarjama beyond oral mediation, as a turjumān (interpreter) typically provides, to a written practice producing artifacts: a faithful Arabic copy (nuskha), rendered in translation (mufassaran).¹²While the anecdote employs collation as a narrative device to authenticate a doctrinal claim, it captures a standard of fidelity and comparison that later Abbasid translators would codify in technical terms.

ʿAlī’s impromptu commissioning directive in the anecdote relayed here frames a request for a faithful translation in written form. It implies a client-translator relationship that encompasses most of the aspects later patron-translator relationships embody: need-driven commissioning of work, remuneration, and insistence on faithfulness and quality. While the circumstances of this early episode align with Gutas’s characterization of Umayyad and early translation activities as “random and ad hoc,”¹³Gutas et al., “Tard̲j̲ama.” contrasting them to later systematization in Baghdad, concerns seem to be shared among the agents active in either time period. Only the scale and motivations were different in the Abbasid period.

Arabic translations of Greek scientific and philosophical works began around the mid-eighth century and continued to the end of the tenth; within this span, the movement reached its high point between 850 and 950.¹⁴Gutas, Greek Thought, Arabic Culture, 1, 124. The systematization distinguished Baghdad’s Arabic translation movement through its unprecedented duration and momentum, diverse patronage, large and sustained funding, and “scholarly methodology and philological exactitude,”¹⁵Gutas et al., “Tard̲j̲ama,” n.p. which was fostered under patronal scrutiny.¹⁶Gutas, Greek Thought, Arabic Culture, 138–39. Ḥunayn b. Isḥāq (d. 260/873), a prominent figure of the Abbasid translation movement, lays out these dimensions in the Risāla, addressed to ʿAlī b. Yaḥyā (d. 275/888–89). The extant Arabic text of the Risāla, translated from Syriac, systematically documents a translation movement that was already well established. Long before Abbasid Baghdad, Greek knowledge was translated and transmitted in Syriac Christian milieus and, at times, under Sasanian patronage via Middle Persian or Pahlavi.¹⁷For an overview of the Syriac translation tradition and its role in transmitting Greek learning eastward, including into Sasanian Iran, see Gignoux, “Syriac Translators” and Russell, “Greek Medicine in Persia”. In ninth- and tenth-century Baghdad, translators came overwhelmingly, though not exclusively, from Syriac-speaking Christian communities.¹⁸Gutas, Greek Thought, Arabic Culture, 3. By Ḥunayn’s time, the Abbasid translation enterprise in Baghdad had moved far beyond earlier ad hoc activities into a more self-conscious and methodologically articulated practice.

Though narrow in scope, providing detailed information about the circumstances of producing translations of Galen’s (d. c. 216 CE) works and a few other Greek scholars, such as Hippocrates (d. c. 375 BCE), Plato (d. 348/347 BCE), and Aristotle (d. 322 BCE) insofar as Ḥunayn’s knowledge went, the Risāla covers valuable dimensions: the quality and availability of “source text” manuscripts, and the methodological and stylistic approaches employed in the translation process.¹⁹Lamoreaux, “Introduction,” xv. In this article, I focus on patronage and translation methodology as documented in the Risāla. I argue that the Risāla outlines an iterative, cross-lingual, patron‑oriented methodology of translation. It provides a portrait of tarjama as a methodologically self‑conscious and self-evaluative enterprise, that is to say, aware of limitations stemming from material conditions, able to recognize them, and active in remedying them: by the time of the Risāla’s composition in Arabic, tarjama had developed its own evaluative language, quality standards, professional hierarchies, all readily discernible in the Risāla. This clearly marks the transformation from improvised linguistic mediation to an institutionalized intellectual enterprise, insofar as the medical/scientific field was concerned. Though al-Jāḥiẓ’s (d. 255/869) prescriptive remarks on tarjama in kitāb al-Ḥayawān suggest that methodological rigor was expected across domains, the Risāla is our clearest descriptive witness for how such rigor was effected in scientific/medical translation and what it consisted of in practice.

