The best working hypothesis that can be proposed at this point is to think of Copernicus's acquaintance with some type of an Arabic astronomical work that contained commentaries on earlier works, such as the work of Quṭb al-Dīn al-Shīrāzī (d. 1311) where one could find proofs of Ṭūsī's theorem like the one reproduced by Copernicus, since the works of Shīrāzī were themselves commentaries on Ṭūsī's work. In addition the works of Shīrāzī also contained the adoption of 'Urḍī's model for the upper planets, which was summarily adopted by Copernicus through Ibn al-Shāṭir's work and where he almost unconsciously adopted 'Urḍī's Lemma. We just said that this model was the chosen model for Shīrāzī against the model of his own teacher Ṭūsī. But then Shīrāzī's work does not have the world-view of Ibn al-Shāṭir. And thus it cannot explain the identical lunar model of Ibn al-Shāṭir that was adopted by Copernicus, nor his deployment of the same Ṭūsī Couple technique as was done by Ibn al-Shāṭir while describing the motions of the planet Mercury. If we follow the route of commentaries, then it will be for historians of Arabic astronomy to find a commentator, such as Shīrāzī, who must have lived after Ibn al-Shāṭir, and who must have probably written a lengthy commentary on Ibn al-Shāṭir and tried to place Ibn al-Shāṭir's works in the context of the earlier works of 'Urḍī and Ṭūsī. But no one knows of such a commentary and the problem is yet to be solved.
The historiographic lesson, though, is the following: Had it not been for those similarities that have now surfaced between the works of Copernicus and the earlier astronomers of the Islamic domain, this problem would not have arisen in the first place, and we would not have even suspected that such commentaries may even exist.
There is yet another line of research that has to be pursued as well, this time taking a direction well rooted in the works of Ṭūsī. The later commentators on Ṭūsī's Tadhkira have already started this research but very few people have pursued it in the modern literature.[329] The research in question has to do with implications of the rupture of the Aristotelian universe by the Ṭūsī Couple. The rupture is in the following sense. As we have already said before, Aristotle had already divided he universe, on the basis of the natural motion of its elements, into two basic divisions: the celestial region (which moved by the natural circular motion of the element ether of which the celestial world is made) and the sublunar region (where linear motion predominated). With Ṭūsī's Couple, one can now demonstrate that circular motion could produce linear motion, and vice versa. Does that mean that the Aristotelian division has to collapse as a result, to what extent, and what can be saved of it if any? Only future research will uncover such repercussions.
Returning to the problem of contacts with Europe, the fundamental question for the intercultural science still remains: How could Copernicus know about those Arabic results, at such a late date, with all the known conditions of the Renaissance? The answer to this question has to presuppose other questions about Copernicus himself. Did he know Arabic at all? Was he in contact with Arabists? How well did the Arabists that he knew know about technical Arabic science? All these questions go to the very core of the intellectual environment during the Renaissance, and will have to be tackled from both sides of the Mediterranean.
So far the assumption had been that Copernicus did not know any Arabic, and since the Arabic sources were not translated into Latin, he must have known about them through some other language to which they were "translated" and that Copernicus could read. Again it was Neugebauer who assumed correctly that Copernicus, like all other well educated Renaissance men, could read both Greek and Latin. Since Latin could not be considered as the language of translation for there was no evidence that the Arabic texts were translated into it, that left Greek as the only possibility, according to Neugebauer's implied reasoning. This train of thought led Neugebauer to examine the Byzantine Greek sources for clues to the solution of the problem of transmission. At the time when Neugebauer first came in contact with the Arabic material that signaled possible contacts with Copernicus, he had already been working on the Byzantine astronomical texts for his own Studies in Byzantine Astronomical Terminology. So at first glance, Byzantine Greek looked like a plausible route for such a transmission. Within few years Neugebauer's diligent search quickly yielded a very interesting fruit in the form of a Byzantine Greek manuscript now kept at the Vatican library as Gr. 211. The manuscript in question contained a Greek version of Ṭūsī's couple, and thus seemed like a good lead to pursue. But now that the same manuscript has been published,[330] we can clearly see that although it seems to have a qualitative description of the Ṭūsī Couple, it does not seem to have the proof of that Couple. And as we have seen before, there were remarkable similarities between the proofs of the Ṭūsī Couple in the Arabic work of Ṭūsī and the Latin work of Copernicus. And we have seen that both proofs depended greatly on the identical usage of the same letters of the alphabet to designate the same geometrical points. So the proofs are essential to explain this phenomenon and to clench the argument of the possible connections between the two. Nor does the Vatican Greek manuscript contain any of the material of 'Urḍī or Ibn al-Shāṭir, which we have seen were very relevant to Copernicus and somehow made available to him.
All those issues and questions go beyond the accepted historiography of science, as it is now generally understood. Copernicus's connection to earlier Islamic material is such a new field of research that it has not yet had the chance to have an impact on the general history of science. But whatever information we now possess, inevitably leads us to the conclusion that there must have been an intimate connection, at least on the theoretical mathematical level, between the works of Copernicus and the works of his predecessors in the Islamic world.
A word of caution is in order so that these issues of contacts between Copernicus and his predecessors in the Islamic world will not be confused with Copernicus's genius idea of heliocentrism. None of the astronomers who worked in the Islamic world, and who have been mentioned so far, had any interest in such concepts as heliocentrism. In my estimate, they were so closely wed to the stubborn, but all comprehensive, Aristotelian cosmology, which dictated a geocentric universe, and which continued to reign supreme in the world of astronomy all the way till the time of Newton. This despite the hints and disparaging remarks one would hear about it from time to time by astronomers working on both sides of the Mediterranean.
In this context too, one is also forced to raise the question of the scientific legitimacy of heliocentrism itself in a pre-Newtonian universe, where no alternative cosmology was yet available. Copernican scholars, who have been busy trying to explain the origins of Copernican heliocentrism, have yet to explain how could Copernicus convince himself that he could shift the center of motion to the sun without having to propose some non- Aristotelian cosmological theory that could hold the world together in the same way the Aristotelian cosmology did.[331] That is, without the benefit of the Newtonian law of universal gravitation, how could he have hoped to maintain the system together?
What Copernicus's predecessors were doing was well within the limits of Aristotelian cosmology. And in that sense they were perfectly consistent in their attempt to replace the Ptolemaic models with alternative models that behaved much more consistently by avoiding the absurdities of the Ptolemaic models. But for Copernicus to complain, in the introduction of his Commentariolus about the equants, a complaint that made sense only in an Aristotelian universe, and then go ahead and abandon all that system, and retain the modified models that avoided the absurdity of the Ptolemaic equant, is very puzzling indeed. The models that were developed in the Islamic world to solve the problem of the Ptolemaic equant, were developed specifically so that the models would be consistent with Aristotelian cosmological considerations. So if one was willing to abandon the Aristotelian universe, then why retain those models? Such problems will have to be left for the Copernican scholars to handle.
329
See the tantalizing hints by Willy Hartner in "Naṣīr al-Dīn al-Ṭūsī's Lunar Theory",
331
For the latest and most convincing attempt to explain the roots of Copernicus's heliocentrism, see Bernard Goldstein, "Copernicus and the Origin of His Heliocentric Universe",