Chapter 1 Aristotle’s Light Analogy in the Greek Tradition

In: Forms of Representation in the Aristotelian Tradition. Volume Three: Concept Formation
Börje Bydén
Search for other papers by Börje Bydén in
Current site
Google Scholar
Open Access

1 Introduction

In De anima 3.5, 430a14–17, Aristotle makes a famous distinction between two kinds of intellect (noûs), one which “becomes all things” and one which “makes all things.” By “things” we should no doubt understand “things in the realm of intellect,” maybe episodes of intellection (or individual thoughts), maybe intelligible objects (or concepts). Aristotle compares the intellect that “makes all things” to “a kind of state (héxis), such as light.” The light analogy is a rare clue to understanding his views about the nature of this intellect – the “active” or “productive” intellect (noûs poiētikós), as it came to be called in the later tradition – and its role in human intellection, whether it is supposed to “produce” individual thoughts or concepts or both (or, for that matter, neither, but something else).1 The pity is that the analogy itself is rather obscure. Aristotle left it to his commentators to explain exactly what the relation is between the secunda comparata or source terms of the analogy, namely, light and colours, that he considers to be similar to that between the prima comparanda or target terms, so as to be able to define more precisely the respective natures and roles of the two kinds of intellect.

In this paper, I will outline some of the problems that face the interpretation of the light analogy and examine some of the responses to these problems offered by the Greek commentators on the De anima. “Commentators” should be understood here in a wide sense. In fact, my main focus will be on the responses offered by Alexander of Aphrodisias (fl. early third century), whose most relevant extant work, also entitled De anima, is a compendium of Aristotelian psychology rather than a commentary proper,2 and on the criticism of Alexander’s responses in two of those line-by-line commentaries that were written in the sixth and seventh centuries by Neoplatonists, namely those by John Philoponus and his spurious namesake, Ps.-Philoponus. Philoponus’ commentary is based on lectures by Ammonius: the part on book three is extant only in a medieval Latin translation covering chapters 4–8 and fragments in the original Greek.3 Only book three of Ps.-Philoponus’ commentary is extant.4 I will only sporadically refer to the paraphrase by Themistius, the fourth-century orator and statesman, and not at all to the line-by-line commentary by Priscian of Lydia (Ps.-Simplicius, early to mid-sixth century).

2 Text and Context

All commentators devote disproportionately large parts of their De anima commentaries (or the equivalent) to chapter 3.5. But the passage of Aristotle’s chapter in which we are now interested is very short, only three lines. So let us begin by looking at a rather literal translation of our focus text, as established by Sir David Ross in his 1961 edition (T2), preceded by the first four lines of the chapter (T1). The texts have been divided into segments for easy reference.5

(T1) (a) Since, in the whole of nature, one thing is matter for each genus – (b) and this is what is potentially all those things – (c) and another thing is what is causative and productive, (d) in virtue of making all things, (e) as is the situation with an art relative to its matter, (f) it is necessary that these divisions obtain also in the soul (430a10–14).6

(T2) (a) And this kind of intellect exists in virtue of becoming all things, (b) whereas the other [exists] in virtue of making all things, (c) in the manner of a kind of state (d) such as light: (e) for in a certain way light, too, makes what are potentially colours into actual colours (430a14–17).7

I will say a few words about the grammar and lexicon of T2 in a moment. Let me first say something about its immediate context (T1) and the way that our commentators understand it.

The text of T1 as transmitted in our manuscripts is ungrammatical: both “epeí” (“[s]ince”) and “hṓsper” (“just as”) in 430a10 mark the beginnings of dependent clauses, each of which would require a main clause to complete its meaning; but there is in fact only one (compound) dependent clause (T1a + c) and one main clause (T1f). To remedy this situation one may delete either “epeí” or “hṓsper.” Since both Alexander’s and Themistius’ paraphrases of T1–T2 begin with “epeí/-dḗ8 and neither of them includes any word or phrase corresponding to “hṓsper” in T1a,9 it was suggested by Ross in his two editions (1956, 1961) that “hṓsper” should be removed.10

Whether or not it is justifiable in the eyes of Textual Criticism to emend Aristotle’s text just to save him from grammatical blunders, there may be an additional reason for following Ross’ suggestion. If the main clause (T1f) is modified by a dependent clause beginning with “hṓsper,” the most natural interpretation of T1 will be that it infers a fact about the soul by analogy with a fact about the whole of nature.11 This is precisely how T1 is interpreted by the Neoplatonist commentators, who all apparently had “hṓsper” in their texts, and who all hold that the human soul partly transcends nature, understood as the realm of change.12 According to Philoponus, for instance, the reason why it is necessary that things are with the soul as they are in the whole of nature is that even the soul is not completely unchangeable (“intransmutabilis,” from “ametáblētos”), since it is changed with respect to the passage from potentiality to actuality, although not with respect to its nature or substance.13 Alexander of Aphrodisias, on the other hand, interprets T1 straightforwardly as setting out a general principle of causation covering everything that comes to be by nature, which is then applied, by subalternation, to the human intellect.14 The principle is that all such causation involves two terms: a material cause and an efficient one.15 It can be extended to those things that come to be by art, where the art is what imposes a form upon the matter.16

It stands to reason that, in natural as well as artificial causation, the material cause, “what is potentially all things,” must be part of nature; but this is not necessarily true of the efficient cause. There seems to be no reason why an entity acting upon a natural entity could not itself be exempt from change, let alone coming-to-be. Still, if T1 sets out a general principle of causation in nature and the principle applies to the human intellect, this can only be because at least the material cause of the human intellect is part of nature. And, according to Alexander, that is precisely what it is. Souls are, as Aristotle says,17 forms of natural bodies potentially possessed of life; and thus they are, according to Alexander, inseparable from these bodies;18 the kind of souls that belong to human beings are rational;19 and “intellect” is just another name for the cognitive capacity of the rational soul.20 This means that all normal human beings are born with what Alexander calls a “material intellect,”21 which, he thinks, is “the part of soul called intellect” that Aristotle describes in the preceding chapter as being “in actuality none of the things that exist, prior to intellection.”22

As the reader may have guessed, however, Alexander does not think that the efficient cause of the human intellect is part of nature. As a matter of fact, he identifies it with the intrinsically immaterial First Cause, which is obviously exempt from change, let alone coming-to-be.23 It is what Aristotle, in Metaphysics 12.7–9, refers to as a divine intellect. And since the human soul in Alexander’s view is entirely natural, he does not think that the efficient cause of the human intellect is part of the human soul either, but rather that it is what Aristotle, in De generatione animalium 2.3, describes as an intellect “from outside.” But hold on a minute: is not what Aristotle says in T1f that there must be both a material and an efficient cause in the soul?

It is indeed, as the later commentators are quick to point out.24 And they interpret this to mean that not only the kind of intellect mentioned in T2a – which, like Alexander, they identify with “the intellect prior to intellection” in De anima 3.4, although they prefer to call it a “potential” rather than a “material” intellect – but also the kind of intellect mentioned in T2b – that is, the “productive intellect,” as it is called by Alexander and Themistius, or the “intellect in actuality,” as it is called by the Neoplatonists25 – is part of the human rational soul.26 Accordingly, both Themistius and Ps.-Philoponus take Alexander to task for disregarding Aristotle’s choice of words.27 For Alexander only infers from the general principle of causation in T1 that there should be these divisions “in the case of the intellect, too.”28

In Philoponus’ and Ps.-Philoponus’ minds, on the other hand, there is no doubt that both the potential and the productive intellect are part of the human rational soul. Both commentators in fact endorse the view that the intel- lect in actuality is the same as the intellect in potentiality,29 but it seems that they disagree about the type of sameness involved. Ps.-Philoponus simply claims that the two intellects are the same in subject (“tôi hypokeiménôi”), although not in time.30 Philoponus, in contrast, while acknowledging that the same intellect is at one time in potentiality and at another time – once it has been actualised – in actuality, insists that the intellect in potentiality must be actualised by an intellect from which it is numerically different: the intellect in actuality, he repeatedly says, is the intellect of the teacher.31 Since he nowhere hints that this is a personal observation (epístasis), we should probably assume that his view reflects Ammonius’ teaching.32 Indeed, there is reason to suspect that it ultimately derives from the commentary of Plutarch of Athens, the teacher of Syrianus and Proclus. Both Philoponus and Ps.-Philoponus present the view they endorse as the last in a series of four pre-existing opinions about the intellect in actuality.33 Philoponus does not mention the names of their authors, but his first three reports correspond relatively closely to the opinions attributed by Ps.-Philoponus to Alexander, Marinus, and Plotinus, and the view that the intellect in actuality is the same as the intellect in potentiality is attributed by Ps.-Philoponus to Plutarch.

Whether we should follow Ps.-Philoponus in thinking that Plutarch considered the intellects in actuality and in potentiality to be the same in subject, or emend his report on the basis of Philoponus, depends partly on our assessment of Ps.-Philoponus’ reliability as a witness, partly on the inherent plausibility of the reported view. As for Ps.-Philoponus’ reliability, it is worth noting that even though he appears to have made direct use of Plutarch’s commentary (he mentions Plutarch’s name more than 40 times in 150 pages, not infrequently in connection with finer points of textual as well as philosophical criticism), his reports are tantalisingly inconsistent. As for the plausibility of the view, it seems to have the consequence – granted that Ps.-Philoponus affirms that the intellect in potentiality is actualised by the intellect in actuality – that individual intellects are actualised by their own future selves.34 In the name of charity, then, it seems preferable to credit Plutarch with the view reported by Philoponus, namely, that each human intellect in potentiality is actualised by another human intellect, which is in actuality beforehand.

This goes to show, I think, that it is wrong to suppose that the later commentators, in insisting that the productive intellect is part of the human rational soul, are simply motivated by their wish to persuade themselves and others of Aristotle’s belief in the immortality of the latter. At least for the Neoplatonists this would be pointless, since in their view Aristotle considered the whole human rational soul, including the potential intellect, to be immortal. In the case of Philoponus such a motive can be definitely ruled out, since he locates the productive intellect in a numerically different soul from that on which it acts. Most of the criticism on the part of the later commentators of Alexander’s identification of the productive intellect with the First Cause is in fact rather closely based on Aristotle’s text.35 I have mentioned their insistence on a literal reading of the phrase “in the soul” in T1f. Two other points of criticism have to do with the comparison of the productive intellect to a kind of state and to light. I shall return to them in the following sections.