Corpus

The Risāla constitutes the earliest systematic documentation of tarjama as professional practice. Written in the ninth century, the original Syriac version is lost.²⁰The two extant recensions derive from a revised Arabic version. Recension B (Ayasofya 3590) may be closer to Ḥunayn’s evolving text, while Recension A (Ayasofya 3631) shows more substantial stylistic corrections and additions (Lamoreaux, “Introduction,” xx, xxii). Lamoreaux confirms Gotthelf Bergsträsser’s conclusion that both recensions were updated after Ḥunayn’s death in 873, likely by students, and notes a higher frequency of editorial intervention in A compared to B (Lamoreaux, “Introduction,” xxiii). The surviving Arabic text was translated from another Syriac text in 241–2AH/855–6 CE, which is also lost, and revised in 250 AH/ 864 CE.²¹Lamoreaux, “Introduction,” xxii. The extant Arabic text was first critically edited by Bergsträsser in 1925 based on the so-called A Recension; Bergsträsser later published corrections and an analysis of another recension (B) in 1932.²²Lamoreaux, “Introduction,” xix–xx. Bergsträsser’s 1925 edition was reprinted by ʿAbd al‑Raḥmān al‑Badawī in 1981 and by Mahdi Muḥaqqiq in 2001, accompanied by a Persian translation.²³Lamoreaux, “Introduction”, xix, n. 34. For more on the publication history of the Risāla, see Lamoreaux, “Introduction”, xviii–xxi. For example, a third witness (recension C), preserved in MS Ayasofya 3593, is a heavily abridged version more closely aligned with the B recension; an edition with German translation has been undertaken by Fabian Käs (Lamoreaux, “Introduction,” xxi).

The Risāla encodes the range of practices involved in the translation movement through a set of forms derived from the verb tarjama, such as turjima, tarjamta-hu, yutarjamu, used to name the act, its sources, and its products, as well as the agents of translation (al-tarājim).²⁴Ḥunain b. Isḥāq, On Galen Translations, 4. The distribution of forms in the Risāla points to the expanded conceptual scope of tarjama in this period, making this text a key witness to the term’s consolidation as a conceptual umbrella for the translation movement within Abbasid translation discourse.

Historical Semantics and Contextual Analysis

Patronage and Network

Translations in the Abbasid period were commissioned by court‑affiliated Christian and Muslim patrons and by practicing physicians, notably Salmawayh b. Bunān (d. 226/840–41), Yūḥannā b. Māsawayh (d. 243/857), and Bakhtīshūʿ b. Jibrīl (d. 256/870).²⁵Lamoreaux, “Introduction,” xv–xvi. The translation movement was to a great extent a patron–client‑driven practice in which specific requests directly initiated projects. For instance, Ḥunayn retranslated Galen’s Pinax—previously rendered into Syriac by Ayyūb al‑Ruhāwī (d. c. 220/835)—into Syriac at the request of Dāwūd al‑Muṭaṭabbib (fl. late second to early third century AH/late eighth to early ninth century CE) and into Arabic at the request of Abū Jaʿfar b. Muḥammad b. Mūsā (d. 259/872).²⁶Ḥunain b. Isḥāq, On Galen Translations, 7, 9. That the Pinax underwent two different translations in Syriac at the request of various patrons suggests that translations held the status of proprietary intellectual goods, not uniformly disseminated nor identical in form to the “source text.” The Risāla itself is another reactive project: before composing the work in Syriac for “a certain man, a speaker of Syriac” (rajulun min al‑suryāniyyīn),²⁷Ḥunain b. Isḥāq, On Galen Translations, 2–3. whose identity remains unknown, Ḥunayn had sought to direct him to Galen’s bibliographic record, the Pinax, arguing that it is “better to seek to understand the books of Galen from Galen rather than from me.”²⁸Ḥunain b. Isḥāq, On Galen Translations, 4. The patron, however, replied with a more specific demand:

