Chapter 3 - Soul’s Self-Motion and Immortality

Summary

In the third chapter, I focus on the concept of self-motion, which is tied to the definition of soul in Plato. Aristotle famously criticises this view in De anima 1.3, showing that the soul is unmoved. I offer the first lengthy discussion of Proclus’ repudiation of Aristotle’s criticism which differs from other Neoplatonist responses. Most importantly, I demonstrate how Proclus develops his own views on self-motion by using Platonic and Aristotelian insights.

3.1 The Problem of Self-Motion

In the previous chapter, I tackled the puzzle of why Proclus adopts both an unmoved intellect and a self-moving soul as principles of motion, when either of these seems prima facie to be sufficient explanations for the origin of motion. As it emerged, Proclus envisions the self-moving soul as a necessary mediator between the unmoved intellect and other-moved bodies since the unmoved mover does not cause directly the motion and being of the other-moved bodies. Given the importance of self-movers, I now turn to a discussion of the nature of self-motion by considering Proclus’ Platonic and Aristotelian background in developing this notion. What does self-motion consist of? Does it denote a single activity of soul or rather a plurality? And are we to understand it as a physical phenomenon, bound to space? The last question specifically has been heatedly discussed since antiquity and is the main focus of this chapter. By ‘spatial motion’ I mean in the following every type of change that requires and occurs in space.

According to Plato, self-motion is not just one characteristic among many of soul but rather its essence (οὐσία), as he emphasises in Phdr. 245e2–4 and Leg. 10.896a3–4.¹ This definition of soul is very consequential, primarily because self-motion is the reason for soul’s immortality in the Phaedrus – a central tenet of Plato’s thought. Moreover, by originating its own motion and that of others, the world-soul is the prime mover in the cosmos. Also, through its self-motion soul maintains ethical independence, as it is able to originate its own actions. Despite the doctrine’s ontological, physical and ethical significance, Plato tells us little about the nature and the workings of self-motion. The picture we get in the Phaedrus, the Timaeus and Laws 10, where Plato characterises soul as self-moved and touches upon the question of self-motion, is complex. Let me briefly set out my own views on this issue which are laid out in greater detail in Marinescu (2021).

The Phaedrus is particularly quiet on this issue. In the immortality proof at 245c1–246e2, which was discussed in the previous chapter, Plato does not further characterise the nature of self-motion. This lack of clarity has led to interpretations of self-motion as either locomotion² or an incorporeal, non-spatial type of motion.³ A possible motive for Plato’s reticence in this dialogue is provided by Griswold (1986: 80): ‘The perception required to state the essence of human and divine soul would seem to transcend human powers; indeed, to try to state the essence of the divine seems hubristic. Perhaps this is why so little is said in the passage [i.e., 245c5–246a2] about the soul beyond the assertion that soul is immortal self-motion.’ Given the poor and inconclusive textual basis in the Phaedrus for either a spatial or non-spatial reading, I believe that judgement needs to be suspended here.⁴

A clearer picture emerges in the Timaeus. Here, Plato’s heavy usage of spatial language to describe the (world-)soul has recently led to a wide consensus among scholars that the soul is spatially extended – either in two or three dimensions – and that its motion is circular locomotion.⁵ This interpretation has supplanted the older view that we ought to take Plato’s language as metaphorical and, thus, not regard the soul as actually extended in space and its circular motion only as a symbolism for rationality.⁶ According to this recent consensus, a more literal interpretation should be preferred not only on exegetical grounds, by taking seriously how precisely Plato describes the composition and working of the world-soul, but also on philosophical grounds. For on such a reading one can avoid a strong dualism and better account for soul–body interaction. Fundamentally, the soul is capable of interacting with the body due to its ontological makeup (i.e., by being composed of a mixture of divisible and indivisible being, sameness and difference) and, more specifically, due to its extension in space – a property it shares with the body. Since the soul and the body are extended in space the soul can transmit its own locomotion to the body and vice versa.⁷ This explains, for instance, how perceptions can affect the soul, as Plato describes in detail (Tim. 43b5–c7).

Unlike in the Timaeus, in Laws X Plato regards self-motion as non-spatial, as I have argued in Marinescu (2021). Through a discussion of Plato’s classification of motion and its context, I reached the conclusion that self-motion cannot be identified with any type of spatial motion (since it is categorically distinct from corporeal types of motion) and that it cannot be regarded generally as requiring space. Most significantly, I demonstrated that there is positive evidence supporting my interpretation. For at 896c5–d5 Plato emphasises that the soul is prior to extension in space and lacks dimensionality. Additionally, Plato’s analogy of rational motion with circular motion at 897d3–e2 only emphasises that soul’s motion is to be understood metaphorically and not as actual spatial motion in a circle.

From this brief discussion of the Phaedrus, the Timaeus and Laws 10 a varied picture of self-motion emerges, whereby the more extensive treatments in the Timaeus and Laws 10 yield contradictory results, as self-motion is presented as spatial in the Timaeus and non-spatial in Laws 10. Generally, Plato provides us in Laws 10 a less corporealist conception of soul and its activity than in Tim. which is grounded in his description of the composition of soul.⁸ Unsurprisingly, this difference between the accounts of the Timaeus and Laws X has supplied later – ancient or modern – commentators with solid exegetical grounds for a spatial or non-spatial understanding of self-motion in Plato. In antiquity, the former view is favoured by Aristotle and some Middle Platonists such as Atticus and Plutarch, while the latter is propounded by the Neoplatonists.

Aristotle’s critique of spatial self-motion is a turning point in the study of this concept as it provides the Neoplatonists with good reasons besides Laws 10 for dismissing a spatial interpretation of self-motion. Aristotle recognises the importance of self-motion and offers a significant critique of Platonic psychology by attacking Plato’s concept of self-motion in De anima 1.3 and by denying that soul in virtue of itself (καθ’ αὑτό) can be seen as in motion. Despite this critique, the idea that the soul is self-moved is universally accepted by the Neoplatonists and taken up by Medieval philosophers.⁹ Thus, for Proclus, self-motion is the essence of soul: τὸ αὐτοκίνητον οὐσία τῆς ψυχῆς (In Tim. 3.328.14 [2.242.20–1]). In addition, self-motion is associated with key Neoplatonist concepts, such as self-reversion (ἐπιστροφὴ πρὸς/εἰς ἑαυτό) and being self-constituted (αὐθυπόστατον). It remains the defining characteristic of soul and the cause of its immortality, as Proclus argues especially in the commentary on the Timaeus and in Elements of Theology.

Yet, how is it possible to maintain the importance of self-motion in light of Aristotle’s criticism, which was well known among Neoplatonists? To answer this question, one needs to focus more on Proclus’ Platonic and Aristotelian sources and his engagement with them. Against this background, Proclus’ theory proves to be far more dependent on both than usually assumed. Thus, I will argue in this chapter that Proclus in his theory of soul as self-moved and immortal is not just indebted to Plato but also to Aristotle’s critique of Plato’s psychology. My discussion is split into four parts.

First (3.2), I discuss Aristotle’s De anima 1.3. Aristotle offers a more general critique of Plato’s self-moving soul, as encountered in the Phaedrus, Laws 10 and the Timaeus, before turning to a specific attack on the latter. Aristotle strongly objects to – what he regards as – Plato’s spatial conception of soul and its motion. This critique is crucial, as it offers strong reasons against associating soul with a spatial motion. After a short discussion of Middle Platonist and early Neoplatonist views on this issue (3.3), I turn to Proclus (3.4). First (3.4.1), I analyse his refutation of Aristotle’ objection in In Tim. Proclus maintains that Aristotle’s reading of Plato is wrong, and that soul and its motion are actually non-spatial and, generally, non-physical in Plato. simultaneously I demonstrate that Proclus in fact agrees with Aristotle’s own view on soul as non-spatial and immaterial. This causes a certain tension between rejecting Aristotle’s reading of Plato and endorsing his understanding of soul. Then, I show that Proclus’ approach to Aristotle is more critical than other Neoplatonists’ (3.4.2) and emphasise the importance of Laws 10 among Neoplatonists and, specifically, Proclus for developing their view on self-motion as non-spatial, which has been overlooked in scholarship (3.4.3). In the last part, I engage with Proclus’ own theory of self-motion (3.4.4). Based on a close reading of ET §§15–17 and §187, I elucidate how Proclus incorporates Aristotle’s critique to argue for soul’s motion as an essentially separate activity from body, which guarantees the immortality of soul. Keeping in line with Plato, Proclus shows why this non-physical motion belongs to soul’s essence and why it guarantees its immortality. For Proclus, self-motion is essential for soul since soul constitutes itself by moving itself. Thus, soul is conceived as self-causing through its specific motion. Most importantly, soul retains a certain ontological and ethical independence from higher beings such as intellect and the One since its own being and actions are grounded in itself.

