Commentators and Doxographers on Xenophanes’ Theology

Abstract

In this paper, I examine the various interpretations of Xenophanes’ theology in antiquity. After distinguishing between the traditions of commentaries and of doxographies, I focus on two unexpected testimonies: Pseudo-Aristotle in On Melissus, Xenophanes and Gorgias and Simplicius. Both attribute to Xenophanes, unlike other authors, the thesis that the god is neither limited nor unlimited and neither moved nor unmoved. I argue that this reading originates from Theophrastus, more specifically from a commentary on Aristotle’s Physics, but that Pseudo-Aristotle is responsible for misinterpreting this claim and adding arguments to justify it. I finally highlight the many sources of Simplicius, who uses not only Theophrastus’ commentary on the Physics and Pseudo-Aristotle, but also another doxographical work, possibly the Physical Opinions of Theophrastus.

I. Introduction

The interpretation of Xenophanes’ thought is particularly disputed. This is because some regard him to be a systematic theologian, while for others he is mostly a poet who criticized the views of other poets without expounding any kind of theology. ¹ Some think that he was the first monotheist, others that he was a polytheist for whom there was a hierarchy with a greater god on top. ² Similar controversies are found in the ancient doxographical accounts, but usually not on the same aspects. In this paper, I shall focus on issues concerning Xenophanes’ theology. While the ancients almost unanimously regard him as a monotheist, they disagree on the characteristics of this one god, and more specifically on his spatial and kinetic predicates. Simplicius summarizes the controversy in his commentary in Phys. 23.14–19 (see T2c below): some say that Xenophanes’ god is spherical and unmoved, a few others that he is infinite and unmoved, but according to Simplicius and the pseudo-Aristotelian treatise On Melissus, Xenophanes and Gorgias (MXG), he is neither limited nor unlimited and neither moved nor unmoved (I will call these two claims ‘antilogies’ for the sake of brevity).

There is no real disagreement on this topic in recent scholarship. In fragment B26 (which is quoted in T2b), Xenophanes explicitly says that the god does not move. And even though no fragment mentions his limitation, most critics think that he did not give any particular shape to his god, ³ but that most doxographical accounts attribute him a spherical form because they assimilate him to Parmenides’ being. There is disagreement, however, over the reasons for this diversity of interpretations in the testimonies, and in particular for the unexpected reading of Simplicius and the MXG. It is obvious to all recent critics that Xenophanes is not the author of this antilogical theology, not only because he explicitly claims that the god is unmoved, but also because of the Eleatic ⁴ aspect of these theses and even more so because of the arguments that are provided to justify them. The few people who believed that this account was accurate had to suppose that Xenophanes wrote his theology after Parmenides. ⁵ But if Xenophanes did not design these antilogies, one may wonder where they come from and how they were attributed to him. This raises the issue of Theophrastus’ role: Simplicius claims that he draws his own reading from him, but many critics are unconvinced. It is usually assumed, since Hermann Diels’ Doxographi Graeci, that Theophrastus is the ultimate source of most doxographers, who almost unanimously say that Xenophanes’ god is limited and unmoved. ⁶ Jaap Mansfeld contests this view and claims that Theophrastus is the source of Simplicius on this topic, not of the doxographies. ⁷

One must then clarify how the reception of Xenophanes diverged and why Simplicius and Pseudo-Aristotle attributed such a complex theory to him. No comprehensive and satisfactory explanation has been proposed so far since most critics fail to account for many aspects of this reading. In particular, the Eleatic arguments that both Simplicius and the MXG provide in support of the antilogical claims are almost never explained, nor is the fact that Simplicius quotes a fragment (B26) that contradicts his own interpretation. Only Mansfeld really tackles this issue: in ‘Theophrastus and the Xenophanes doxography’, he claims that Simplicius and Pseudo-Aristotle were indeed inspired by Theophrastus (a view that I support) and provides an explanation for this interpretation that I will partly revise. Mansfeld does not, however, account for the arguments that Simplicius and Pseudo-Aristotle give to justify the antilogies. In ‘Compatible alternatives: Middle Platonist theology and the Xenophanes reception’, he explains the juxtaposition of a positive presentation of Xenophanes’ god and an antilogical one by drawing a parallel with Eudorus’ treatment of the Pythagoreans: as a consequence, he supposes that Simplicius and Pseudo-Aristotle were under the influence of a common Middle-Platonic text. This reading, as I will show, finds little support in the text.

This paper aims to explain these conflicts within the tradition on Xenophanes’ theology by retracing the origins of the various doxographical accounts, and especially of Pseudo-Aristotle’s and Simplicius’ interpretation. This reading will prevent some misinterpretations of Xenophanes’ thought, by retracing the origins of errors in these accounts, but will also shed some light on the way doxographies, especially those originating from Theophrastus, were transmitted, combined and reinterpreted in antiquity. I will first distinguish between the doxographies and the commentaries which, as we will see, have no other information on Xenophanes other than that which Aristotle provides (section II). I will then turn to Simplicius and the MXG (section III) and give an explanation of the antilogies (section IV) and of their arguments (section V). On this basis, I will claim that Pseudo-Aristotle is the author of these arguments and restore Diels’ thesis (recused by most recent critics) that he is one of Simplicius’ sources (section VI). Finally, I will reconsider some aspects of Simplicius’ testimony and offer a hypothesis concerning Theophrastus’ role (section VII).

II. Multiple traditions

i. Earliest mentions of Xenophanes’ theology

Plato and Aristotle provide our first reports on Xenophanes’ theology. In Soph. 242d, Plato makes him the first representative of the Eleatic school, whose thesis he formulates as ‘everything is one’. This testimony shows how early Xenophanes was ‘Eleatized’, that is his theology assimilated to the Eleatic ontology, and more precisely to Parmenides’. ⁸

In his analysis of causes in Metaphysics A, Aristotle also considers Xenophanes one of the Eleatics:

T1: Oὐ μὴν ἀλλὰ τοσοῦτόν γϵ οἰκϵῖόν ἐστι τῇ νῦν σκέψϵι. Παρμϵνίδης μὲν γὰρ ἔοικϵ τοῦ κατὰ τὸν λόγον ἑνὸς ἅπτϵσθαι, Μέλισσος δὲ τοῦ κατὰ τὴν ὕλην (διὸ καὶ ὁ μὲν πϵπϵρασμένον ὁ δ’ ἄπϵιρόν φησιν ϵἶναι αὐτό). Ξϵνοφάνης δὲ πρῶτος τούτων ἑνίσας (ὁ γὰρ Παρμϵνίδης τούτου λέγϵται γϵνέσθαι μαθητής) οὐθὲν διϵσαφήνισϵν, οὐδὲ τῆς φύσϵως τούτων οὐδϵτέρας ἔοικϵ θιγϵῖν, ἀλλ’ ϵἰς τὸν ὅλον οὐρανὸν ἀποβλέψας τὸ ἓν ϵἶναί φησι τὸν θϵόν. ⁹

However, that much is appropriate for our investigation [of causes]: Parmenides seems to grasp unity according to definition and Melissus according to matter ¹⁰ (this is why one says that it is limited, the other unlimited). But Xenophanes, the first among them to ‘unicize’ ¹¹ (for Parmenides is said to have been his pupil), made nothing clear and does not seem to have grasped the nature of either of them [the formal and material cause]. But looking at the whole universe, he says that the one is god. (Arist. Metaph. A.5, 986b17–25) ¹²

Aristotle is more informative than Plato: he claims that Xenophanes was Parmenides’ master and shared his doctrine of unity, but that he was unclear and mostly said that ‘the one is god’. This text will be of crucial importance for understanding the various doxographical accounts, especially those of Simplicius and the MXG.

ii. Doxographies

Xenophanes is regularly mentioned in ancient texts out of interest for his theology, his physics or his epistemology. ¹³ Many authors deal with the characteristics of his god: some in the context of a commentary on Aristotle’s Physics or Metaphysics (Alexander of Aphrodisias, Philoponus, Simplicius, Asclepius), others within a doxography. ¹⁴ Most critics oppose Simplicius and Pseudo-Aristotle, according to whom Xenophanes’ god has antilogical properties, Cicero’s Nat. D. 1.28 and Nicolaus of Damascus (as mentioned in T2c), who claim that he is unlimited and unmoved, and the other doxographical accounts, for whom he is limited and unmoved. I think, however, that distinctions must be made within this third group, in particular between the doxographies and the commentaries.

The doxographies present information on various topics of Xenophanes’ thought: his theology or ontology, his epistemology and his physics. Most of the information does not come from Aristotle but from other sources. Aristotle does not mention Xenophanes’ epistemology at all, and his only account of his physics is that the earth is unlimited under our feet (Cael. II.13, 294a21–28). We find, however, detailed accounts concerning his astronomy, meteorology or the origins of the sea in Pseudo-Plutarch (in Euseb. Praep. evang. 1.9.4), Hippolytus of Rome (Haer. 1.14.2–6), Diogenes Laertius (IX.19) and in various chapters of Aetius. ¹⁵ A striking contrast between the doxographies and Aristotle is that according to many doxographies, Xenophanes says that everything comes from earth, ¹⁶ while Aristotle claims that no philosopher ever considered earth as the only element (Metaph. A.8, 989a3–10).

