Copernicus and Ibn al-Shatir: how close are their Mercury models?
Two models of Mercury’s movement look strikingly alike. Which matches suggest knowledge was passed on, and which leave its route unproven?
Where the Mercury models match
Nikfahm-Khubravan and Ragep compared the Mercury models of Ibn al-Shatir and Copernicus in technical detail. The research case rates at 96 the claim that Copernicus's mature Mercury model in De revolutionibus has very high geometrical correspondence with Ibn al-Shatir's earlier one. At 90 it adds that this correspondence strengthens a shared-ancestry hypothesis without identifying who carried the model. Arabic Sciences and Philosophy ↗
Close, but not copied constants
The same record is explicit that parameter correspondence is mixed: several component values are close while other radii differ materially. A blanket claim that Copernicus used identical copied constants is graded unsupported, at a confidence of 96 that it is unsupported.
What is not enough
The research case rates at 95 that the Tusi couple, diagram lettering, or a broad similarity of circular models cannot alone establish direct transmission, and at 92 that no rare shared error has yet been found strong enough to act as a smoking gun. Ragep's study of the Uppsala notes is the kind of close reading that would have to supply one. Journal for the History of Astronomy ↗
An early model that failed
Copernicus's early Mercury model in the Commentariolus fails. The research case treats that as evidence of a developmental discontinuity at 94, and records that reading it as a copied misunderstanding is disputed. It also notes that the early construction belongs to the same technical family of physical Tusi-couple variants used in Ibn al-Shatir's solid-sphere Mercury model.
The case against direct copying
Blasjo's critique of the arguments for Maragha influence on Copernicus is part of the record. Two hypotheses stay open and graded as hypotheses: that tropical-year timing in the early model preserves a residual geocentric fossil, and that the order of construction in the Commentariolus preserves lineage information. Journal for the History of Astronomy ↗
Which match is too specific to ignore?
If two models match closely in structure but differ in several numbers, which kind of match is harder to produce independently? State what each outcome would predict before looking at the parameters.