فكان من جوابك في ذلك أن قلت أنه وإن كان الأمر على ما وصفت فإنّ بنا وسائر أهل هذا الغرض ممن يقرأ الكتب بالسريانية والعربية حاجة إلى أن نعلم ما تُرجم من هذه الكتب إلى اللسان السرياني والعربي وما لم يُترجم، وما كنت أنا المتولي لترجمته دون غيري، وما تولى ترجمته غيري، وما سبقني إلى ترجمته غيري، ثمّ عدت فيه فترجمته أو أصلحته، ومن تولى ترجمة كتاب كتاب من الكتب التي تولى ترجمتها غيري ومبلغ قوة كل واحد من أولئك المترجمين في الترجمة ولمن ترجمت، ومن الذين ترجمت أنا لهم كل واحد من تلك الكتب التي توليت ترجمتها؟ وفي أي حد من سني ترجمته؟ لأن هذين أمرين قد يحتاج إلى معرفتهما إذ كانت الترجمة إنّما تكون بحسب قوّة المترجم للكتاب والذي تُرجم له، وأي تلك الكتب مما لم يُترجم إلى هذه الغاية وُجدت نسخته باليونانية؟ وأيّها لم توجد له نسخة أو وُجد البعض منه؟ فإن هذا أمر يحتاج إليه ليعنى بترجمة ما قد وُجد منها ويطلب ما لم يوجد.

You answered as follows: “Even if the matter is as you describe, we and the rest of those concerned with this subject who read books in Syriac and Arabic need to know which of these books have been translated into Syriac and Arabic and which have not been translated; which of them I alone undertook to translate, which others translated, and which others translated before me and to which I later returned, whether to translate or to correct them; who undertook the translation of each of the books not translated by me, the extent of each of those translators’ competence in translation, for whom they were translated; and for whom I myself translated each of the books I undertook to translate, and at what age I translated them—for these are two matters whose knowledge may be needed, since translation depends on the ability of the book’s translator and of the one for whom it was translated; and which of those hitherto untranslated books has had a Greek copy discovered, and for which no copy or only a partial copy has been found. For this is a necessary matter, so that attention may be given to translating what has been found, and what has not been found may be sought.”²⁹B. Isḥāq al-ʿIbādī, Risālat Ḥunayn, 2–3.

The patron clarifies that his interest is not in the breadth of Galen’s oeuvre per se, but in accessible, contextualized knowledge of the state of translation and the network of translators and patrons responsible for disseminating Greek knowledge—here, Galen’s medical treatises—into Syriac and Arabic.³⁰Ḥunain b. Isḥāq, On Galen Translations, 5. The Risāla does exactly that, highlighting a social multi‑node translation ecosystem, including patrons (often scholars), translators, and students. The anonymous patron is particularly attentive to quality concerns, requesting that Ḥunayn assess his peers and enumerate the revision history of works he corrected or retranslated. He also asks Ḥunayn to identify translation gaps: works for which Greek manuscripts exist but remain untranslated, and works whose Greek copies must be sought so that they may be translated.³¹On the ideological use of Greek book procurement within Abbasid courtly narratives, see Vagelpohl and Sánchez, “Why Do We Translate?”, 5 and Gutas, Greek Thought, Arabic Culture, 28–60. Ḥunayn meets this demand, proceeding to document the translation output of his time and the entire enterprise of the translation movement known immediately to him: its translators, its patrons, the quality of the output, and the material gaps.

Collation and Retranslation

To meet the expectations of patrons, who offered significant remuneration for translation work,³²Gutas, Greek Thought, Arabic Culture, 138. the translator’s role came to require unprecedented linguistic, intellectual, and methodological rigor.³³For adab’s role in shaping the intellectual economy of the Abbasid court and the social capital of books and translation, see al-Rahim, “Arabic Literary Prose and Adab”; cf. Montgomery, In Praise of Books. Ḥunayn b. Isḥāq’s career exemplifies these demands and can be read as a response to them, an effort to preempt criticism by seeking specialized knowledge. Bilingual in Syriac and Arabic, Ḥunayn trained in medicine with Yūḥannā b. Māsawayh (d. 243/857) in Baghdad until a falling‑out prompted his departure.³⁴Lamoreaux, “Introduction,” xiii. He later returned having acquired intimate knowledge of the Greek language and culture,³⁵Gutas, Greek Thought, Arabic Culture, 139–40. resumed study with Ibn Māsawayh, soon “established himself as a master of the ancient works” and acquired a reputation as a renowned translator.³⁶Lamoreaux, “Introduction,” xii–xiii. In this sense, Ḥunayn’s training directly informed his method. It also exemplifies broader pressures on translators at the time.