The doxographies also present information on Xenophanes’ theology that Aristotle seems to ignore. Some appears to be authentic, for example the thesis that the god perceives and thinks as a whole (Diogenes Laertius, Pseudo-Plutarch, Hippolytus), which fragment B24 confirms. ¹⁷ Other elements might rather be the product of interpretation, and in particular Eleatization: this is especially the case for the claims that the god is all alike (in Hippolytus) or spherical (according to Cicero in Luc. 118, Sextus Empiricus in Pyr. 1.225, Diogenes Laertius, Hippolytus and Theodoret). If one excepts an unexpected polytheistic report in Pseudo-Plutarch, the doxographers are fairly unanimous on most of the god’s predicates: he is eternal, one, limited, unmoved and unchangeable. ¹⁸

There are two divergent accounts regarding the god’s limitation. Cicero in Nat. D. 1.28 and Nicolaus of Damascus (according to Simplicius, see T2c below) both say that Xenophanes’ god is unlimited. Cicero is probably taking up an Epicurean doxography, since he put his speech in the mouth of the Epicurean Caius Velleius; it actually contrasts with the information he provides in Luc. 118, where he claims that Xenophanes’ god is spherical. One way to explain this claim is through confusion: Mansfeld raises the possibility that Cicero’s doxography might confuse Xenophanes with Anaximander or Anaxagoras. ¹⁹ However, since Xenophanes is more commonly associated with Parmenides than with those thinkers, there is no particular reason why this would have happened, even though one cannot exclude it. One should note, however, that the Epicurean doxography deduces the god’s infinity from the universe’s, which is assimilated to the god. ²⁰ The thesis would then have been attributed to Xenophanes by someone who thought that his universe was unlimited. There are two possible reasons for this: either because he claimed that the earth is unlimited under our feet (B28) or because of the view (attributed to him but for which there is no fragment) that there are infinite worlds. ²¹

The origin of Nicolaus’ claim is more uncertain, since his opinion is mentioned only by Simplicius and we have no context at all. Some critics believe that he conflated Xenophanes and Melissus, ²² but again I see no reason why this would be the case. Xenophanes was said to be Parmenides’ master (T1), not Melissus’, and none of the predicates of his god might justify confusing him with Melissus rather than with Parmenides. I would therefore rather support the same interpretation as for Cicero: Nicolaus assimilated the god with the whole and was led to think that Xenophanes’ universe is unlimited either because for him the earth is infinite or because the worlds are infinitely many. ²³ Mansfeld recuses the idea that Nicolaus was influenced by Aristotle’s claim of an infinite earth, saying that ‘it is hard to credit him with an equation of this earth and Xenophanes’ God’. ²⁴ But many doxographers assimilate Xenophanes’ universe and the god, probably because of Aristotle’s claim in the Metaphysics (T1) that ‘looking at the whole universe, [Xenophanes] says that the one is god’. ²⁵ If one thinks that for Xenophanes, on the one hand, the god and the universe are the same thing and, on the other, his universe is infinite, one can logically deduce that his god is unlimited.

In conclusion, with the exceptions of Cicero’s Nat. D. and Nicolaus of Damascus, there is relative agreement on Xenophanes’ theology among the doxographers, especially on the point that his god is limited and unmoved. Most importantly, they do not rely on Aristotle’s work, or at least not on our Aristotelian corpus, for their information. Despite a strong Eleatization, which might explain why Xenophanes’ god is presented as spherical and homogenous, these doxographies ultimately had access to some genuine knowledge of Xenophanes on many topics, since it is partially confirmed by fragments. This is not the case, as we will see, for Aristotle’s commentators.

iii. The commentators

The commentaries on Aristotle diverge depending on whether they are concerned with the Metaphysics or the Physics: hence, we are dealing with two different traditions. Alexander of Aphrodisias (29.20–30.7, 42.22–28 and 44.6–10), Asclepius (40.23–27 and 41.27–42.4) and the recensio altera of Alexander’s commentary ²⁶ comment on our text T1. They mostly agree with Aristotle in saying that Xenophanes, like Parmenides and Melissus, thinks that being is one. Since Aristotle discusses Xenophanes’ thought in the context of his examination of those who think that the principle is one and unmoved, they often assume that his principle is also unmoved. ²⁷ Concerning Aristotle’s claim that Xenophanes ‘does not seem to have grasped the nature of either of them’, Alexander refers to form and matter. However, Asclepius understands it as meaning that ‘he did not consider [the god] as limited nor unlimited’ (οὔτϵ γὰρ ἄπϵιρον ὑπέθϵτο οὔτϵ πϵπϵρασμένον, 41.29–30) and the recensio altera argues that ‘he did not say that [the principle] was material nor formal because he neither said that it is unlimited, like Melissus, nor limited, like Parmenides’ (οὔτϵ γὰρ ὡς ὑλικὸν ϵἶπϵν αὐτὸ οὔτϵ ὡς ϵἰδικόν, διὰ τὸ μήτϵ ἄπϵιρον ϵἰπϵῖν αὐτὸ ὡς ὁ μέλισσος, μήτϵ πϵπϵρασμένον ὡς ὁ Παρμϵνίδης). All these readings can find a justification in Aristotle’s text, depending on how strongly one connects the opposition between form and matter and that between limited and unlimited, and do not add any details concerning Xenophanes’ theology.

Commentaries on the Physics mostly differ from those on the Metaphysics on the issue of the god’s limitation. Aristotle does not mention Xenophanes in the Physics. However, at the beginning of Phys. I.2, he presents a taxonomy of the ancient opinions on principles, depending on their number and whether they move or not. ²⁸ On this occasion, he names Melissus and Parmenides as those claiming that the principle is one and unmoved (I.2 184b15–16). Alexander (as mentioned by Simplicius at 23.16, see T2c), Philoponus (22.15–23) and Simplicius add Xenophanes to the list and claim that the principle is, according to him, one, limited and unmoved. Simplicius supports this interpretation, just like Alexander and Philoponus, at 28.4–8 and 29.5–12 (before correcting it at 29.12–14 with his antilogical reading of Xenophanes’ theology). ²⁹ Xenophanes is associated with Parmenides, on the grounds that they both think that the principle is limited, and he is also contrasted with Melissus, ³⁰ who claims that it is unlimited. Alexander even specifies that Xenophanes’ god is spherical.

For Xenophanes, the idea that the principle (the god) is limited has no support in Aristotle’s Metaphysics or Physics. It might be explained by Xenophanes’ Eleatization: as Parmenides’ master (T1), he was thought to hold the same doctrine, and his god inherited the predicates of Parmenides’ being. ³¹ However, it is unlikely that these commentators used more developed doxographies like those described previously since they had no knowledge of Xenophanes other than what they found in Aristotle: they do not mention any other predicate of the god, nor Xenophanes’ epistemology or physics. Simplicius even admits in in Cael. 522.4–10 that he does not have access to the verses on earth’s infinity and cannot say whether the earth itself or what is beneath the earth, aether, is unlimited. He would have no such doubts had he known a doxography like Hippolytus’ or Pseudo-Plutarch’s, who claim that the earth is not surrounded by air or fire.

Hence, there is no evidence of communication between commentaries and doxographies on Xenophanes. They probably all concluded that Xenophanes’ god is limited either in independent ways, by assimilating his thought to Parmenides’, or because this reading is ancient and was already considered standard when the various commentaries and doxographies were written. We already find traces of it in Timon (third century BC): according to him, Xenophanes’ god is ‘throughout equal’ (ἶσον ἁπάντῃ, fr. 60 Di Marco), which echoes Parmenides’ depiction of the being’s spherical form (μϵσσόθϵν ἰσοπαλὲς πάντῃ). Therefore, I suggest that at least by the time of Timon, Xenophanes’ god was already considered to have the same form as Parmenides’ being. In any case, no commentator seems to have had access to the content of the doxographical accounts described in the previous section.

III. The MXG and Simplicius

Let us now turn to the most problematic testimonies, Pseudo-Aristotle’s MXG 3 and Simplicius’ in Phys. 22.26–23.20. For the sake of brevity, I quote the latter in its entirety:

T2a. Ἀνάγκη τοίνυν τὴν ἀρχὴν ἢ μίαν ϵἶναι ἢ οὐ μίαν, ταὐτὸν δὲ ϵἰπϵῖν πλϵίους, καὶ ϵἰ μίαν, ἤτοι ἀκίνητον ἢ κινουμένην. καὶ ϵἰ ἀκίνητον ἤτοι ἄπϵιρον, ὡς μέλισσος ὁ Σάμιος δοκϵῖ λέγϵιν, ἢ πϵπϵρασμένην, ὡς Παρμϵνίδης Πύρητος Ἐλϵάτης, οὐ πϵρὶ φυσικοῦ στοιχϵίου λέγοντϵς οὗτοι, ἀλλὰ πϵρὶ τοῦ ὄντως ὄντος. μίαν δὲ τὴν ἀρχὴν ἤτοι ἓν τὸ ὂν καὶ πᾶν καὶ οὔτϵ πϵπϵρασμένον οὔτϵ ἄπϵιρον οὔτϵ κινούμϵνον οὔτϵ ἠρϵμοῦν Ξϵνοφάνην τὸν Κολοφώνιον τὸν Παρμϵνίδου διδάσκαλον ὑποτίθϵσθαί φησιν ὁ Θϵόφραστος ὁμολογῶν ἑτέρας ϵἶναι μᾶλλον ἢ τῆς πϵρὶ φύσϵως ἱστορίας τὴν μνήμην τῆς τούτου δόξης· τὸ γὰρ ἓν τοῦτο καὶ πᾶν τὸν θϵὸν ἔλϵγϵν ὁ Ξϵνοφάνης.

