Aratus’ Mirror

This paper will demonstrate that what has been presumed in Aratus of Soli’s Phaenomena to be an error by the poet in his description of the Kneeler constellation is in fact an intentional play on the names of the constellation and an allusion to Eudoxus and Plato. The supposed error occurs at v. 70, where Aratus follows Eudoxus (fr. 17) in saying that it is the right foot of the constellation which lies above the head of Draco and not the left, which contradicts the orientation convention chosen by both texts. Manuscripts unanimously convey δεξιτεροῦ ποδός.

Kidd notes the apparent astronomical error and also Hipparchus’ judgement at In Arat. 1.4.9 that the mistake was more likely a simple oversight, παρεωρακέναι, rather than any genuine error, διημαρτηκέναι (Kidd 1997, 204). This mistake, however—in either Aratus’ science or in his poetic expression thereof—has wider implications: Partly on the basis of this passage, many modern scholars have assumed Aratus to be astronomically inexpert (cf. van Noorden 2009, 256; Natunewicz 1973; Marrou 1956, 254). Was Aratus incompetent in the scientific material of his poem, or simply careless? There have been attempts to amend the error and so rehabilitate the passage. Martin, for instance, following the lead of Attalus and other ancient commentators, changes the adjective’s case to δεξιτερῷ so that it modifies Draco’s head rather than the Kneeler’s foot (“the Kneeler’s foot lies above the middle of the right side of Draco’s head”: Martin 1998, 2:184–86).

I propose instead that Aratus intentionally preserves Eudoxus’ error in order to allude to and comment on contemporary uncertainties regarding astronomical models and their relationship to real celestial bodies; he thereby makes a window reference to Platonic cosmology via equally Platonic ideas of image and reflection.

The key question is, in fact, not the “erroneous” account given at 70, but the relationship of 70 to 272, where Aratus gives the correct orientation of the Kneeler, by way of the figure’s same right leg. Martin’s approach, as we have seen, is to bring 70 into line with 272 through emendation—a violent solution given the unanimity of the tradition. It is crucial that Aratus uses two names for this constellation, instead of Eudoxus’ one: the latter gives “the Kneeler”, ὁ ἐν γόνασι, but Aratus refers to it in both 70 and 272 as εἴδωλον. In the second passage at 272, he fails to use the name ὁ ἐν γόνασι at all and εἴδωλον is the sole identifier. Why? The famous passage of Plato’s at Timaeus 46a3–46b concerning the “εἰδωλοποιίαν”—the “image-making” of mirrors—gives the clue: “δεξιὰ δὲ φαντάζεται τὰ ἀριστερά”, “left appears right” in mirrors. Aratus thus makes the “Mirrored” constellation, “εἴδωλον”, itself mirrored. Aratus, either following Eudoxus’ lead in a discussion no longer extant—perhaps in the lost Enoptron, or “Mirror”, itself—or subverting a genuine error to his own ends, appears in either case to use the Mirrored constellation to evoke a Platonic lens for his own poetic model of the universe.