3.2 Aristotle’s Critique of Soul’s Self-Motion in DA 1.3

Aristotle not only criticises Plato’s self-moved soul in Physics 8, where the emphasis lies on showing that there has to be a higher principle of motion than soul, but also in De anima 1.3–4, where he specifically attacks Plato’s concept of soul in the Phaedrus, Laws 10 and the Timaeus, demonstrating that ‘it is impossible that motion belongs to the soul’ (1.3.406a2).¹⁰ Insofar as soul can be said to be moved, this occurs only incidentally (καθ’ ἕτερον) by being in a moving body but not per se, in virtue of itself (καθ’ αὑτό). His main point of contention is that Plato and other philosophers who maintain that the soul is self-moved wrongly attribute motion to soul because they conceive the soul as body-like. According to Aristotle, Plato maintains that by being extended in space and moving like a body, soul is able to impart its own locomotion to the body (1.3.406b26–8). Aristotle objects to this conception of soul and to framing psychophysical interaction in dualist terms. Instead, he proposes his own famous definition of soul as the first actualisation of an organic body, whereby soul is ‘neither a body nor without a body’ (2.2.414a19–20).¹¹

Aristotle’s treatment of Plato is part of a larger doxographical overview in De anima 1, which in earlier scholarship had been rather overlooked and not regarded as philosophically challenging (e.g., Ross 1961: 19). However, more recent publications by Menn (2002), Carter (2017; 2019) and Ferro (2022) have argued for the importance of De anima 1 regarding the development of Aristotle’s own psychology (as is suggested by 1.2.403b20–5).¹² Menn (2002) specifically sees the psychology of De anima as the ‘result of … an internal critique of Platonic approaches to the soul’ (86). In this light, Aristotle’s engagement with Plato’s definition of soul is of particular significance for understanding his psychology.¹³ Most importantly for my current undertaking, De anima 1.3 proved to be highly influential for later Platonists who reacted to Aristotle’ critique in differing ways but generally agreed with its result, that is, that soul and its activity cannot be extended in space.

Aristotle proceeds to criticise Plato’s theory of a self-moving soul more generally (1.3.405b31–406b25), before attacking specifically the account of the Timaeus (406b25–407b11). I will leave out the latter¹⁴ which is addressed in Section 3.4 on Proclus. In his more general attack on self-motion, Aristotle does not target Plato explicitly, but it clearly emerges from the details and language of his critique that his teacher is among the addressees: ‘We ought first to inquire into motion. For it is presumably not merely false that soul’s essence (οὐσίαν) is such as claimed by those who say that the soul is ‘that which moves itself’ (τὸ κινοῦν ἑαυτὸ), or is ‘that which is able to move itself’ (δυνάμενον κινεῖν), but it is, rather, impossible that motion belong to the soul’ (405b31–406a2). The two expressions concerning the essence of soul as self-moving are similar to Phdr. 245c7 (τὸ αὑτὸ κινοῦν) and Leg. 10.896a1–2 (τὴν δυναμένην αὐτὴν αὑτὴν κινεῖν κίνησιν), and undoubtedly should be seen as references to these dialogues.¹⁵ This is significant, as it shows that Aristotle is acquainted besides Phdr. 245c also with Laws 10.¹⁶ Most importantly, he takes Plato to have a unified theory of self-motion in these texts and interprets him accordingly.

Among the objections Aristotle presents here against Plato’s theory of the self-moving soul, four turn out to be particularly relevant:

1. (1) A thing causing motion does not need to be in motion itself (406a3–4). From this it follows that a self-moving soul does not need to be the principle of motion.

2. (2) If soul moves itself, it must do so with at least one of the four types of motion: locomotion, alteration, growth and decay (406a12–22). However, these cannot be attributed to soul due to its lack of extension.

3. (3) If soul’s essence is self-motion and every motion is a ‘displacement’ (ἔκστασις), then soul departs from its essence (406b11–15).

4. (4) What Plato refers to as soul’s motions – being pained, pleased and so on – in fact belong to the human being using a soul. (408b1–4)

Argument (1), already mentioned in De anima 1.2, refers to Aristotle’s discussion in Physics 8, particularly chapter 5. There, Aristotle argues against a self-moving soul as responsible for the eternal motion of the cosmos and, instead, presents his theory of the unmoved mover. The argument of Physics 8 has been discussed at some length in the previous chapter, where I argued that Proclus fundamentally accepts this Aristotelian insight by backing it up with Neoplatonist metaphysics. This does not imply that Proclus accepts all parts of the argument, such as Aristotle’s analysis of self-motion.

According to (2), if soul moves itself, it has to do so with one (or more) of the types of physical motion.¹⁷ This means essentially that soul moves like a body, having the same kind(s) of motion, which also explains Aristotle’s supposition that soul transmits its own motion to the body (406a30–1). Since undergoing any kind of motion requires place, according to Aristotle’s theory of motion, soul would be then located in a place. Aristotle rejects this since soul has no magnitude and can therefore not be located somewhere. Also, soul would have a natural and an unnatural motion which Aristotle regards as equally absurd.¹⁸ From this it becomes clear that Aristotle takes the motion of soul described by Plato in the Phaedrus, Laws 10 and the Timaeus to be spatial – which, as I have argued in Section 3.1, seems at least for Laws 10 not to be the case. In the latter’s classification of motion, Plato counts the four types of motion mentioned by Aristotle as subtypes of corporeal other-motion, which requires space and is in turn categorically distinct from psychic self-motion.¹⁹ Aristotle here imputes his own concept of motion to Plato by listing only the types of motion he himself recognises.²⁰ Thus, although Aristotle is acquainted with Laws 10, he wrongly characterises soul’s motion as spatial.

Argument (3) takes its starting point from Phdr. 245c7–8, again showing Aristotle’s reliance on Plato in De anima 1.3.²¹ There Plato states: ‘it is only what moves itself that never desists from motion, since it does not leave off being itself (οὐκ ἀπολεῖπον ἑαυτό)’. Plato’s argument is that, since soul’s essence consists in moving itself (or, more precisely, in being a self-mover) and it always moves itself, soul always acts according to its essence. In other words, soul is always itself. On the contrary, Aristotle maintains, if Plato’s definition is taken at face value soul actually departs from its own essence due to its motion:

Moreover, if in fact the soul moves itself, it would itself be moved as well. So, if every motion is a dislodging (ἔκστασις) of the moved in the respect in which it is moved, the soul too would be dislodged from its essence (ἐξίσταιτ’ ἂν ἐκ τῆς οὐσίας), if, that is, it does not move itself co-incidentally, but motion belongs to its essence in virtue of itself.

(406b11–15)

This objection is of crucial importance, as Plato’s ever-moving soul is construed by Aristotle as continuously departing from its own essence. As a consequence, soul would no longer be soul.²² The objection hinges of course on the Aristotelian premise that motion equals displacement of the moved in the respect in which it is moved.²³ Since motion on this view is an essential and not accidental characteristic of soul, states Aristotle, soul does not change accidentally but essentially – which fits Plato’s description of soul in both the Phaedrus and Laws 10.

After his discussion of the Timaeus (406b25–407b11), Aristotle raises another possible objection (4) to his view that the soul is moved only incidentally as part of a body by attributing to it certain motions: ‘we say that the soul is pained and pleased, is confident and afraid (λυπεῖσθαι χαίρειν, θαρρεῖν φοβεῖσθαι), and further that it is angry and also that it perceives and thinks. But all of these seem to be motions. On this basis, one might suppose that the soul is in motion (1.4.408b1–4)’ Although missed by many ancient and modern commentators,²⁴ this passage is clearly a reference to Laws 10. For the pairs λυπεῖσθαι χαίρειν, θαρρεῖν φοβεῖσθαι appear at 896a2–3 as χαίρουσαν λυπουμένην, θαρροῦσαν φοβουμένην where they are regarded as ‘primary-work’ motions of soul. The objection claims that soul has certain mental attitudes that are motions, and that therefore the soul is in motion. Aristotle quickly rejects this objection by maintaining that ‘it is perhaps better not to say that the soul pities or learns or thinks, but that the human being does these things with the soul’ (408b13–15). While it is not entirely clear here whether the subject of the motion is the composite human being or the soul insofar as it is embodied, it is obvious that the soul in virtue of itself (καθ’ αὑτό) does not undergo the motion but rather in virtue of another (καθ’ ἕτερον).²⁵

This is an interesting critique in its own right, as Aristotle rightly points out that the motions of which soul’s self-motion consists in Leg. 10.896e–897a seem to arise from an embodied state of the soul and not from the soul directly, as would a purely cognitive activity like theoretical thinking (νόησις). But it also points towards a more important Aristotelian objection: if soul’s self-motion consists of activities intrinsically linked with the body, how can it guarantee its separability from the body and, thus, immortality, as Plato emphasises in the Phaedrus? For in order to be separable, the soul requires a separate activity from the body, as Aristotle rightly points out in DA 1.1.403a3–16. In Section 3.4.4, I will show how Proclus deals with these concerns.