Then necessarily, the principle is either one or not one, that is many, and if it is one, it is either unmoved or moved. And if it is unmoved, it is either unlimited, as Melissus of Samos seems to say, or limited, as Parmenides of Elea, son of Pyres, [seems to say]. But they do not speak about a natural element, but about the true being. Xenophanes of Colophon, Parmenides’ teacher, assumed that the principle is one, or being and the whole is one, and that it is neither limited nor unlimited nor moved nor still, according to Theophrastus, who admits that mentioning his opinion belongs to another investigation rather than to the one on nature. For Xenophanes said that this one and whole is god.

T2b. Ὃν ἕνα μὲν δϵίκνυσιν ἐκ τοῦ πάντων κράτιστον ϵἶναι. Πλϵιόνων γάρ, φησίν, ὄντων ὁμοίως ὑπάρχϵιν ἀνάγκη πᾶσι τὸ κρατϵῖν· τὸ δὲ πάντων κράτιστον καὶ ἄριστον θϵός. ἀγένητον δὲ ἐδϵίκνυϵν ἐκ τοῦ δϵῖν τὸ γινόμϵνον ἢ ἐξ ὁμοίου ἢ ἐξ ἀνομοίου γίνϵσθαι· ἀλλὰ τὸ μὲν ὅμοιον ἀπαθές φησιν ὑπὸ τοῦ ὁμοίου· οὐδὲν γὰρ μᾶλλον γϵννᾶν ἢ γϵννᾶσθαι προσήκϵι τὸ ὅμοιον ἐκ τοῦ ὁμοίου· ϵἰ δὲ ἐξ ἀνομοίου γίνοιτο, ἔσται τὸ ὂν ἐκ τοῦ μὴ ὄντος. καὶ οὕτως ἀγένητον καὶ ἀίδιον ἐδϵίκνυ. οὔτϵ δὲ ἄπϵιρον οὔτϵ πϵπϵρασμένον ϵἶναι, διότι ἄπϵιρον μὲν τὸ μὴ ὂν ὡς οὔτϵ ἀρχὴν ἔχον οὔτϵ μέσον οὔτϵ τέλος, πϵραίνϵιν δὲ πρὸς ἄλληλα τὰ πλϵίω. παραπλησίως δὲ καὶ τὴν κίνησιν ἀφαιρϵῖ καὶ τὴν ἠρϵμίαν. ἀκίνητον μὲν γὰρ ϵἶναι τὸ μὴ ὄν· οὔτϵ γὰρ ἂν ϵἰς αὐτὸ ἕτϵρον οὔτϵ αὐτὸ πρὸς ἄλλο ἐλθϵῖν· κινϵῖσθαι δὲ τὰ πλϵίω τοῦ ἑνός· ἕτϵρον γὰρ ϵἰς ἕτϵρον μϵταβάλλϵιν, ὥστϵ καὶ ὅταν ἐν ταὐτῷ μένϵιν λέγῃ καὶ μὴ κινϵῖσθαι

ἀϵὶ δ’ ἐν ταὐτῷ μίμνϵι κινούμϵνον οὐδέν,

οὐδὲ μϵτέρχϵσθαί μιν ἐπιπρέπϵι ἄλλοτϵ ἄλλῃ,

οὐ κατὰ τὴν ἠρϵμίαν τὴν ἀντικϵιμένην τῇ κινήσϵι μένϵιν αὐτό φησιν, ἀλλὰ κατὰ τὴν ἀπὸ κινήσϵως καὶ ἠρϵμίας ἐξῃρημένην μονήν.

He shows that he is one from the fact that he is the most powerful of all. For necessarily, he says, if there were more, power would belong similarly to all of them. And the god is the most powerful and best of all. And he showed that he did not come to be from the fact that what comes to be has to come to be either from what is like or from what is unlike. But what is like is unaffected by what is like. For it no more fits what is like to generate than to be generated by what is like. And if it came to be from what is unlike, being would come to be from not-being. This is how he showed that he did not come to be and is eternal. And [according to him,] he is neither unlimited nor limited, since not-being is unlimited (for it has no beginning nor middle nor end), and multiple things are limited against each other. In a similar way he rejects both movement and stillness. For not-being is unmoved, since nothing else could come in its place nor could it come in another’s place. And things that are more than one are moved; for one thing changes into another. Therefore, when he says that it remains in the same place and does not move:

He always remains in the same place without moving at all,

And it does not suit him to wander at different moments to different places (B26),

he says that it ‘remains’ not in the sense of stillness, the opposite of movement, but in the sense of a permanence that transcends movement and stillness.

T2c. Νικόλαος δὲ ὁ Δαμασκηνὸς ὡς ἄπϵιρον καὶ ἀκίνητον λέγοντος αὐτοῦ τὴν ἀρχὴν ἐν τῇ Πϵρὶ θϵῶν ἀπομνημονϵύϵι, Ἀλέξανδρος δὲ ὡς πϵπϵρασμένον αὐτὸ καὶ σφαιροϵιδές· ἀλλʼ ὅτι μὲν οὔτϵ ἄπϵιρον οὔτϵ πϵπϵρασμένον αὐτὸ δϵίκνυσιν, ἐκ τῶν προϵιρημένων δῆλον, πϵπϵρασμένον δὲ καὶ σφαιροϵιδὲς αὐτὸ διὰ τὸ πανταχόθϵν ὅμοιον λέγϵιν. καὶ πάντα νοϵῖν δέ φησιν αὐτὸ λέγων

ἀλλ’ ἀπάνϵυθϵ πόνοιο νόου φρϵνὶ πάντα κραδαίνϵι. ³²

Nicolaus of Damascus reports in his On Gods that for him, the principle is unlimited and unmoved, Alexander, that it is limited and spherical. But, as is clear from what we just said, he shows that it is neither unlimited nor limited, but that it is limited and spherical because he says that it is everywhere alike. And he says that it thinks everything by claiming:

But he shakes everything with his mind, without work of his thought (B25).

Despite some differences, this account unmistakably parallels that of the MXG. They first share the same theses: Pseudo-Aristotle claims that Xenophanes’ god is eternal, one, homogenous, spherical, neither limited nor unlimited and neither moved nor unmoved, while Simplicius claims that he is one, eternal, neither limited nor unlimited and neither moved nor unmoved, and he discusses homogeneity and sphericity when he examines Alexander’s reading (T2c). Second, both authors provide quite developed arguments for each predicate, and while Simplicius’ version is always shorter than the MXG’s, they are virtually the same, with similar formulations, as we will see in section VI. These strong similarities imply either that they share a source or that Simplicius copies the MXG. ³³ Since Simplicius attributes this interpretation to Theophrastus, the most natural hypothesis is that they both rely on him.

Let us first examine the structure of these two presentations. Like most critics, I distinguish between the positive part of the account and the negative or antilogical one. The claim that Xenophanes’ god is eternal, one, homogenous and spherical is well attested in other doxographical texts. In particular, Hippolytus says: τὸν θϵὸν ϵἶναι ἀίδιον καὶ ἕνα καὶ ὅμοιον πάντῃ καὶ πϵπϵρασμένον καὶ σφαιροϵιδῆ καὶ πᾶσι τοῖς μορίοις αἰσθητικόν (Haer. 1.14.2), to be compared with MXG 3, 977b18–19: τὸν θϵόν ἀίδιόν τϵ καὶ ἔνα, ὅμοιόν τϵ καὶ σφαιροϵιδῆ ὄντα. Pseudo-Aristotle also mentions perception in all parts, as Hippolytus does, when he argues for homogeneity. ³⁴ Therefore he reports the same predicates as Hippolytus except limitation. Moreover, some arguments are paralleled in the doxography. In particular, both Pseudo-Aristotle and Simplicius justify the god’s unity by arguing that he is more powerful than anything else, and gods, if there are many, cannot all be the most powerful. ³⁵ The idea that nothing is more powerful than the god also appears in Pseudo-Plutarch, ³⁶ although he does not use it to demonstrate that the god is one, since he attributes polytheism to Xenophanes. ³⁷ Therefore, even though it is much more fully argued than any other doxographical account, since those usually provide the theses without any justification, the positive report on Xenophanes’ god stems from the same tradition as other doxographies.

