Ibn al-Shatir's Nihayat: which manuscript counts?
A fourteenth-century Damascus astronomer left a planetary theory that survives in several manuscripts. Learn why a manuscript on his astrolabe is not evidence for his Mercury model.
Start with the critical edition
Penchèvre's 2021 critical edition of Ibn al-Shatir's Nihayat al-sul fi tashih al-usul establishes the Arabic text from several manuscript copies. The research case rates it a verified fact, at 98, that the work survives through a multi-manuscript Arabic tradition, which makes a census and stemma of the witnesses the next priority. Ibn al-Shatir, Erwan Penchèvre (critical editor) (2021) ↗
A witness from the author's lifetime
Saliba's 1987 study of Ibn al-Shatir's work uses the Bodleian manuscript Marsh 139 as its principal reference. The research case describes it as a lifetime-era witness copied in AH 768, about 1366 to 1367 CE, and lists a core set of five witnesses used in specialist work: Marsh 139, Marsh 290, Marsh 501, Hunt 547, and Leiden Or. 194. Journal for the History of Astronomy ↗
A manuscript is not the whole work
British Library Add MS 9598 is a composite astronomical-instrument manuscript containing Ibn al-Shatir's astrolabe treatise, the Takhbir al-ahbab. One section is dated 1497. The case notes it is not a witness of the planetary-theory text, and states a methodological rule at 99: an astrolabe or zij manuscript cannot be evidence for the transmission of his mature Mercury model unless the planetary content is actually present. British Library Archives and Manuscripts Catalogue ↗
What an early copy does and does not show
A second rule, also at 99: the existence of an early Arabic witness such as Marsh 139 constrains textual ancestry but does not establish any European read or access event before Copernicus. Where a text was copied is not the same as who read it.
A question to carry forward
If a library catalogue lists 'a work by Ibn al-Shatir', which questions would you ask of the entry before using it as evidence for a particular theory?