In conclusion, Aristotle attributes here to Plato a spatial understanding of soul and its motion. Based on my analysis of Plato’s concept of self-motion, I submit that while his spatialist interpretation might fit the Timaeus,²⁶ he is wrong in imputing this view more generally to Plato, since we receive a different picture in Laws 10 (and, perhaps, in the Phaedrus). Plato offers us different perspectives on self-motion in these dialogues: in the Timaeus the perspective is physical, while in Laws 10 psychological and theological.²⁷ Aristotle proves to be quite a superficial reader of Laws 10, although it is clearly Plato’s most elaborate version of his theory of self-motion.²⁸ Thus, unlike Carter (2019), I am more reserved about the quality of his critique, which partly is eristic in nature as well as based on assumptions about motion that Plato does not share.²⁹

3.3 The Legacy of Aristotle’s Critique in Middle Platonism and Neoplatonism

Platonists did not receive Aristotle’s scathing and, at times, captious criticism well. Some early evidence for this tendency can be encountered in the second-century Platonist Atticus who often polemicises against Aristotle. Atticus refers to the significance of the doctrine of the soul’s immortality for ethics, cosmology and epistemology, concluding that ‘absolutely all of Plato’s doctrines are fixed to (ἐξηρτημένων) and dependent on (ἐκκρεμαμένων) the divinity and immortality of the soul – and anyone who does not agree with this overturns the whole of Plato’s philosophy’ (fr. 7.5). After these emphatic remarks, he rebukes Aristotle for rejecting the ‘primary-work motions’, that is, ‘deliberation, thought, anticipation, memory, calculation’ and thus self-motion of soul (fr. 7.6–12).³⁰ Rather, according to Atticus, Aristotle maintains that soul does not possess these motions and that the aforementioned activities can be attributed only to the human as whole – that is, the soul-body compound (see argument (4) in the previous section). But in denying that they belong to soul directly, ‘he would seem to have left us no evidence for its existence or purpose’ (fr. 7.12). Atticus concludes that in regard to soul’s immortality and motion it is not worth consulting a philosopher who has ‘slain’ (ἀποκτιννύντος) the soul and denied any type of motion to it. Atticus rejects here Aristotle’s criticism by making extensive use of the Timaeus, the Phaedrus and Laws 10.

This example demonstrates that, already in the Imperial Age, Plato’s and Aristotle’s views on the soul’s nature and activities were contrasted, especially by focusing on the soul’s self-motion. However, as I argue, this does not result necessarily in a serious and intellectually honest engagement by the Middle Platonists with Aristotle’s criticism in De anima 1.3 or with Aristotle’s psychology as a whole, as the polemical character of fr. 7 makes plain. In spite of the Aristotelian objections, Atticus, as well as Plutarch, for instance, regards soul’s motion as physical³¹ and pays no attention to the problems this view causes, which Aristotle set out in detail.

This is quite different from the Neoplatonists’ more mature treatment of Aristotle. For in Neoplatonism, Aristotle’s insight that spatial motion cannot be attributed to soul is universally accepted.³² Since self-motion needs to be preserved as an essential characteristic of the Platonic soul, it is consequently always taken to be non-spatial. While the ‘founder’ of Neoplatonism, Plotinus, is less focused on self-motion and its significance for soul’s immortality,³³ the issue reappears at the centre of debates on psychology from Iamblichus onwards, as interest in Aristotle’s De anima increases.³⁴ Indeed, Iamblichus sees Aristotle as bringing the study of soul to completion (Ps.-Simpl. In DA 1.10–11). Proclus singles out Porphyry and Iamblichus as breaking up with earlier spatialist interpretations of Plato’s soul (In Tim. 3.145.4–150.12 [2.104.17–108.14]).³⁵ Many later Neoplatonists, who are well-versed in Aristotelian psychology, try to harmonise Aristotle with Plato on the issue of self-motion by claiming that Aristotle objected only to a superficial interpretation of Plato’s words, while being well aware that Plato used the term motion also in a non-physical manner. Proclus, however, stands out as more critical by attacking Aristotle explicitly in this respect and maintaining that Aristotle misunderstood Plato. It is important to note the diverse and nuanced reception of Aristotle’s objections to Plato, as the Neoplatonist response is sometimes wrongly depicted as uniformly positive and harmonistic, ignoring general tendencies in the Alexandrian and Athenian schools as well as more specific differences between individual philosophers.³⁶

In the following section, I demonstrate that (1) Proclus takes a non-harmonist stance on Aristotle by analysing his refutation of Aristotle’s critique in DA 1.3 and contrasting it with other Neoplatonist interpretations of the passage. Next, I argue that (2) Aristotle’s criticism shapes the development of certain psychological views in Proclus and thus serves a positive function.

3.4 Proclus

3.4.1 Proclus’ Refutation of Aristotle’s Critique

Proclus studied psychology extensively.³⁷ According to his biographer Marinus (VP §12.9–11), Proclus began his studies in Athens by reading Plato’s Phaedo and Aristotle’s De anima under the guidance of the scholarch Plutarch of Athens who wrote a now lost commentary on the latter. This proved fruitful since Proclus produced a commentary on the Phaedo as well as the treatise On the Three Arguments Through Which the Immortality of the Soul is Proven by Plato (Περὶ τῶν τριῶν λόγων δι’ ὧν παρὰ Πλάτωνι τὸ τῆς ψυχὴς ἀθάνατον ἀποδείκνυται), where he discusses Plato’s proofs of immortality at Phd. 105b5–107a1, Resp. 10.608c1–611a9 and Phdr. 245c5–246a2.³⁸ The last proof was particularly important for Proclus.³⁹ Through his study of De anima, Proclus also became aware of Aristotle’s criticism of Plato’s psychology and, especially, self-motion. Moreover, Proclus had knowledge of Aristotle’s rejection of the self-moving soul as origin of motion in Physics 8 which he studied intensively, as his work Elements of Physics proves (see Chapters 1 and 2).

Apparently, Proclus felt so provoked by Aristotle’s objections to the Timaeus in De anima 1.3 that he dedicated a (now lost) treatise to refuting these. The treatise probably was entitled Investigation of Aristotle’s Objections to the Timaeus (Investigation; Ἐπίσκεψις τῶν πρὸς τὸν Τίμαιον ὑπὸ Ἀριστοτέλους ἀντειρημένων) and dealt with Aristotle’s criticisms in Metaphysics 12.8, De anima 1.3 and De caelo 1.2–4, 1.10, 2.1 and 3.7–8.⁴⁰ Since Proclus refers to it in his commentary on the Timaeus – which he wrote at the age of twenty-seven according to Marinus – the treatise must be one of his earliest works. While Proclus engaged there with various claims made by Aristotle, one feature of Aristotle’s critique was particularly striking for him (as it is also for some modern scholars): the idea that the Platonic soul is a spatial magnitude (μέγεθος).⁴¹ Based on this literal, non-allegorical interpretation, Aristotle objects to the world-soul’s portrayal as extended and possessing physical motion.⁴² This issue is crucial for comprehending the nature of self-motion. As has been shown earlier, the question of whether Plato conceived the soul and its motions as spatially extended in the Timaeus (and elsewhere) has recently sparked a new interest among scholars with the majority sympathetic to a spatialist reading. In contrast, Proclus argues throughout his commentary on the Timaeus against this spatialist interpretation, and specifically against Aristotle’s analysis. Even though Proclus refers here primarily to the world-soul (since this is the focus of the psychogonia), his conclusions about the non-spatiality of self-motion apply generally to all types of soul, as will be seen below in this section.

In Text A and Text B – taken from commentary on the Timaeus but referring directly or indirectly to his Investigation – he offers a few arguments against Aristotle’s interpretation of the Timaeus.

Text A:

Therefore it is necessary to conceive of this life-engendering (ζῳογόνον) shape of the soul as shapeless (ἀσχημάτιστον) and lacking in extension (ἀδιάστατον), unless we intend to infect ourselves and the theory of Plato with many absurd consequences (πολλῆς ἀλογίας), such as those which Aristotle introduced. (i) He assumed that the soul is a magnitude according to Plato because of the circle, and then demonstrated that, as such, it is incapable of being intellectual (νοεράν), for intellect is indivisible and such as to cognise indivisible intelligible beings. (ii) In addition, if the soul is the sort of thing with magnitude, then it will only be divisible, and not in any way indivisible – much less will it be composed of being that is indivisible in addition to being divisible. [A magnitude], whether it be a circle or a ring, has only a divisible nature and will be in no way indivisible.