Aware of this competitive nature of his enterprise, Ḥunayn displayed specialized, intimate knowledge of Galen’s work, partly to distinguish himself from other translators, yet mainly to satisfy the requirements of tarjama as an interpretive and explicatory act. He reports appending a section to his translation of the Pinax, clarifying that Galen had omitted mention of some works, listing many, and explaining the reasons for the omissions.³⁷Ḥunain b. Isḥāq, On Galen Translations, 8–9.

The Greek-Arabic translation movement focused predominantly on non-literary, non-historical writings.³⁸Gutas, Greek Thought, Arabic Culture, 1. Within this corpus, which included arithmetic, geometry, physics, botany, philosophy, and medicine, among other subjects, tarjama evolved through sustained engagement with disciplinary challenges. As Vagelpohl and Sánchez have already shown, medical translation demanded terminological precision to avoid lethal errors; philosophy required sophisticated conceptual mapping, as evidenced by al‑Fārābī’s (d. 339/950) extensive treatment of Aristotelian terminology; and religious works necessitated safeguards against both accidental misinterpretation and deliberate heretical interpolation, further conditioned by the doctrine of the Qurʾan’s inimitability.³⁹Vagelpohl and Sánchez, “Why Do We Translate?” 258–260. Arabic sources also distinguish between word-for-word and sense-for-sense translation and debate whether to transliterate Greek technical terms or coin Arabic neologisms.⁴⁰Vagelpohl and Sánchez, “Why Do We Translate?” 318–319, 330–331. Translators generally achieved distinction through philological rigor, and to attain that distinction one needed to meet strict prerequisites, which al‑Jāḥiẓ articulated in al-Ḥayawān.

Al‑Jāḥiẓ insists that translators possess deep subject-matter competence and be equally fluent in source and target languages (أن يكون أعلم الناس باللغة المنقولة واللغة المنقول إليها).⁴¹Jāḥiẓ, Ḥayawān, 76. Al-Jāḥiẓ adds that “the more difficult (aʿsar) and specialized (aḍyaq) the field, and the fewer its experts (wa-l-ʿulamāʾ bihi aqall), the more challenging (ashadd) it is for the translator, and the more likely (ajdar) the translator is to err.”⁴²Jāḥiẓ, Ḥayawān, 77. He turns to religious texts to fully articulate his portrait of the ideal translator (he employs the words kamāl/yakmul to signify perfection).⁴³Jāḥiẓ, Ḥayawān, 77–9. “When a translator does not fully meet [the conditions of] perfection,” al‑Jāḥiẓ concludes, “his errors are proportionate to his shortfall from perfection.”⁴⁴Jāḥiẓ, Ḥayawān, 78. Consequently, he continues, such shortcomings are made worse by the very material practices associated with translation: errors introduced in translation propagate through copying, are compounded by scribal mistakes, and gradually distance the text from its intended meaning (maʿnā). In some cases, he observes, rewriting may be easier than repairing an error‑ridden text.⁴⁵Jāḥiẓ, Ḥayawān, 78–9. Al‑Jāḥiẓ is generally pessimistic, treating the profile as practically unattainable, never to be found.⁴⁶Jāḥiẓ, Ḥayawān, 77. To this ideal, however, Ḥunayn’s profile comes close, especially in the medical field. Evidence for this may be sought in his methodology.

Ḥunayn records several editorial patterns he engaged in: revision, correction, quality‑driven retranslation, and patron‑directed assessment. His translations were refined iteratively. Ḥunayn’s translation of Sects, for instance, went through several rounds of correction, revision, and eventual retranslation, distributed across different stages of his life:

وقد كان ترجمه قبلي إلى السرياني رجل يقال له ابن شهدا من أهل الكرخ وكان ضعيفًا في الترجمة، ثمّ أنّي ترجمته وأنا حدث من أبناء عشرين سنة أو أكثر قليلًا لمتطبّب من أهل جندى سابور يقال له شيريشوع بن قطرب من نسخة يونانيّة كثيرة الأسقاط، ثمّ سألني بعد ذلك وأنا من أبناء أربعين سنة أو نحوها حبيش تلميذي إصلاحه بعد أن كانت قد اجتمعت له عندي عدّة نسخ يونانيّة قابلت تلك بعضها ببعض حتّى صحّت منها نسخة واحدة، ثمّ قابلت بتلك النسخة السرياني وصحّحته، وكذلك من عادتي أن أفعل في جميع ما أُترجمه. ثمّ ترجمته من بعد سنيّات إلى العربية لأبي جعفر محمّد بن موسى.

