Chapter 2 - The Relationship of the Unmoved Mover and the Self-Mover

Summary

The second chapter concerns the origin of motion in the universe. While Plato assumes a self-moving soul as origin, Aristotle posits an unmoved intellect. Proclus brings these two views together by regarding the unmoved intellect as ultimate source of motion and the self-moving soul as an intermediary entity. I demonstrate that his harmonisation effort goes beyond previous Platonist attempts due to the philosophical reasoning he provides. I also defend Proclus’ assumption of both unmoved intellect and self-moving soul as sources of motion against concerns brought up in scholarship.

2.1 What Is the Origin of Motion?

An analysis of Elements of Physics in the previous chapter has shown how Proclus engages creatively with Aristotle’s Physics and De caelo. Yet, the conclusion of EP reached in §2.21 seems anti-Platonic: How can an unmoved mover be the origin of motion? Plato plainly states in the Phaedrus and Laws 10 that the principle of motion is a self-moving soul. In contrast, Aristotle attacks Plato’s concept of self-motion and posits an unmoved intellect, ontologically superior to souls, as the ultimate origin of motion. Aristotle criticises Plato’s view chiefly in Physics 8.5 and De anima 1.3, maintaining that the soul remains essentially unmoved. His main critique, to put it roughly, focuses on two aspects of essentially the same issue: (1) the origin of motion and (2) the nature of self-motion. Not only is Plato’s concept of self-motion flawed, Aristotle argues, but soul cannot be regarded as the prime mover. As will be seen, this debate has far-reaching consequences for physics, metaphysics and psychology.

Both points of contention are discussed by later Platonists who make explicit references to the texts named above, since they pose a serious threat to the supposed agreement between Plato and Aristotle.¹ While I deal with the nature of self-motion in Chapter 3, here I focus on the first aspect of the debate: What is the ultimate origin of motion in the cosmos? I answer this question by tracing the legacy of this debate in antiquity. In the Imperial Age, both Platonic and Aristotelian accounts of motion are brought together, leading for instance the Middle Platonist Alcinous to talk of intellect and soul as origin of motion. First, I show that this Middle Platonist appropriation is problematic, as it creates a tension between intellect and soul which is left undiscussed. Secondly, I argue that a solution to this problem is offered by Proclus who develops a clear triadic system – different from the binary systems of Plato and Aristotle – of unmoved mover, self-mover and other-moved. This reconciliation shows us how Proclus makes use of Aristotle and sheds light on our understanding of the relationship envisaged by him between Plato and Aristotle.

After a discussion of the Platonic (2.2) and Aristotelian (2.3) views on the origin of motion, I present the contradictory claims made by Alcinous on this topic (2.4) and Proclus’ reception and solution of the problem (2.5).

2.2 The Platonic Background: Self-Mover and Other-Moved

Plato argues in the Phaedrus and Laws 10 for a binary system of movers, consisting of internally and externally moved entities, that is, self-mover and other-moved, whereby the active self-mover is causally superior to the passive other-moved.² The self-mover, that is, an entity causing its own motion, is the archē (ἀρχή) or aitia (αἰτία) of a chain of moving things in the cosmos. Plato reaches this conclusion in a famous argument from the Phaedrus (245c1–246e2) concerning the immortality of the soul.³ The main structure of his argument is the following:

• (P1) Soul is that which is its own source of motion.

• (P2) That which is its own source of motion is immortal.

• (C) Soul is immortal.

This argument, as Bett (1986) has convincingly shown, is in fact dependent on two sub-arguments for (P2).⁴ The first – and arguably the more important – of these is found at the beginning of the proof:

Ψυχὴ πᾶσα ἀθάνατος. τὸ γὰρ ἀεικίνητον ἀθάνατον· τὸ δ’ ἄλλο κινοῦν καὶ ὑπ’ ἄλλου κινούμενον, παῦλαν ἔχον κινήσεως, παῦλαν ἔχει ζωῆς. μόνον δὴ τὸ αὑτὸ κινοῦν, ἅτε οὐκ ἀπολεῖπον ἑαυτό, οὔποτε λήγει κινούμενον, ἀλλὰ καὶ τοῖς ἄλλοις ὅσα κινεῖται τοῦτο πηγὴ καὶ ἀρχὴ κινήσεως.

All soul is immortal. This is because whatever is always in motion is immortal, while what moves something and is moved by something, stops living when it stops moving. So it is only what moves itself that never desists from motion, since it does not leave off being itself. In fact, this self-mover is also the spring and principle of motion in everything else that moves.

(245c5–d1; tr. Nehamas and Woodruff, modified)

The sub-argument in this passage is:⁵

• (P1) Soul is self-moved.

• (P2) Whatever is self-moved is always in motion.

• (P3) Whatever is always in motion is immortal.

• (C) Soul is immortal.

Soul is characterised as a self-mover since it does not depend on any external cause for its own motion but rather causes it itself (P1). Causing this motion is identical to the soul’s essence, as Plato later clarifies (245e2–4): qua its essence it moves itself and others. If the soul would stop moving, it would desist being itself (c7–8: ἅτε οὐκ ἀπολεῖπον ἑαυτό). Thus, soul is defined as something being moved by itself.⁶ As source of motion, it moves the bodies in which it inheres: ‘for every bodily object that is moved from outside has no soul, while a body whose motion comes from within, from itself, does have a soul, that being the nature of a soul’ (245e4–6). Plato thus contrasts the self-motion of soul with the other-motion of bodies – an opposition which is also encountered in the Timaeus and the Laws (see below in this section). As something self-moved, Plato maintains that soul is always in motion (P2). Furthermore, he closely links motion with life: as long as something moves, it is alive. This motion refers to the activity of soul – what type is not specified here – so that being in motion implies being active and this in turn being alive. Thus, that which always moves (ἀεικίνητον) is always alive, that is, immortal (P3), and is for bodies not just a principle of motion but also a principle of life.

That Plato refers here not just to individual souls but also to the soul of the cosmos, the world-soul, he clarifies in the following way: for if the soul were not immortal and would cease to exist as cause of motion ‘all heaven and that which comes to be (γένεσιν)⁷ would collapse, come to a stop, and never have cause to start moving again’ (245d8–e2).⁸ Through its unceasing self-motion, the world-soul sustains the eternal motion and, thus, the existence of the cosmos – an idea that resurfaces at Leg. 10.895a5–b7. As will be seen in the next section (2.3), this makes Plato’s account very similar to the role of Aristotle’s unmoved mover in Physics 8.

Plato also states in the Phaedrus that soul is not just a proximate cause of motion, deriving its causative power from a higher source, but the ultimate principle of motion:

ἀρχὴ δὲ ἀγένητον. ἐξ ἀρχῆς γὰρ ἀνάγκη πᾶν τὸ γιγνόμενον γίγνεσθαι, αὐτὴν δὲ μηδ’ ἐξ ἑνός· εἰ γὰρ ἔκ του ἀρχὴ γίγνοιτο, οὐκ ἂν ἐξ ἀρχῆς⁹ γίγνοιτο.