By contrast, the antilogies find absolutely no echo in the rest of the literature. They also use a different kind of argument than those that are provided for the positive predicates. The positive predicates are demonstrated as predicates of the god. In particular, the demonstrations of unity and homogeneity only make sense if they apply to the god: he is one because he cannot be dominated, and his homogeneity is connected with the fact that he perceives in all his parts. ³⁸ Hence, even though it has an Eleatic aspect ³⁹ and the god is often assimilated with being, the reasoning is still mainly designed for the god. This is not the case for the arguments of the antilogies: they can apply to any kind of being. As I will show in section V, they are very similar to Parmenides’ and Melissus’ arguments. In Pseudo-Aristotle’s account, this difference between the positive arguments and the antilogical ones even threatens the validity of Xenophanes’ reasoning. For while the demonstration of the god’s unity shows that there is only one god, not that the god is the only thing in existence, the arguments against limitation and movement clearly rest on the undemonstrated premise that he is the one being. ⁴⁰

Therefore, the MXG and Simplicius combine two kinds of text: a doxography that is derived from the same tradition as the others (let us call it D1), and a text that attributes antilogical predicates to the god (D2). The two sources disagree in one obvious aspect: while, according to D1, the god is spherical, according to D2, he is neither limited nor unlimited. Both Pseudo-Aristotle and Simplicius observe the issue, but they deal with it in different ways: while Pseudo-Aristotle uses this tension as an easy way to criticize Xenophanes, ⁴¹ Simplicius tries to make the two claims compatible: this point will be further discussed in section VII.

IV. Origins of the antilogies (D2)

How were these antilogies attributed to Xenophanes? A first issue is whether we should believe Simplicius when he says that Theophrastus himself supported this interpretation. There are two main reasons for doubting this. ⁴² First, according to most doxographers, as we have seen, Xenophanes’ god is limited and unmoved, and since Diels’ Doxographi Graeci, one usually assumes that most of them (in particular Hippolytus) ⁴³ source their information from Theophrastus. Therefore, he cannot be responsible for the antilogies. Second, Xenophanes obviously did not support this antilogical theology, and it is difficult to understand how Theophrastus, who generally has good knowledge of the Presocratics, might have attributed such an unlikely opinion to him. ⁴⁴

Against the first objection, one may provide two answers: either the doxographies did not get their information from Theophrastus (at least not on the god’s spatial and kinetic characteristics), ⁴⁵ or they drew it from a text other than the one Simplicius is quoting. I will deal with this question in section VII. A solution to the second issue is developed by the many critics who accept Simplicius’ testimony. They separate part T2a of my text, where Simplicius says that according to Theophrastus, Xenophanes’ god is neither limited nor unlimited and neither moved nor unmoved, from the argumentative part in T2b (let us call these two sources D2a and D2b). If one does not make Theophrastus responsible for the demonstration in T2b, one can argue that he did not attribute a strong antilogical theology to Xenophanes, but simply thought that Xenophanes was unclear about the spatial and kinetic characterizations of his god. He would thus have said that Xenophanes did not say whether the god is limited or unlimited and moved or unmoved, but this would have been misunderstood as meaning that Xenophanes said that he is neither limited nor unlimited nor moved nor unmoved. ⁴⁶ The arguments of T2b would then have been introduced to justify this misinterpretation of Theophrastus.

I think this approach is the most promising one. On the one hand, it is hard to understand why Theophrastus, who is usually quite reliable on the Presocratics, would not only attribute a complex negative theology to Xenophanes, but also invent various Eleatic arguments to justify it. On the other hand, it is difficult to explain why Simplicius would have attributed a claim to Theophrastus that he did not support. ⁴⁷ Diels suggests the hypothesis that he mistook the MXG for a Theophrastean work, ⁴⁸ but this is incomprehensible if one supposes that Simplicius had access to one or several of Theophrastus’ treatises, especially the doxographical ones. ⁴⁹ Some critics, like Diels and John McDiarmid, try to solve this issue by saying that Simplicius did not have Theophrastus’ book but only knew him through quotations by Alexander; this is quite unlikely, however, since in other passages, Simplicius uses Theophrastus’ interpretation of the Presocratics and Plato to attack Alexander’s. ⁵⁰ Moreover, he attributes to Theophrastus the claim that studying Xenophanes ‘belongs to another investigation than the one on nature’: the MXG says no such thing, which shows that Simplicius does not derive his whole account from this work. ⁵¹ By contrast, there are many parallels between T2a and Aristotelian texts, reinforcing the thesis that it is of Theophrastean origin. For the rejection of Xenophanes’ thought from physics echoes many passages in Aristotle. ⁵² Theophrastus could also draw the claims that Xenophanes was Parmenides’ master and that being and the god are the same thing from T1, ⁵³ although this information is not in the MXG.

As a consequence, one should trust Simplicius when he says that he draws the content presented in T2a from Theophrastus. We still have to explain, however, why Theophrastus would claim that Xenophanes was either unclear or ambiguous regarding the god’s limitation and movement. It is especially surprising in the case of movement since, as I have already mentioned, Xenophanes explicitly claims that the god is unmoved in fragment B26, which is quoted by Simplicius himself.

Some critics look for a justification in Xenophanes’ work. Peter Steinmetz, who argues that Xenophanes is nothing like a systematic philosopher but presented different claims in different poems, thinks that he may have said in some verses that the god is infinite and in others that he is limited, and similarly for movement. ⁵⁴ Concerning limitation, John Palmer suggests the hypothesis that this claim was attributed to Xenophanes because the earth, according to fr. B28, is limited on one side and unlimited on the other. ⁵⁵ Concerning movement, a popular interpretation consists in claiming that this reading originates from the idea that the god is unmoved but the world, to which he is assimilated, is moved. ⁵⁶ However, these explanations would justify saying that, for Xenophanes, the god is both limited and unlimited and moved and unmoved, not that he is neither. Aryeh Finkelberg thinks that Theophrastus was first uncertain about Xenophanes’ belief on movement, since in fragment B26 he only says that the god does not change place, and Theophrastus would rather wonder whether he was impervious to change in general. ⁵⁷ After this moment of doubt, he would have collected evidence from different texts in order to ascertain that Xenophanes’ god was completely unmoved, which would explain the content of the other doxographies. However, it would be quite strange for Theophrastus to read the clause κινούμϵνον οὐδέν in B26 and claim that Xenophanes did not say whether god is κινούμϵνον or ἠρϵμοῦν in another sense of the word. Moreover, it is unclear why Simplicius and Pseudo-Aristotle would have kept the first part of the reasoning and not the second. ⁵⁸

A more promising approach consists in claiming that Theophrastus did not formulate his opinion by reading Xenophanes’ work, but by interpreting Aristotle. This is quite clear in the case of limitation: in T1, as we saw, Aristotle says that it is unclear whether Xenophanes’ unity is formal (which he connects with the fact that the principle is limited) or material (which would mean that the principle is unlimited). ⁵⁹ Theophrastus could then reasonably, like many other interpreters, draw from the Metaphysics that Xenophanes did not say whether his god is limited or unlimited. A problem remains for understanding how this could apply to movement. Unlike Michael Stokes, ⁶⁰ I think that it is insufficient to draw a parallel with the antilogy on limitation: one should explain why Pseudo-Aristotle and Simplicius chose to make another antilogy concerning movement. First, most people agree that Xenophanes’ god is unmoved and, second, immobility was undisputed among the Eleatics. While Melissus says that being is unlimited and Parmenides that it is limited, they both think that it is unmoved, and the impossibility of movement is regarded as one of the main Eleatic claims. It is indeed paradoxical, on the one hand, to Eleatize Xenophanes and, on the other, to claim that his principle is neither moved nor unmoved.

Only Mansfeld has tried to provide an explanation in line with that for limitation: according to him, while in Metaph. A.5, Aristotle includes Xenophanes among those who reject motion, in A.3, 984a27–b8, he says that Parmenides was the first to ‘bother about the explanation of motion’. ⁶¹ Theophrastus would then have deduced that, on the one hand, Xenophanes denies movement and, on the other, he does not discuss this issue. He would then have interpreted Aristotle’s claim that ‘he made nothing clear’ in T1 as also applying to movement. However, in A.3, Aristotle does not claim that Parmenides was the first to see that movement is problematic for monists, but that, among the monists who saw the problem and therefore rejected the possibility of movement, he was the first to provide a solution, which consists in distinguishing two causes, a material one (earth) and a moving one (fire). Xenophanes would then logically belong to the group of monists who simply denied movement, and there is no incompatibility with Metaph. A.5.

In order to support Mansfeld’s reading, one should suppose Theophrastus misinterpreted Aristotle’s claim in A.3. But even if it were the case, the question remains: why would Theophrastus have presented Xenophanes as unclear not only on limitation (where he could rely on a certain interpretation of Aristotle’s T1) but also on movement? Indeed, as Mansfeld claims, when Aristotle says in T1 that Xenophanes ‘made nothing clear’, one can either understand it as meaning that he was not clear on the issue of whether the god is one in form and limited or one in matter and unlimited, or as meaning that he is unclear in general. But in either case, there is no particular reason for Theophrastus to focus on the issue of movement, especially by relying on a text that does not mention Xenophanes at all.