(Galen’s On Sects) was translated into Syriac before me by a certain Ibn Šahdā from al-Karḫ, who was a weak translator. When I was a young man of twenty or a little older, I translated it for a physician from Ǧondēšāpūr named Šīrīšūʿ b. Quṭrub, from a very faulty Greek manuscript. Later, when I was about forty, my pupil Ḥubayš asked me to correct the translation. Meanwhile, a number of Greek manuscripts had accumulated in my possession. I collated these manuscripts and thereby produced a single correct copy. Next, I collated the Syriac text with it and corrected it. I am in the habit of doing this with everything I translate. A few years later, I translated the Syriac text into Arabic for Abū Ǧaʿfar Muḥammad b. Mūsā.⁴⁷B. Isḥāq al-ʿIbādī, Risālat Ḥunayn, 5; Vagelpohl and Sánchez, “Why Do We Translate?” 314–315.

Ḥunayn’s account underscores a sustained commitment to accuracy through iterative refinement and active manuscript curation: following an initial translation, Ḥunayn would later return to the text as better evidence became available, collating Greek witnesses until a sound Greek base had been established and then using that stabilized exemplar to emend earlier translations. A similar pattern, which may be classified as patron-directed assessment and improvement, emerges in his collaboration with Salmawayh b. Bunān as they worked to correct Sergius of Rēšʿaynā’s (d. 536) Syriac translation of Galen’s Methodus medendi, a project that ultimately led to a new translation.⁴⁸See Lamoreaux’s Introduction to Ḥunayn’s Risāla for a discussion regarding the active role some of Ḥunayn’s patrons played in producing the translations, including manuscript procurement, collation, and stylistic revisions: Lamoreaux, “Introduction”, xvi–xvii. The report involving Salmawayh is also selected and briefly discussed in Vagelpohl and Sánchez, “Why Do We Translate?”, 15–16.

It is crucial here to emphasize the horizontal nature of the refinement process (i.e., multilingual) as opposed to a vertical (i.e., concerned with one language). Ḥunayn moves between Greek, Syriac, and later Arabic, emending the Syriac version and afterward producing an Arabic translation when a new patron and context demand it. As this horizontal movement unfolds over a stretch of years, source and target texts are treated as provisional: Greek witnesses are improved through ongoing collation, and Syriac translations are rectified or reworked in light of better evidence, while Arabic translations are produced later when the source is considered textually stable—at least in this particular case. Any given text is, therefore, only provisionally finished. Such iterative correction was not unique to Ḥunayn, however: in the Arabic tradition, scientific works were often retranslated and revised several times as part of their transmission history.⁴⁹Ibn al‑Nadīm’s Fihrist presents Euclid’s (fl. c. 300 BCE) Elements (Stoicheia), for instance, as a text repeatedly translated, revised, and corrected in Arabic. A comparable translation history is attested for Dioscorides’ (d. c. 90 CE) De materia medica (On Medical Material). See Vagelpohl and Sánchez, “Why Do We Translate?”, 270–271, 278

The passage also confirms the social economy of translation as proprietary, commissioned labor: what is translated, when, and for whom is shaped by patronage so that retranslation functions both as scholarly emendation and as the production of a new, owned textual commodity. This model shows, more importantly, that what counts as a “correct” (ṣaḥīḥa) translation is materially contingent. In other words, it was largely dependent on access to witnesses and on the scholarly networks that secured them, exemplified here by Ḥunayn’s network. In these conditions, the translator’s authority, we may conclude, indeed rested on curatorial labor as much as it did on linguistic competence and the ability to negotiate between languages and patrons.

Conclusion

Tarjama’s semantic field spans interpretation and translation; its historical trajectory runs from interpretive mediation with collateral material operations, as demonstrated by the Kitāb Sulaym anecdote of ʿAlī commissioning a turjumān and ordering an Arabic nuskha, to a methodologically rigorous, patron‑driven, and professionally stratified and competitive practice. Such competition, as Gutas has shown, fed on “the incentives provided by the munificence of… sponsors, a munificence which in turn was due to the prestige that Baghdad society attached to the translated works and the knowledge of their contents.”⁵⁰Gutas et al., “Tard̲j̲ama,” n.p. These conditions shaped the agendas of translators and patrons alike and initiated comprehensive bibliographic surveys, such as the Risāla.