So, while it is very likely that the Neoplatonists had non-exegetical reasons for attributing to Aristotle, as they do, the view that the whole human rational soul is immortal, this view does not seem to entail that the productive intellect is part of the human rational soul. The only non-exegetical reason for thinking that it is that seems to be safely attributable to the Neoplatonists is the methodological principle, invoked in this context already by Plutarch of Athens, as reported by Ps.-Philoponus, that philosophical works have their unique themes, which should as far as possible govern their interpretation.36 For the consensus is that the unique theme of the De anima is the souls of natural beings, so the only rational soul within its scope is human.37

Conversely, however, the view that the productive intellect is part of the human rational soul does seem to entail that at least part of the human rational soul is immortal. That is to say, given what Aristotle says about the productive intellect in the remainder of De anima 3.5, the immortality of at least part of the human rational soul is a necessary but not a sufficient condition for the productive intellect’s being part of that soul. Consider, for instance, the following text:

(T3) And this intellect is separable, unaffected and unmixed, being in its essence actuality (430a17–18).38

T3 follows immediately on T2, so it seems inevitable that the phrase “this intellect” in 430a17 should refer to the productive intellect, introduced in T2b. A few lines further down it appears – notwithstanding some difficult interpretative issues – as though Aristotle goes on to say that the productive intellect is the only thing that is immortal and eternal:

(T4) And once it is separated, [this intellect] is precisely what it is and nothing else, and this, and nothing else, is immortal and eternal […] (430a22–23).39

In view of this it goes without saying that it is impossible for Alexander, who thinks that the human soul is wholly inseparable from its natural body and therefore mortal, to identify the productive intellect with any part of it.40

It is worth mentioning also how the Neoplatonists deal with T3 and T4. Philoponus’ approach is to take “this intellect” in 430a17 to refer to the sum total of potential and productive intellect, that is, the whole human rational soul, and to explain the fact that it is characterised in 430a18 as “in its essence actuality” by saying that this is not because it is not at all potential, but because like everything else it is characterised in accordance with its form rather than its matter.41 A similar explanation is afforded by Ps.-Philoponus.42 If this is accepted, there is nothing to prevent taking the subject of T4, too, to be the whole human rational soul. And hey presto, what Aristotle is saying here is simply that the rational soul – as a whole – is the only thing that is immortal in human beings. Or rather, according to Ps.-Philoponus, that it is the only thing in human beings that is both immortal and eternal, “for the non-rational and the vegetative soul […] are immortal but not eternal.”43

3 Grammar and Lexicon

But let us now look at our text (T2), and let me begin by making some brief remarks on matters of grammar. The compound sentence in T2a–d (430a14–15) presents a few syntactic problems. I would like to draw attention especially to T2c–d, which is where the light analogy appears.

T2c is made up of the phrase “hṓs héxis tis” (“in the manner of a kind of state”). This can be taken either as a predicative complement of the subject of the sentence, “ho dé” (“the other [intellect]”) in T2b, or as an adverbial modifier of “poieîn” (“making”) in the same segment. On the former option, it is the productive intellect itself that is said to be like a kind of state such as light; on the latter, which is compatible with several different construals of T2b as a whole,44 it is rather the manner in which the productive intellect makes (all) things that is compared to the manner in which a kind of state such as light makes colours. Scholars have been divided over the issue. Thomas Aquinas read the phrase as a predicative complement: this construction was forcefully defended by Franz Brentano, who was followed by Georges Rodier.45 To the extent that it is possible to ascertain the way that the Greek commentators construe the sentence, it seems that this is also their preferred option,46 whereas modern interpreters and translators, at least from Robert Drew Hicks onwards, tend to take the phrase as an adverbial modifier of “poieîn,”47 which is what I have also done in the translation above.

The good news is that, despite Brentano’s protestations to the contrary, it really makes little difference which construction we choose. For even if we follow the predicative option, it must be because or in so far as it makes (all) things (“tôi pánta poieîn,” instrumental dative) that the productive intellect is said in T2b–d to be like a kind of state such as light. On either construction, then, it is most natural to take the point of the comparison to be to qualify the manner in which the productive intellect is “productive,” namely, in the same restricted sense in which a kind of state such as light makes colours. For we are told in T2e that light makes colours only in a restricted sense (“in a certain way”). Again, to the extent that it is possible to ascertain the opinions of the Greek commentators, they seem to agree that this is indeed the point of the comparison.48 Most if not all of them also agree that when Aristotle says that light “makes what are potentially colours into actual colours,” what he really means is that light makes actual colours, which are potentially visible, actually visible.49

A second syntactic problem is whether the appositive “hoîon phôs” (“such as light”) in T2d is restrictive (answering to an omitted “toiaútē” in T2c), so as to specify a unique kind of state as the secundum comparatum, or non-restrictive, so that it simply offers one among many possible examples of the kind of state – or héxis – to which the productive intellect is compared. But if the point of the comparison is to qualify the manner in which the productive intellect is “productive” as that which is described in T2e, the scales seem already for semantic reasons to be tipped in favour of the former alternative. For it is not any random héxis that “makes” things in the sense in which light makes colours: the art of building, for instance, is a héxis, and a productive one to boot; but this héxis brings buildings into existence and light does not, on Aristotle’s theory, bring colours into existence. In fact, as we shall later see, it does not even act upon them: rather, it allows them to act, that is, to pass from second potentiality to second actuality. The art of building, in contrast, imposes the form of a building on some building materials, which are thereby subjected to genuine change (kínēsis), which is a development from first potentiality to first actuality.

Accordingly, if the point of the comparison is to qualify the manner in which the productive intellect is “productive” as that which is spelt out in T2e, it is clear that the real secundum comparatum must be light, so that it is with good reason that we speak of Aristotle’s “light analogy” rather than his “state analogy.” In that case, the question is only why Aristotle would have bothered to mention that light is a héxis in the first place (if, indeed, he did: as we shall see, there may be cause for doubt). To be better equipped to answer this question, we shall have to inquire briefly into the concept of héxis. What role could it possibly play in the context, and how did the commentators understand it?

Héxis” is the action noun of the verb “échein.” But “échein” can be used both transitively, so as to mean “to have – or hold, or contain – (something),” and intransitively, usually with an adverbial modifier, so as to mean “to be (somehow) disposed.” And “héxis” is the action noun of both transitive and intransitive “échein.” As a result, it is radically ambiguous.50 It is so because it is impossible to reduce either of the two verbal actions – on the one hand, the having of something, on the other, a disposition – to the other, or indeed both to a common core. Every time “héxis” is translated, then, the ambiguity between the two verbal actions has to be resolved – unless it is presumed that the author conflates them.51

Aristotle uses “héxis” in both these senses. And he seems to be aware of it. In fact, it seems to me (pace Stephen Menn52 ) that he makes precisely the distinction between the two verbal actions in his discussion of the different ways in which a thing can be said to be a héxis in Metaph. 5.20, 1022b4–14. Moreover, I think he shows in this discussion that he has as good a grasp of the cause of the ambiguity as could be expected from someone in possession only of a very rudimentary grammatical conceptual apparatus.53 At any rate, he there makes a basic distinction between (1) “a kind of activity of the possessor and the thing possessed,”54 and (2) a kind of disposition (“diáthesis”),55 and points out, as a characteristic feature of (1), that it is impossible, on pain of an infinite regress, to have such a héxis.56 But this impossibility is characteristic of all and only transitive héxeis.

The definition of “héxis” in Categories 8 as the more enduring subspecies of disposition (“diáthesis”) – dispositions being one species of quality besides natural capacities and incapacities, affections, and figure or shape – clearly relates to the verbal action of intransitive “échein.” This is the sense familiar from Aristotelian ethics, in which moral excellence is a kind of héxis in virtue of which we are well disposed relative to our affections.57 Likewise, opinion, reasoning, scientific knowledge, and intellect, “by which we grasp the truth,” according to Posterior Analytics 2.19, 100b5–6, are héxeis in this sense.58Héxis” in this sense is standardly translated as “(positive) state.” Again, there is nothing impossible about having a héxis in this sense.59

So far, so uncontroversial. Some scholars seem to suppose,60 however, that the sense in which a héxis is opposed to a privation must relate to the verbal action of transitive “échein.” After all, the opposite of lacking a thing is having it. But this may not be as straightforward as it appears. For it may well be the case that within this opposition, too, a héxis is on the basic level simply conceived of as a state, the absence of which, in a subject in which one may expect it to be present, is a privation (stérēsis). But since (in Greek, but not in English) a natural way to distinguish a subject in which the state is present from one in which it is absent is to say that the former “has” (English: “is in”) the state, whereas the latter does not, it is but a small step to referring, in the same context, to the having of the state, too, as a “héxis.” The step may be all the more tempting as it allows “héxis” to be opposed to “stérēsis” even when what is present or absent is not a state, but, for instance, an activity. In this way, the privation will be opposed in one way to the state itself and in another way to the having of that state,61 but the Greek-speaking philosopher will express both relations by the term pair “héxis”–“stérēsis.”62 Thus “héxis” will be used alternately of states and of havings – and of havings, indeed, that are not even necessarily of states.63

So héxeis such as moral excellence and scientific knowledge are states. They can be possessed, but are not themselves possessings. What these two héxeis also have in common is that they enable activities. That is to say, they are “first actualities,” as these are described, notably, in De anima 2.5, 417a21–b2. In the Greek commentators from Alexander onwards, “héxis,” in the sense of “state,”64 is so frequently – and sometimes formally65 – opposed, on the one hand, to “epitēdeiótēs” (“suitability” or “first potentiality”), and, on the other hand, to “enérgeia” (“activity” or “second actuality”), that it seems fair to say that it is a technical term for “first actuality.” Accordingly, all Greek commentators refer to the intellect in second potentiality described in De anima 3.4, 429b5–9, as being in possession of its héxis (or “on the level of héxis”: “kath’ héxin”).66

What kind of héxis is it, then, to which the productive intellect is compared in T2c? Is it a having or a state? Is it a “first actuality”? In T2d it is specified as light. So it should be the kind of héxis that light is. That light is the actuality (enérgeia, entelécheia) of a transparent body qua transparent is expressly stated in De anima 2.7, 418b9 and 419a11. It is also implied in the same chapter that light is a héxis that belongs to a transparent body:

(T5) We have said, then, what the transparent is and what light is: it is neither fire nor in general a body – nor an emanation from any body, for in that case, too, it would be some kind of body – but the presence of fire or something similar in the transparent. For, to begin with, it is not possible for two bodies to be in the same [place]. Moreover, light is considered to be contrary to darkness. But darkness is a privation of this kind of héxis from something transparent. It follows, clearly, that light is also the presence of this [sc. fire or something similar in the transparent].67

For the argument to be valid, the phrase “this kind of héxis” (418b19) must refer either to “fire or something similar” – in which case darkness is contrary to light in the sense of being the privation, in something transparent, of that of which light is the presence – or to “the presence of fire or something similar” – in which case darkness is contrary to light in the sense of being, simply, the privation of light. The former alternative can be excluded out of hand: a héxis cannot be a body.68 It follows that the kind of héxis we have to do with in this passage is the presence (“parousía”) of one body in another body, although, admittedly, not in the strictest sense of the word, since one body cannot in the strictest sense be present in another body: they would be in the same place, and that is ruled out as impossible in the following line.