A principle is ungenerated. Necessarily everything that comes to be comes to be from a principle, but the principle itself does not. For if a principle would come to be from something, then all that comes to be would¹⁰ not come to be from a principle.

(245d1–3; tr. mine)

Plato claims that if a principle comes to be from something else, it is not a principle. Being a principle entails not being generated by something else and, hence, not being dependent on something else in its existence. This has consequences for soul, conceived as a principle of motion. For if soul were generated by another cause, it would not be a principle in the strict sense. In that case, the ultimate cause of motion – which here includes generation¹¹ – would be the cause of soul’s coming to be. Soul would be only a proximate cause of motion in the bodies which it inhabits. But clearly, Plato wants to prevent this view here: since soul is not generated and thus has no superior cause, it is, in the Phaedrus, the ultimate principle of motion and, hence, of generation.

Since Plato’s proof as well as its psychology and cosmology are for various reasons problematic and partly in tension with remarks from other dialogues, I would like to emphasise that Plato’s seemingly superficial discussion in the Phaedrus is due to the context of the proof, which is primarily practical and not theoretical.¹² The proof is part of Socrates’ palinode where the main objective is to describe the impact of erotic madness on the soul. In order to assess this properly, a discussion of soul’s nature is necessary which in turn includes a treatment of its immortality (245b7–c4). Unlike in theoretical treatises on cosmology, such as the Timaeus, or on theology, such as Laws 10, Plato does not put a focus here on discussing these issues in greater detail.

This brings us to Plato’s late work Laws 10 where a similar picture emerges.¹³ His objective is to show that the ‘origin of all motion is … the one that moves itself’ (895b3–5). First, he offers a dihairesis of ten types of motions where self-motion plays a prominent role (893b4–895b8). The ninth kind of motion, other-motion, is in fact a genus of which the first eight non-self-motions are different species. These eight types of other-motion are grouped as pairs. Thus, Plato ultimately distinguishes here, as in Phdr. 245c–d, between self-motion and other-motion and therefore between self-movers and other-moved entities.¹⁴ The former is portrayed as a source of motion to itself and to others; a potentially infinite chain of moved movers is brought to a halt by the introduction of a self-mover (894e4–895a3). As such, self-motion is ‘most powerful (ἐρρωμενεστάτην) and radically effective (πρακτικὴν διαφερόντως)’ (894d1–2) and has ontological priority over other kinds of motion.¹⁵ What self-motion precisely consists of, is discussed in the next chapter. Most importantly, self-motion is associated with soul who is the ‘first cause of generation and destruction of all things’ (891e5–6: πρῶτον γενέσεως καὶ φθορᾶς αἴτιον ἁπάντων; cf. 896d8, 899c7), while other-motion is secondary, as it depends on self-motion, and is identified with bodies. Thus, soul is moved internally (by itself), while bodies are moved externally (by soul).

As in the Phaedrus, this dichotomy leads to a subordination of corporeal motion, that is, other-motion, to psychic motion, that is, self-motion, which Plato expresses by calling the former secondary-work (δευτερουργοί) motion and the latter primary-work motion (πρωτουργοί) (897a4–5). Both terms are hapax legomena and apparently coined by Plato. This distinction emphasises that the primary causal force lies in the soul that initiates motion, not in the body, as Plato claims that ψυχὴν μὲν προτέραν γεγονέναι σώματος …, σῶμα δὲ δεύτερόν τε καὶ ὕστερον, ψυχῆς ἀρχούσης, ἀρχόμενον κατὰ φύσιν (896c1–3). Plato stresses the causal priority of soul over body in accordance with the general purpose of his discussion of motion in Laws 10 which intends to prove that ‘soul [i.e., world-soul] drives all things in the heavens and on earth and in the sea through its own motions’ (896e8–9). Clearly, the account of motion in Laws 10 is very similar to and compatible with the theory found in the Phaedrus.

Although the picture is more complicated in the Timaeus, there is compelling evidence – contrary to what Bett (1986: 23–6) and, recently, Corcilius (2018: 62) claim – for a distinction between self-movers and other-moved objects as well.¹⁶ Soul is described as self-moved (37a6–7: κινουμένη διὰ πάσης ἑαυτῆς; b5: τῷ κινουμένῳ ὑφ᾽ αὑτοῦ) and bodies as other-moved (46e1: ὅσαι δὲ ὑπ’ ἄλλων μὲν κινουμένων). The two types of motion are contrasted at 89a1–3. Plato also suggests that soul’s motion is unceasing.¹⁷ This minimises the difference in this respect between the psychology of the Timaeus, on the one hand, and the Phaedrus and Laws 10, on the other.¹⁸

In conclusion, Plato is committed to the view that the origin of motion can only be something that is itself in motion. This, I submit, is not so much based on ignorance, that is, Plato is simply not considering the possibility of an unmoved mover – as is sometimes suggested.¹⁹ In fact, it has been argued by Brown (1998: 199–200) and Crubellier (2017: 20–4) that the origins of the concept of an unmoved mover can be found in Plato. This view is based on their specific reading of Soph. 248d–249b, where the forms act by not being affected. Indeed, as I would like to add, this view can be also found in Socrates’ palinode, where the goal of the souls’ procession is the vision of the forms (246d6–248c2). Here the forms seem to act as unmoved movers for the souls which desire them as ‘appropriate nourishment for their best [i.e., rational] part’ (248b7). Yet, how this position can be harmonised with the notion of the self-moving soul remains obscure.²⁰ Similar to Aristotle, Plato probably does not believe that a desire triggered by something external precludes the possibility of the soul still moving itself, that is, by preserving a certain causal autonomy such as the ability to choose whether to pursue the object of desire. Regardless of this, in maintaining soul’s self-motion as origin of motion Plato bases his view on (either of) two fundamental pre-supposition(s) of his theory of causation:

x can only cause the property F in y if x itself is F.

Ex.: fire can only cause hotness in the pot if fire itself is hot.

Or

x can only cause the property F in y if x is not un–F.

Ex.: fire can only cause hotness in the pot if fire is not un-hot (i.e., cold).

While (1), dubbed ‘transmission theory of causation’, has been shown to lead to some obvious problems (e.g., must the cause of death be dead?) and is perhaps not always applied by Plato, the weaker claim (2) seems to be generally accepted as one of Plato’s laws of causation.²¹ In regard to his theory of motion, I take it that Plato upholds (1), while (2) would at any rate provide a further reason for assuming that the principle of motion is not unmoved. Just as ugliness cannot cause beauty, the unmoved cannot cause the moved. It is thus because of these underlying philosophical commitments, I believe, that Plato posits a self-moving entity as ultimate cause of motion. Since these two presuppositions offer a straightforward explanation for Plato’s choice for the principle of motion, I see little reason for assuming, as Menn (2012a: 57) proposes, the less economical solution that Plato had in mind an argument later used by Sextus Empiricus: ‘what moves [something] is acting, what acts is in motion, therefore what moves [something] is in motion …’ (Against the Physicists 2.76), or, contrapositively, what is not in motion is not acting or doing anything, and therefore cannot move something else, since moving something is an activity.