This work offers a snapshot depicting translation as a competitive market of expertise, which had acquired evaluative overtones as translators and patrons sought to classify who translated well or poorly and assessed translated works. In this context, quality was a function of competence, a point reflected in al‑Jāḥiẓ’s ideal profile of the translator and echoed in the methodological choices of translators themselves. Translations produced by Ḥunayn b. Isḥāq and his associates were widely recognized for their quality and were often contrasted with those produced by the circle of the philosopher al‑Kindī (d. 256/870).⁵¹Adamson, “Al-Kindī and Greek Philosophy”, 49 n. 3; for an in-depth look at the translation methodology of Ḥunayn b. Isḥāq, including his innovative techniques and the sociolinguistic context of his work, see Osman, “Sheikh of Translators”. Compared to those produced in al‑Kindī’s circle, Ḥunayn’s translations were consistently more methodical and terminologically stable, whereas some of al‑Kindī’s translations took the form of paraphrase, exhibited structural rearrangements vis‑à‑vis their sources, and included insertions by the translators themselves (Adamson, “Al-Kindī and Greek Philosophy,” 32–33). For a detailed analysis of the methodologies of translation during the Middle Ages, particularly the shift from literal to sense-for-sense (ad sensum) translation and the system of intra-translator corrections, see Osman, “Translation and Interpreting in Arabic”. Recording standards of fidelity, revision, and retranslation within a patronage economy that commissioned and evaluated work for intelligibility and exactness, the Risāla illuminates how Abbasid Baghdad transformed translation into an intellectual discipline embedded in institutional economic structures. Crucially, however, Ḥunayn’s iterative, multilingual collation and patron‑oriented retranslations highlight quality as contingent on the availability of witnesses, alongside the networks that sought and collected them, including the translators themselves.

The meticulousness of Ḥunayn’s method, particularly with regard to collating source text manuscripts and updating the translations in light of new evidence, leads to an important reflection, one that can be summarized in the following questions: Where Greek exemplars were scarce or corrupt and could not be fully stabilized, at least to the standards Ḥunayn repeatedly describes, to what extent did knowledge transferred into Arabic remain fragmentary and epistemically uneven (this is not to imply inferiority) compared to its Greek source counterparts? And when later collations improved individual source texts, for how long did that unevenness last?

By the third/ninth century tarjama already stood as a mature discipline with methodological concerns and competing centers of authority (for example, Ḥunayn vs. al-Kindī). Moreover, the circle of reception undoubtedly widened through the technological advances of the book revolution, with paper becoming “the catalyst for [this] revolution” and the driver of its momentum.⁵²Gruendler, Arabic Book, 12. Crucially, however, the practice of tarjama was inextricably linked to the material condition of its Greek and Syriac Vorlagen.

The Risāla itself attests to this, It stands as a work that evaluates both source and target language texts, illustrating that the effort of patrons and practitioners was distributed between translation and verification/canonization of source material. Conceptualized through the term tarjama from an early stage, translation in this context was as much concerned with transferring meaning as it was with anchoring itself to a text to be conveyed in a different medium. This echoes the term’s Aramaic origins, where translation was inseparable from making a text clear, which is a deliberate act of explanation,⁵³Kaufman et al., Aramaic Lexicon. not mere lexical substitution, or naql. Indeed, by the tenth century, Abū Ḥayyān al-Tawḥidī (d. 414/1023) employs naql and tarjama in different senses: in evaluating Ibn Zurʿa (d. 398/1008), he describes him as ḥasan al-tarjama, ṣaḥīḥ al-naql (good at translation, sound of transmission),⁵⁴Al-Tawḥīdī, al-Imtāʿ, 33. assigning the interpretive quality of ḥusn to tarjama and reserving technical standard of correctness (ṣiḥḥa) to naql. Further research is needed to identify the extent to which these terms imply specific methodological rigor and which scholarly profile either term entailed.