Does this also mean that it is the one body’s “having” the other body? Or, more precisely, is it because illuminated air or water “has” fire or aether that Aristotle calls light a “héxis”? An alternative interpretation may be that what is described as the presence of fire or aether in a transparent body is the effect on the transparent body of being acted upon by a contiguous body of fire or aether. If this is what light is, then Aristotle’s rationale for calling it a “kind of héxis” may have been that he wanted to do justice to the fact (as he saw it) that this effect is a state, a being-somehow-disposed, of the transparent body. But this is already somewhat speculative, and one may worry about whether light is enduring enough to qualify as a héxis in this sense anyway. Other states in Aristotle seem to correspond to the definition in Categories 8: at least they do not have the tendency of sublunary light to go on and off once a day or more.69

On the sole evidence of De anima 2.7, then, it may seem better to accept that light is a héxis because illuminated air or water “has” fire, either in the sense that fire is at its disposal, or – as suggested both by the prepositional phrase “in the transparent” (418b16–17) and by the parallel passage in De sensu 370 – in the sense that it contains fire.71 This would imply that “héxeōs” in 418b19 is opposed to “stérēsis” not as any enduring state is opposed to the corresponding privation, nor indeed as the simple and unqualified having of an enduring state or of anything else is opposed to the corresponding privation, but as the containing of a body is opposed to the corresponding privation, which is a non-containing of the selfsame body (which might be why Aristotle qualifies it as “this kind of héxis”).

The obvious objection to this is that when Aristotle in T2 is comparing the productive intellect to a héxis such as light, he cannot possibly be using “héxis” in the sense of the containing of a body or even in that of having a body at one’s disposal. And if the choice is between contending that the productive intellect is compared to the containing of a body and accepting that Aristotle conceived of light as an enduring state, there is no doubt that the latter option, as worrying as it may be, is much to be preferred.

4 Héxis in the Commentators

Neither Alexander nor Themistius ever speaks of light as a héxis. Philoponus, however, being the author of a line-by-line commentary, is duty-bound to take note of Aristotle’s usage in T5, which he does at in de An., 344.13–17. It seems likely from a comparison with in de An., 341.14–16 that he understands the héxis of which darkness is said to be a privation as a héxis in the sense of state, which to him as to the other commentators implies that it is a first actuality.72 For while it may be true that light is the second actuality (or enérgeia) of the transparent body qua transparent,73 it is only when illuminated that the transparent body is susceptible of being acted upon by colour, which means that light is the first actuality of the transparent body qua transmissive of colour:

(T6) For light is a héxis of what is transparent, but colour is such as to perfect the actuality on the level of héxis. For when the colour is present, that which is transparent, in turn, becomes in actuality such as to transmit the colours.74

In his commentary on De anima 2.5, Philoponus defines perfect actuality (as opposed to imperfect actuality, i.e., change) as “the instantaneous projection of the héxis […],” and provides an interesting example: “[…] such as the projection of light is: for all that is suitable is instantaneously illuminated at the same time as the illuminating [body] appears.”75 Presumably the héxis in the definition is the subject of the projection – that is, the illuminating body’s ability to illuminate – and light in the example is the (internal) object of the projection, that is, a second actuality (enérgeia). If so, there may be an echo in Ps.-Philoponus, who must have forgotten that the art of building is also a héxis: since a héxis, he says, projects actuality (enérgeia) but not substance, Aristotle’s comparison of the productive intellect to a héxis shows that the productive intellect cannot be, as Alexander thought, the divine intellect, which projects both actuality and substance.76 However, Ps.-Philoponus also seems to share Philoponus’ understanding of the sense in which light is a first actuality, since he says that it projects the actuality of colours.77

From Philoponus’ and Ps.-Philoponus’ point of view, then, light is a héxis only in so far as it is a first actuality. Accordingly, if the productive intellect is compared to a héxis such as light, it is compared to a first actuality.78 But at the end of the day, a first actuality is only a second potentiality. Philoponus takes the comparison to confirm that Aristotle’s productive intellect is not, as Alexander thought, the divine intellect, for the divine intellect is no héxis: it is “from the outset […] actuality without potentiality.”79 On the other hand, as Aristotle says in the text to which Philoponus refers in the cited passage (de Interpretatione 13, 23a21–26), an actuality that is combined with potentiality is posterior to it in time. So the comparison can also be taken to suggest that whatever possesses the productive intellect must previously have had a potentiality for possessing it. Still, it is not necessarily the case that Philoponus thinks it must have had a first potentiality, for, as we shall see, he distinguishes between different degrees of second potentiality. Again, Ps.-Philoponus aligns himself with his non-spurious namesake by adding, quite explicitly, that a héxis is a second potentiality, whereas the First Cause is not in any kind of potentiality.80 Arguably, however, he breaks ranks a couple of pages later, when he explains that the intellect in actuality stands to the potential intellect “just as the héxis stands to the first potentiality (for it perfects it).”81 For in contrast to Philoponus, Ps.-Philoponus is adamant that the potential intellect is in first potentiality. I shall return to this point towards the end of the chapter.

As has been seen, Philoponus and Ps.-Philoponus both argue that the comparison of the productive intellect to a héxis rules out the possibility that Alexander was right to identify it with the First Cause. Their arguments are based on the assumption that the relevant kind of héxis is a first actuality, an assumption that sits well with their own view that the productive and the potential intellect are one and the same, whether numerically (Ps.-Philoponus) or not (Philoponus). But does the criticism hit the mark? Inasmuch as Alexander shares the assumption that the relevant kind of héxis is a first actuality – or even that the relevant kind of héxis is inseparable from the thing that has it82 – and inasmuch as he believes – as Philoponus suggests, and as seems reasonable anyway – that Aristotle in T3 is referring to the productive intellect, it will certainly not be easy for him to explain how the productive intellect can be a héxis. Still, the only thing that one can reasonably demand that he explain, in his capacity of an interpreter of Aristotle, is how it can in some respect be comparable to a héxis. This should be a more manageable task, so to that extent Philoponus’ and Ps.-Philoponus’ criticisms are irrelevant. But there may be an even stronger reason for dismissing them. Let us see which.

Perhaps the most intriguing feature of the Greek reception of Aristotle’s light analogy is this: not only do Alexander and Themistius omit from their paraphrases the comparison of the productive intellect to a héxis, they do so under such circumstances that scholars have been led to suspect that they may have had a different text from ours.83 Both of them speak, in paraphrasing T1–T2, of the héxis of the intellect, but not as a secundum comparatum of the productive intellect: on the contrary, this héxis is the result of the productive intellect’s action on the potential intellect. In other words, it is the first actuality of the human intellect described by Aristotle at de An. 3.4, 429b5–9. Thus Alexander says in the immediate sequel to his paraphrase of T1 that

(T7) since there is a material intellect, there must also be a productive intellect, which becomes the cause of the héxis of the material intellect.84

Similarly, Themistius affirms that by promoting the potential intellect to actuality “the intellect in actuality completes the intellect on the level of héxis.”85 And according to the first section of Alexander(?), De intellectu (Sharples’ “A” section), the productive intellect, which Aristotle is said to have compared to light, “makes the potential and material intellect into an intellect in actuality by imposing the intellective héxis on it.”86

Pierluigi Donini suggested that the “singularity” of Alexander’s interpretation could have been obtained by “simply modifying the word order,”87 so as to read:

(T2’) (a) And this kind of intellect is, in virtue of becoming all things, (c) like a kind of state, (b) whereas the other [is], in virtue of making all things, (d) similar to light.88

Such transpositions, he claimed, are an interpretative method well attested in Alexander’s commentaries: he referred especially to in Metaph. 221.34–222.3.89 But in that passage Alexander (1) expressly says that the clause order of the transmitted text makes for obscurity, and (2) only suggests an alternative clause order that makes the sense of the transmitted text clearer, not one that alters the sense. That T2’, in contrast, has a different sense from T2 can hardly be denied: for how could it otherwise explain the “singularity” of Alexander’s interpretation?

To my mind, it is scarcely conceivable that Alexander would have interfered with T2 in the way suggested by Donini unless he considered the text to be corrupt. If he did, his reasons for emendation will doubtless have been stated in his commentary proper; but since this, alas, is lost, we can only speculate. The later commentators’ silence on the matter may perhaps be taken as an indication that there was no such interference with T2 on Alexander’s part. On the other hand, the “singularity” not only of Alexander’s interpretation, but also of those of Themistius and Alexander(?) in the “A” section, would be equally well explained on the hypothesis that T2’ was in fact the text transmitted in the manuscripts available to these authors. And there are independent reasons to suspect that this may have been the case.

The results of our inquiry so far have shown that “hṓs héxis tis” is problematic in its current location (T2c). That the productive intellect cannot be both a héxis and what is described in T3 and T4 is perhaps no insurmountable difficulty: T2b–d is after all a comparison, which may focus on some specific common feature that does not involve separability or mode of existence. Still, the obvious candidate for such a common feature is the way in which the terms of comparison “produce” things, and in this respect the productive intellect is not comparable to héxeis (in the sense of states) in general, since many héxeis produce substances and the productive intellect apparently does not. As we have seen, there is reason to doubt whether Aristotle really conceives of light as a héxis in the sense of state. But even if he does, he can hardly think that this is more than coincidental to the fact that the productive intellect is comparable to light. So it remains unclear why he should mention in this connection that light is a héxis.