2.3 Aristotle’s Response: Unmoved Mover and Other-Moved

Like Plato, Aristotle also presents us a binary classification of movers but one quite different from his teacher’s, as it consists of an entity lacking motion and one externally moved, that is, an unmoved mover and an other-moved. Interestingly enough, he preserves the autoreferential aspect of the prime mover, insofar as it is self-thinking according to Metaphysics 12.7.²² Unlike Plato, Aristotle does not trace back the origin of motion to another motion but to something that, while itself lacking motion, can impart it to others. Self-movers in a strict, that is, Platonic, sense do not exist, since they are actually constituted of an unmoved mover (soul) and an other-moved part (body). The view that soulless bodies are other-moved is clearly taken over from Plato.²³ Aristotle thus claims that Plato’s concept of self-motion is flawed, as is his understanding of the origin of motion. I focus here on Aristotle’s critique of Plato’s position in Physics 8,²⁴ since it is more relevant on the origin of motion, whereas De anima 1.3 criticises the concept of self-motion, which is discussed in the next chapter.²⁵

In Physics 8, Aristotle proves that all physical motion ultimately stems from an unmoved mover.²⁶ After concluding in the first part of his investigation (8.1) that motion has to have always existed in the universe, Aristotle reaches another stage at which he goes on to identify the origin of this motion. First, he makes the fundamental claim that everything in motion needs to be moved by something (8.4). Since this line of argumentation runs the risk of an infinite regress, as also shown by Plato at Leg. 10.894e4–895a4, that is, the motion of a moved mover could be traced to another moved mover and then to another and so on, Aristotle concludes that the motion in the world needs to originate either in an entity itself in motion, that is, a self-mover, or in a being that is not in motion, that is, an unmoved mover.

Aristotle infers that it has to be the latter via an analysis of self-motion (8.5) which – due to its crucial role in his argumentation – he discusses at great length:²⁷

But surely, if it should be necessary to examine whether the self-mover (τὸ αὐτὸ αὑτὸ κινοῦν) or the thing moved by another (τὸ ὑπ’ ἄλλου κινούμενον) were the cause (αἴτιον) and principle (ἀρχὴ) of motion, everyone would say the former. For what is a cause in its own right is always prior to what is itself a cause through another. So we must examine this question by making another beginning: if something moves itself, how and in what way does it cause motion?

(8.5.257a27–33)

In 8.5 Aristotle argues that self-motion cannot be conceived as something moving itself as a whole. Rather, self-motion can be reduced to an unmoved mover in the self-mover, that is, animate being: ‘Of the whole [i.e., self-mover], therefore, one part will cause motion while remaining unmoved, and one part will be moved’ (258a1–2). Thus, a self-mover consists of an agent (mover) and a patient (moved) part and these two need to be distinct: in an animal the soul is the mover and the body the moved. Only in this derivative sense can we say that an animal is a self-mover.

Aristotle essentially presents two connected arguments for this view. (1) The same thing cannot be simultaneously both agent and patient of the same motion: ‘For [the self-mover] would be transported as a whole, and it would transport with the same motion, being one and indivisible in form, and it would be altered and alter’ (257b3–4). (2) He also excludes that something is potential and actual in regard to the same aspect, that is, motion: when an entity capable of motion is actualised and moves, it has to be actualised by something already possessing this quality, that is, a mover: ‘[the moveable] is in motion through potentiality (δυνάμει), not through actuality (ἐντελεχείᾳ), and the potential is in process to realisation (ἐντελέχειαν), and motion is the incomplete realisation (ἐντελέχεια) of the moveable. But the mover is already actual (ἤδη ἐνεργείᾳ)’ (257b6–9).

After having thus established that a self-mover is in fact made up of two parts, he asks whether a self-mover thus understood could be the ultimate origin of motion in the cosmos. Since the unmoved part is accidentally set in motion when the self-mover moves, he rules out the possibility that the prime mover is a self-mover: ‘if something belongs to the class of things that are unmoved but move themselves accidentally, it is unable to cause continuous motion’ (8.6.259b20–2). One reason is that self-movers are dependent on external stimuli in order to cause motion (259b1–19). But it had been already established that there needs to be an eternal and continuous motion in the cosmos, on which the generation of animals depends (258b16–259a6). If then the prime mover cannot be a something accidentally moved, it needs to be absolutely unmoved, possessing only the capacity to cause motion, without being part of a self-mover (259b22–31). This, I think, poses a strong argument against associating the prime mover with the soul (or something analogous) of the cosmos, as some ancient and modern interpreters of Aristotle do.²⁸

Regarding the nature of the prime unmoved mover, Aristotle maintains in the last chapter, 8.10, that it is ‘without parts and without magnitude’, that is, indivisible²⁹ and (spatially) unextended. He shows this by a number of arguments, for example, that a magnitude cannot have an infinite power which would be needed to cause the eternal motion.³⁰ This again, is an implicit critique of Plato, as Aristotle understands the Platonic world-soul as spatially extended which would thus be unable – so Aristotle – to cause the cosmos’ motion. Generally, it seems that Aristotle regards Plato’s world-soul as prime mover and disregards the causal function of the demiurge, seeing the latter as merely mythical (apparently like Plato’s successors Speusippus and Xenocrates).³¹

In the two preceding sections I have argued that, for Plato, an entity itself in motion such as the world-soul is the prime mover. In contrast, for Aristotle the cosmos’ eternal motion can only be caused by something that is absolutely unmoved such as the divine intellect. Aristotle does not dismiss the world-soul as a possible prime mover solely on the ground that he rejects the – supposedly! – Platonic concept of a spatially extended world-soul. Rather, Aristotle would regard the world-soul together with the cosmos as a composite, whereby the unmoved world-soul would be set in motion accidentally by the cosmos’ motion. Regardless of whether Aristotle’s criticism is correct, clearly a combination of both positions is problematic, as they seem to be exclusive alternatives: the principle of a chain of motion must be either self-moving or motionless.

2.4 The Middle Platonists: Intellect as Unmoved Mover and Soul as Self-Mover

In the Imperial Age, some Platonists start to make heavy use of Aristotelian theories.³² Motion is no exception, as references to an unmoved intellect as principle of motion suggest. Yet, this area of harmonisation has so far not been studied in its own right in Middle Platonist scholarship despite its interesting philosophical character and its importance for later Neoplatonists.³³ It is in Middle Platonism, I maintain, that a change occurs from a binary system of movers, as seen in Plato and Aristotle, to a triadic one, where both intellect and soul are regarded as sources of motion and bodies as externally moved. I discuss here Alcinous as a fitting example of such an appropriation in order to show that some Middle Platonists followed Aristotle in accepting an intellect as origin of motion. As will be seen, this clashes with the Platonic view of soul as ultimate cause of motion, leading to an inconsistency which is not adequately dealt with. Only later in Proclus, as I argue, we find a more satisfying solution.