(In Tim. 3.339.5–16 [2.250.8–19])⁴³

In this passage, Proclus presents certain ‘absurdities’ arising from Aristotle’s reading of the Timaeus in DA 1.3.406b26–407b11. As Proclus makes clear, Aristotle regards the Timaean soul as a magnitude ‘because of the circles’ (3.339.10 [2.250.12]), that is, because Aristotle takes the circles of the same and the different in the world-soul (Tim. 36b–c) to be spatially extended. However, if soul is understood as a magnitude in the Timaeus, (i) it would be divisible and, thus, unable to know indivisible entities such as the forms (presumably because there must exist an ontological likeness between subject and object of thinking, as Proclus emphasises)⁴⁴ and, moreover, (ii) it would be exclusively divisible and not indivisible; but this explicitly disagrees with the Timaean psychogonia according to which soul is made of both divisible and indivisible being.⁴⁵ The very essence of the world-soul, according to Proclus, is an ‘intermediary between the following extremes: the one cosmic intellect and the entirety of the divisible being that has come to be in the realm of bodies’ (In Tim. 3.192.14–16 [2.141.13–14]). Aristotle is thus portrayed here as a bad interpreter who attributes to Plato a spatialist understanding of soul that otherwise does not fit the overall psychology of the Timaeus. Let us consider the next passage.

Text B:

And since I know what has been written in the Aristotelian objections (ἀντιλήψεις) to the generation of the soul (ψυχογονίαν) and the alleged solutions (λύσεις) of Platonists in response to them, I don’t think it is necessary to expend much effort [over them]. Otherwise, refutation (ἀντιλογίας) would bring us totally astray, for the soul is not a circle like a magnitude (μέγεθος). Neither is it necessary when this supposition has been refuted to think that one has thereby laid hold on Plato’s doctrine (καθάπτεσθαι τῆς Πλατωνικῆς θεωρίας).⁴⁶ For this reason, it seems to me that it is surely acceptable to pass over these matters, for I know that an investigation of them has been undertaken in the book I published specifically on this, Investigations of Aristotle’s Objections to the Timaeus (τῶν πρὸς τὸν Τίμαιον Ἀριστοτέλους ἀντιρρήσεων ἐπισκέψεις).

(In Tim. 3.377.12–21 [2.278.27–279.4])

At the start of text B, Proclus emphasises that he does not want to spend too much time on Aristotle’s objections to the Timaean psychogonia, since he and other Platonists have already dealt with these extensively. Who these other Platonists are, is open to debate. One could be inclined to count Atticus among them, as he criticised Aristotle objections from De anima 1.3 in fr. 7 (quoted in Section 3.3) and is cited by Proclus quite often in his commentary on the psychogonia (cf. e.g., 3.159.2 [2.115.1], 3.208.20 [2.153.29], 3.411.13 [2.306.1]). However, as seen, Atticus (like Plutarch) has a spatialist view of soul with which Proclus disagrees. It seems more probable that Proclus is referring to Plotinus and Porphyry, as they also reacted negatively to De anima 1.3 and defended Plato.⁴⁷

Proclus mentions then four related Aristotelian interpretations of the Timaeus with which he has dealt more extensively in his Investigation:

In it I have dealt with these issues at length and shown that (1) magnitude is not correctly ascribed in the case of the soul according to the Timaeus. As a result of this, I also show that (2) it is no more possible that the soul should cognise the indivisible intelligibles by means of divisible magnitude than that one should make the indivisible fit over the top of the divisible. In addition I show that (3) the motions of the heaven are not identical to the motions of the soul, but rather, according to the teachings of the Timaeus, the former have been made to exist as a result of the latter. Neither (4) is it impossible that the soul should often cognise the same object by means of the same thing, but it is even necessary in the case of discursive thought – if it is indeed the case both that the intelligibles have been limited and also that cognition takes place by means of a circle. So therefore it seems to me that these matters can be set aside at present because I have dealt with them at greater length in the book just mentioned.

(3.377.21–378.9 [279.4–16])

The four claims made by Aristotle in De anima 1.3, which Proclus aims to refute, are:⁴⁸

1. (1) Soul is a magnitude. (DA 1.3.407a3–5)

2. (2) Soul cannot know intelligible entities. (407a10–12)

3. (3) Heaven’s motions are identical with soul’s motions. (407a1)⁴⁹

4. (4) Soul cannot know the same object by means of the same thing. (407a21–32)⁵⁰

In short, Proclus provides four responses:

1. (1) Soul is not a magnitude, since it is not a spatially extended circle.

2. (2) Soul knows intelligible and sensible entities, since it is made of indivisible as well as divisible being which corresponds to the intelligible and sensible realm.

3. (3) Soul produces, and is thus different from, the heaven’s motions.

4. (4) Soul is able to think the same things, as intelligible entities are limited and thinking is circular.

These responses show that, for Proclus, Aristotle’s objections are based on a fallacious understanding of Plato’s text, as also seen in Text A.

Proclus concurs with Aristotle that a spatialist understanding of the Timaean soul really leads to ‘absurd consequences’ (3.339.7 [2.250.10]). Yet, unlike Aristotle, Proclus maintains that this is not a correct interpretation of Plato’s text but rather a misunderstanding of Plato’s teaching based on an erroneous, literalist interpretation of the Timaeus. In this way, Proclus’ position is strikingly close to some modern assessments of De anima 1.3, such as that of Cherniss (1944: 405–6) and Nuyens (1973: 230, n. 34), who regard its interpretation as too literal and thus unfair towards Plato.⁵¹ It is noteworthy that Aristotle himself seems to refer to this interpretative strategy, that is, taking metaphors literally:

It is possible also to argue captiously (συκοφαντεῖν) against the user of metaphorical expression, as though he had used it in its literal sense (ὡς κυρίως εἰρηκότα); for the definition stated will not apply, e.g. in the case of temperance [as a harmony]; for harmony is always found between notes.

(Top. 6.2.139b35–8; tr. Pickard-Cambridge)

However, as I have argued in the discussion of texts A and B, Aristotle has only refuted – according to Proclus – a superficial reading of Plato’s text, without reaching its proper meaning.⁵² Proclus portrays Aristotle as directly attacking Plato’s position and not just one possible reading of the Timaeus. As I show in the next section, this differs from the reading of Ps.-Simplicius and Philoponus.

While Proclus refutes Aristotle’s interpretation, his refutation nevertheless has a constructive function: Proclus – implicitly at least – concedes that Aristotle rightly shows the absurdity of the spatial reading and, consequently, correctly rejects it. Thus, Proclus actually endorses the result of Aristotle’s critique, that is, that neither soul nor its motions are spatial.⁵³ Proclus himself emphasises the ridiculousness of the literalist reading – thus, again, implicitly endorsing and backing up Aristotle’s objections:

Nor is there anything that requires us to accept what is said by those who take the soul’s shape to be genuinely composed from two circles. For if circles are without breadth, how is it possible to split the one from the other without their having breadth? And if they are particular rings, how is the soul composed from them supposed to be interwoven ‘from the middle all the way to the furthest reaches of heaven’ [Tim. 36e2]? For in what manner can rings be stretched through the whole of a spherical body?

(In Tim. 3.338.20–339.1 [2.249.31–250.3])⁵⁴

In this sense, Aristotle’s criticism can be used to understand why a good Platonist should not entertain such a superficial reading but instead should look for a deeper meaning in Plato’s text. Moreover, it points towards a more fundamental agreement between Proclus and Aristotle beyond the correct exegesis of Platonic texts.⁵⁵ While Aristotle misunderstood the meaning of Plato’s text, he still held the correct belief about soul’s nature.⁵⁶

3.4.2 Comparison with Other Neoplatonist Exegeses of DA 1.3

While Proclus explicitly attacks Aristotle for misunderstanding Plato’s text, other Neoplatonists have a more conciliatory attitude according to which Aristotle is aware that he refutes only an apparent meaning of the Timaeus. The difference between them and Proclus lies in their view of Aristotle’s interpretation. A comparison between the different Neoplatonist answers makes Proclus’ diverging hermeneutical strategy stand out more clearly.

Neoplatonists mainly employ two hermeneutical strategies when encountering Aristotle’s criticism. (1) Aristotle criticises a literal interpretation of a Platonic text, knowing well that Plato had a metaphorical sense in mind. Aristotle does this mainly to refute incompetent interpreters of Aristotle who only take Plato literally. (2) Aristotle did not recognise Plato’s metaphorical language and criticised a literal meaning which he took to be Plato’s opinion.⁵⁷ Strategy (1) is characteristic of the Alexandrian school under Ammonius and is found specifically in the works of his pupils, Simplicius and the early Philoponus, who thus show that Aristotle actually does not disagree with the true meaning of Plato’s words, only with their misguided interpretation.⁵⁸ Golitsis (2018) who discusses this strategy at length calls it the ‘preventive function of philosophical criticism’, as it serves to hinder students of Plato to understand his texts in a certain way. Proclus follows (2), as he believes that Aristotle directly criticises Plato on soul’s nature and motion.⁵⁹ Since most commentators use strategy (1), Proclus stands out in his approach to Aristotle.