Doubts about our text may also be encouraged by the fact that Theophrastus (fr. 320B), according to Heinze’s Greek text of Themistius, in his “investigations concerning Aristotle’s productive intellect,” asked himself what the consequences are “if the potentiality (hē dýnamis) is like a héxis.”90 This question certainly sounds as if it were prompted by T2’ rather than by T2. According to Dimitri Gutas, however, the text on which Isḥāq ibn Ḥunayn’s translation was based must have read “if it (sc. the productive intellect) is like a héxis or a potentiality (ē dýnamis).”91 And the three Neoplatonic commentators clearly had our text in front of them, as did Marinus, if Ps.-Philoponus’ report is anything to go by.92

The long and the short of it is that while the indirect tradition supports the reading of our manuscripts from the fifth century onward, in its earlier stages it never seems to connect a héxis with the productive intellect, but rather with the potential intellect, as in T2’. There is one apparent exception. This is Alexander(?), De intellectu 113.4–6 (Sharples’ “C” section), where an anonymous philosopher is reported to have said that

(T8) one should also adapt the text in the third book on the soul to these [doctrines] and bring the héxis and the light to bear upon this [intellect], the one that is everywhere.93

According to the doctrines referred to by the anonymous philosopher – and rejected by Alexander(?) in the following section94 – the intellect in actuality always pervades the whole corporeal realm: whenever an individual human body develops an organ, or instrument, suitable for use by it, the intellect in actuality latches on to this and individual human intellection ensues. This instrument is what Aristotle calls a potential intellect. The intellect in actuality is compared by the anonymous philosopher to an artisan whose art is exercised sometimes with, sometimes without an instrument.95 Thus conceived, the intellect in actuality can be separable, in its essence actuality, and eternal, in compliance with T3 and T4.96 But relative to individual human intellection it will be a héxis, that is, a second potentiality, which is activated by the presence of a suitable instrument, that is, a potential intellect. It does not fit this conception to compare the potential intellect, even when wielded by the intellect in actuality, to a héxis, since the idea is – as Alexander(?) complains in his reply97 – that the real agent of human intellection is the intellect in actuality, whose activity, in this specific case, is only channelled through the potential intellect.

It may be argued that the anonymous philosopher only means to recommend that T2c should be reinterpreted, not emended by transposition. But the only conceivable reinterpretation that does not presuppose a standard interpretation according to which “the héxis and the light” are the secunda comparata of two distinct intellects is one that simply substitutes the anonymous philosopher’s conception of the productive intellect for any other conception, and it would be strangely superfluous for the anonymous philosopher to end a prolonged argument in favour of his own conception by recommending that this particular passage should be reinterpreted accordingly, unless the passage presented some particular obstacle to this conception.

T8 is really only comprehensible on the presumption that in the standard interpretation “the héxis and the light” are the secunda comparata of two distinct intellects. Such an interpretation is also implied by the paraphrases of Alexander and Themistius, as well as Alexander(?)’s “A” section. It is difficult to see how it could have become standard unless the transmitted text corresponded to T2’ rather than to T2. And on the most natural reading of T8 it recommends emendation of the text in such a way as to make both “the héxis and the light” secunda comparata of the productive intellect. I would submit, therefore, that Alexander(?)’s “C” section should be added to the group – otherwise consisting of Alexander, Themistius and Alexander(?)’s “A” section – of paraphrastic witnesses that testify to a text of De anima 3.5, 430a14–15 corresponding to T2’ rather than to T2. Since all these witnesses are older than the entire direct tradition, their testimony should carry considerable weight.

It can hardly be claimed, however, that emending in accordance with T2’ would instantly resolve all the interpretative problems relating to T2. To begin with, it may seem to create a new one. As we have seen, T2 applies, either by analogy or by subalternation, the principle of causation set out in T1 to the human intellect. We should expect T2’ to do the same. Accordingly, “this kind of intellect” in T2’a should be the instantiation “in the soul” of the material cause invoked in T1. A suitable candidate for such a role is the kind of intellect mentioned in de An. 3.4, 429a22–24 and compared in 429b30–430a2 to a blank writing-tablet, since it is in first potentiality (contrast 429b5–9). But whereas the intellect mentioned in T2a can be identified with the intellect in first potentiality, the intellect mentioned in T2’a cannot, since a first potentiality is not (comparable to) a héxis. The intellect in first potentiality only comes to be “like a kind of state” by becoming all things, that is, as a result of being acted upon by the productive intellect. And then it is already the kind of intellect in second potentiality discussed by Aristotle in de An. 3.4, 429b5–9, before the excursion on “thises” and essences leading up to the two puzzles at the end of that chapter. It is entirely, I think, within the realm of possibilities for “this kind of intellect” in T2’a to refer back to this passage,98 but rather more disputable – to say the least – whether the intellect in second potentiality can instantiate the material cause invoked in T1.

On the positive side, however, emending in accordance with T2’ might go some way towards resolving a problem that we have not yet really dealt with, namely, how the productive intellect can be compared to light.

5 Light in the Commentators

As I said above, we should expect “this kind of intellect” in T2’a to be an instantiation of the material cause invoked in T1. By the same token, we should expect “the other [intellect]” in T2’b to be an instantiation of the efficient cause also invoked in T1. In T2’, as in T2, “the other [intellect]” is compared to light. We have seen that the point of the comparison is most naturally taken to be to qualify the way in which this intellect is “productive,” that is, the sense in which it is an efficient cause. Light, Aristotle says in T2e, is productive in the sense of actualising colour. Whether or not “colour” needs to be corrected into “the visibility of colour,” in accordance with the Greek commentators’ suggestion, the “production” in question can only amount to, as Thomas Aquinas pointed out, making the transparent body susceptible of being acted upon by colour.99 In fact, light is this susceptibility, which enables colour to pass from first to second actuality. If one accepts that light is a héxis in the sense of state, one can argue that the action of the colour upon the transparent body promotes the latter from its illuminated state to the corresponding second actuality, which is the actual transmission of the colour: this train of thought was exemplified in T6. But it remains a fact that the action is caused not by the light but by the colour. In short, it would seem that, on Aristotle’s theory, light does not really act upon colour: it merely enables it to act upon the transparent body.100 But this does not seem to qualify it for the role of efficient cause in the sense adumbrated in T1.

Consequently, if the productive intellect makes things in the manner in which light “makes” colours, that is, by enabling them to act, it is not really an efficient cause in the sense adumbrated in T1. It definitely cannot be compared to an art (e.g., that of building), which imposes a form upon some suitable matter, thereby changing what is an F in first potentiality into an actual F. But our reason for thinking that there is a productive intellect in the first place is that (according to T1) the slot for efficient cause must be filled, as “in the whole of nature,” so also “in the soul.”

There is another side to the coin. If the potential intellect is actualised in the manner in which colours (or their visibility) are, it is itself a causal agent and cannot be a first potentiality. In which case, of course, it makes perfect sense to compare it to – indeed to say that it is – a kind of héxis.

When viewed solely in terms of the light analogy, then, T2’ makes perfect sense. “This kind of intellect” in T2’a would refer back to de An. 3.4, 429b5–9. The productive intellect would “make (all) things” in the sense of enabling the potential intellect’s activity. The potential intellect would “become (all) things” in the sense in which – on Aristotle’s view – a field of lilies becomes multicoloured at sunrise. The productive intellect’s role on such an interpretation is not to promote an intellect from first potentiality to héxis (first actuality) but to allow episodes of intellection (second actuality) to happen. But on such an interpretation, of course, “this kind of intellect” in T2’a cannot instantiate the material cause invoked in T1.

Alexander, as we have seen, takes T1 seriously. He infers from the principle invoked by Aristotle that there is one intellect that is literally material (for change) and one that imposes its form on the material intellect. As a result, the material intellect is promoted to first actuality. Accordingly, if Alexander’s manuscripts did exhibit T2’, he would have been forced to read the clause T2’a–c as if it stated, not that “this kind of intellect” is a héxis, but that it is made into a héxis by the productive intellect, as suggested by his paraphrase (and even more clearly by Alexander(?), de Int., 107.29–30).101 But this would not have helped him with the light analogy. Fortunately, Alexander had the resources to make sense of the light analogy even within the framework of the interpretation suggested by his paraphrase. As we shall see below, he displays but does not actively deploy them in the De anima.

It has often been pointed out that Aristotle’s light analogy owes a debt to the analogy of the sun in Plato’s Republic 6 (507a–509b). It has often been noted, too, at least from Themistius onwards,102 that, in spite of this, the two analogies are different. We saw above that Philoponus and Ps.-Philoponus take the comparison of the productive intellect to a héxis to imply that Aristotle cannot have meant to identify the productive intellect with the First Cause, as Alexander thought. They take the comparison of the productive intellect to light to imply the very same thing, and for the very same reason: light, too, produces actuality, not substance. At de Int., 57.57–58, for instance, Philoponus says that

(T9) if he [sc. Aristotle] were speaking in this passage of the creative intellect, it would have been more reasonable to compare it to the sun than to light.103

And Ps.-Philoponus makes a similar statement at in de An., 537.27–28. In effect, then, both commentators reproach Alexander for misinterpreting Aristotle’s light analogy along the lines of Plato’s sun analogy. At in de An., 539.35–39, Ps.-Philoponus suggests that it was this misinterpretation that led Alexander to misidentify the productive intellect, and that the misinterpretation was in turn based on the failure to notice that Aristotle only says that light makes all things in a certain way. Philoponus’ diagnosis, again, is similar.104

To a certain extent, I suppose Alexander reaps what he has sown here. Having first argued, in the immediate sequel to T7, for identifying the productive intellect with an intrinsically immaterial form,105 he proceeds to add, in apparent emulation of Plato (Republic 6, 509b), that the First Cause, being the cause of all other things (and this time “things” should no doubt be taken in the widest possible sense), is also “productive” in the sense of generating “every object of intellection”106 – a phrase that in this context must refer to the enmattered intelligible forms or essences that, according to Alexander’s theory, will only subsequently be promoted from potential to actual intelligibility in the act of intellection (see below). This addition left an indelible (and I think unfortunate) mark on the twentieth-century discussion, since it convinced Paul Moraux that the productive intellect’s task in Alexander’s De anima is to produce potentially intelligible objects, the actualisation of which is then taken care of by the individual human intellects.107

But this is not the role assigned by Alexander to the First Cause qua productive intellect. On the contrary, this role is to promote the material intellect to its héxis, as he says in T7. For clearly, when he goes on to say, at de An., 89.6–7, that “if there did not exist something intelligible by nature, nor would any other thing become intelligible,”108 what he means is that if there did not exist something actually intelligible by nature – for if something is intelligible by nature, it is actually intelligible109 – nor would any other thing become actually intelligible. In this regard, then, neither the Neoplatonists’ nor Moraux’s criticism seems deserved.