Platonists of this period, who focus primarily on Tim., generally associate the demiurge with a transcendent νοῦς.³⁴ Their system is usually characterised by three principles: god (as νοῦς), forms and matter.³⁵ The forms are often identified with the thoughts of god which has led Sharples (1995) to the observation that the Middle Platonists have actually two and a half principles.³⁶ Moreover, god is sometimes characterised as unmoved mover. The motivation of the Middle Platonists to associate νοῦς with the origin of motion is not just influenced by Aristotle, but, more importantly, grounded in their reading of the Timaeus (mediated through certain interpretations by the Old Academy). While the self-motion of soul seems to be maintained in the Timaeus, the world-soul is shown to be causally dependent on the demiurge who fashioned and connected it with the body of the cosmos (34b–c). Crucially, the whole cosmos, including the world-soul, is set in motion (37c6: κινηθέν) by the demiurge. This presumably is part of the explanation why some Middle Platonists saw the doctrine of an unmoved intellect as genuinely Platonic.

Clear evidence of the Platonist adoption of the intellect as prime mover is provided by Alcinous who – among the Middle Platonists – was exceptionally well versed in Aristotle’s philosophy.³⁷ In a theological passage of the Didaskalikos he discusses the nature of the highest god:

Since (1) intellect (νοῦς) is superior to (2) soul, and superior to (1c) potential (ἐν δυνάμει) intellect there is (1b) actualised (κατ’ ἐνέργειαν) intellect, which cognises everything simultaneously and eternally, and finer than this again is the cause of this and whatever it is that has an existence still prior to these, the (1a) primal God (ὁ πρῶτος θεός), being the cause of the eternal activity (ἀεὶ ἐνεργεῖν) of the intellect of the whole heaven. It acts on this while remaining itself unmoved (ἐνεργεῖ δὲ ἀκίνητος), as does the sun to vision, when this is directed towards it, and as the object of desire moves desire (ὸ ὀρεκτὸν κινεῖ τὴν ὄρεξιν), while remaining unmoved (ἀκίνητον) itself. In just this way will this (1a) intellect move (κινήσει) the (1b) intellect of the whole heaven. (Didask. 10.2.164.18–27)³⁸

In this complex passage Alcinous presents us an ascending ontological hierarchy, made up of (2) soul, (1c) potential intellect (i.e., of a human being), (1b) active intellect (i.e., of the cosmos) and a (1a) first god. Thus, there is a threefold distinction of human, cosmic and divine intellect. It is generally agreed upon that the active or cosmic intellect is the intellect of the world-soul.³⁹ Most importantly, the first god is also an intellect – as is stated towards the end of the passage (l.26) – which, unlike the human and cosmic intellect, is not immanent in a substrate but rather transcendent like the Aristotelian intellect or – on some reading at least – the Platonic demiurge.

The divine intellect is described not only with a reference to the Platonic sun simile from Resp. 6.508a–b (which is repeated at 164.39–40 and 165.21–3) but also in Aristotelian terms borrowed from Met. 12.7: the divine intellect causes motion qua being the object of desire, while remaining itself unmoved but engaged in ἐνέργεια. It is noteworthy that other contemporary Platonists describe god in similar terms but, crucially, leave out the aspect of motive causality.⁴⁰ ‘Unmoved’ is understood in relation to place, that is, locomotion, as well as qualitative change, as Alcinous clarifies: ἀκίνητος ἂν εἴη κατὰ τόπον καὶ ἀλλοίωσιν (10.7.165.38).⁴¹ A later passage explains more precisely how the first intellect moves:

He is Father because he is the cause of all things and bestows order on the heavenly intellect and cosmic soul in accordance with himself and his own thoughts. By his own will he has filled everything with himself, rousing up the cosmic soul and turning it towards himself (ἐπεγείρας καὶ εἰς ἑαυτὸν ἐπιστρέψας), being the cause of its intellect. It is this latter that, set in order (κοσμηθεὶς) by the Father, itself imposes order (διακοσμεῖ) on all of nature in this world.

(10.3. 164.40–165.4)

As is evident, the first intellect does not cause physical motion primarily, like the Aristotelian intellect,⁴² but instead causes the activity (164.22) and order (165.3) of the cosmic intellect of the world-soul who in turn arranges the cosmos.⁴³ Alcinous emphasizes that the first god did not ‘make’ (14.3.169.36: ποιεῖ) the world-soul. In this way, Alcinous sticks close to Plutarch (De An. Proc. 1014B–C, 1016C–D) and Atticus (fr. 11).

This picture, however, gets more complicated in a later passage of the Didaskalikos. For at 25.4 he identifies the origin of motion with a self-moved mover using language and concepts from the Phaedrus and Laws 10:⁴⁴

Furthermore, that which is self-moving primordially (αὐτοκίνητον ἀρχικῶς) is eternally moving (ἀεικίνητον), and such a thing is immortal; but the soul is self-moving. Again, that which is self-moving is the first principle of all motion and generation (ἀρχὴ πάσης κινήσεως καὶ γενέσεως); and a first is ungenerated (ἀγένητον) and indestructible (ἀνώλεθρον); so both the soul of the universe and the soul of man would be such, since both partake in the same mixture. Plato says that the soul is self-moving, because it has life as something innate in it, eternally active in itself.

(178.15–23)

Alcinous characterises soul as principle of all motion and generation, remaining faithful to Plato. This is a very strong statement. As in Plato, soul’s self-motion can be contrasted with the bodies’ other-motion which Alcinous maintains in 11.2 by describing bodies as purely passive, while only the incorporeal is active. Unlike what has been claimed by Dillon (1996: 316), γενέσεως (178.18) is neither a ‘significant addition’ nor ‘development’ to Plato’s definition of soul as ἀρχὴ κινήσεως in the Phaedrus. As mentioned above (Section 2.2), Plato characterises generation as a kind of motion caused by soul. Furthermore, generation is explicitly named as one of the types of motion in Leg. 10.894b11 of which soul is the origin.

Alcinous’ remarks on the nature of soul contradict his earlier statements in 10.2 regarding the function of god, as we now have an unmoved intellect and a self-moving soul as origins of motion. If the world-soul is the self-moving principle of all motion, how can the divine intellect be described as setting the cosmic intellect of the world-soul in motion (164.26–7: ὁ νοῦς κινήσει τὸν νοῦν τοῦ σύμπαντος οὐρανοῦ)? Alcinous unfortunately nowhere answers this question. Thus, his ‘reconciliation’ of Plato and Aristotle occurs without dealing with the different philosophical premises behind their claims, as is realised – to differing degrees – by latter Platonists such as Proclus, Hermias and Simplicius. Neither Whittaker (1990) nor Dillon (1993) seem to be aware of this conflict in their commentaries on this passage, although it is amply discussed by Aristotle and later Platonists; Dillon arguably even aggravates the problem by mistakenly translating ἀρχή at 178.18 as ‘first principle’ instead of just ‘principle’ or ‘origin’. It should be clear that the self-moving soul is somehow dependent on the higher unmoved intellect which can rightly be called a ‘first principle’.

