In this short monograph Theokritos Kouremenos proposes an unorthodox reading of Aristotle’s de Caelo and the theory of the fifth element. He argues that Aristotle initially introduced the aither as matter for the “fixed stars” alone, and not the Sun, Moon, and planets, which Aristotle “probably” thought were made of fire. Kouremenos proposes a developmental hypothesis according to which Aristotle expands the role of the aither to include the Sun, Moon, and planets only after realizing certain insurmountable problems his theory faced. Kouremenos also argues that in Cael. Aristotle did not subscribe to a physically realized version of the homocentric spheres model of the cosmos discussed in Metaphysics Λ.8, and that he probably did not seriously entertain it as a physical theory even there. I do not find Kouremenos’s arguments regarding the limited role of the aither in Cael. to be persuasive, nor do I agree with the developmental approach he takes.
The first of the book’s three chapters (“Aristotle’s cosmology”) serves as an introduction for the arguments of the second (“The stuff of the heavens”) and third (“Aristotle and the theory of homocentric spheres”).
Chapter one provides a summary of the orthodox view of Aristotelian cosmology. Kouremenos emphasizes two features of this cosmology that play an important role in his later argument. First, he stresses that the elemental transformations that occur in the sublunary world must counterbalance each other, such that, on a cosmological scale, the “mass ratios” between the elements remains constant. An imbalance of the elements could lead to the absurdity of one element completely overtaking the others (e.g., fire burning everything in the cosmos). The need for balance between the elements underlies one of Aristotle’s arguments, in Meteorology. A.3, for the existence of a fifth element capable of filling the cosmos’s vast spaces while not overpowering the others.
The second feature pertains to the motions of the celestial bodies. While the sphere of the fixed stars possesses a single, uniform motion, the other celestial bodies each additionally possess much slower annual motions, oppositely directed to that of the fixed stars. These “zodiacal” motions are non-uniform: their angular rates appear to vary. As we shall see, Kouremenos makes much of the consequences this zodiacal anomaly has for Aristotle’s cosmological theory.
In Cael. Aristotle introduces the aither as matter of the fixed stars and the sphere in which they are embedded. Unlike the sublunary elements, aither is absolved from all change save for change in place: it moves naturally in a circle. This circular motion is, importantly, uniform. If the Sun, Moon, and planets are also made of aither, then their apparent non-uniform motions must be analyzable into some combination of uniform circular motions. This of course is the basis for the famous (and, by Kouremenos’s lights, likely apocryphal) story of Plato’s challenge to “save the phenomena.” Thus enters the theory of homocentric spheres attributed to Eudoxus. Kouremenos’s discussion of the geometry of the homocentric spheres is dense and would have been aided by the inclusion of a diagram or two (his includes none), though even with diagrams it is devilishly hard to envision how the motions of the rotating spheres combine. (Henry Mendell has a useful web page with a number of flash presentations showing how this works).
Kouremenos stresses that the Eudoxan model is not a complete solution to the problems presented by the phenomena. It not only lacks predictive power, but it also fails to account for the zodiacal anomalies discussed above, a shortcoming that Callipus’ modifications were perhaps meant to remedy (Aristotle does not specify this in Λ.8). According to Kouremenos, Eudoxus “seems to have solely aimed at giving an idea of how some very broad aspects of the apparent motions of the Moon, the Sun and the planets can be reproduced geometrically” (p. 38). Indeed, the Eudoxan model’s inability to reproduce the motions of these celestial bodies exactly, with or without Callipus’ modifications, together with Aristotle’s hesitancy in Λ.8 to vouch for it, leads Kouremenos to conjecture that Aristotle never seriously adopted it as a physical model of the cosmos. It is offered, Kouremenos suggests, “simply in order to give an example of how astronomy may contribute, with its study of celestial motions, to precisely determining the multitude of the unmoved movers that might be needed to explain the eternity of motion and change in the cosmos” (p. 46).
In chapter two, Kouremenos mounts his case for the claim that in Cael. Aristotle introduced the aither as matter of the fixed stars alone. His strategy is to look at passages where Aristotle discusses the properties of the aither and to see whether he explicitly attributes these properties to any celestial bodies other than the stars. Finding no such explicit attribution, Kouremenos concludes that Aristotle believed they were not composed of aither. For example, Kouremenos reads the argument in Cael. A.2 for the existence of a simple body that is separated from us and of a higher nature “in proportion as it is removed from all things existing here” (269b13-17, tr. Kouremenos) as implying that the aither’s “spatial separation from the center of the cosmos” is “the largest possible” (p. 55). According to this reading, the aither makes up only those bodies that occupy the outermost place of the cosmos, namely the fixed stars. But the passage need not exclude the Sun, Moon, and planets, which are themselves “removed from all things existing here” by a very great distance. And even if the argument does apply only to the fixed stars, that need not also imply that the Sun, Moon, and planets are not composed of aither. Must Aristotle’s argument for the existence of a fifth element cover every instance of that element in the cosmos?
Kouremenos’s strategy is to marshal a number of such examples and hope that together they make a compelling case. To his credit, he does find many passages that could be so read, but in general I found the arguments in this section underwhelming. At the least, Kouremenos certainly overstates his case. Even weaker, in my view, are his arguments for the alternative cosmology he finds in Cael.. According to Kouremenos, “Aristotle’s implicit view in the main bulk of the de Caelo is that the planets, the Sun and the Moon are in the uppermost part of the cosmic place which is filled with fire, and consist of this simple body” (p. 71). But just as Aristotle never explicitly states in Cael. that the Sun, Moon, and planets are not composed of aither, so also he never states that they are composed of fire. The justification for this “implicit” view rests largely on Kouremenos’s interpretation of Cael. B.4. There (287a2ff.) Aristotle argues that the region of the cosmos in which the Sun, Moon, and planets reside is spherical because it is continuous with the aitherial sphere of the fixed stars. That Aristotle felt it necessary to argue for the sphericity of this region indicates to Kouremenos that he did not believe it was composed of aither. Rather, Kouremenos concludes that Aristotle must have thought that the region is composed of that element immediately below the first, namely fire.