Moraux also brought other charges of Platonism against Alexander, this time referring to Alexander’s actual argument for identifying the productive intellect with an intrinsically immaterial form. This is again the application of a general principle, namely:

(P) in every set of things with a certain property F, what is strictly and eminently F is the cause of the F-ness of the other members of the set.110

Thus it is reasonable, Alexander says, that what is strictly and eminently intelligible should be the cause of the intellection of other (less intelligible) objects too.111 And since all intelligible objects are forms, but enmattered forms are potentially intelligible before being rendered actually so, whereas intrinsically immaterial forms are always actually intelligible, what is strictly and eminently intelligible must be an intrinsically immaterial form – granted that there is such a form, of course.112 According to Moraux, P is an illegitimate (Platonist) conversion of the orthodox Aristotelian principle that the cause is greater than its effect.113 There is no need to discuss here the legitimacy of the conversion, except to say that it was defended, I think rightly, by A. C. Lloyd.114

When Alexander suggests that the intrinsically immaterial form is the cause of the intellection (“nóēsis”: de An. 89.5) of other objects, the italicised word is not necessarily a mistake for “intelligibility,” as one might suspect. It appears from several passages in the De anima and elsewhere that it is Alexander’s view that nothing can be actually intelligible unless it is actually being intelligised.115 Ultimately, this is the reason why enmattered intelligible forms are in themselves incapable of causing intellection in a way analogous to that in which enmattered perceptible forms cause sense perception. In order to be an object of intellection, an enmattered form must be separated from its material environment by the intellect, and this separation is already an act of intellection.116 It makes no difference, then, whether what is eminently intelligible is said to be the cause of the actual intelligibility of the less intelligible objects or of their actually being intelligised. If, on the other hand, Alexander had meant that what is eminently intelligible is the cause of the potential intelligibility of the less intelligible objects, as Moraux maintained, the word “intellection” would have had to be a mistake.

However, if this is Alexander’s view, a problem looms. For the less intelligible objects would then have to be promoted from potentiality to actuality for each new episode of intellection; in other words, it would be impossible for concepts to be retained as such. This is a consequence that Alexander seems to accept a bit later in the text:

(T10) For to be sure, universal and common items have their existence in particular and enmattered things, but it is when they are intelligised without matter that they become common and universal […]. Accordingly, when separated from the intellect that intelligises them, they pass away, assuming that their being consists in being intelligised.117

But the evidence is ambiguous. At de An., 86.5–6, for instance, Alexander speaks of the intellect on the level of héxis, that is, in first actuality, as “in a way” a storehouse of concepts “at rest.”118 Exactly how this metaphor is to be understood – for instance, whether there is a role here for the faculty of phantasía – will have to be deferred to another discussion.

Alexander adduces two other examples of P: light, which, being strictly and eminently visible, is the cause of the visibility of other visible objects, and the eminently good, which is the cause of the goodness of other good things,

(T11) for the other things are deemed good on account of being conducive to this [sc. that which is eminently and primarily good].119

At first glance, neither of these examples seems to be a valid application of P within an Aristotelian theoretical framework.120 To begin with the light-and-visibility example (de An., 89.1–2), it presumes that light is strictly and eminently visible in the same sense in which colours are visible. This does not seem to be an orthodox Aristotelian presumption, since light on the Aristotelian theory is not visible in itself in the sense of having an intrinsic cause of visibility, but only on account of extraneous colour.121 Alexander, on the other hand, repeatedly says, and so presumably thinks, that light is what is strictly and eminently visible.122 This puts him in a position to interpret the light analogy in a way that may seem to honour the principle of causation set out in T1, but that Aristotle could not, I think, have intended.

The goodness example (de An., 89.2–4) is a more complex case. Again, there is no need to discuss the details here: Moraux condemned what he saw as a relapse into a Platonic theory of participation; others have spoken in Alexander’s defence, most successfully, perhaps, Accattino and Donini, who drew attention to the correspondences between Alexander’s example and Eudemian Ethics 1.8, 1218b7–24.123 The fact that Aristotle in the latter text speaks of the eminently good as the final cause of human actions does not necessarily undermine the relevance of these correspondences, since the general principle P, which Alexander’s examples are meant to illustrate, is a principle for identifying an unspecified type of cause of any determinate property. Accordingly, there is no reason to expect the examples to be restricted to efficient causes only. It is perhaps more of a worry that there are no other indications that Alexander was conversant with the Eudemian Ethics.124

Be that as it may, it is worth noting that both of these examples, the goodness example as well as the light-and-visibility one, are introduced as special instances of P; but P is only a principle for identifying causes, employed by Alexander to identify the cause of intellection. That is to say, neither example is part of a paraphrase of Aristotle’s comparison of the manner in which the productive intellect produces things to the manner in which light makes colours. That is not to say, however, as I have already hinted, that the light-and-visibility example does not lend itself to such a paraphrase. If the productive intellect is what is strictly and eminently intelligible and light is what is strictly and eminently visible, then the productive intellect and light cause other things to be, respectively, intelligible and visible, by the same principle, namely P. That is, they “make (all) things” by being, respectively, the eminently intelligible and the eminently visible object. As for Alexander’s interpretation of the light analogy, it is probably reflected in Alexander(?), de Int., 107.31–108.2 (Sharples’ “A” section) and especially 111.32–36 (Sharples’ “B” section).

In the final section I shall briefly discuss how some of Aristotle’s Neoplatonist readers tried to exploit the light analogy to the full by extending its implications to the intellect compared by Aristotle in de An. 3.4, 429b30–430a2 to a blank writing-tablet.

As was argued above, if “this kind of intellect” in T2a is an instantiation of the material cause invoked in T1, it is reasonable to identify it with the intellect compared to a blank writing-tablet, since the latter is in first potentiality.125 Not unexpectedly, some Neoplatonist readers disputed this rather literal interpretation of the writing-tablet analogy. According to a report in Ps.-Philoponus, Iamblichus maintained that the whole point of the analogy is that the souls of children do contain the rational principles (lógoi) of things, albeit faintly and non-manifestly.126 A bit earlier in Ps.-Philoponus’ commentary, the same interpretation is attributed to Plutarch of Athens, who apparently for this reason redesignated the intellects of children as being “on the level of héxis.”127 For this he was criticised, Ps.-Philoponus says, by Ammonius, who may – if 519.37–520.6 is part of the criticism introduced at 518.32–33 – have protested that he foisted on Aristotle what is really a distinctively Platonic view. If this reconstruction is correct, it is to Ammonius’ credit that he took issue with what is evidently a rather strained interpretation of the writing-tablet analogy, despite being, presumably, as sympathetically disposed towards the Platonic view as he was towards the idea that the two philosophers are in fundamental agreement.128

Strained as it may be, this interpretation allows for the intellect in T2a to be identified with the intellect compared to a blank writing-tablet and still be “produced” in much the same way in which colour is “produced” by light, that is, by being enabled to act. The light analogy, as Philoponus points out, is grist to the mill of the Platonisers.129 Although the evidence is again somewhat ambiguous, it seems as though it was Philoponus’ idea to make Iamblichus’ and Plutarch’s interpretation seem more sensible in the following ingenious way.130 Assuming, with Aristotle, that the world is eternal and that an actual infinity of immortal souls is impossible, Philoponus argues that if the rational soul is immortal, it must (at birth) possess the forms “on the level of héxis,” that is, as a second potentiality; and, by contraposition, if it possesses the forms only as a first potentiality it must be generated and thus – since everything generated is perishable – mortal.131 But there can be no doubt, he says, that Aristotle considers the rational soul to be immortal.132

The apparent contradiction between the conclusion of this argument and Aristotle’s statement that the intellect prior to intellection is all its objects potentially but none actually133 is resolved by introducing a distinction between two degrees of second potentiality, illustrated by, on the one hand, a sleeping geometer and, on the other, a waking one, and suggesting that when Aristotle describes the intellect prior to intellection as being in a first potentiality,134 what he has in mind is the first of these two degrees: “the intellect that emerges in the world of becoming is comparable to a sleeping or raging person.”135 When he describes the intellect posterior to “learning or discovering” as being in a second potentiality,136 he has in mind the second degree. Philoponus finds support for this interpretation in Aristotle’s light analogy, inasmuch as the rising sun does not give subsistence to colours, but makes already subsisting colours manifest. In the same way,

(T12) intellect which is in actuality perfects intellect which is in potentiality and brings it to actuality not by putting into it forms which are not there, but by bringing to light forms which are non-manifest and hidden because of the state of swoon which is the effect of birth.137

And thus, once the intellect in the writing-tablet analogy has taken half a step forward and the one in the light analogy half a step backward, the two are indistinguishable. The individual human intellect is at birth in potentiality, not in the sense of being a mere suitability for receiving the intelligible forms, but in the sense of lacking the intellectual transparency required in order for the existing forms to be actually intelligised. This state may equally well be described as a first potentiality of the second degree or a second potentiality of the first degree. It is a héxis disabled by the circumstances.

This is where the teacher comes in (cf. above, pp. 39–40), whose role is simply to remove the opacity from the student’s intellect.138 Even though individual human intellects are not from birth in a position to launch themselves into second actuality, on account of “the state of swoon” that they are in, the fact that they are born with non-manifest and hidden forms dispenses with the need for an explanation as to how they have been promoted into first actuality. And assuming, with Aristotle, that the world is eternal, there will always have been teachers around to disperse the fog.139

One might have expected that the same role could also be played by experience.140 At de Int., 56.31–40, however, Philoponus tries to forestall the objection that since we can find out things by ourselves, the teacher is superfluous, by insisting that it is only when we have received the principles and the héxis from the teacher that we can find out things by ourselves. This looks like a throwback to a “transmission-model” understanding of the productive intellect’s action upon the potential intellect, and so it is tempting to speculate that it reflects Ammonius’ teaching.141

But it could also be a symptom of unresolved tension in Philoponus’ interpretation. For even if, admittedly, the analogy between Philoponus’ teacher’s intellect and Aristotle’s light is about as perfect as they come and, furthermore, there is nothing to prevent Philoponus’ student’s intellect from being identical with the intellect prior to intellection, since they are both understood to be in the first degree of second potentiality, there is still no way in which T2 on Philoponus’ interpretation can be the application of the principle of causation set out in T1. For on Philoponus’ interpretation, “this kind of intellect” in T2a is not really a material cause and “the other” intellect in T2b is not really an efficient cause.

Nor does it help to retain “hṓsper” and suppress “epeí” in T1a, since the divisions that must “obtain also in the soul” (T1f) are supposed to be identical with those “in the whole of nature” (T1a) regardless of the nature of the relation between the two realms and the reason for the inference. It is worth reflecting upon, however, that Philoponus’ justification of what he thinks is Aristotle’s analogy between nature and soul (see above, p. 37), namely that the soul, too, is “changed” with respect to the passage from potentiality to actuality, also places a limitation on the degree to which the divisions in the two realms can be identical.