The cause of this inconsistency lies in the general trend among Middle Platonists to integrate into their essentially Platonist system certain doctrines from different philosophical strains such as Aristotelianism and Stoicism. This combination is not always successful but can indeed create considerable tensions.⁴⁵ Until Proclus, Neoplatonists do not seem to offer a solution for this tension – at least based on our extant evidence. This is also because only with Proclus and other late Neoplatonists we get a systematic theory of unmoved, self-moved and other-moved beings.

2.5 Proclus: Unmoved Mover, Self-Mover and Other-Moved

Like earlier Platonists, Proclus tries to combine the accounts of Plato and Aristotle in explaining the origin of motion.⁴⁶ His main contribution lies in offering a more systematic view of this issue than previous Platonists like Plotinus and Iamblichus which is in constant interaction with the Platonic and Aristotelian sources. Prima facie he does not seem to follow Plato as closely in his theory of motion as one would expect, since for Proclus self-movers are ultimately dependent on an unmoved mover which is ontologically prior: ‘the unmoved is prior to the things that are moving and moved’ (EP §2.19).⁴⁷ The unmoved mover is identified with intellect and the self-mover with soul. Crucially, motion is transmitted to the physical realm via self-movers, that is, souls. As intermediaries these guarantee the connection between the metaphysical and the physical sphere, and, thus, play a central role in Proclus’ theory of motion. The triadic structure of movers which Alcinous has only foreshadowed is here made manifest: ‘everything is unmoved, self-moved, or other-moved’ (ET §14; cf. In Tim. 3.176.12–13 [2.128.20–2]). When Proclus follows Aristotle regarding the question of the origin of motion, he implicitly accepts Aristotle’s criticism of Plato. However, this impression is to some degree deceiving. For Proclus, as for all Neoplatonists and some modern scholars such as Hackforth (1965), Menn (1994) and Karfík (2004), a transcendent intellect is part of Plato’s metaphysics. Aristotle’s theory of the unmoved mover only spells out what already is in Plato – so Proclus. Proclus would thus consider the idea of an unmoved intellect as prime mover as genuinely Platonic and not Aristotelian.

Proclus primarily argues for the necessary existence of the unmoved mover in EP which offers us remarkable evidence for the adoption of Aristotelian philosophy and its reconciliation with Plato. While I discussed EP at length in Chapter 1, I focus here on the unmoved mover and self-mover which seems absent in EP. This is surprising given its significance not only for Aristotle who reaches his conclusion that an unmoved mover is the origin of motion via an analysis of self-motion but also for Proclus who in his other systematic treatise, Elements of Theology, discusses the self-moving soul and its relationship with the unmoved intellect extensively (§§14–20). My discussion of the relationship of the unmoved mover and the self-mover is twofold, split between EP and ET. I show that his arguments for the existence of both are problematic in these works. Yet, by looking at Proclus’ larger system I maintain against Opsomer (2009) that these inconsistencies can be solved.

2.5.1 Primacy of Unmoved Mover in EP

Does Proclus include self-movers in his discussion of the origin of motion in EP? Or is his account of motion so thoroughly Aristotelian that he accepts a binary system of unmoved and moved movers? While Ritzenfeld (1912: VIII) noted the difference between the accounts of motion in EP and ET, consisting of the former’s lack of self-movers, Dodds (1963: 201) argues that both discussions are actually compatible and that self-motion is implied in the general treatment of EP. The discussion focuses particularly on §2.19: ‘the unmoved is prior to the things that are moving and moved’ (τῶν κινούντων καὶ κινουμένων ἡγεῖται τὸ ἀκίνητον) and the correct understanding of the terms κινούντων and κινουμένων. My argument is a via media between Ritzenfeld and Dodds: while explicit references to self-movers are indeed missing, the account of EP is still compatible with ET. First, we need to look more closely at the proposition.

In §2.19 (which is based on Phys. 8.5) Proclus establishes the priority of the unmoved mover as origin of the universe’s eternal motion. In doing so he excludes the possibility either (1) that this eternal motion is caused by a finite series of things in motion moving each other in a circle, that is, A moving B moving C moving again A and so on, or (2) that eternal motion is due to an infinite series of moving things. Option (2) is rejected, since an infinite number of magnitudes (or an infinitely large magnitude) is irreconcilable with the idea of a finite cosmos, as Proclus had demonstrated earlier in §2.15 that an infinite magnitude cannot exist. Regarding (1), Proclus answers that ‘if the motion is in a circle, one of the things which are sometimes moved (τῶν ποτὲ κινουμένων) will be the cause of the eternal motion, if all move and are moved by each other in a circle. But this is impossible; for that which produces eternal motion is eternal (τὸ γὰρ τὴν ἀίδιον κίνησιν κινοῦν ἀίδιόν ἐστιν)’ (§2.19.56.23–6). Proclus excludes the possibility that the cause of the eternal motion are things causing motion only intermittingly. Thus, Proclus has shown that the eternal motion of the universe can be caused neither by an infinite series of movers nor by a finite series of movers moving each other. Rather, he concludes, a single mover must always be causing it. This had also been demonstrated in the previous proposition §2.18: ‘That which produces an eternal motion is eternal’.

Yet, it might seem puzzling why Proclus leaves out another option (3): a self-mover accounting for the eternal motion, that is, a self-moved world-soul setting the universe in motion. For this would be the obvious solution for a Platonist following Plato’s account in the Phaedrus and, particularly, Laws 10 where Plato clarifies that a self-mover as prime mover averts an infinite regress (see Section 2.2).⁴⁸ Proclus thus clashes here with Plato’s argument in Laws 10.

Having said this, it first needs to be clarified what Proclus means by τῶν κινούντων καὶ κινουμένων in §2.19 and whether these terms imply a self-mover. The most accurate translation, preferred by Ritzenfeld (1912: 57) and Opsomer (2009: 195–6), takes both terms together as ‘moved movers’ or literally ‘the things that are moving and moved’. This expression is vague: while it clearly implies other-moved entities, self-movers could be referred to as well since they move other things and are moved internally.

However, there has been another term for self-movers proposed by Dodds which Proclus uses towards the end of the proposition: ‘From this becomes clear that … not everything is sometimes at rest and sometimes moved (for there is also something eternally moved as well as something always unmoved).’ (2.19.56.28–58.5). Proclus distinguishes between ‘something eternally moved’ (τὸ ἀιδίως κινούμενον) and ‘something always unmoved’ (τὸ ἀεὶ ἀκίνητον). While the latter is clearly to be identified with the prime mover, Dodds (1963: 201) argues that the former expression relates to a self-mover. Dodds presumably has in mind the world-soul which is always in motion (ἀεικίνητος) according to the Phaedrus and causes the eternal motion of the heaven. While this might be correct, given the Aristotelian background of EP it could much rather be a reference to the eternal motion of the cosmos as being externally caused by the unmoved mover. I thus conclude that Proclus does not explicitly engage with the concept of self-motion in 2.19 and generally in EP.