Similarly, a little later in B.4 (287a30ff.), Aristotle offers another argument for the sphericity of the cosmos, but this time begins not with the outermost region, but the innermost. Since the surface of the water upon the Earth forms a sphere (as everyone agrees that is does), then so also do the other elements form spherical layers, and thus, by the same argument, so too do the “upper bodies” ( ta anô sômata). Kouremenos argues that Aristotle’s use of the plural “upper bodies ” suggests that the superlunary region of the cosmos contains more than one “elemental type of matter” (p.68, n31), and thus that this region is, like the sublunary world, stratified, with the outermost region composed of aither, and that immediately below it composed of fire. I disagree that the phrase “the upper bodies” need refer to more than one type of matter. Kouremenos states that it “cannot possibly be identified with the shells of the first simple body which make up the heavens according to the cosmology Aristotle outlines in Metaph. Λ.8” (p. 66, n31). Perhaps not (though I doubt it “cannot possibly” be done), but it seems more natural to read the phrase as referring to the actual celestial objects themselves – the Sun, Moon, planets, and stars. Here again Kouremenos objects that if Aristotle held that the Sun, Moon, and planets were made of aither, this argument would be otiose. But this second argument is offered as an alternative to the first, one that does not require the belief in a spherical aither, and thus is not redundant.
This picture of the cosmos, with fire extending to the outermost sphere of the fixed stars, is specifically rejected by Aristotle in Meteo. A.3. In fact, he there calls it a “childish opinion” ( paidikês doxês, 339b34). Nevertheless, Kouremenos holds that it was only after writing Cael. that Aristotle came to realize its childishness and revised his cosmological theory accordingly. Kouremenos suggests that at the time of composing Cael. Aristotle either was ignorant of the relevant astronomical facts that reveal the vast size of the cosmos (even he admits this is unlikely), or had simply failed to take proper account of them (p. 75). That is, Aristotle simply failed to think through the consequences of his cosmological theory. Those of us who spend our time poring over every word of “the Philosopher” may find this last conclusion difficult to accept, even if infallibility is a virtue we must withhold from him.
But does Aristotle really never explicitly say in Cael. that the Sun, Moon, and planets are composed of aither? Well, no – as Kouremenos concedes, Cael. B.7 does indeed state this. In fact, Kouremenos adds, in his typically emphatic manner, that B.7 “leaves no doubt” (p. 50) that this is the case. To accommodate this troublesome passage, Kouremenos is forced to bracket B.7 as a late addition (p. 53). Here we find one of the many unfortunate hallmarks of the developmental hypothesis: bracket, exclude, mark as spurious, reject. Such an interpretive tool must be wielded with great care, and I am not convinced this case merits its application.
In the final chapter Kouremenos considers whether the theory of homocentric spheres is anywhere present in Cael.. The order of argumentation is somewhat odd. Kouremenos first considers a passage in Cael. B.6 that some believe implies a cosmology based on the homocentric spheres (he argues it doesn’t), then one in B.12 (he argues it is spurious), after which he fills out the sketch of the alternative cosmology begun in chapter two, based on passages from B.10, B.12, A.4, and B.2-3 (in that order), before turning to Metaph. Λ.8. As mentioned above, I find this alternative cosmology implausible, and I do not believe the additional arguments in chapter three add much credence to the claim. For example, Kouremenos uses the notion of a resisting medium of fire that is carried around by the fixed stars to make sense of an odd claim in Cael. B.10 regarding the diminishing influence the fixed stars have on celestial bodies closer to Earth. He argues that such a resisting medium does not conflict with Aristotle’s arguments in B.9 regarding the tremendous noise movement through such a medium would create, because in B.9 Aristotle is concerned with movement through a stationary medium, not a moving one. While I grant his point that movement through a medium that is itself moving in the same direction (like the boat moving through the river discussed at 291a9ff.) will not create the awful din with which B.9 is concerned, I find it unbelievable that Aristotle should then state that certain of his predecessors erred in believing that the celestial bodies move in a medium of air or fire (291a20), and then not go on to specify that their error lay in positing a stationary medium, rather than a moving one.
Kouremenos’s arguments regarding the spuriousness of the explicit reference to the homocentric spheres in Cael. B.12 are also not without their problems, but, like Kouremenos, I do find it odd that Aristotle should make a passing allusion to the homocentric spheres there, while not treating them thematically elsewhere in Cael.. Kouremenos is not the first to suggest that at least portions of Cael. were written without the homocentric spheres in mind 1, but it is a far step from there to deeming the passage a “thoughtlessly appended addition to the original version of the chapter by someone other than Aristotle himself “ (p. 110).
In summary, I am not convinced by Kouremenos’s arguments for the more limited role he sees the aither playing in Cael.. Scholars of Aristotelian natural philosophy and the history of astronomy may find in the book a welcomed opportunity to look closely at a number of important passages and to reconsider their import from a rather unorthodox perspective, but I suspect that many readers will not be won over. Kouremenos may get a better hearing for his arguments regarding the homocentric spheres in Cael., and the possible spuriousness of certain puzzling passages. However, the trend in Aristotelian scholarship away from the sort of developmental hypothesis he suggests may cause more than a few readers to turn a deaf ear.
1. See e.g. Easterling, H. J. “Homocentric Spheres in de Caelo ”, Phronesis, Vol. 6, No. 1-2 (1961), pp. 138-154.