Ps.-Philoponus takes a much stricter view of the potential intellect. He repeatedly insists that it is not “on the level of héxis” but a mere suitability, containing no rational principles, and sharply rebukes anyone who argues otherwise.142 It is surprising, therefore, to find that his elucidation of the light analogy is very similar to that of Philoponus:

(T13) For just as light does not itself make the colours, but makes those already existing manifest, so the intellect in actuality does not make things, but imprints and engraves those already existing on the potential intellect.143

It is tempting to read this in the light of the above-mentioned report of Iamblichus, and especially a passage in which Ammonius seems to be paraphrased to the effect that the potential intellect

(T14) has the intellection of all things, just as the underdrawing in a picture has the outlines of all the things [the picture] will receive, even though they are not manifest.144

As noted above, Ammonius is also reported by Ps.-Philoponus to have criticised Plutarch precisely for ascribing to Aristotle the view that children have in their intellects the rational principles of things although they do not yet cognise things.145

Are our Neoplatonists just being inconsistent? Desperate as it may seem, let us make an effort to clear at least Ps.-Philoponus of that suspicion. The metaphor of engraving reappears a couple of pages down, where Ps.-Philoponus is trying to show that “making all things” in T2b can be a property of the human intellect. It can, he says, because “making all things” means “inscribing the imprints of all things in the potential intellect.” Thus Aristotle “puts [the intellect in actuality] down as a scribe.”146 As employed by Ps.-Philoponus, the metaphor seems to allow that the forms should be conceived of as present in the potential intellect only in first potentiality (as characters are on a blank writing-tablet), while they actually pre-exist in the “intellect in actuality.” So the metaphor by itself does not seem incompatible with Ps.-Philoponus’ strict view of the potential intellect.

But how on earth is it to be combined with the light analogy? Does Ps.-Philoponus think of colours as somehow contained in and projected by light? Since no commentary by Ps.-Philoponus on De anima 2.7 is extant, we should obviously exercise caution, but we saw above (p. 51) that he describes light as “projecting” the activity of colours,147 only not, presumably, onto potentially visible surfaces, but rather onto potentially seeing eyes. It is possible, then, that the light analogy has been stood on its head in Ps.-Philoponus, so that the action of the “intellect in actuality” on the potential intellect is compared to the effect of light on the visual sense rather than on the potentially visible object. And this is perhaps not so unreasonable. For as we have seen, light does not, properly speaking, act on the potentially visible object any more than it does on the visual sense: it only enables the former to act on the latter.

It is, however, a presupposition of any interpretation of the light analogy according to which the action of the productive intellect is compared to the effect of light on the visual sense that Aristotle in T2e means to say not only that light promotes potentially visible objects to actuality, as all Greek commentators say he does (see above, n49), but that it promotes them to second actuality, that is, to being seen. For the only visible objects that exist in the visual sense are in second actuality.148 Whether this kind of interpretation is in fact endorsed by Ps.-Philoponus is not so easy to ascertain: “manifest” (“phanerá”) in T13 could refer to a first-actuality visibility as well as a second-actuality one. It is obviously difficult to square with his claim that the intellects in actuality and in potentiality are the same in subject, although not in time, but it might work with Philoponus’ identification of the former with the teacher’s intellect, understood on the “transmission-model.”

6 Conclusion

In broad outline, the following picture has emerged from the above discussion. One of the major challenges faced by the Greek commentators on Aristotle’s De anima 3.5, 430a10–17 was to negotiate the tension between the principle of causation from which the existence of a “potential” and a “productive” intellect is supposed to follow and the light analogy by which the relation between the two intellects is meant to be illustrated. The principle of causation suggests (1) that the potential intellect stands to the productive one as a material cause stands to an efficient cause. That is to say, it suggests that the potential intellect is promoted from first potentiality to first actuality by the productive intellect. The light analogy, in contrast, suggests (2) that the productive intellect merely enables the activity of the potential intellect, which must then already be in first actuality independently of the productive intellect – as colours, on Aristotle’s theory, are actual independently of light.

In Alexander’s interpretation, all the stress is on (1). Alexander may not have been particularly troubled by its conflict with (2), since his conception of light as the eminently visible object allowed a different interpretation of the way in which light can be “productive,” namely of vision rather than mere visibility. It remains the case, however, even if the “patient” of the “action” of light is understood to be the visual sense rather than colour, that this “patient” must be in first actuality independently of light. Philoponus, on the other hand, embraces the innatist implications of (2) for the interpretation of Aristotle’s view of “the intellect prior to intellection,” mentioned in De anima 3.4. He improves upon earlier Neoplatonic accounts by introducing a distinction between degrees of second potentiality intended to facilitate the identification of the potential intellect mentioned in 430a14–15 with “the intellect prior to intellection.”

Philoponus’ and Ps.-Philoponus’ criticism of Alexander’s interpretation is principally aimed at his identification of the productive intellect with the First Cause. This identification, they claim, is incompatible with Aristotle’s express requirements that the productive intellect should be (a) in the soul, (b) like a state (héxis), and (c) productive in the same way as light. Although based on Aristotle’s text, their criticism is arguably irrelevant. This is particularly the case with (b), since there is reason to believe that the productive intellect was not compared to a héxis in Alexander’s text of Aristotle. In fact, there seems to be no record of such a comparison in paraphrases and discussions of De anima 3.5, 430a10–17 before the fifth century CE. In these sources – including Alexander – it is instead the potential intellect that is said to become a héxis by the agency of the productive intellect. This suggests that they are based on a slightly different text from ours. Since this different text would readily lend itself to an innatist interpretation of the potential intellect, however, it seems unlikely that it was known to the Neoplatonic commentators, in whose accounts it has left no trace.

Appendix: Aristotle’s Light Analogy in the Late Byzantine Paraphrases


On the following pages I will briefly report and analyse the passages dealing with Aristotle’s light analogy in three Late Byzantine paraphrases of the De anima, namely those by Sophonias (fl. c.1285), Theodore Metochites (1270–1332), and George Scholarios (Patriarch Gennadius II, c.1400–after 1472). I have searched in vain for a discussion of the light analogy in George Pachymeres’ (1242–after 1307) Philosophia (book 7, part 3, ch. 5–8),149 a compendium that draws, for the relevant chapters, rather heavily on Priscian’s commentary and more lightly on that of Ps.-Philoponus.


As was noted above (p. 35n3), Sophonias’ paraphrase draws either on the lost commentary on De anima 3 by John Philoponus or perhaps, as Arnzen has argued,150 on a paraphrase of the De anima (*Ψ) closely related to both Philoponus’ and Ps.-Philoponus’ commentaries that has been lost in the Greek original but is preserved in an Arabic adaptation (with additions from other sources).

Since the parallels between the Arabic paraphrase and Sophonias are in some passages closer not only than those between the Arabic paraphrase and the late antique commentaries on which it ultimately depends but also than those between Sophonias and Philoponus, Arnzen concludes that *Ψ must have been still accessible to Sophonias.151 *Ψ must have been composed after c.575, regardless of the authorship of Ps.-Philoponus’ commentary, since the author was also familiar with works by the late sixth-century commentators David and Elias, and before c.830, when the Arabic adaptation was executed (Arnzen ascribes it to Yaḥyā ibn al-Biṭrīq, allowing for later redactional interventions152 ). The Arabic text was later translated into Persian by Afḍaladdīn Kāshānī (d. before 1268).

A clear idea of the nature of Sophonias’ paraphrase of De anima 3.4–8 can be gained from van Riet’s table of sources.153 Over the course of the fifteen pages devoted to these five chapters, there are about two lines per page that are not verbatim quotation or close paraphrase of either Philoponus’ commentary, Aristotle’s text, or, in two cases, other identifiable sources, one of which is Priscian’s commentary.154 Sophonias mentions the light analogy twice. The first time is in the course of his thirteen-line treatment of De anima 3.5, which appears, somewhat unexpectedly, in the middle of his section on De anima 3.4.155 It consists in a reproduction of Aristotle’s text with a few minor subtractions and additions, most notably the explanation, interpolated between T2d and e,156 that the creative intellect is similar to the sun, but our human intellect in actuality to light.157 As we have seen, this point is made by Themistius and repeated, as part of their criticism of Alexander, by Philoponus and Ps.-Philoponus.158 The nearest parallel in Philoponus’ commentary to Sophonias’ phrasing is probably de Int., 57.57–61.159

Sophonias addresses the question of whether the intellect in actuality is internal or external to the individual human intellect at in de An., 136.6–24. By way of reply he summarises the reports of the four views on the productive intellect in Philoponus, de Int., 43.18–45.59 and 48.28–32.160 His second mention of the light analogy is part of the description of the second view, according to which the intellect in actuality is “second to the first and divine [intellect], but also immediately superordinate to us and illuminating our intellect: this is also that to which, in [the proponents’] view, the example of the light refers.”161

Sophonias’ description of the fourth view is based on the report in de Int., 48.28–32, to the exclusion of that in 45.53–59, and his understanding of the way in which the teacher’s intellect works tends decidedly towards the “transmission model” suggested by de Int., 56.31–40, rather than the “illumination model” suggested elsewhere by Philoponus.162 Interestingly, he then seems to combine the report of the fourth view in de Int., 45.53–59 with Ps.-Philoponus’ interpretation of it163 into a fifth distinct view, which he himself endorses:

Others claim that this [sc. the intellect in actuality] is that of the teacher, which perfects the potential intellect in another person by depositing the theorems and concepts of the sciences, and which has itself once been brought from potency to actuality. A fifth view besides these, which I believe is closer to the truth, is the one which states that the potential intellect and the intellect in actuality are one and the same, and not external, but internal to one and the same soul and one and the same individual human being, differing from itself in respect of perfection and imperfection.164

Still, he ends by generously allowing that “the other views are also true if attuned to different interpretations.”165

Theodore Metochites

Theodore Metochites’ paraphrase of the De anima is part of a collection of paraphrases covering all of Aristotle’s extant works on natural philosophy, probably published around 1312–13.166 An edition of the De anima paraphrase is currently being prepared by myself.167 References to it in the following are to codex Vat. gr. 303.