How could this absence of self-movers in EP be understood, given that it is so significant in Physics 8 as well as in other Proclean works? It should be pointed out that although Proclus leaves out this crucial step in Aristotle’s argument, he still retains the same conclusion, that is, that an unmoved mover (and not a self-mover) is the origin of motion. Two reasons, I argue, explain his omission: (1) Proclus regards the inclusion of self-movers as an unnecessary complication of his argument, since – in line with the Aristotelian doctrine expounded in EP – he takes it for granted that self-movers depend on an unmoved mover (ET §20). In order to accomplish his goal in EP to demonstrate the existence of the unmoved mover he does not require a detour via an analysis of self-movers. Nevertheless, his account of self-movers is compatible with EP as his discussion in ET will show. (2) More importantly, the proper context of discussing self-movers is, for Proclus, metaphysical and not physical as in EP, since self-movers are souls and transcend the strictly physical realm.⁴⁹ Since EP was probably designed as a textbook for students mastering Aristotle’s natural philosophy, it is not surprising that the more metaphysical background is left out, as students were not yet introduced to it. Similarly, it is a common feature of Neoplatonist commentaries to adapt to the presupposed knowledge of the students, as they were designed in the context of the school’s curriculum. This explains why some commentaries on Aristotle are lighter on intricate metaphysical questions and why Proclus defers his discussion of self-movers to ET where non-physical motion plays a crucial role.

2.5.2 Unmoved Mover and Self-Mover in ET

In ET, we find a detailed account of the relationship and hierarchy of the heterokinēton (other-moved), autokinēton (self-moved) and akinēton (unmoved), as exemplified by §14:⁵⁰ ‘Every being is either unmoved (ἀκίνητον) or moved (κινούμενον) and if moved, either by itself or by another; and if by itself, self-moved (αὐτοκίνητον); if by another, other-moved (ἑτεροκίνητον). Thus, everything is either unmoved, or self-moved, or other-moved.’⁵¹ Later, in §20, Proclus identifies ἀκίνητον with intellect, αὐτοκίνητον with soul and ἑτεροκίνητον with body.⁵² As I emphasise below in this section, intellect, soul and body are here collective terms which denote classes of beings and not specific entities. Moreover, it is important to note that the context here differs from EP. In ET motion serves to describe the relationship of different metaphysical entities, primarily soul(s) and intellect(s), and their activity, for example – in the case of soul – discursive thinking, willing, opining and so on. Bodies play only a marginal role: as other-moved they have no motive and, generally, no causative force on their own but rather derive this from a higher, non-physical source, that, soul and, ultimately, intellect.⁵³

A few remarks on the terminology are required. Plato already differentiated in, for example, Leg. 10.894b8–c1 between things moved by others and things moved by themselves, which here forms the background for Proclus’ own distinction between αὐτοκίνητον and ἑτεροκίνητον. Earlier evidence for hierarchies of movers, quite similar to our example, are also found in other Neoplatonist sources, too.⁵⁴ Among these, it is only in Proclus’ teacher Syrianus that the same terminology and triadic structure of ἀκίνητον, αὐτοκίνητον and ἑτεροκίνητον occurs, albeit not all three in the same context.⁵⁵ His students Proclus and Hermias (who uses the same terminology frequently in his commentary on the Phaedrus, e.g., 110.23, 120.22–3, 126.28) took over the terminology and triadic structure of movers and transmitted it to later Neoplatonists such as Damascius, Simplicius and Philoponus. Syrianus seems to be the first Neoplatonist to use ἑτεροκίνητον for bodies – a term which cannot be found in his authorities Plato and Aristotle. However, the idea that bodies are moved externally is well grounded in Plato as well as Aristotle and is a commonplace among Platonists.⁵⁶ Unlike what has been claimed by Longo (2020: 124), Syrianus is not the first philosopher to use this term. Rather, we find it in Alexander as part of the triad ἀκίνητον, αὐτοκίνητον and ἑτεροκίνητον (In Phys. 8.5.599.9–10). Plato makes no mention of αὐτοκίνητον, but it does occur once in Aristotle (Phys. 8.5.258a2).⁵⁷ In fact this singular appearance in Aristotle recently has led Rashed (2011: 556) to claim that the sentence including the term was originally a gloss from Alexander, who often employs it, as by the Imperial Age αὐτοκίνητον is established as a technical term.⁵⁸

To what entity does Proclus actually refer here when he states that νοῦς is the prime mover? Unlike Aristotle who refers to the intellect of the outermost sphere, Proclus in fact uses the term collectively and circumscribes a class of beings – which is rarely emphasised enough in scholarship. This is made clear by the formulation of prop. 14 quoted above which claims that ‘everything’ can be exhaustively divided into unmoved, self-moved and other-moved beings. Clearly, anything higher than soul cannot be described as self-moved or other-moved, which leaves as the only possible description of these beings the unmoved. It follows that in §14 and §20 the whole intelligible realm – including forms, demiurge and other gods – is described by the term νοῦς and ‘prime mover’. Proclus makes thus a very general claim about the origin of motion and does not specify the precise beings involved or the details of the process. The same goes for his use of soul and body in these propositions. To a certain degree this is similar to Plato’s discussion at Phdr. 245c5 where the term ‘soul’ is also used collectively.

While this collective usage of the term might be less troublesome given the introductory nature of these propositions, the identification of νοῦς with the intelligible realm causes more difficulties in Proclus. This has historical reasons. Unlike Plotinus who had a clearcut division between the different layers One – intellect – soul, Proclus takes over this structure and analyses it into distinct aspects. What for Plotinus represented the intellect – the intelligible realm – is split by Proclus into three aspects which form a triad and a hierarchical structure: being, life and intellect.⁵⁹ These three are both simultaneous aspects of a single reality as well as successive stages in the procession from the One. In accordance with this division, intellect is thus – strictly speaking – no longer identical with the whole intelligible realm, like in Plotinus, but rather only with one of its layers. Thus, at ET §103 Proclus clearly distinguishes intellect as one aspect of the triad:

πάντα ἐν πᾶσιν, οἰκείως δὲ ἐν ἑκάστῳ· καὶ γὰρ ἐν τῷ ὄντι καὶ ἡ ζωὴ καὶ ὁ νοῦς, καὶ ἐν τῇ ζωῇ τὸ εἶναι καὶ τὸ νοεῖν, καὶ ἐν τῷ νῷ τὸ εἶναι καὶ τὸ ζῆν, ἀλλ’ ὅπου μὲν νοερῶς, ὅπου δὲ ζωτικῶς, ὅπου δὲ ὄντως ὄντα πάντα

All things are in all things, but in each according to its proper nature: for in Being there is life and intellect; in Life, being and intellect; in Intellect, being and life; but each of these exists upon one level intellectually, upon another vitally, and on the third existentially.⁶⁰

I believe this apparent inconsistency can be explained by the character of ET as a στοιχείωσις. In the early propositions Proclus offers a fundament and basic structure of reality which he then further develops. Applied to our problem this means that in some way one can consider the whole intelligible realm as νοῦς, but in a more precise and strict way the νοῦς is only one of its aspects.

