Translation as philosophical creation

A word can change a question

Imagine translating a passage about why things happen. Every time it mentions a cause, you write “the person responsible.” A stone falling from a cliff now reads like someone’s deliberate act. Your sentences are easy to read, but you have changed the question the passage asks.

The case exaggerates, but the difficulty is real: a translator has to understand an argument well enough to carry its distinctions into another language. Remember Aristotle’s different kinds of explanation: knowing what something is made of does not answer every question about why it exists. The words a translator chooses can shape which questions readers learn to ask.

A shared language of inquiry

In ninth-century Baghdad, scholars translated Greek learning into Arabic, sometimes by way of Syriac. Syriac had an established Christian scholarly tradition, and its scholars had already studied and interpreted Greek works. Arabic became a shared language of inquiry for people of different religious commitments. [4]

Falsafa, an Arabic word derived from the Greek word for philosophy, names a tradition shaped by this encounter. Translators, patrons and philosophers did different kinds of work within it, and two figures make those differences visible: the Christian physician and translator Hunayn ibn Ishaq and the Muslim philosopher al-Kindi, whose activities belonged to distinct circles within a wider intellectual world. [3]

Hunayn revisits a translation

In his Epistle, Hunayn describes his work on Galen’s On Sects, a medical text. He reports that he first translated it from Greek into Syriac when he was about twenty, working from a defective manuscript. Some twenty years later his pupil Hubaysh asked him to revise it. By then Hunayn had several Greek copies, so he compared them, established a corrected Greek text and used it to correct his Syriac translation. Later still he translated the Syriac into Arabic for the patron Muhammad ibn Musa. [1]

Comparing manuscripts to find where they differ is called collation. For Hunayn, translating involved deciding what the source text should say before deciding how to render it, so a better translation could begin with a better source text. The book that reached its new readers carried the results of that earlier judgment and revision.

What would prove a connection?

Hunayn names a work, languages, people and a sequence of revisions, so his testimony supports a specific account of transmission. It does not show that every Arabic translation passed through Syriac, and it cannot independently confirm everything he says about his own skill.

To test such an account further, scholars compare surviving texts, vocabulary, distinctive errors, dates and reports of who had which books, and a proposed connection grows more convincing as these clues agree. [4] Similarity alone is weaker evidence: two thinkers might ask the same question independently, or both might draw on a third source.

Even a proven connection leaves a further question open. Intellectual reception concerns what readers do with inherited work: which parts they select, how they understand them and what new use they make of them. Establishing that a text arrived does not settle how it was received.

Translating, explaining and extending an argument

Translation renders a text in another language. Commentary explains or examines it. Original argument develops or adapts reasons for a conclusion, going beyond a report of what an earlier writer said. These distinctions between activities stay useful even when a single passage combines them.

Hunayn’s practice shows why the boundaries need care. As Uwe Vagelpohl explains, a translator might clarify a term or supply information that new readers need, so we have to compare the texts before deciding whether a given sentence translates, explains or adds a claim. [5]

A name on a book can reshape reception too. An Arabic adaptation of parts of Plotinus circulated as the Theology of Aristotle. Remember Plotinus’s search for a first principle, the source on which everything else depends: material associated with that search could now be read under another philosopher’s name, and the title alone cannot tell us who composed its arguments. [4]

Al-Kindi asks what philosophy is for

Al-Kindi wrote philosophical works in Arabic while working with a circle of translators. In the opening chapter of On First Philosophy, he presents learning as cumulative, since no single person could discover everything unaided, and he holds that truth can come from people distant from us in place or community. Learning from earlier thinkers does not require believing that they got everything right. [2]

He also presents the study of truth, including inquiry into God, as compatible with prophetic teaching. That is his philosophical position within a Muslim setting; it does not speak for every Muslim or erase disagreement about what philosophy can establish. [3]

Here the inherited material serves a pressing question: can anyone give a reason for rejecting philosophical inquiry without practicing it?

Can reason argue against itself?

Al-Kindi argues that an opponent who says philosophy is unnecessary must explain why, and that giving the explanation already involves the very inquiry the opponent rejects. The critic ends up depending on the practice they are trying to dismiss. [2]

The argument has a serious limit. Someone can accept the need to give reasons and still reject a philosopher’s account of God or nature. Al-Kindi’s argument is strongest against a wholesale rejection of reasoned inquiry. It does not prove every conclusion reached in its name.