Metochites understands the subject matter of De anima 3.5 to be the productive intellect, just as that of the preceding chapter was the potential intellect. For his exposition he makes eclectic use of Themistius’ paraphrase and Priscian’s commentary. Like Priscian,168 he understands the relation between the matter invoked in T1169 and the potential intellect to be one of analogy. Despite this, he follows Themistius170 in explaining that nature would be acting in vain if the potential intellect were not brought to perfection; but since nothing is brought to perfection by itself, there must be another intellect which brings the potential intellect to perfection. Since it does so “in virtue of its combination with the other [intellect], [it must] be understood to be a kind of héxis of it.”171

For the last phrase, Metochites has clearly referred back to Aristotle’s text, since it is the productive intellect that he describes as a kind of héxis, whereas the only héxis mentioned by Themistius is the result of the action of the productive intellect on the potential one.172 But he retains an element of the Themistian paraphrase, since he describes it as a kind of héxis of the potential intellect. He must himself have felt that the effect is to divest the productive intellect of whatever priority it needs to fulfil its causal role, since he hastens to introduce the light analogy as a correction of the description of the productive intellect as a kind of héxis, just as Priscian does,173 and with roughly the same justification:

Or rather, he says, [it must be understood] to be such as light is in what are potentially colours. For this is more appropriate for purposes of illustration, lest someone should believe, in accordance with the characteristic property of a state, that it is entirely insubstantial and [only] found to exist in something else. For light, he says, while itself being something, makes the potentially existing colours into colours that are actually present, and is in a sense productive of them.174

Thereupon Metochites reverts to the Themistian interpretation,175 stressing the unity of the productive and the potential intellects and pointing out that comparing the productive intellect to an art would also be misleading, inasmuch as an art is external to its own appropriate matter.176

As is seen, the light analogy on Metochites’ interpretation serves to qualify, first, the comparison of the productive intellect to a héxis, since a héxis is inseparable from its subject, and second, the comparison of the productive intellect to an art (which is not, it should be noted, directly drawn by Aristotle177 ), since an art is external to its matter. On Metochites’ interpretation, then, the productive intellect is a substantial part of the individual human soul, and its substance is identical to its activity: it is always thinking all of its objects at once.178 This interpretation seems to ascribe to Aristotle a view like that of Plotinus, who famously holds that a part of the individual human soul remains undescended.179 When Ps.-Philoponus criticises Plotinus’ opinion about the intellect in actuality,180 this is the view he has in mind, apparently unconcerned that it may not have been intended as an interpretation of Aristotle at all.181 Whether Metochites took notice of the similarity between his own interpretation and that attributed to Plotinus is not clear.

George Scholarios

Metochites’ paraphrase was later epitomised by George Scholarios (date uncertain, but before 1450).182 The relevant passage is found in Adnotationes in Aristotelis opera diversa, 451.20–33.183 Scholarios’ epitome departs from the original in leaving out both those comparisons (art, héxis) that on Metochites’ interpretation the light analogy is there to qualify. Instead it says that while the potential intellect is analogous to the matter of natural and artificial things, the productive intellect is analogous to their form.184 It is likely that the substitution of “form” for héxis is influenced by Thomas Aquinas, whose commentary Scholarios translated into Greek (c.1435).185

While the light analogy itself is retained, Scholarios adds his own emphasis here as well: the productive intellect is, he says,

like a light which is connatural but latent, which shoots up and intertwines with the potential intellect and illuminates it, and becomes entirely unified with it.186

So there it is, the final distillate of everything the ancient and medieval Greek-speaking world had to say about Aristotle’s comparison of the productive intellect to light. Whether Scholarios was aware of it or not, the adjective “connatural” (sýmphytos) connects his exposition with the very first contributions to that discussion, namely Theophrastus’ fragments 320AB (if the productive intellect were connatural – sýmphytos – it should have been present at once and always),187 and 307A (how it is possible for the productive intellect, if it is external, to be connatural – symphyḗs – all the same?). Theophrastus is, of course, quoted by Themistius.188

And, basically, it is Themistius’ interpretation that is encapsulated in Scholarios’ exposition, except that its obscurity regarding the prehistory of the encounter of the potential and the productive intellects has been cleared up in a way that suggests the influence of those Neoplatonic authors (especially Ps.-Philoponus) who hold that the two intellects are numerically identical. Thus, according to Scholarios, if I understand him correctly, the productive intellect is part of our souls from birth. It is always active, but its interaction, indeed union, with the potential intellect begins at a determinate point in the latter’s development, and it is only after this that its actions become manifest (i.e., I suppose, conscious).


Preliminary studies for this paper were presented at seminars in Gothenburg and Oslo. I have benefited greatly from the comments of a number of readers, notably Peter Adamson, Frans de Haas, Stephen Menn, Miira Tuominen, and several members of Representation and Reality. My heartfelt thanks are due to them and to the organisers (Thomas Johansen in Oslo) and participants of the seminars.


Shelves of articles and books have been written on the nature and role of the productive intellect. A brief introduction to its history of interpretation will be found in Fred D. Miller Jr., “Aristotle on the Separability of Mind,” in The Oxford Handbook of Aristotle, ed. C. Shields (Oxford: Oxford University Press, 2012), 306–40, esp. 320–40; for a “survey of earlier interpretations” (which omits, e.g., all the Greek Neoplatonists), see Franz Brentano, Die Psychologie des Aristoteles, insbesondere seine Lehre vom νοῦς ποιητικός (Mainz: Franz Kirchheim, 1867), 5–36 (trans. R. George as “Nous Poiētikos: Survey of Earlier Interpretations,” in Essays on Aristotle’s De anima, ed. M. C. Nussbaum and A. Oksenberg Rorty (Oxford: Oxford University Press, 1992), 313–41).


Cf. Alexander, De anima, ed. I. Bruns (Berlin: Georg Reimer, 1887), 2.4–9. According to Paolo Accattino and Pierluigi Donini, trans., Alessandro di Afrodisia, L’anima (Rome: Laterza, 1996), vii–viii, the De anima depends closely on Alexander’s lost commentary on Aristotle’s De anima. The section on intellect (de An., 80.16–92.11) roughly follows the plan of Aristotle’s work (de An. 3.4–5; 3.8); from 88.17 to 91.6 it follows Aristotle’s de An. 3.5 so closely as to be to a large extent, in effect, a paraphrase. For the correspondences between the two works, see M. Bergeron and R. Dufour, eds. and trans., Alexandre d’Aphrodise, De l’âme (Paris: Vrin, 2008), 15–18. The account of intellect in the so-called Mantissa, ed. I. Bruns (Berlin: Georg Reimer, 1887), 106.18–113.24, henceforth Alexander(?), De intellectu (or de Int.), has been held to differ subtly but significantly from that in Alexander’s De anima. This has led some modern scholars to conclude that it is probably spurious (see especially the arguments in Paul Moraux, Alexandre d’Aphrodise: Exégète de la noétique d’Aristote (Liège: Bibliothèque de la Faculté de Philosophie et Lettres de l’Université de Liège, 1942), 135 and 140–41 with notes), whereas others think it may reflect an earlier phase in the development of Alexander’s views (e.g., Paul Moraux, Der Aristotelismus bei den Griechen von Andronikos bis Alexander von Aphrodisias, vol. 3, Alexander von Aphrodisias (Berlin: De Gruyter, 2001), 392–94; Robert W. Sharples, ed., Alexander Aphrodisiensis, De anima libri mantissa (Berlin: De Gruyter, 2008), 2, 148–49, 151, and 154–55). A minority view worthy of consideration is that it is a later attempt by Alexander to fill in the details left unstated in the De anima (Bernardo C. Bazán, “L’authenticité du ‘De intellectu’ attribué à Alexandre d’Aphrodise,” Revue philosophique de Louvain 71 (1973): 468–87). Bruns’ edition of the Mantissa was superseded by Sharples, ed., Alexander Aphrodisiensis, De anima libri mantissa, but the page and line numbering of the former is retained in the latter.


William of Moerbeke’s translation of chapters 4–8 is usually referred to as the De intellectu (John Philoponus, Commentaire sur le De Anima d’Aristote: Traduction de Guillaume de Moerbeke, ed. G. Verbeke (Louvain-la-Neuve: Publications Universitaires de Louvain / Paris: Béatrice-Nauwelaerts, 1966)). Numerous extracts of the Greek original have been preserved in Sophonias’ paraphrase (late thirteenth century), as documented by Simone van Riet, “Fragments de l’original grec du ‘De Intellectu’ de Philopon dans une compilation de Sophonias,” Revue Philosophique de Louvain 63 (1965): 5–40. Sophonias’ paraphrase may derive from an intermediary paraphrase, also lost in the Greek but preserved in an Arabic adaptation (see Rüdiger Arnzen, Aristoteles’ De anima: Eine verlorene spätantike Paraphrase in Arabischer und Persischer Überlieferung (Leiden: Brill, 1998), 104–7). Extracts of the Greek original have also been preserved in the margins of cod. Laur. Plut. 87,20, as announced by Carlos Steel, “Newly Discovered Scholia from Philoponus’ Lost Commentary on De anima III,” Recherches de Théologie et Philosophie médiévales 84 (2017): 223–43. For Philoponus’ authorship, see the introduction to the English translation by William Charlton (John Philoponus, On Aristotle on the Intellect (London: Duckworth, 1991)).


For arguments in favour of attributing this commentary to Stephanus (fl. c.610?), see the introduction to the English translation by William Charlton (“Philoponus,” On Aristotle on the Soul 3.1–8 (London: Duckworth, 2000), 1–12); for a recent attempt to defend Philoponus’ authorship, see Pantelis Golitsis, “John Philoponus’ Commentary on the Third Book of Aristotle’s De anima, Wrongly Attributed to Stephanus,” in Aristotle Re-Interpreted: New Findings on Seven Hundred Years of the Ancient Commentators, ed. R. Sorabji (London: Bloomsbury, 2016), 393–412. For a reassessment of the evidence concerning Stephanus’ identity and date, see Mossman Roueché, “A Philosophical Portrait of Stephanus the Philosopher,” in Aristotle Re-Interpreted: New Findings on Seven Hundred Years of the Ancient Commentators, ed. R. Sorabji (London: Bloomsbury, 2016), 541–63.


For summaries of Aristotle’s discussion of intellect in de An. 3.4–5, see Sten Ebbesen’s and Pavel Gregoric’s Introduction, 4–5, and Ana María Mora-Márquez’ “Abstraction and Intellection of Essences in the Latin Tradition,” 181–82, in this volume.


(a) Ἐπεὶ δ’ [ὥσπερ] ἐν ἁπάσῃ τῇ φύσει ἐστὶ [τι] τὸ μὲν ὕλη ἑκάστῳ γένει ([b] τοῦτο δὲ ὃ πάντα δυνάμει ἐκεῖνα), (c) ἕτερον δὲ τὸ αἴτιον καὶ ποιητικόν, (d) τῷ ποιεῖν πάντα, (e) οἷον ἡ τέχνη πρὸς τὴν ὕλην πέπονθεν, (f) ἀνάγκη καὶ ἐν τῇ ψυχῇ ὑπάρχειν ταύτας τὰς διαφοράς· (de An., ed. W. D. Ross (Oxford: Clarendon Press, 1961), 3.5, 430a10–14).


(a) καὶ ἔστιν ὁ μὲν τοιοῦτος νοῦς τῷ πάντα γίνεσθαι, (b) ὁ δὲ τῷ πάντα ποιεῖν, (c) ὡς ἕξις τις, (d) οἷον τὸ φῶς· (e) τρόπον γάρ τινα καὶ τὸ φῶς ποιεῖ τὰ δυνάμει ὄντα χρώματα ἐνεργείᾳ χρώματα. (De An. 3.5, 430a14–17.)


Alexander, de An., 88.17–24; Themistius, in de An., ed. R. Heinze (Berlin: Georg Reimer, 1899), 98.12–24.