For instance, Philoponus makes it clear that Aristotle knew that he merely refuted a fallacious interpretation of Plato’s text: ‘But here, too, Aristotle, as he always does, refutes only what appears at face value, so that someone who is unable to perceive what is being conveyed through these riddles would stay at the level of the apparent’ (In DA 116.26–8; tr. van der Eijk). Thus, according to Philoponus, Aristotle did not disagree with the true meaning of Plato’s text.⁶⁰ Of the same opinion is Ps.-Simplicius in his respective commentary:⁶¹

It is this alone [i.e., heaven’s physical motion] that this philosopher [i.e., Aristotle] calls motion, and he contradicts Timaeus about his ascription of a divided extension and activity to the soul, lest we, following the customary use (τῇ συνήθει) of words, should so understand Plato, or think it to be a magnitude or motion in a bodily manner

(40.20–4; tr. Urmson).⁶²

The distinction here between Aristotle’s ‘customary usage of words’ and Plato’s more technical language is common among Neoplatonists.⁶³ According to this passage, Aristotle’s criticism is helpful as it prevents the reader from understanding the Timaeus wrongly. Philoponus’ and Ps.-Simplicius’ statements are part of their general harmonist agenda, derived from their teacher Ammonius, which intends to show that Plato and Aristotle fundamentally agree in their philosophy.⁶⁴

In summary, both Proclus, on the one hand, as well as Philoponus and Ps.-Simplicius, on the other, claim that Aristotle refutes only a superficial reading of the Timaeus. However, the crucial difference between them lies in the intention they attribute to Aristotle. While Philoponus and Ps.-Simplicius maintain that Aristotle is aware that he is not refuting Plato’s teaching but a misguided interpretation, Proclus thinks that Aristotle actually intended to criticise Plato.⁶⁵ He claims that Aristotle rejected what he took to be Plato’s position and, thus, failed to grasp the true meaning of the Timaeus. Given the enormous importance of this dialogue for all Platonists, this amounts to a significant failure on Aristotle’s part. Therefore, Proclus has no qualms in presenting Aristotle as openly dissenting from Plato – something from which the other two would refrain. In this way, it is incorrect to claim, as Carter (2017) does, that the ancient commentators ‘unanimously took Aristotle to have offered in DA 1.3 a strong refutation of a literal interpretation of a mythological Timaeus (and not a refutation of what he took to be Plato’s own views)’ (52). Other prominent authors, such as Gerson (2005), Hadot (2015) and d’Hoine (2016), overlooked this difference.

Yet, regardless of these exegetical differences, it is crucial for all Neoplatonists to conceive soul as self-moved in accordance with Plato’s definition in the Phaedrus and Laws 10. Instead of denying motion of soul altogether, like Aristotle, they in fact maintain, as I will show for Proclus, that self-motion is non-spatial and connected with a central Neoplatonist term – self-reversion. Regarding self-motion’s non-spatiality, Aristotle had a formative influence, as noted. However, it is not just him but also a specific reading of Plato’s dialogues that gives rise to this view.

3.4.3 Why Did the Neoplatonists Conceive Self-Motion as Non-Spatial?

The dialogues that lead to this position are the Phaedrus, the Timaeus and Laws 10. As seen earlier, the majority of modern scholars prefer – just like Aristotle but contrary to Proclus – a literal understanding of Timaeus’ spatial language. While the Timaeus strongly suggests this, the Phaedrus is free of any discussion of the nature of self-motion, except for establishing its priority. Laws 10, however, insinuates quite clearly, as I noted, that self-motion is non-spatial and, thus, radically different from corporeal other-motion.

Since the Neoplatonists have a systematic and unitarian understanding of Plato’s philosophy, which goes against the apparent inconsistency of Plato’s views, they usually form their views by choosing one or two dialogues as authoritative on a certain issue and interpret others in accordance with them.⁶⁶ In the case of self-motion, the most important texts were the Phaedrus and Laws 10, as ample evidence among Middle Platonists and Neoplatonists suggest.⁶⁷ This is not meant to downplay the role of the psychogonia of the Timaeus, which was extremely influential for the development of Platonist psychology – especially in regard to the composition of soul – but less so for self-motion.⁶⁸ The influence of Phaedrus among Platonists as a whole, and the palinode and its immortality proof in particular, is unquestionable.⁶⁹ For instance, in a testimony from his lost commentary on the Phaedrus, Proclus emphasises the superiority of the immortalityproof of the Phaedrus to the proofs in the Phaedo, since it deals with the essence of soul, that is, self-motion, and not just its activity (ap. Philop. De aet. 253.17–254.3).

My choice of Laws 10, however, needs to be defended, as the Laws were neither among the most important dialogues for the Middle Platonists nor later part of the Neoplatonist curriculum.⁷⁰ In consequence, there are almost no scholarly assessments on the influence of Laws 10 on imperial and late antique Platonism. To my knowledge, there is as yet no overview of the influence of Laws 10 on the Middle Platonists and only a short article on the Laws generally in Neoplatonism by Dillon (2001), which offers merely a cursory overview. Additionally, the latter fails to mention the significance of Laws 10 for the Neoplatonist doctrine of self-motion. Yet, references to Laws 10 are abundant among Platonists from the first century AD onwards. Although the Laws were not part of the Neoplatonist curriculum (just as, e.g., the Republic), they were still widely read.⁷¹ Proclus’ teacher Syrianus even wrote a no longer extant commentary on Laws 10.⁷² As a survey of the evidence indicates, for the Middle Platonists, the treatise is important particularly due to the theory of the evil soul⁷³, but for the Neoplatonists, primarily because of the theory of (self-)motion and divine providence.⁷⁴ For instance, Simplicius cites the Phaedrus and Laws 10 in his discussion of self-motion (In Phys. 1247.26–8) and then emphasises the role of the latter: ‘And that he would have the soul be what is self-moved in the proper sense, he indicates by giving its definition in the Phaedrus, and even more clearly in the tenth book of the Laws’ (1248.10–12; tr. Share and Chase). Likewise, Hermias states that Plato clarified in Laws 10 that soul’s self-motion is distinct from corporeal motions (In Phdr. 110.2–4).

In Proclus, there are around 130 references to Laws 10 found in most of his major works.⁷⁵ No other Neoplatonist refers to this work so often. For instance, he cites Laws 10 (besides the Phaedrus) as a source for the view that soul is the origin of motion by being self-moved (In Crat. 53.1–3). He emphasises the significance of Laws 10 for theology due to its treatment of providence (PT 1.5.24.20–1) and other divine attributes.⁷⁶ In consequence, he discusses the work extensively in PT 1.13–16. The three demonstrations of Laws 10 concerning the existence of the gods, their providence and their immutability are, according to Proclus, ἁπάντων … τῶν ἐν θεολογίᾳ δογμάτων ἀρχοειδέστερα (PT 1.13.59.21–2).⁷⁷ In his commentary on the Parmenides, he offers a discussion of the dihairesis of motion and, particularly, of self-motion (7.1155.12–1158.26), which is based on the Laws 10. More specifically, when he treats self-motion in ET, he borrows arguments from Laws 10, as I demonstrated in Chapter 2. This rich evidence points towards the crucial importance of this work for the theory of motion in Proclus, as well as other late Neoplatonists, and makes a separate study on this issue highly desirable.

3.4.4 Self-Motion in ET

After having discussed Proclus’ refutation of Aristotle’s objections to the Timaeus, I now turn to Proclus’ own theory of self-motion. While Proclus does mention self-motion in his commentary on the Timaeus (e.g., 3.323.17–21 [2.239.1–5], 5.229.16–230.5 [3.335.10–23]),⁷⁸ the clearer and philosophically more challenging discussion of self-motion is found in ET §§15–17 and §§186–9.⁷⁹ In the following, I not only discuss these propositions in greater detail than has been done by earlier scholars⁸⁰ but also show that they partly represent an implicit reaction to Aristotle’s criticism of Plato’s psychology, similar to the one found in the commentary on the Timaeus. I focus here primarily on self-motion’s significance for soul’s immortality and how it is conceived by Proclus in reaction to Plato and Aristotle.

In ET, self-motion is part of a discussion of metaphysical and psychological doctrines that lie at the heart of late Neoplatonism. Following Plato, Proclus claims that self-motion is the essence of soul (ET §20.22.8; PT 5.18.64.21–2; In Tim. 3.328.14 [2.242.20–1]).⁸¹ In brief, he argues that self-motion implies self-reversion (epistrophē eis/pros heauto), since by moving itself, soul reverts to itself (§17). Self-reversion, in turn, implies self-constitution, so that whatever reverts to itself constitutes itself (authupostaton) (§§42–3).⁸² Ultimately soul causes itself through its self-motion. The term is thus of fundamental ontological importance.⁸³ But whatever constitutes itself is ungenerated, imperishable and eternal (§§45–9). In this way, Plato’s definition of soul’s essence as self-motion occupies a central role in Proclean psychology and is explained through Proclus’ elaborate and innovative metaphysics. The two crucial terms that Proclus introduces here to account for soul’s immortality are self-reversion and self-constitution. Self-reversion accounts for soul’s (1) incorporeality (§15) and (2) separability from body (§16), while self-constitution for its (3) self-causation (§42). I will discuss these in the following by starting with §17 and a discussion of reversion generally, before moving on to §§15–16 and §42. The discussion is then summed up via §187.