Nor did he invent this defense from nothing. His modern translator, Alfred Ivry, traces it to an older Aristotelian argument, which al-Kindi puts to work defending philosophical inquiry before its critics. Translation makes arguments available; philosophical judgment decides what to do with them.

Try the distinction

Leila translates an old sentence as “Learning requires time.” In a note she explains that “time” here means sustained practice, not merely growing older. Then she argues that a school should give students fewer tasks and more time to revise them.

Which part translates, which comments and which makes a further argument? And what more would you need before claiming that another writer’s similar proposal came from Leila?

Read a possible response

The sentence is presented as a translation, and checking its accuracy would require the source. The note interprets a term, so it is commentary. The school proposal is a further argument that needs reasons: sustained practice might support revision, but it does not by itself prove that fewer tasks are best. As for the later writer, a similar proposal would not establish transmission; evidence that they read Leila, together with distinctive shared wording or reasoning, could strengthen the case. The exercise asks what each claim requires, without asking you to endorse either proposal.

Carry this forward

Translation can preserve a distinction, repair a text or carry an old argument into a new dispute, and its philosophical importance comes into focus when we can say exactly what changed and what evidence shows the connection. In What would a necessary being be?, Avicenna distinguishes what something is from whether it exists. We will see how he uses that distinction to argue about dependence, and where an objector could challenge the reasoning.

Sources and limits

[1] Hunayn ibn Ishaq, Epistle on Galen’s works, selected English passages in Uwe Vagelpohl and Ignacio Sánchez, “Why Do We Translate? Arabic Sources on Translation”, in Dimitri Gutas (ed.), Why Translate Science?, Brill, 2022, for Hunayn’s account of his own translating. Both selections were read with their introductions and displayed notes, together with Part One’s essay:

• The preface: Part Two III.1.
• The entry on Galen’s treatise: III.3, the On Sects entry, no. 3, from Bergsträßer’s Arabic edition, p. 4, line 19 to p. 5, line 9. This translation is credited to Rosenthal, The Classical Heritage, p. 20, adapted by the chapter’s authors.

These passages give Hunayn’s reported practice, not an independent comparison of manuscripts.

[2] Al-Kindi, On First Philosophy, chapter one, translated by Alfred L. Ivry in Al-Kindi’s Metaphysics, SUNY Press, 1974, printed pp. 55–60, for al-Kindi’s view of learning and his defense of philosophy. The complete chapter was read in page images, along with:

• Notes and synopsis: the displayed footnotes and the first-chapter synopsis.
• Ivry’s account of the text and translation: pp. 7–8.
• Ivry’s commentary: pp. 115–133, up to the next chapter’s notes.

The defense appears at pp. 59–60, Arabic edition 105.2–9, and Ivry’s note on 105.2 traces its earlier ancestry. The Arabic numbers are page-and-line references to the edition; no second translation was checked against them. The larger work survives incompletely; it was not read in full.

[3] Peter Adamson, “al-Kindi”, Stanford Encyclopedia of Philosophy (SEP), revised 25 April 2024, for the historical setting, al-Kindi’s works and his philosophical program. The introduction and §§1.1, 1.2, 2.1 and 2.2 were read complete.

[4] Cristina D’Ancona, “Greek Sources in Arabic and Islamic Philosophy”, SEP, revised 23 March 2026, for specialist context on Syriac scholarship, the evidence for transmission and the Arabic Plotinus tradition. The introduction, §1, the opening and selected later paragraphs of §3 and the opening of §4 were read; this does not amount to a reading of the whole article or of the manuscripts and studies it cites.

[5] Uwe Vagelpohl, “In the Translator’s Workshop”, Arabic Sciences and Philosophy 21(2), 2011, pp. 249–288, for examples of the overlap between translation and explanation. Three sections were read completely: the abstract, the Introduction and “Hunayn on Translation and Medicine.” The rest of the article and its appendix were not read in full, and the Galen passages underlying its examples were not independently read.

The falling stone, Leila and the possible response are original teaching material, and the limit proposed for al-Kindi’s defense is an argument for the reader to assess. No independent specialist or learner review has been performed. No Greek, Syriac or Arabic collation, systematic second translation or learner timing was performed. Understanding the positions described here does not require sharing their religious commitments.