I will propose an interpretation that solves the difficulty and gives a coherent explanation for both antilogies, which are obviously linked. One should take into consideration the fact that this claim appears within a commentary on Aristotle’s Physics. More precisely, Simplicius mentions Xenophanes in his commentary on Phys. I.2, 184b15–22, where Aristotle presents a division on the number of principles:

T3: The principle is necessarily either one or many. If it is one, it is either unmoved, like Parmenides and Melissus say, or moved, as the physicists say, some of them claiming that the first principle is air, others that it is water. If there are many [principles], they are either limited or unlimited, and if they are limited but more than one, they are either two or three or four or any other number. And if they are unlimited, either, as Democritus says, they are one in kind but differ in shape, or they differ in form, or they are even opposite.

Aristotle’s division is asymmetrical: on the side of monism, he further divides depending on whether the principle is moved or unmoved, and on the side of pluralism, depending on whether it is limited or unlimited in number:

Simplicius, however, develops the scheme and introduces the division into moved and unmoved on the pluralists’ side and the one into limited and unlimited on the monists’ side. ⁶² While no pluralist ever professed that his principles are unmoved, the question of limitation (in extension, not in number) is pertinent for the monists (in Phys. 22.15–20). In particular, as Simplicius notes, Parmenides claimed that being is limited while, according to Melissus, it is unlimited. Thus, Simplicius presents the following division for the monists:

This development of the Aristotelian division shows no innovation by Simplicius. For we already find a similar scheme in the commentaries of Themistius (2.27–30) and Philoponus (21.6–25.10). Simplicius even quotes Eudemus (in Physics 22.15–16) as mentioning the possibility of multiple unmoved principles. Mansfeld argues that Theophrastus also had a more complete division in mind. ⁶³ Even if Theophrastus did not present a complete scheme, he could easily have integrated, at the very least, the indications of Metaph. A.5 on the limitation of the Eleatic being into his commentary on Phys. I.2.

If one has such a division in mind and tries to make the content of Metaph. A.5 fit into this framework, it raises the issue of Xenophanes’ place. As we saw in section II.iii, most commentators on the Physics group him with Parmenides and say that his god is limited and unmoved. But Aristotle implies in the Metaphysics that Xenophanes did not express any opinion on the question of limitation. One could still argue that he seems to count him among the immobilists, since the whole section is dedicated to those who think that there is only one unmoved being. However, Aristotle does not explicitly say that Xenophanes made his one god unmoved. Indeed, he only attributes one predicate to the Xenophanean god, unity: he says that Xenophanes was ‘the first among [the Eleatics] to unicize’ and that, according to him, ‘the one is god’. It is possible to interpret this text as meaning that the only attribute of Xenophanes’ god is unity, and that he said nothing else about him.

My hypothesis is that this is exactly what Theophrastus did. It is likely from the content of T2a that he was commenting on Aristotle’s Physics: this explains why he says that studying Xenophanes does not belong to natural science. It is also justified by Aristotle’s own statement in Phys. I.2, 184b25–185a5 that examining the Eleatics is not fit for an investigation of nature, since those who say that being is one necessarily deny the existence of principles, and physics is a study of the principles of nature. ⁶⁴ One can thus imagine that Theophrastus tried, in this commentary, to find a place for Xenophanes in Aristotle’s scheme. He would have looked for indications in the Metaphysics, and the only certain thing that he could deduce from T1 was that Xenophanes’ principle is one and that concerning movement and limitation, he said nothing.

We can find some support for this interpretation in Asclepius’ commentary. He comments on T1 as follows: ὁ μὲν γὰρ Παρμϵνίδης ἔλϵγϵν ἓν ϵἶναι τὸ ὂν καὶ ἀκίνητον καὶ πϵπϵρασμένον, ὁ δὲ Μέλισσος ἓν καὶ ἀκίνητον καὶ ἄπϵιρον, ὁ δὲ Ξϵνοφάνης ἓν μόνον, ‘there were others who talked about the whole as if it were one nature. For Parmenides said that being is one, unmoved and limited, Melissus that it is one, unmoved and infinite, and Xenophanes only that it is one’ (in Metaph. 40.22–27). ⁶⁵ If I am right, it is quite likely that Asclepius was relying, directly or indirectly, on Theophrastus’ commentary.

The whole claim ascribed to Theophrastus in T2a (μίαν δὲ τὴν ἀρχὴν ἤτοι ἓν τὸ ὂν καὶ πᾶν καὶ οὔτϵ πϵπϵρασμένον οὔτϵ ἄπϵιρον οὔτϵ κινούμϵνον οὔτϵ ἠρϵμοῦν) can then be drawn from Metaph. A.5 and justified by an attempt to give Xenophanes a place in the division of Phys. I.2. This hypothesis has the advantage of accounting for both antilogies in the same way.

We can then draw the following conclusions: 1) Theophrastus may have arrived at the idea that Xenophanes did not say whether his god is limited or unlimited or moved or unmoved by commenting on Phys. I.2 with the help of Metaph. A.5. Simplicius was then probably relying on his commentary to the Physics, not on one of his doxographical works such as the Physical Opinions (see section VII). ⁶⁶ 2) Theophrastus developed this interpretation without any particular knowledge of Xenophanes’ work. He does not indeed add anything to what Aristotle says, but just insists on Xenophanes’ lack of clarity concerning the properties of his principle.

V. The arguments for the antilogies (D2b)

I have presented an explanation for the origin of Theophrastus’ claim about Xenophanes, which only consists in asserting that Xenophanes did not say anything about the god’s movement or limitation. Both the MXG and Simplicius go much further, though: first, they say that for Xenophanes, the god has antilogical predicates: he is neither limited nor unlimited and neither moved nor unmoved. Second, they provide arguments to justify this claim. Most critics think that these arguments are not Xenophanean; they do not, however, inquire into their origins, but simply regard them as forgeries. ⁶⁷ This matter should be investigated.

Let me first quote the version of the arguments we find in Pseudo-Aristotle’s MXG:

T4: Ἀίδιον δὲ ὄντα καὶ ἕνα ⁶⁸ καὶ σφαιροϵιδῆ οὔτϵ ἄπϵιρον οὔτϵ πϵπϵράνθαι. ἄπϵιρον μὲν ⁶⁹ τὸ μὴ ὂν ϵἶναι· τοῦτο γὰρ οὔτϵ μέσον οὔτϵ ἀρχὴν καὶ τέλος οὔτ’ ἄλλο οὐδὲν μέρος ἔχϵιν, τοιοῦτον δὲ ϵἶναι τὸ ἄπϵιρον· οἷον δὲ τὸ μὴ ὄν, οὐκ ἂν ϵἶναι τὸ ὄν· πϵραίνϵιν δὲ πρὸς ἄλληλα, ϵἰ πλϵίω ϵἴη. τὸ δὲ ἓν οὔτϵ τῷ οὐκ ὄντι οὔτϵ τοῖς πολλοῖς ὡμοιῶσθαι· ἓν γὰρ <ὂν> ⁷⁰ οὐκ ἔχϵιν, πρὸς ὅτι πϵρανϵῖ.

τὸ δὴ τοιοῦτον ἓν, ὃν τὸν θϵὸν ϵἶναι λέγϵι, οὔτϵ κινϵῖσθαι οὔτϵ ἀκίνητον ϵἶναι· ἀκίνητον μὲν γὰρ ϵἶναι τὸ μὴ ὄν· οὔτϵ γὰρ ἂν ϵἰς αὐτὸ ἕτϵρον οὔτ’ ἐκϵῖνο ϵἰς ἄλλο ἐλθϵῖν. κινϵῖσθαι δὲ τὰ πλϵίω ὄντα ἑνός· ἕτϵρον γὰρ ϵἰς ἕτϵρον δϵῖν κινϵῖσθαι. ϵἰς μὲν οὖν τὸ μὴ ὂν οὐδὲν ἂν κινηθῆναι· τὸ γὰρ μὴ ὂν οὐδαμῆ ϵἶναι. ϵἰ δὲ ϵἰς ἄλληλα μϵταβάλλοι, πλϵίω αὐτὸν ⁷¹ ϵἶναι ἑνός. διὰ ταῦτα δὴ κινϵῖσθαι μὲν ἂν τὰ δύο ἢ πλϵίω ἑνός, ἠρϵμϵῖν δὲ καὶ ἀκίνητον ϵἶναι τὸ οὐδέν. τὸ δὲ ἓν οὔτϵ ἀτρϵμϵῖν οὔτϵ κινϵῖσθαι· οὔτϵ γὰρ τῷ μὴ ὄντι οὔτϵ τοῖς πολλοῖς ὅμοιον ϵἶναι.