In §17, Proclus identifies the self-mover as belonging to the class of entities that are self-reverting:

Πᾶν τὸ ἑαυτὸ κινοῦν πρώτως πρὸς ἑαυτό ἐστιν ἐπιστρεπτικόν.

εἰ γὰρ κινεῖ ἑαυτό, καὶ ἡ κινητικὴ ἐνέργεια αὐτοῦ πρὸς ἑαυτό ἐστι, καὶ ἓν ἅμα τὸ κινοῦν καὶ τὸ κινούμενον. […] εἰ δὲ ἓν καὶ ταὐτὸν κινεῖ καὶ κινεῖται, τὴν τοῦ κινεῖν ἐνέργειαν πρὸς ἑαυτὸ ἕξει, κινητικὸν ἑαυτοῦ ὄν. πρὸς ὃ δὲ ἐνεργεῖ, πρὸς τοῦτο ἐπέστραπται. πᾶν ἄρα τὸ ἑαυτὸ κινοῦν πρώτως πρὸς ἑαυτό ἐστιν ἐπιστρεπτικόν.

Everything originally self-moving is capable of reversion upon itself.

For if it moves itself, its motive activity is directed upon itself, and mover and moved exist simultaneously as one thing. … And if one and the same thing moves and is moved, it will (as a self-mover) have its activity of motion directed upon itself. But to direct activity upon anything is to turn towards that thing. Everything, therefore, which is originally self-moving is capable of reversion upon itself.

(18.21–20.2)

Proclus makes explicit by the expression ‘originally self-moving’ that the subject is soul and not the ensouled body or living being which is only in a secondary sense self-moving.⁸⁴ Moreover, he clarifies that a true self-mover moves and is moved in the same respect (ἓν ἅμα τὸ κινοῦν καὶ τὸ κινούμενον) and does not consist of distinct parts, whereby one moves and the other is moved.⁸⁵ The latter, of course, is Aristotle’s understanding of self-motion in Physics 8 which Proclus rejects for non-bodily motion. Proclus then concludes that self-motion implies self-reversion, since aiming motion towards oneself means turning towards oneself. What is the relationship between self-motion and self-reversion? I take it that the latter is more fundamental and a condition of self-motion.⁸⁶ Everything self-moving is capable of self-reversion, as Proclus states, but not vice versa. Notably, intellect is capable of self-reversion but not self-moving. The class of self-reverting entities is therefore larger and includes self-moving beings as a sub-class.

3.4.4.1 Excursus: (Self-)Reversion

In order to grasp Proclus’ concept of self-reversion, we need first to understand what the term reversion (epistrophē) means generally. Proclus and likeminded Neoplatonists commonly describe the constitution of being through the triad monē (μονή) – proodos (πρόοδος) – epistrophē (ἐπιστροφή).⁸⁷ This triadic structure is the backbone of Proclus’ metaphysics, delineating the process of causation as a circular activity (ET §33.36.11–12: κυκλικὴν … τὴν ἐνέργειαν) and accounting for the well-orderedness of reality. The triad is found fully expressed in Iamblichus and is to a certain extent present in Plotinus, but its precise historical roots beyond the latter are obscure. Fundamentally, it expresses the idea that an effect remains in its cause,⁸⁸ proceeds from it, and returns to it (§35: Πᾶν τὸ αἰτιατὸν καὶ μένει ἐν τῇ αὐτοῦ αἰτίᾳ καὶ πρόεισιν ἀπ’ αὐτῆς καὶ ἐπιστρέφει πρὸς αὐτήν). An entity ‘remains’ in its cause insofar as the effect is already potentially present in its cause and insofar as, when the effect is realised, it still maintains a similarity to its cause. Were the effect not to remain in its cause, there would be no similarity between cause and effect, as Proclus argues (§30). While μονή describes the similarity or identity of the effect to its cause, πρόοδος denotes the dissimilarity or difference between them. For if the effect remained only in its cause, there would be no causation. Thus, the effect needs to ‘proceed’ from its cause by differentiating itself from it. The third term, ἐπιστροφή, describes the return or reversion of the effect to its cause, as Proclus maintains that every being capable of it desires to return to its origin (§31).⁸⁹ This return, however, does not negate the difference between cause and effect. Rather, the return halts a procession into infinity and, more importantly, guarantees the goodness or ‘well-being’ of the effect, since Proclus takes (in this context at least) the final cause to be identical with the producing cause: ‘[t] hrough that which gives it being it attains its well-being; the source of its well-being is the primary object of its desire; and the primary object of its desire is that upon which it reverts’ (§31.34.34–36.2).⁹⁰ Damascius, thus, fittingly calls the procession ousiopoion (οὐσιοποιόν) and the reversion teleiopoion (τελειοποιόν) (De princ. 2.125.15–16). In order to return to its cause, the effect needs to revert through as many causes as it has proceeded through, as Proclus explains in ET §38. For instance, the soul needs to return to the One/Good via the intellect, since it is caused by the One only in mediation through the intellect. As Gerson (1997: 21) succinctly notes: ‘ἐπιστροφή is the abstract term that refers to the process or event that consists in what the creature does to fulfil its desire for the good’.

3.4.4.2 Soul’s Self-Reversion

From this more general notion of ἐπιστροφή, ἐπιστροφή πρὸς/εἰς ἑαυτό needs to be distinguished. The latter is a crucial term for Neoplatonism and heavily inspired by Hellenistic concepts of introspection as well as by certain Platonic (e.g., Charm. 167a1–7) and Aristotelian (e.g., Met. 12.9) passages on self-reflexivity.⁹¹ In certain higher beings, such as intellect and soul, this reversion occurs via a reversion to itself and then to their higher cause. This is because these beings are not just caused by a higher being but also by themselves. Thus, intellect is caused by the One as well as itself, while soul is caused by intellect (and thus indirectly by the One) as well as itself. Since they also cause themselves they must return to themselves according to the rule that every effect returns to its cause. Self-reversion is thus intimately connected to the notion of self-constitution or self-causation (§§40–51), which I discuss in Section 3.4.4.3.

What is the relevance of self-reversion for soul’s immortality? In brief, entities capable of self-reversion meet two of three crucial requirements for immortality, as Proclus makes explicit later (§§186–7): they are (1) incorporeal (ἀσώματον) (or, more generally, indivisible), and have (2) a separable essence (χωριστὴ οὐσία) from body. The first claim about (1) incorporeality and indivisibility is made in §15:

Πᾶν τὸ πρὸς ἑαυτὸ ἐπιστρεπτικὸν ἀσώματόν ἐστιν.

οὐδὲν γὰρ τῶν σωμάτων πρὸς ἑαυτὸ πέφυκεν ἐπιστρέφειν. εἰ γὰρ τὸ ἐπιστρέφον πρός τι συνάπτεται ἐκείνῳ πρὸς ὃ ἐπιστρέφει, δῆλον δὴ ὅτι καὶ τὰ μέρη τοῦ σώματος πάντα πρὸς πάντα συνάψει τοῦ πρὸς ἑαυτὸ ἐπιστραφέντος· τοῦτο γὰρ ἦν τὸ πρὸς ἑαυτὸ ἐπιστρέψαι, ὅταν ἓν γένηται ἄμφω, τό τε ἐπιστραφὲν καὶ πρὸς ὃ ἐπεστράφη. ἀδύνατον δὲ ἐπὶ σώματος τοῦτο, καὶ ὅλως τῶν μεριστῶν πάντων· οὐ γὰρ ὅλον ὅλῳ συνάπτεται ἑαυτῷ τὸ μεριστὸν διὰ τὸν τῶν μερῶν χωρισμόν, ἄλλων ἀλλαχοῦ κειμένων. οὐδὲν ἄρα σῶμα πρὸς ἑαυτὸ πέφυκεν ἐπιστρέφειν, ὡς ὅλον ἐπεστράφθαι πρὸς ὅλον. εἴ τι ἄρα πρὸς ἑαυτὸ ἐπιστρεπτικόν ἐστιν, ἀσώματόν ἐστι καὶ ἀμερές.

All that is capable of reverting upon itself is incorporeal.

For it is not in the nature of any body to revert upon itself. That which reverts upon anything is conjoined with that upon which it reverts: hence it is evident that every part of a body reverted upon itself must be conjoined with every other part, since self-reversion is precisely the case in which the reverted subject and that upon which it has reverted become identical. But this is impossible for a body, and universally for any divisible thing: for the whole of a divisible thing cannot be conjoined with the whole of itself, because of the separation of its parts, which occupy different positions in space. It is not in the nature, then, of any body to revert upon itself so that the whole is reverted upon the whole. Thus if there is anything which is capable of reverting upon itself, it is incorporeal and without parts.