What is the role of αὐτοκίνητον? According to Proclus, a self-mover is the proximate cause of other-moved beings⁶¹ and acts as a kind of ‘middle’ (ET §14.16.25: μέσον) term between an unmoved mover and an other-moved, mediating the motion between these two, since it has both their active and passive aspects.⁶² This does not mean that a self-mover in Proclus is made up of two distinct parts, unmoved and moved, like in Aristotle. Instead of having two locally distinct parts (as one could assume for Aristotle), the soul qua self-mover can be analysed into two conceptual parts or aspects.⁶³ Faced with the possible dichotomy between an unmoved mover and an other-moved, Proclus states that there must be an intermediate being between these two: ‘For since there are things other-moved it is necessary that there is also something unmoved, and an intermediate being which is self-moved’ (ET §14.16.13–14). What its self-motion precisely consists of, is discussed in the next chapter.

Why is the existence of a self-mover ‘necessary’ (ἀνάγκη)? Before Proclus explains this, he discusses the origin of motion, that is, the unmoved mover:

For if every other-moved is moved by something set in motion externally, then we either (i) have a circuit of communicated motion or (ii) an infinite regress. But neither of these is possible, inasmuch as all beings are limited by a principle (ὥρισται τῇ ἀρχῇ τὰ ὄντα πάντα) and the mover is superior to the moved (τὸ κινοῦν τοῦ κινουμένου κρεῖττον). There must, then, be something unmoved which is the first mover.

(§14.16.15–19)

Both consequences (i) and (ii) are identical to the options set out in EP §2.19 and thus ultimately derived from Aristotle’s Physics 8.5. So too is his conclusion, that is, that the first mover must be unmoved. But his explanations, as I show, are quite different from the ones found in EP, since they must be based on earlier propositions. This is due to the axiomatic structure of ET where each proposition is – or should be theoretically at least – deducible from an earlier one, just as in EP (see Section 1.3). Proclus thus provides here Platonist and not Aristotelian explanations. This is similar to the phenomenon treated in Chapter 1: when discussing material or conclusions from the Physics or the De caelo in other treatises than EP, Proclus offers Platonist arguments which tend to be more in-depth and based on a specific understanding of Plato. In ET §14 his two claims that are supposed to refute both alternatives are the following: (1) ‘all beings are limited by a principle’ (16.17) and (2) ‘the mover is superior to the moved’ (16.17–18).

(2) rests on a fundamental Neoplatonist concept, expressed in §7: ‘Every productive cause is superior to the nature of its product’, that is, the cause is greater than its effect.⁶⁴ Thus, an efficient cause cannot bring about an effect that is either equal or superior in nature to it. Since Proclus discusses motion in terms of efficient causality, he applies this earlier proposition to his argument in §14. It clearly clashes with (i), that is, the idea of a circular structure of things moving and being moved by others simultaneously: Seen as efficient causes, mover A would produce a lesser effect (motion) B which in turn acting as a cause would generate an even lesser effect C and so on. The obvious problem encountered in such a finite circle of entities being simultaneously cause and effect or mover and moved is that at some point a lesser effect would need to cause the motion of a higher cause which is impossible, since it lacks the causative power to do so, that is, the causally weak C would need at some point to cause the motion of the causally potent A.

The danger of an infinite regress of movers (ii) can be refuted by a recourse to (1) which is based on §11: ‘all beings proceed from a single, first cause’. Proclus uses an epistemological argument, borrowed from Aristotle (Met. 2.2), to argue against an infinite regress of motive causes (12.25–8): In case of an infinite chain of causes ‘all things will be unknowable. For nothing infinite can be apprehended; and the causes being unknown, there can be no knowledge of their effects’.

According to Proclus, these two arguments show that the origin of motion has to be an unmoved mover. But this is fallacious since the motion could also be generated by a self-mover: his counterarguments (1) and (2) show only that the first mover does not have to be set in motion externally. Yet, they do not exclude the possibility that this principle sets itself in motion. Proclus’ conclusion is deficient, unless we grant that he takes Aristotle’s analysis of self-motion as a given, whereby an absolute self-mover in the Platonic sense does not exist but rather is made up of unmoved and moved aspects. I do not suppose this is the case in ET since he explicitly distinguishes here between three kinds of movers and claims that ‘if the mover be one part and the moved another, in itself the whole will not be self-moved, since it will be composed of parts which are not self-moved: it will have the appearance of a self-mover, but will not be such in essence’ (§17.18.25–8). This plainly goes against the Aristotelian conception of self-movers.

After having thus established the necessity of an unmoved mover, Proclus turns towards the self-mover:

But if so, there must also be something self-moved. For imagine all things to be at rest: what will be the first thing set in motion (κινούμενον)? Not the unmoved, by the law of its nature. And not the other-moved, since it is moved from without (ὑπ’ ἄλλου). It remains, then, that the first thing moved is the self-moved, which is in fact the link (συνάπτον) between the unmoved and the other-moved things. At once mover and moved, the self-moved is a kind of mean term (μέσον πως) between that which merely moves [i.e., unmoved mover) and that which is merely moved [i.e., other-moved]. Every being, therefore, is either unmoved, or self-moved, or other-moved.

(§14.16.20–7)

In a hypothetical state of absolute rest, highly reminiscent of Phdr. 245d8–9 and Leg. 10.895a6, neither the unmoved nor the other-moved would be first set in motion but rather the self-moved.⁶⁵ The elimination of the unmoved is obvious, but why is the other-moved excluded as first thing moved? Proclus’ only explanation seems to be that it is moved externally (16.22: ὑπ’ ἄλλου γὰρ κινεῖται) and as such cannot be the first thing set in motion. From this we must conclude that the primary moved has to be something not moved externally but moved by itself, that is, a self-mover. Proclus maintains that the self-mover connects the unmoved with the other-moved by mediating the motion. Thus, it seems that the unmoved mover cannot move the other-moved directly and instead requires the mediation via the self-moved. How it does so is here not explained. I will try to give an answer in Section 2.5.3.

But first let us consider the background of the argument in 16.20–7 which is Laws 10. After determining that self-motion has priority over other kinds of motion and is indeed the origin of all motion (894e7–895a3), Plato provides another argument for these claims:

If somehow everything were to come to a standstill (σταίη), just as most of those men venture to say, which motion of the ones we spoke of [i.e., the ten motions] would necessarily be the first to come to be among them? Surely one that moves itself (ἑαυτὴν … κινοῦσαν). For it would never be changed by another that is prior (ἔμπροσθεν), since there is among them no prior change.

(895a6–b3)

Plato’s argument here is quite different from Proclus’. Plato talks about the first motion (πρώτην κίνησιν) to arise (γενέσθαι) in this hypothetical state of rest and, unlike Proclus, not about the first thing moved. According to Plato, self-motion is primary precisely because it is not brought about by something else, as there is no unmoved mover for Plato which could produce this change. This idea is reflected in Proclus’ exclusion of the other-moved as first thing set in motion since it is moved externally. Yet, unlike Proclus, Plato wants to establish by this argument that the self-mover is the prime mover.