Since [the god] is eternal, one and spherical, he is neither unlimited nor limited. On the one hand, not-being is unlimited. For it has no middle nor beginning and end nor any other part, and such is what is unlimited. And being cannot be like not-being. On the other hand, if there were many, they would be limited against each other. But the one has no similarity either to not-being or to the many. For since it is one, it has nothing against which it is limited.

So, the one like this (which he says is the god) is neither moving nor unmoved. For, on the one hand, not-being is unmoved, since nothing else could come in its place nor could it come in another’s place. On the other hand, beings that are more than one are moving, since one thing has to move in another’s place. Now, nothing could move into not-being’s place, for not-being is nowhere. And if things changed places with one another, he would be more than one. For these reasons, two things (or more than one) could move, while what is nothing is at rest and unmoved. But the one is neither still nor moving. For it is similar neither to not-being nor to the many. (MXG 3, 977b2–18)

The arguments rest on the premise that the god is the one being. They consist in opposing, on the one hand, being to not-being and, on the other, the one to the many. Not-being, according to this reasoning, is both unlimited, since it has no beginning or end, and unmoved, because it cannot go anywhere. By contrast, the many are necessarily limited, since they limit one another, and moved, because they have some place to move to. But the god ‘has no similarity either to not-being or to the many’. Hence, he cannot be unlimited and unmoved, like not-being, nor limited and moved, like the many. An important (and obviously problematic) premise of this reasoning is that opposite things cannot have the same predicates: if not-being is unlimited, being cannot be unlimited.

Olof Gigon and Jürgen Wiesner underline the Eleatic character of the arguments. ⁷² According to Gigon, they are directly derived from Parmenides’ poem, but Wiesner argues that they combine elements from Melissus’ and Parmenides’ treatises. The resemblance to Eleatic claims, and in particular Melissus’, is indeed undeniable. In general, the opposition between being and not-being is characteristic of Parmenides, and Melissus also opposes the one and the many (especially in fragment B8, where he calls being τὸ ἕν, ‘the one’). Part of the arguments also comes from Melissus: for he argues in B5–6, as does Xenophanes according to Pseudo-Aristotle and Simplicius, that the many cannot be unlimited because they limit each other. Moreover, Xenophanes supposedly claims that not-being is unmoved because nothing can go into it, just as Melissus says in fragment B7.

Even though part of the reasoning is directly drawn from Parmenides and Melissus, it is not completely Eleatic. Firstly, some of the arguments are not in our fragments: neither Parmenides nor Melissus claims, for example, that the many are necessarily moved. And although the spirit of the arguments is clearly Eleatic, in particular because it rests on a radical opposition between being and not-being, they also contradict some of their claims. For example, Melissus says that being has no beginning or end and is therefore unlimited (B2), but in the argument attributed to Xenophanes, having no beginning or end is characteristic of not-being, and being cannot be unlimited. Immobility also becomes a predicate of not-being instead of being, against Parmenides’ and Melissus’ theses. In general, it appears very un-Eleatic to attribute any kind of property, even negative, to not-being: one of Parmenides’ main claims is that nothing can be said about not-being or known about it, and that one should hold back from any kind of investigation into it. To claim that not-being is unmoved and unlimited violates this requirement. Hence, the reasoning presented in the MXG and Simplicius’ commentary is not simply derived from the Eleatics: it takes inspiration from them but also contradicts their theses in many aspects.

I will argue that this kind of reasoning is characteristic of Gorgias. For Gorgias, in his treatise On Not-Being, ⁷³ refutes the main Eleatic (and more generally Presocratic) ⁷⁴ theses: that there is a being, that it can be known and that it can be said. But he does so by using Eleatic arguments (mostly from Zeno and Melissus). This is explicitly said by Pseudo-Aristotle: ‘he attempts to show [his claim] partly like Melissus, partly like Zeno’ (MXG 5, 979a22–23). For example, Gorgias demonstrates that if being is eternal, it is necessarily unlimited, by relying on Melissus’ argument in B2. ⁷⁵ As a consequence, Gorgias also uses Eleatic arguments for his own purpose.

Moreover, the premise that the predicates of a thing cannot belong to its opposite does not appear anywhere in the work of the Eleatics, but it is assumed by Gorgias, if one follows Pseudo-Aristotle and Sextus Empiricus:

T5: However, if not-to-be is, then to-be, its opposite, he says, is not. For if not-to-be is, to-be should not be. (MXG 5, 979a28–30)

If not-being is, being will not be. For they are opposite to one another, and if to-be belongs to not-being, not-to-be belongs to being. (Sext. Emp., Math. VII.67)

According to Gorgias, then, if not-being has a predicate, then being, since it is its opposite, cannot. This premise is assumed again in Sextus’ summary (VII.80) to show that if being is thought, not-being cannot be thought. There again, the reasoning that is attributed to Xenophanes rests on a premise (that being cannot be like not-being nor the one like the many) that is explicitly assumed by Gorgias.

One may add that, contrary to Parmenides and his followers, Gorgias does not refuse to discuss not-being: for he develops a whole reasoning based on the hypothesis that not-being is, ⁷⁶ and also claims that not-being is not thought. Within a Gorgian framework, then, there is no difficulty in saying that not-being is unlimited or unmoved.

Most importantly, one of the main particularities of Gorgias’ reasoning is to rest on antilogies: he makes a reductio ad absurdum of a thesis by demonstrating that it leads to the conclusion that something is neither x nor not-x. So, in order to assert that nothing exists, he shows that neither not-being nor being are, and to demonstrate that being cannot be, he argues that it is neither born nor eternal and neither one nor many (MXG 5, 979a18–21). This kind of reasoning appears to be an innovation of Gorgias: we find no trace of it in Parmenides and Melissus. Zeno presents a similar kind of antilogical argument, inasmuch as he refutes claims by showing their logical conclusion is that something is both x and not-x. For example, in fragment B3, he claims that if there were many things, they would be both limited and unlimited in number. ⁷⁷ But we only find the negative and antilogical structure in Gorgias’ work. ⁷⁸ The arguments that are attributed to Xenophanes apparently have the same structure as those regularly used by Gorgias.

So many similarities with Gorgias’ reasoning can hardly be a coincidence: I think, on the contrary, that the author of the arguments (D2b) was inspired by him. This means that he would have read Theophrastus’ claim that Xenophanes did not say whether his god is moved or unmoved and limited or unlimited. He interpreted it to mean that he is neither moved nor unmoved and neither limited nor unlimited, and added those Gorgian arguments. There are two possibilities: first, Gorgias himself could have developed these arguments and they were thought to match Theophrastus’ account and therefore attributed to Xenophanes. It is wholly possible that Gorgias did not just show that being is neither born nor eternal and neither one nor many, but also that it is neither limited nor unlimited and neither moved nor unmoved; we have no evidence for that, however. The second possibility is that the arguments were forged for Xenophanes in the style of Gorgias: since the antilogical claims attributed to him echo Gorgias’, it would be quite logical to use him as a model for designing these arguments especially if, as I will now argue, they were invented by the author of the MXG himself.

VI. Role of the MXG

I have distinguished three sources so far for the presentation of Xenophanes’ thought in the MXG and Simplicius’ commentary:

• D1 is a doxography on Xenophanes, which reports a succession of predicates of the god with arguments: the god is eternal, one, homogenous and spherical. Since it is comparable to most other doxographies (especially Hippolytus’), it probably derives from the same source.

• D2a is Theophrastus’ commentary on Aristotle’s Physics, where he would have claimed that Xenophanes’ principle is one, but did not say whether it is moved or unmoved and limited or unlimited.

• D2b incorrectly interprets Theophrastus’ claim as antilogical, that is, as meaning that Xenophanes’ god is neither limited nor unlimited and neither moved nor unmoved, and produces Gorgian arguments to justify these claims.

Now, I will argue that Pseudo-Aristotle is probably the author of D2b and that Simplicius directly used the MXG, not a source common to both treatises. I do not wish to claim that Simplicius had the whole MXG in the form we now possess. He might have had only the part on Xenophanes (we have no evidence that he knew the part on Melissus or Gorgias), have read it quoted in another text or even used a slightly different version of it, either earlier or later than the text we have. In any case, I will show that the text Simplicius used is very similar to our MXG.

The question of whether Simplicius used the MXG or the same source as Pseudo-Aristotle has been little studied, but most recent critics think that they had a common source. ⁷⁹ For the content of Simplicius’ account in T2 is not paralleled in the MXG in its entirety, in particular the claim that Xenophanes’ thought is not relevant to an investigation of nature (T2a) or, even more importantly, the two quotations of Xenophanes, B25 and B26. It is not necessary, however, for Pseudo-Aristotle to be Simplicius’ only source. I have already claimed that Simplicius relied on Theophrastus’ commentary for T2a. Concerning the quotations, I will discuss their presence more precisely in the next section, but it should already be noted that B26 is in direct contradiction with the claims presented in T2a and in T2b. The fragment explicitly says that the god is unmoved, while in T2a–b, Simplicius shows that he is neither moved nor unmoved. As I will elaborate in the next section, I believe that Simplicius got the fragments neither from Theophrastus’ commentary nor from the MXG.