(16.30–18.6)

Proclus here sets out to prove that no divisible magnitude, including bodies, can revert to itself. His reasoning is rather straightforward.⁹² Self-reversion requires a complete conjunction (16.32 and 18.2: συνάπτεται) of subject and object which leads to them becoming one/identical (16.35). In the case of any divisible substance this is excluded, since its parts occupy a separate place due to their extension in space and cannot each join each other.⁹³ Thus, if one divides a line in half and joins one half with the other, part A will be joined with its opposite A’ and part B with B’. Yet, it is impossible for part A to overlap with both A’ as well as B and B’. This, of course, goes as well for surfaces and solids. In this way, a spatially extended magnitude cannot meet the essential requirement of complete conjunction which constitutes self-reversion. Proclus thus concludes that an entity reverting to itself must be incorporeal and spatially indivisible. It should be noted, however, that Proclus refers here to spatially extended parts; the soul is not partless in a general sense (it has, e.g., different faculties) but only physically.

The reason for rejecting spatial extension for self-reverting beings is the impossibility of an identity between spatially extended subjects and objects in self-reversion. According to Coope (2020), ‘Proclus’s point here is that self-reversion explains both the unity of that which has reverted and the unity of that on which it has reverted’ (129), that is, in the process of self-reversion both subject and object of the reversion become one. This, however, can only occur if they are not spatially extended.

Why does Proclus stress in the title and the conclusion of the proposition that an entity capable of self-reversion must be incorporeal, if his argument actually makes the stronger claim that no divisible entity – that is, neither lines, nor planes, nor bodies – can revert to itself? This needs to be emphasised in order to understand the argument fully. Arguably, Proclus focuses on bodies here due to the Stoic background of the notion of ἐπιστροφή.⁹⁴ The Stoics emphasise the importance of turning inwards in order to perfect one’s character and attain happiness (Epictetus, Ench. 10.1–6; Seneca, Ep. 7.8). At the same time, they adhere to a materialistic physics where the soul is conceived as corporeal so as to interact with the body. It is the latter view that Proclus wants to refute here. According to Proclus, the ethical insight of the Stoics is correct, but it must be paired with a psychology that regards soul as spatially unextended and, hence, incorporeal.

However, scholars have ignored that Proclus argues here more generally against any spatialist conceptions of soul and intellect, since also a geometrically or physically extended soul (or intellect) could not revert to itself.⁹⁵ For instance, Coope (2020: 127) declares that §15 ‘attempts to establish that being a self-unifier (and hence, a self-maker) is incompatible with having bodily parts’.⁹⁶ Yet, Proclus talks here more generally of spatially divisible entities. This is significant, as such views regarding soul were prevalent among Platonists before Plotinus. For instance, Severus, Plutarch and Atticus regarded soul as spatially extended and should be counted among Proclus’ targets here as well.⁹⁷ Moreover, this also includes a rejection of Aristotle’s understanding of the Platonic soul and again emphasises how important it was for Proclus to refute Aristotle’s interpretation.

In the next step (§16), Proclus demonstrates that whatever reverts to itself is not just spatially indivisible but also has (2) an essence separate from body. This is a crucial step in his argument since one could accept the incorporeality/indivisibility argument (1) but still maintain that self-reverting entities such as soul and intellect are dependent on a body as a substrate and require it for their existence and activity. It should thus not be left out in the discussion of self-reversion as for example Coope (2020) does. Proclus argues in the following way:

Πᾶν τὸ πρὸς ἑαυτὸ ἐπιστρεπτικὸν χωριστὴν οὐσίαν ἔχει παντὸς σώματος.

εἰ γὰρ ἀχώριστον εἴη σώματος οὑτινοσοῦν, οὐχ ἕξει τινὰ ἐνέργειαν σώματος χωριστήν. ἀδύνατον γάρ, ἀχωρίστου τῆς οὐσίας σωμάτων οὔσης, τὴν ἀπὸ τῆς οὐσίας ἐνέργειαν εἶναι χωριστήν· ἔσται γὰρ οὕτως ἡ ἐνέργεια τῆς οὐσίας κρείττων, εἴπερ ἡ μὲν ἐπιδεής ἐστι σωμάτων, ἡ δὲ αὐτάρκης, ἑαυτῆς οὖσα καὶ οὐ σωμάτων. εἰ οὖν τι κατ’ οὐσίαν ἐστὶν ἀχώριστον, καὶ κατ’ ἐνέργειαν ὁμοίως ἢ καὶ ἔτι μᾶλλον ἀχώριστον. εἰ δὲ τοῦτο, οὐκ ἐπιστρέφει πρὸς ἑαυτό. τὸ γὰρ πρὸς ἑαυτὸ ἐπιστρέφον, ἄλλο ὂν σώματος, ἐνέργειαν ἔχει χωριζομένην σώματος καὶ οὐ διὰ σώματος οὐδὲ μετὰ σώματος, εἴπερ ἥ τε ἐνέργεια καὶ τὸ πρὸς ὃ ἡ ἐνέργεια οὐδὲν δεῖται τοῦ σώματος. χωριστὸν ἄρα πάντῃ σωμάτων ἐστὶ τὸ πρὸς ἑαυτὸ ἐπιστρέφον.

All that is capable of reverting upon itself has an essence separable from all body.

For if there were any body whatsoever from which it was inseparable, it could have no activity separable from the body, since it is impossible that if the essence be inseparable from bodies the activity, which proceeds from the essence, should be separable: if so, the activity would be superior to the essence, in that the latter needed a body while the former was self-sufficient, being dependent not on bodies but on itself. Anything, therefore, which is inseparable in its essence is to the same or an even greater degree inseparable in its activity. But if so, it cannot revert upon itself: for that which reverts upon itself, being other than body, has an activity independent of the body and not conducted through it or with its co-operation, since neither the activity itself nor the end to which it is directed requires the body. Accordingly, that which reverts upon itself must be entirely separable from bodies.

(18.7–20)

Proclus starts by maintaining that if anything capable of reversion is inseparable from body, then this entails that its essence⁹⁸ is inseparable from body since it belongs to the essence of that thing to be embodied. But if the latter is the case, then its activity must also be inseparable from body, since the activity arises from the essence (18.11: τὴν ἀπὸ τῆς οὐσίας ἐνέργειαν). Therefore, the activity of an essentially embodied thing is inseparable from body. Yet, clearly, this contradicts the result of the previous proposition since self-reverting entities are incorporeal and have an activity separate from the body. Is Proclus’ view that an inseparable essence entails an inseparable activity valid, that is, is it the case that a thing with an embodied essence must have an embodied activity, as some commentators, such as Steel (2006: 242), readily assume?

Proclus’ claim that activity proceeds from essence can be understood by illuminating its Neoplatonist background. When talking about soul and intellect, Neoplatonists commonly use the triadic scheme ousia (οὐσία) – dunamis (δύναμις) – energeia (ἐνέργεια) to refer to their characteristics, whereby δύναμις is understood as active power or capacity (i.e., the Aristotelian δύναμις τοῦ ποιεῖν), not potentiality (i.e., δύναμις τοῦ πάσχειν),⁹⁹ and thus regarded as ontologically prior to activity.¹⁰⁰ Only by having a certain capacity can, for instance, soul act in a certain way. Both capacities and activities of soul derive from its essence which is their ontological foundation.

With this in mind we can better understand Proclus’ argument. Since a thing’s activity depends ultimately on its essence, its activity must be closely related to its essence. Thus, if only the essence would be embodied but not the activity, absurd consequences would follow, as the ‘activity would be superior (κρείττων) to the essence, in that the latter needed a body while the former was self-sufficient (αὐτάρκης), being dependent not on bodies but on itself’ (§16.18.12–14). However, Proclus continues, since it has been shown that whatever is capable of self-reversion is incorporeal, its activity must be incorporeal as well. And if it is incorporeal, it must be separable. This means that entities capable of self-reversion have an incorporeal and separable essence and activity which occurs neither through (διά) nor with (μέτα) a body and, generally, does not require (οὐδὲν δεῖται) a body.¹⁰¹

Because soul is one of the entities envisaged here by Proclus, demonstrating the separability of self-reverting things is relevant for proving the immortality of soul. This is also the case in Plato and Aristotle. Separability is a requirement for soul’s immortality in Plato’s Phaedo where death is defined as a ‘release and separation of soul from body’ (67d4: λύσις καὶ χωρισμὸς ψυχῆς ἀπὸ σώματος).¹⁰² It is also a central concern for Aristotle who discusses separability prominently in the opening chapter of De anima:

There is also a puzzle about the affections of the soul, concerning whether all are common to what has the soul as well or whether there is something peculiar to the soul itself. … If, then, some one of the functions or affections of the soul is peculiar to it, it would be possible for the soul to be separated; but if there is nothing peculiar to it, it would not be separable.

