10/25/2005

Osiander and Anti-realism

This is another post from the frontlines of the class I'm teaching on Early modern philosophy and the scientific revolution. For those who haven't ever looked at Copernicus's On the Revolutions of the Heavenly Spheres, the book's first preface is written by a man named Andreas Osiander (though this preface was left unsigned in the original work).

In this preface, Osiander advocates for (what today would be called) an anti-realist conception of astronomy: the aim of astronomy is not to arrive at "true or even probable hypotheses," but rather to construct a mathematical model that will generate accurate predictions of the observed apparent locations of the celestial bodies.*

Osiander has come in for a lot of criticism, both from his contemporaries (like Rheticus, who entrusted the publication of Copernicus's book to him) as well as current commentators. However, I think the justifications Osiander offers for his view that we should not take astronomical models as literally true are not crazy. First, he notes that, if Ptolemy's model is correct, Venus's apparent size in the sky should change a great deal more than it actually does. That is obviously an empirical argument that Ptolemaic models do not reveal the true structure of the cosmos -- even though these models do make accurate predications about the location of Venus in the nighttime sky. Second, Osiander claims that there are genuine incompatible theories that both account equally well for the phenomena: he asserts that the Sun's observed motion can be modelled using an eccentric circle as basis or using an epicycle. (Unfortunately, I don't know anything about the details of this example.) If this is a genuine example of inconsistent but observationally equivalent theories, then Osiander has as good an argument against interpreting astronomical theories as literally (approximately) true as any argument given by an anti-realist motivated by underdetermination arguments.

Finally, note that these reasons for anti-realism are specific to astronomy. Thus we should not take Osiander to be advocating a general anti-realism towards all of science. To borrow the terminology of Magnus and Callander's recent "Realist Ennui" paper in Philosophy of Science, Osiander is not offering a "wholesale" argument for anti-realism, but a "retail" one, i.e., one specific to our pretensions to knowledge of the true physical structure of the universe.

________
* Tagging Osiander with various forms of anti-realism has been contested; see Barker and Goldstein's 1998 "Realism and Instrumentalism in Sixteenth Century Astronomy: A Reappraisal," in Perspectives on Science. They do agree, however, that Osiander considers knowledge of the true physical characteristics of the cosmos to be forever beyond human reach -- which strikes me as something a modern anti-realist might say. They also make the last point in the above post -- Osiander's skepticism is restricted to astronomy.

10/11/2005

Astrology, Astronomy, and the Scientific Revolution

One of the large-scale questions in academic discussions of the Scientfic Revolution concerns the relationship of the developments we today consider scientific to traditions we today consider pseudo-scientific or mystical, e.g. alchemy, astrology, and magic. People who make pronouncements like "There was no such thing as the Scientific Revolution" often justify such a claim by identifying and stressing continuities between mystical/ magical traditions and various new ideas that we now deem 'scientific.'

It is undeniable that significant continuities and similarities exist between pre-revolutionary views of nature and later ones. But I have often had the gut feeling that people sometimes overstate the case. Here is one example, from a brilliant historian of science, Allen Debus:
Some of the scholars, whose work contributed to our modern scientific age, found magic, alchemy, and astrology no less stimulating than the new interests in mathematical abstraction, observation, and experiment. Today we find it easy -- and necessary -- to separate "science" from occult interests, but many could not. (Man and Nature in the Renaissance)
This seemed overblown to me, because from Ptolemy up through Renaissance astrologer-astronomers such as Girolamo Cardano, the distinction between astrology and astronomy is explicitly drawn, and the historical figure often argues for the location of the boundary. So Debus's claim that students of nature during the Scientific Revolution 'could not separate science from occult interests' struck me as demonstrably false -- they could, and they did (at least in the case where the science is astronomy and the occult field is astrology).

But, as I have been working on my Magic, Medicine, and Science class (discussed last post), I've started thinking that there is something very right about Debus's idea, even if I would not couch the matter exactly as he does. What struck me is that, in the Ptolemy-Cardano scheme, astrology is classified as part of physics (in the Aristotelian sense, i.e. the study of nature), for it studies the physical influences of the sun, moon, planets and stars upon the Earth and its inhabitants. (Some astrologers thought the celestial bodies also had non-physical influences on us and our environs.) Astronomy, as mentioned in my last post, was classified as part of mathematics. Ptolemy, for one, states very clearly in Tetrabiblos that astrology studies physical, material causes associated with celestial bodies, whereas astronomy does not. And Cardano writes that astrology, unlike astronomy, studies "how lower things are linked to the higher ones."

So what is right about the Debus quotation? From the point of view of the Ptolemy-to-Cardano distinction between astronomy and astrology, the people working in the 17th C on a new physics of the celestial realm were apparently doing astrology, not astronomy. When Kepler is attempting to discover the physical cause of the planetary orbits, under the older taxonomy, that can't be astronomy, since astronomy does not deal with physical, material affairs. Thus what Kepler is doing (since it's still about the celestial realm) would naturally be classified as astrology. (And perhaps, though this is wild and irresponsible speculation, that partially explains why Kepler's theory, which appeals to entities like the Sun's 'motive soul,' has elements strongly reminiscient of earlier astrology.)

One possible problem with this idea: is there perhaps, in the Ptolemy-to-Cardano classification scheme, a separate heading for works like Aristotle's De Caelo, which does not appear to be straightforwardly astrological? That is, just because the old taxonomy won't count Kepler as astronomy, that doesn't imply that a celestial physics must be astrology: there could be some third category under which De Caelo and Kepler fall. Gentle reader, do you have any information to guide me here?