I claimed in section IV that Simplicius is using Theophrastus’ commentary on the Physics. It is likely that T2a reports more or less the whole content of Theophrastus’ account. It explains Xenophanes’ place in Aristotle’s division of Physics I.2 and justifies his rejection from physics, which is quite enough for a commentary, ⁸⁰ but it does not give any details on the doctrine and arguments of the thinker. ⁸¹ Simplicius, who is always interested in the Presocratics, may then have looked for more information on Xenophanes, and found the MXG, which provides arguments and details on Xenophanes’ thought that fit Theophrastus’ commentary. In this scenario, he would also have been influenced by the MXG in misinterpreting Theophrastus’ commentary.

If one examines the arguments in T2b, they are indeed not only strictly parallel to those of the MXG, but if one excepts the quotation of B26 and its commentary, they say nothing more than the MXG: Simplicius only presents a shorter version of Pseudo-Aristotle’s arguments. Let us compare the two versions of the most developed argument, that on generation (I underline the similar passages):

T6: Ἀνάγκη γὰρ ἤτοι ἐξ ὁμοίου ἢ ἐξ ἀνομοίου γϵνέσθαι τὸ γϵνόμϵνον· δυνατὸν δὲ οὐδέτϵρον· οὔτϵ γὰρ ὅμοιον ὑφ’ ὁμοίου προσήκϵιν τϵκνωθῆναι μᾶλλον ἢ τϵκνῶσαι (ταὐτὰ γὰρ ἅπαντα τοῖς γϵ ἴσοις καὶ ὁμοίως ὑπάρχϵιν πρὸς ἄλληλα). Οὔτ’ ἂν ἐξ ἀνομοίου τἀνόμοιον γϵνέσθαι. ϵἰ γὰρ γίγνοιτο ἐξ ἀσθϵνϵστέρου τὸ ἰσχυρότϵρον ἢ ἐξ ἐλάττονος τὸ μϵῖζον ἢ ἐκ χϵίρονος τὸ κρϵῖττον, ἢ τοὐναντίον τὰ χϵίρω ἐκ τῶν κρϵιττόνων, τὸ ὂν ἐξ οὐκ ὄντος ⁸² ἂν γϵνέσθαι. (MXG 3, 977a15–22)

For necessarily, what has come to be has done so either from what is similar or from what is dissimilar. And neither of these is possible. For it does not befit what is similar to be begotten any more than to beget what is similar (for equal things, indeed, have all the same properties, and they are in a similar relationship to each other). Nor should what is dissimilar come to be from what is dissimilar. For if the stronger came to be from the weaker, or the larger from the smaller, or the better from the worse, or on the contrary the worse from the better, being would come to be from not-being, which is impossible.

Ἀγένητον δὲ ἐδϵίκνυϵν ἐκ τοῦ δϵῖν τὸ γινόμϵνον ἢ ἐξ ὁμοίου ἢ ἐξ ἀνομοίου γίνϵσθαι· ἀλλὰ τὸ μὲν ὅμοιον ἀπαθές φησιν ὑπὸ τοῦ ὁμοίου· οὐδὲν γὰρ μᾶλλον γϵννᾶν ἢ γϵννᾶσθαι προσήκϵι τὸ ὅμοιον ἐκ τοῦ ὁμοίου· ϵἰ δὲ ἐξ ἀνομοίου γίνοιτο, ἔσται τὸ ὂν ἐκ τοῦ μὴ ὄντος. (Simpl. in Phys. 22.33–23.3, see T2 for translation)

Simplicius cuts out parts of the MXG’s reasoning or summarizes them, ⁸³ but he also copies many sentences, and nothing in his account is original. I therefore disagree with Finkelberg, who claims that ‘the differences in wording and even, to an extent, in the setting of the arguments, between Simplicius and the MXG suggest their independent derivation from a common source rather than the former’s direct dependence on the latter’. ⁸⁴ The difference of setting he refers to is that Simplicius demonstrates the god’s unity before his eternity, unlike the MXG. But Simplicius might just have changed the order because unity is more important for him than eternity, which plays no role in Aristotle’s division. It appears, then, that Simplicius is either directly copying the MXG, or copying a source that is extremely similar to it. If both were adapting a source with significant differences from the MXG, there would most probably be details in Simplicius that are not in the MXG.

If Simplicius’ source is the MXG (or some text very similar to it), then Pseudo-Aristotle could be the initiator of the addition of Gorgian arguments. This is, as I will argue, the most likely hypothesis. Pseudo-Aristotle closely connects the three Eleatic authors he deals with in his treatise, Melissus, Xenophanes and Gorgias. ⁸⁵ He gives Xenophanes an intermediary role among them: while he thinks that the one being has positive predicates, as Melissus does, he also presents antilogies, which are characteristic of Gorgias. It was then logical for Pseudo-Aristotle to be interested in a version of Xenophanes’ theology that contains antilogies and to consider Gorgias as a model to provide arguments for them.

It is difficult to ascertain whether Pseudo-Aristotle got the idea that Xenophanes’ god is neither limited nor unlimited and neither moved nor unmoved directly from Theophrastus’ commentary on the Physics or from some intermediary source that had already misunderstood Theophrastus’ claim. But since he was particularly inclined to find some connection between Xenophanes and Gorgias because of the structure of his treatise, it is more likely that he is also responsible for the misreading. ⁸⁶

Moreover, one can explain why Pseudo-Aristotle was not only liable to attribute antilogies to Xenophanes because of the similarity it created with Gorgias, but also added arguments. He always provides a justification for the theses of the Eleatics in his treatise, so that he might criticize these demonstrations. Indeed, Pseudo-Aristotle refutes the authors he deals with not by proving that their theses are wrong, but only that their conclusions do not follow from their premises and that their reasoning is invalid. As a consequence, he needed Xenophanes not only to have clear theses, but also deductive arguments to justify them. While he inherited some of these arguments from the tradition (for unity at least, cf. section III), he also had to create new ones, in particular for the antilogies, for which Gorgias provided a perfect model. ⁸⁷

VII. Simplicius’ sources

I have claimed so far that Simplicius has two sources: Theophrastus’ commentary on Aristotle’s Physics and the MXG. I now need to examine two strange aspects of his account: first, the two quotations of Xenophanes that he provides, and second, his combination of the antilogical interpretation with the standard one, according to which Xenophanes’ god is limited and unmoved. I think the two issues are linked. Obviously, Simplicius has access to an account that does not fit his presentation of Xenophanes from Theophrastus and the MXG and that relies on quotations of Xenophanes (let us call it D3). Instead of rejecting one account for the other, he tries to interpret them as compatible.

There are two texts in which Simplicius presents another version of Xenophanes. First, in Phys. 28.4–8, when he introduces Leucippus:

T7: Leucippus the Eleatic or Milesian (both have been said about him) associated with Parmenides in philosophy, but he did not follow the same way as Parmenides and Xenophanes about beings, but the opposite one, as it seems. For while they made the whole one, unmoved, ungenerated and limited, and agreed not to even look for not-being, he …

And then in Phys. 29.5–14, when he tries to show that the Eleatics were talking about the intelligible being:

T8: For some talked about the intelligible and first principle, like Xenophanes, Parmenides and Melissus, Xenophanes and Parmenides saying that it is one and limited … Except that Xenophanes considers it to be the cause of everything, rising above everything and beyond movement and rest and so to say every dichotomy, like Plato in the first hypothesis of the Parmenides.

In T7, Simplicius attributes to Xenophanes’ principle the same predicates as Parmenides’: it is, according to him, not only one and ungenerated, but also limited and unmoved. Even more strikingly, Xenophanes is also held to reject any investigation of not-being. In T8, Simplicius combines both versions of Xenophanes’ thought: he first considers his principle as one and limited, but then he rectifies his claim by saying that it is beyond movement and rest and ‘so to say every dichotomy’.

Mansfeld regards this combination of a positive and a negative account as Middle-Platonic. ⁸⁸ The Middle-Platonists studied the many ways one may talk about the god, one of them being only negative, uia negationis, and another positive, uia eminentiae. On this interpretation, Xenophanes’ god would be limited and unmoved and at the same time neither limited nor unlimited and neither moved nor unmoved. However, these two descriptions are compatible since they correspond to two ways of speaking about the god. Mansfeld attributes this interpretation to Eudorus because of parallels with his account of the Pythagoreans, and thinks that he was the source of both the MXG and Simplicius.

Nevertheless, while Simplicius does indeed try to make the two readings compatible, I do not see any evidence of such an approach in Pseudo-Aristotle. First, he does not attribute immobility to Xenophanes’ god. ⁸⁹ Second, he certainly emphasizes in his criticism (978a20–24) the issue that the god is spherical but not limited, but he does not try to explain how these two predicates could be compatible. Only Simplicius accommodates the claims that Xenophanes’ god is limited and unmoved and that he is beyond polarities. Hence, one must reject Mansfeld’s supposition that the source of Simplicius and Pseudo-Aristotle must have combined the standard and antilogical interpretations of Xenophanes’ theology and thought that they were somehow compatible. Only Simplicius made the second move, and it might be an original interpretation. There is no need for him to have drawn this reading from Eudorus or from Middle-Platonism in general. If one admits that he sourced the antilogies from Theophrastus and the MXG, one may more simply explain his interpretation by the fact that he was aware of the standard interpretation of Xenophanes’ god through another source (which I called D3). This source was then quite similar to what I called D1, the standard account that we find in the MXG. Simplicius must have trusted it, since he does not reject the claim that the god is spherical and unmoved but tries to accommodate the two accounts.