As noted by Opsomer (2009: 204–6), two problems arise here which deal with the simultaneous existence of an unmoved mover and a self-mover. (1) The first concerns the necessary existence of the self-mover, which has not been sufficiently proved in §14 and also cannot be inferred from an earlier proposition, despite the supposed ‘geometric’ make-up of Proclus’ work. Why cannot an unmoved mover cause the motion of the other-moved directly?⁶⁶ (2) Additionally, if we accept Proclus’ proof of the self-mover as an internally moved entity which in turn causes the motion of the other-moved, what role does the unmoved mover play? For one could then be content with accepting a self-mover as causing the motion of the cosmos. Hence, (1) and (2) question the triadic structure of movers in Proclus, since a binary system of either unmoved and other-moved or self-moved and other-moved seems sufficient. Both difficulties can be solved only by looking at Proclus’ philosophical system.

2.5.3 Proclus’ Solution in the Commentary on the Parmenides

In the following, I propose to defend Proclus’ triadic system of movers against these objections by considering a passage from his commentary on the Parmenides. In short, Proclus provides there an explanation for the existence of self-movers in terms of producing something: something unmoved can only bring about something unmoved, while something moved only something moved.⁶⁷ This in fact is very close to Plato’s view on causation, outlined in Section 2.2. A self-mover bridges the gap between the two, unmoved and moved, since its ousia (essence) is unmoved, its energeia (activity) moved. But let us look more closely at Proclus’ argument.

In a text from his commentary on the Parmenides, Proclus deals with the question of why the ultimate source of motion must be an unmoved intellect and not a self-moved soul:

It is said that all things produced by an unmoved cause (ἀκινήτου αἰτίας) are unmoved (ἀκίνητα) and unchangeable (ἀμετάβλητα), but those that come about from a moved (κινουμένης) cause are, on the contrary, mobile (κινητά) and changeable (μεταβλητά), being sometimes in one state and sometimes in another. And if this is true, all things that are eternal in essence and unchangeable are produced by an unmoved cause. For if they come from a mobile cause, they will be changeable, which is impossible. All unmoved things, therefore, come from an unmoved cause, that is, if they come into being at all.

(3.795.7–13)⁶⁸

Proclus first presents a dichotomy of unmoved and moved things. To these correspond two kinds of causes which resemble their effects, that is, unmoved and moved. In other words, an unmoved entity cannot be caused by a moved cause and vice versa.⁶⁹ This abstract division into unmoved and moved is then exemplified by certain entities:

The genuinely unmoved beings, consequently, are those that are unchangeable both in their essence (κατ’ οὐσίαν) and in their activity (κατ’ ἐνέργειαν). Such are the intellective beings (νοερά); second come those that are unmoved in essence but mobile in activity, such as souls (ψυχικά); third are those that are invisible but inseparable from visible things, like the natural beings (φυσικά); and last are the visible forms (ἐμφανῆ) that exist distributively in sensible objects.

(796.4–796.8)

The intellect is unmoved in essence and activity, while soul is only unmoved in essence but moved in activity which both together make up its self-motion.⁷⁰ This distinction between essence and activity in soul, which is elaborated in ET §191, helps Proclus in situating the two different characteristics. In this way, the make-up of soul accounts for an uninterrupted transition of the different ontological layers and also explains how soul is immortal and possesses a temporal existence at the same time.⁷¹ It should be noted that Proclus means by soul here only the rational soul which is the true soul.⁷² Only the rational soul is self-moved.⁷³ Soul’s essence is eternal, while its activity is temporal (ET §191). This does not mean that soul has two distinct parts, an unmoved essence and a moved activity; Proclus already rejected this in ET §17 (see Section 2.5.2).⁷⁴ Instead, these are two different aspects of the same thing, both involved in the self-motion of soul. Due to its composition, soul can cause through its activity moved or mobile effects such as the motion of the cosmos. But since soul has an unmoved essence, there needs to be a superior unmoved cause. This, I maintain, is the explanation for the puzzle concerning the reason for excluding the unmoved mover as a direct cause of the other-moved. It also provides an argument for regarding soul as a mediator between the unmoved and other-moved, which was not evident in ET §§14–20 and completely absent in EP. It should be noted that Proclus provides such a defence for his triadic structure of kinetic beings not only in the commentary on the Parmenides but also in ET §76 and Platonic Theology 1.14.⁷⁵ The idea that soul has a double nature and mediates between the immobile and mobile realm is clearly grounded in Proclus’ reading of Tim. 35a1–b3, where soul is described as something intermediate. Proclus picks this up in ET §190: ‘[e]very soul is intermediate between the indivisibles and those which are divided in association with bodies’.

2.6 Conclusion

The passage above explains why Proclus prefers a triadic over a binary system of movers, that is, either (1) an unmoved mover and an other-moved or (2) a self-moved and an other-moved. Regarding the Aristotelian model (1), Proclus recurs to a common principle according to which like causes like, that is, the unmoved causes the unmoved and the moved causes the moved. Since the other-moved belongs to the category of moved things, its cause needs to be likewise moved. This has to be thus the self-moved in order to exclude an infinite regress. Regarding the Platonic model (2), Proclus maintains that, while soul is responsible for the cosmic motion, it is in turn dependent on a higher principle which is unmoved and the cause of its essence. This in fact is a very Platonic idea and can be reconciled with the description of the demiurge in the Timaeus who fashions the world-soul and the whole cosmos. In a certain way, Proclus prefigures here modern solutions on the compatibility of the demiurge’s causation of soul and soul’s self-motion such as Vlastos’ and Brisson’s.⁷⁶ ‘Mover’ needs to be understood differently in the case of the unmoved mover and the self-mover respectively. While soul directly causes physical motion, intellect is a cause of motion only insofar as it causes the unmoved essence of soul and guarantees that the cosmos has an eternal principle of motion. Thus, only in a mediated way (via soul) does the unmoved mover move the cosmos. In providing such a reasoning, Proclus goes beyond earlier Platonists who lack a sufficient argument for assuming both intellect and soul as principles of motion, as exemplified by Alcinous.

Lastly, how Aristotelian is Proclus’ account? While the adoption of the intellect as unmoved mover seems prima facie Aristotelian, Proclus’ motivation as well as the philosophical context are indeed very Platonic. For Proclus it is a Platonic commonplace to regard the demiurge as an intellect, based on a long exegetical tradition reaching back to the Old Academy. Once the demiurge is understood in this way, it is only a small step for Proclus to call him unmoved mover, as the demiurge – according to the Neoplatonist interpretation at least – is lacking physical motion (= unmoved)⁷⁷ and is somehow causally efficacious towards the cosmos (= mover). Proclus superficially agrees with Aristotle’s conclusion that the eternal motion of the cosmos is caused by an unmoved mover and even dedicates a treatise to this deeply Aristotelian question. But, in his more metaphysical and Platonic works it becomes clear that the picture is complicated by the mediating force of the self-moved soul. As Proclus clarifies, the cosmos would reach a standstill without the world-soul, which thus plays a crucial role in causing its eternal motion. This can be contrasted with Aristotle where the prime mover is the direct cause of the cosmos’ motion and the function of souls is limited to causing inner-cosmic motions.