(403a3–12)¹⁰³

In short, Aristotle maintains here that if and only if soul has a peculiar characteristic (ἴδιον) which it does not share with body it could it be separable from body. It is important that Aristotle uses at 403a11 a potential optative ἐνδέχοιτ’ ἄν which indicates that separability is still only a possibility and not a necessity. If soul lacks a specific characteristic and has every affection and function in common with body, it cannot be separated from body and is essentially connected to it.¹⁰⁴ This Aristotelian insight remains crucial for Proclus who refers to it explicitly in On Providence §15.¹⁰⁵

Proclus does not have just a definitional separation in mind (i.e., can anything self-reverting be defined separately from a body or is body necessarily part of the definition?), as often seems to be the case in Aristotle’s De anima, but rather uses the term in an ontological sense, similar to Plato. That is, Proclus wants to know if anything self-reverting depends for its existence on a body or if it can exist independently of it. He shows that the latter is the case and that self-reverting entities qua self-reverting are independent of the body. In comparison to Aristotle, who ultimately argues that only the nous is separable, Proclus focuses on the ontological make-up of soul (and intellect) by singling out the faculty of self-reversion.

3.4.4.3 Soul’s Self-Constitution

Does Proclus’ argument in §16, that soul has a separable essence and activity, mean that it necessarily continues to exist after the demise of its body for an infinite period of time? This point should not be underestimated, since separation does not automatically yield the immortality of soul, as soul could stop existing once it is separated from body or only exist for a short period of time – as Cebes suggests at Phd. 70a2–6. That is, one could maintain that soul is indeed separable from body in essence and activity but that this does not guarantee it an everlasting existence. Separation is thus a necessary but not sufficient condition of soul’s immortality.

Proclus has an interesting answer to this. In §§42–3, he connects self-reversion with the notion of self-constitution.¹⁰⁶ I take this to be the third requirement of soul’s immortality, after (1) incorporeality and (2) separability. The term authupostaton (αὐθυπόστατον), literally that which hypostasises itself or, more commonly, the self-constituted, refers to something which causes itself without being its sole cause.¹⁰⁷ Self-causation – a perennially popular but divisive issue –¹⁰⁸ is adopted here in a restricted sense. It refers chiefly to intellect and soul, as these are beings that, according to Proclus, do not require a substrate in order to exist. They both ultimately derive from the One but also cause their own being. According to Proclus, self-reversion and self-constitution are coextensive. Thus, everything which reverts to itself also constitutes itself and vice versa, as he sets out in §42. This is because the origin of the procession coincides with the goal of the reversion: something self-reverting only reverts to itself because it proceeds from itself. The argument here focuses on the teleological nature of the reversion, as outlined in §31. While procession guarantees a thing’s being, its reversion realises its well-being: ‘for every cause can bestow upon its product, along with the being which it gives, the well-being which belongs to that being: hence it can bestow the latter upon itself also, and this is the proper good of the self-constituted’ (§42.44.16–19).

3.4.4.4 Soul’s Immortality

Taking the three requirements of (1) incorporeality, (2) separability and (3) self-constitution together, Proclus proves the immortality of soul at §187:

Πᾶσα ψυχὴ ἀνώλεθρός ἐστι καὶ ἄφθαρτος.

πᾶν γὰρ τὸ ὁπωσοῦν διαλύεσθαι καὶ ἀπόλλυσθαι δυνάμενον ἢ σωματικόν ἐστι καὶ σύνθετον ἢ ἐν ὑποκειμένῳ τὴν ὑπόστασιν ἔλαχε· καὶ τὸ μὲν διαλυόμενον, ὡς ἐκ πολλῶν ὑπάρχον, φθείρεται· τὸ δὲ ἐν ἑτέρῳ εἶναι πεφυκὸς τοῦ ὑποκειμένου χωριζόμενον ἀφανίζεται εἰς τὸ μὴ ὄν. ἀλλὰ μὴν ἡ ψυχὴ καὶ ἀσώματός ἐστι καὶ ἔξω παντὸς ὑποκειμένου, ἐν ἑαυτῇ οὖσα καὶ πρὸς ἑαυτὴν ἐπιστρέφουσα. ἀνώλεθρος ἄρα ἐστὶ καὶ ἄφθαρτος.

Every soul is indestructible and imperishable.

For all that is capable of being in any way dissolved or destroyed either is corporeal and composite or has its being in a substrate: the former kind, being made up of a plurality of elements, perishes by dissolution, while the latter, being capable of existence only in something other than itself, vanishes into non-existence when severed from its substrate. But the soul is both incorporeal and independent of any substrate, existing in itself and reverting upon itself. It is therefore indestructible and imperishable.

(162.24–31)

A preliminary note on the terminology: anōlethros (ἀνώλεθρος) appears often in the Phaedo, where soul is said to be ἀνώλεθρος (indestructible), if it is athanatos (ἀθάνατος; immortal) (106e2–3).¹⁰⁹ Aphthartos (ἄφθαρτος) is not attested for Plato and seems to be of Aristotelian origin here; but the compound adiaphthartos (ἀδιάφθαρτος) is found in the immortality proof of the Phaedrus (245d3–4), where ungeneratedness (ἀγένητον) is said to imply necessarily imperishability (ἀδιάφθαρτος). For Aristotle, ἀγένητον and ἄφθαρτος are coextensive, as he emphasises in a lengthy discussion in DC 1.11–12. Proclus’ maintains that the De caelo passage is based on the aforementioned Phaedrus passage (In Tim. 2.133.4–16 [1.295.27–296.12]). Proclus does not treat ἄφθαρτος and ἀνώλεθρος as synonyms: instead, ἄφθαρτος stands for the inability of being destroyed by dissolution, while ἀνώλεθρος indicates the inability of being destroyed by separation from the substrate. Together, however, ἀνώλεθρος and ἄφθαρτος indicate that soul is immortal.¹¹⁰

Proclus argues that something can only be destroyed if it is (i) corporeal and composite or (ii) dependent on a substrate. (i) can be excluded since soul has been shown to be (1) incorporeal. (ii) can be rejected on grounds of soul’s (2) separability and (3) self-constitution. While (1) and (2) are implied by soul’s self-reversion, which is named here as a reason for soul’s immortality (162.31),¹¹¹ (3) seems prima facie to be absent in Proclus’ argument, as self-constitution is not explicitly mentioned. Yet, the other reason for rejecting (i) and (ii), next to self-reversion, is that soul exists in itself (162.30: ἐν ἑαυτῇ). This has to be opposed to existing in another (162.28: ἐν ἑτέρῳ), that is, in a substrate. I take the former to be a specific reference to soul’s self-constitution, as according to Proclus ‘all that exists in itself is self-constituted’ (§41.42.31: πᾶν δὲ τὸ ἐν ἑαυτῷ ὂν αὐθυπόστατόν ἐστι). Proclus thus infers not only from (1) and (2) to the immortality of soul but also from (3). Therefore, Menn (2012a: 58) is not right in claiming that ‘Proclus infers from soul’s self-motion to its incorporeality (Elem. Theol. §15) and to its separability from bodies (§16) and thus immortality’.

3.5 Conclusion

The goal of this chapter was to illuminate the Proclean concept of self-motion and its importance for the soul’s immortality by examining its Platonic and Aristotelian sources. This investigation closes a lacuna in scholarship since Neoplatonist conceptions of immortality are seldom studied and often ignored even by more recent treatments of this topic in ancient philosophy such as Long (2019). First, I argued that Laws 10 lent later Platonists, including Proclus, support for a non-spatial, non-physical understanding of self-motion. In a second step, I demonstrated how crucial Aristotle’s criticism of the concept of self-motion as a spatial phenomenon was in further strengthening the view of the Neoplatonists that self-motion is non-spatial. In this context, I showed that Proclus criticises Aristotle’s De anima 1.3 for misunderstanding Plato’s characterisation of the world-soul in the Timaeus. In this open attack on Aristotle, Proclus differs from other Neoplatonists who, in a harmonising spirit, claim that Aristotle only intended to refute what he believed to be a superficial meaning of the Timaeus, but not the true meaning of the text. Finally, in the last part of this chapter I illustrated how Proclus explains soul’s self-motion as self-reversion and self-constitution/-causation which allows him to account for soul’s immortality. In arguing for this, Proclus incorporates the results of his exegesis of Plato and Aristotle within his distinctively Neoplatonist framework. Of the three criteria for soul’s immortality, two go back to Plato and find their mature formulation in Aristotle: incorporeality/lack of spatial extension and separability. The third requirement, self-causation, is Proclus’ specific Neoplatonist contribution. In taking these three requirements together, Proclus is able to offer a convincing proof of the soul’s immortality.