He has to explain, then, how Xenophanes’ god can be limited and unmoved and at the same time neither limited nor unlimited and neither moved nor unmoved. Both issues are dealt with in T2. Concerning limitation, as he responds to Alexander’s statement that Xenophanes’ principle is limited and unmoved (T2c), Simplicius says: ‘but, as is clear from what we just said, he shows that it is neither unlimited nor limited, but that it is limited and spherical because he says that it is everywhere alike’. ⁹⁰ Simplicius does not really reject Alexander’s interpretation, but he tries to explain what Xenophanes meant by πϵπϵρασμένον καὶ σφαιροϵιδὲς (that the god is ὅμοιον). ⁹¹ Limitation and sphericity are then, according to Simplicius, ways to refer to homogeneity, and are therefore compatible with the claim that the god is neither limited nor unlimited.

Concerning movement, Simplicius is faced with an even more difficult task, since he must explain fragment B26, which was probably in D3 too, and which explicitly states that the god is unmoved. He answers this difficulty as follows: ‘he says that it “remains” (μένϵιν) not in the sense of stillness, the opposite of movement, but in the sense of a permanence that transcends movement and stillness (κατὰ τὴν ἀπὸ κινήσϵως καὶ ἠρϵμίας ἐξῃρημένην μονήν)’. Again, his solution consists in reinterpreting the terms: μένϵιν means not ‘to be still’ but ‘in a permanence that transcends movement and stillness’. It is clear that these reinterpretations of the words ‘limited’ and ‘unmoved’ are quite contorted, which shows how eager Simplicius is to make his two sources agree. It must also be noted that he does not solve the issue at all by referring to different ways of speaking about the god, which again weakens Mansfeld’s claim that this interpretation must have Middle-Platonic origins. ⁹²

To sum up, according to my interpretation, Simplicius would have had access to three sources on Xenophanes: first, Theophrastus’ commentary on the Physics (D2a); second, the MXG itself (D2b), which he used to complete what he found in Theophrastus. A third source is D3, according to which Xenophanes, like Parmenides, claimed that being is one, ungenerated, limited and unmoved, and which probably also contained the two quotations that Simplicius copies (B25 and B26). Since, as mentioned in section II.iii, Simplicius has no further knowledge of Xenophanes’ physics or epistemology, one may suppose that D3 only dealt with his theology. It may also have been a source for the other commentaries on the Physics, since they all associate Parmenides and Xenophanes, as Simplicius does in T7 and T8.

Who was D3? Any answer would be extremely conjectural, but I will suggest that it was Theophrastus himself, not in his commentary on the Physics but in another work, maybe the Physical Opinions. ⁹³ Wiesner argues at length that the doxography presented in T7 belongs to Theophrastus, and I will take up some of his points. ⁹⁴ The most striking is the parallel with Alexander in Metaph. 31.7–14:

T9: Πϵρὶ Παρμϵνίδου καὶ τῆς δόξης αὐτοῦ καὶ Θϵόφραστος ἐν τῷ πρώτῳ Πϵρὶ τῶν φυσικῶν οὕτως λέγϵι ‘τούτῳ δὲ ἐπιγϵνόμϵνος Παρμϵνίδης Πύρητος ὁ Ἐλϵάτης’ (λέγϵι δὲ [καὶ] ⁹⁵ Ξϵνοφάνην) ‘ἐπ’ ἀμφοτέρας ἦλθϵ τὰς ὁδούς. Καὶ γὰρ ὡς ἀίδιόν ἐστι τὸ πᾶν ἀποφαίνϵται καὶ γένϵσιν ἀποδιδόναι πϵιρᾶται τῶν ὄντων, οὐχ ὁμοίως πϵρὶ ἀμφοτέρων δοξάζων, ἀλλὰ κατ’ ἀλήθϵιαν μὲν ἓν τὸ πᾶν καὶ ἀγένητον καὶ σφαιροϵιδὲς ὑπολαμβάνων, κατὰ δόξαν δὲ τῶν πολλῶν ϵἰς τὸ γένϵσιν ἀποδοῦναι τῶν φαινομένων δύο ποιῶν τὰς ἀρχάς, πῦρ καὶ γῆν, τὸ μὲν ὡς ὕλην τὸ δὲ ὡς αἴτιον καὶ ποιοῦν’.

About Parmenides and his doctrine, Theophrastus says in the first book of On Natural Things: ‘coming after him (he means Xenophanes), Parmenides the Eleatic, son of Pyres, took both ways. For he both claims that the whole is eternal and tries to explain the generation of beings. He does not think in the same way about both, however, but he supposes that from the point of view of truth, the whole is one, ungenerated and spherical, while from the point of view of the opinion of most people, in order to explain the generation of perceptible things, he makes the principles two, fire and earth, the one as matter and the other as cause and agent’.

This text is explicitly attributed to Theophrastus, makes a connection between Parmenides and Xenophanes, and talks about ‘ways’, ὁδοί, just like T7. The same image of the ways occurs in Pseudo-Plutarch’s Stromates, a representative of the standard doxographical tradition, according to whom ‘Xenophanes of Colophon followed his own way (ἰδίαν τινὰ ὁδὸν), which changed everything that had been said before’ (from Euseb. Praep. evang. 1.8.4). All these texts insist on the relationship between master and disciple and compare the doctrine of each philosopher with those of his predecessors by representing the different philosophical options as ‘ways’. They are not preoccupied, as Simplicius is in T2, by the place those thinkers should have in Aristotle’s division of Phys. I.2, ⁹⁶ but rather present them within, as Steinmetz says, ‘some kind of story of the development of philosophy’. ⁹⁷

We may think, then, that Simplicius relies on Theophrastus both in T7 and T2a; but there is an obvious contradiction between these two texts, since Theophrastus says in one of them that Xenophanes’ principle is limited, but in the other that he dose not say whether it is limited or not. Most critics infer that one of them is not from Theophrastus. But it is entirely possible for Simplicius to have had two different works of Theophrastus in hand. One, in the context of a commentary on Aristotle’s Physics, may have only used Aristotelian content, that is, mostly Metaph. A.5 (T1), and claimed that Xenophanes did not say anything about the god’s limitation or movement, while the other was much more informed, since it provides quotations of Xenophanes. In this second text, Theophrastus may have claimed that Xenophanes’ god is spherical, limited and unmoved. ⁹⁸

It is possible that Simplicius had direct access to this second work of Theophrastus, since he obviously knew quite a number of his works. But he may also have drawn his knowledge of this other work from Alexander, whom he mentions as a supporter of the claim that Xenophanes’ god is spherical and unmoved. This second hypothesis would explain why Simplicius is completely ignorant about Xenophanes’ other doctrines, but also why he blames Alexander for misunderstanding the terms: he would not have understood Theophrastus’ testimony.

My interpretation explains one of the strangest aspects of Simplicius’ presentation: his eagerness to keep both versions of Xenophanes, that with the antilogies and that with a limited and unmoved god, and to make them compatible. Since we know that the first originates from Theophrastus, it could appear strange that Simplicius does not just consider the second reading to be wrong: for he often rejects alternative interpretations of philosophers by referring to Theophrastus. ⁹⁹ Why does he give credit to the claim that Xenophanes’ god is limited and unmoved? A simple reason could be that it also came from Theophrastus: this is why he could not reject one version for the other, but had to interpret Theophrastus’ positive account (D3) as compatible with his negative one (D2). This reading also makes it possible for Theophrastus to be the source of both Simplicius and the other doxographers.

VIII. Conclusion

I started this article by highlighting the gap between the commentaries and the doxographies on Xenophanes’ theology: most of the information that we find in the latter is unknown to Alexander, Philoponus or Asclepius. There are, however, two accounts that mix these two kinds of content. According to my interpretation, the author of the MXG used a standard doxography on the predicates of the god (D1) ¹⁰⁰ and Theophrastus’ reading of Xenophanes in his commentary of Physics I.2 (D2a). Simplicius appears to have an even more complicated story: he relies on Theophrastus’ commentary and uses the MXG to complete it, but he also knows another more informed doxography, maybe from Theophrastus himself, and tries to render all of these texts compatible despite obvious discrepancies on the topic of the god’s limitation and movement.

This investigation leads to the following schema:

This schema reveals the remarkable work of Pseudo-Aristotle and Simplicius, who both combined several sources and tried to interpret and complete them in order to provide a coherent and comprehensive account of Xenophanes’ theology. But this leads to a portrait of Xenophanes that is extremely misleading, which highlights how careful one should be when dealing with testimonies on the Presocratics.