Alexander’s “ὥσπερ” at de An., 88.20 corresponds to “οἷον” in T1e. So also Themistius’ “ὡς” at in de An., 98.24.


See the brief discussion in Sir David Ross, ed., Aristotle, De anima (Oxford: Clarendon Press, 1961), 296.


Philoponus thinks that “ἐν ἁπάσῃ τῇ φύσει” means “in the whole realm of nature,” although he admits that the phrase is used loosely, so as to exclude the heavenly bodies (de Int., 54.86–90). Ps.-Philoponus instead suggests (in de An., ed. M. Hayduck (Berlin: Georg Reimer, 1897), 539.20–24) that it means “in each individual nature.” Alexander’s and Themistius’ paraphrases do not allow any inferences as to whether they took “ἁπάσῃ” “collectively” or “distributively.” The important thing, however, is that all commentators took “τῇ φύσει” literally, as implying the potentiality for change.


Philoponus, de Int., 48.33–35, 54.90–94, 55.00–3, and 55.7–11; Ps.-Philoponus, in de An., 539.15–18. Cf. Priscian, in de An., ed. M. Hayduck (Berlin: Georg Reimer, 1882), 241.35–242.4, 242.8–9.


Philoponus, de Int., 55.7–11; cf. id., in de An., 24.23–27. It is clear that Philoponus is thinking of the passage from second potentiality (héxis) to second actuality (activity): cf. Ps.-Philoponus, in de An., 558.30–31.


Alexander, de An., 88.17–24.


Alexander speaks of the second term as “something productive.” The use of the expression “productive cause” to mean “efficient cause” is standard in Alexander as well as in all the other commentators. Aristotle never seems to use it, but definitely thinks of the “producer” (i.e., “agent”) or “productive (i.e., active) factor” as an efficient cause (cf. Phys., ed. W. D. Ross (Oxford: Clarendon Press, 1950), 2.3, 195a21–23; GC, ed. C. Mugler (Paris: Les Belles Lettres, 1966), 1.7, 324b13–14; Sens., ed. W. D. Ross (Oxford: Clarendon Press, 1955), 4, 441a8–9; GA, ed. H. J. Drossaart Lulofs (Oxford: Clarendon Press, 1965), 1.21, 729b13–14).


Alexander, de An., 88.20–22.


De An. 2.1, 412a19–21.


Alexander, de An., 17.9–15.


Alexander, de An., 29.23–30.6, 73.14–16, and 80.20–24.


Alexander, de An., 81.5–12.


Alexander, de An., 81.26–82.3.


De An. 3.4, 429a22–24. Cf. 429b30–430a2; contrast 429b5–9; cf. Alexander, de An., 84.21–85.1.


Alexander, de An., 88.24–89.21.


Although, as has often been argued (e.g., by Victor Caston, “Aristotle’s Two Intellects: A Modest Proposal,” Phronesis 44 (1999): 205–7), Aristotle may well use the preposition “ἐν” with the dative in the more abstract sense of “in the case of” (LSJ s.v. ἐν A.I.7).


One might have supposed that “intellect in actuality” in this usage is shorthand for “intellect only in actuality” (cf. Themistius, in de An., 98.20–22). But this is not the case: neither Philoponus nor Ps.-Philoponus accepts that the intellect mentioned in T2b is only in actuality. Note that Philoponus also speaks of the intellect “on the level of héxis” as “in actuality” (i.e., first actuality) in his comments on de An. 3.4 (de Int., 18.43–44).


Ps.-Philoponus, in de An., 539.15–19. Cf. Priscian, in de An., 223.4–11, 240.2–5. The reason why I have resorted to the seemingly pleonastic expression “human rational soul” is that, according to Philoponus, it is not strictly true that the intellect is (as Aristotle says at 429a10) a part of the soul, since, if it were, “either the whole soul would be immortal or the whole would be mortal, since a part is of one substance with the whole” (Philoponus, de Int., 2.33–37, trans. Charlton).


Themistius, in de An., 102.36–103.6; Ps.-Philoponus, in de An., 537.17–21.


Alexander, de An., 88.22–23.


Philoponus, de Int., 45.53–59; Ps.-Philoponus, in de An., 535.13–16.


Ps.-Philoponus, in de An., 534.31–32, 536.7–10, 537.35–37, 540.19–20, and 554.21–23.


Philoponus, de Int., 50.79–81, 55.4–7; cf. ibid., 10.34–37, 45.53–59, 48.28–32, 56.31–37, 58.99–3, and 91.42–49. For the teacher having (not being) a héxis and leading the student to it, see Philoponus, in de An., 94.20–27.


On Philoponus’ editions of Ammonius’ lectures and the significance of epistáseis within them, see now Pantelis Golitsis, “Μετά τινων ἰδίων ἐπιστάσεων: John Philoponus as an Editor of Ammonius’ Lectures,” in Aristotle and His Commentators: Studies in Memory of Paraskevi Kotzia, ed. P. Golitsis and K. Ierodiakonou (Berlin: De Gruyter, 2019), 167–93.


Philoponus, de Int., 43.18–48.32; Ps.-Philoponus, in de An., 535.1–539.10.


As far as I can see, Ps.-Philoponus never says that the human intellect is actualised by another human intellect, as suggested by Henry J. Blumenthal, “Neoplatonic Elements in the de Anima Commentaries,” in Aristotle Transformed: The Ancient Commentators and Their Influence, ed. R. Sorabji (London: Duckworth, 1990), 315.


For a recent discussion of Philoponus’ arguments against the views of his opponents (especially Alexander’s) and in favour of his own, see Frans A. J. de Haas, “Intellect in Alexander of Aphrodisias and John Philoponus: Divine, Human or Both?” in The History of the Philosophy of Mind, vol. 1, Philosophy of Mind in Antiquity, ed. J. E. Sisko (New York: Routledge, 2019), 306–11.


Ps.-Philoponus, in de An., 536.2–5.


Philoponus, in de An., 20.30–31, 55.7–13; de Int., 46.80–85; Ps-Philoponus, in de An., 446.5–18. Cf. Priscian, in de An., 172.4–11, 187.16–17, and 191.5–8. On this application of the methodological principle, see Carlos Steel, trans., “Simplicius,” On Aristotle on the Soul 3.6–13 (London: Bristol Classical Press, 2013), 7–9. It should be noted, however, that Philoponus accepts that, towards the end of his treatises on natural philosophy, Aristotle usually “elevates himself also to the transcendent causes of natural things,” and that he does so also in the De anima (Philoponus, in de An., 20.31–34, 55.13–19; cf. ibid., 261.32–35).


καὶ οὗτος ὁ νοῦς χωριστὸς καὶ ἀπαθὴς καὶ ἀμιγής, τῇ οὐσίᾳ ὢν ἐνέργεια. (De An. 3.5, 430a17–18.)


χωρισθεὶς δἐστὶ μόνον τοῦθὅπερ ἐστί, καὶ τοῦτο μόνον ἀθάνατον καὶ ἀΐδιον […]. (De An. 3.5, 430a22–23.)


Cf. Philoponus’ criticism at de Int., 44.20–23 and 58.82–84. Ironically, Themistius (in de An., 103.9–15) sees Alexander’s position as incompatible with de An. 3.5, 430a22–23, since Aristotle could not possibly have meant that the productive intellect is without qualification the only thing that is immortal and eternal, although it is the only part of the soul that is so.


Philoponus, de Int., 57.71–58.96; cf. ibid., 53.50–62.


Ps.-Philoponus, in de An., 538.10–32, 540.6–13.


Ps.-Philoponus, in de An., 541.9–10.


(1) ὁ δὲ [ἔστιν] τῷ πάντα ποιεῖν; (2) ὁ δέ [ἐστιν τοιοῦτος νοῦς] τῷ πάντα ποιεῖν; (3) ὁ δέ [ἐστιν νοῦς] τῷ πάντα ποιεῖν. Each of the three construals presupposes a somewhat different interpretation of T2a.


Thomas Aquinas, Sent. de An., ed. Fratres Praedicatores, 3.4, 218b20–23; Brentano, Die Psychologie des Aristoteles, 169; Georges Rodier, ed. and trans., Aristote, Traité de l’âme (Paris: Ernest Leroux, 1900), 1:181; cf. 2:459.


See Philoponus, de Int., 43.8–9, 56.43, 56.47–51, and 57.65–66; Ps.-Philoponus, in de An., 539.26–27; cf. ibid., 534.28–30.


R. D. Hicks, ed. and trans. Aristotle, De anima (Cambridge: Cambridge University Press, 1907), 500–501. Of recent translators into English, D. W. Hamlyn (Aristotle, De anima, Books II and III (with Passages from Book I), 2nd ed. (Oxford: Clarendon Press, 1993), 60) and C. D. C. Reeve (Aristotle, De Anima (Indianapolis: Hackett, 2017), 55) definitely do so; Christopher Shields’ (Aristotle, De anima (Oxford: Clarendon Press, 2016), 61) and Fred D. Miller’s (Aristotle, On the Soul and Other Psychological Works (Oxford: Oxford University Press, 2018), 57) renderings are more ambiguous.


See especially Ps.-Philoponus, in de An., 539.31. Cf. Priscian, in de An., 242.39–243.6.


Philoponus, de Int., 40.31–34, 43.8–10, 56.47–50; Ps.-Philoponus, in de An., 539.28–29; cf. Alexander, de An., 89.1–2; Alexander(?), de Int., 107.31–32; Priscian, in de An., 242.36–243.3. In this they seem to follow a standard interpretation of Aristotle’s theory of colours, necessary in order to avoid circularity in the definition of vision, which dates back at least to Alexander’s commentary on Sens. 3, 439a13–16. See Todd Stuart Ganson, “What’s Wrong with the Aristotelian Theory of Sensible Qualities?” Phronesis 42 (1997): 263–82, and Alexander, in Sens., 1.14–18, 41.15–18. Themistius, however, adheres to Aristotle’s wording (in de An., 98.35–99.1), although he follows Alexander’s interpretation in his paraphrase of de An. 3.2 (in de An., 83.35–84.2): cf. Ganson, “What’s Wrong…?” 269–70. Aristotle’s wording might still be defended to the extent that he means that light makes what are colours in second potentiality (which are visible in first potentiality) into colours in second actuality (which are visible in first actuality; when they are visible in second actuality they are, of course, actually seen: de An. 3.2, 426a15–26).


This ambiguity is reflected in the main division of the entry “ἕξις” in LSJ and other Greek lexica, but also in Hermann Bonitz, ed., Index Aristotelicus (Berlin: Georg Reimer, 1870), 260b31–261b4. See also Pierre Chantraine, La formation des noms en grec ancien (Paris: Edouard Champion, 1933), 283–89.


It has been fashionable recently to render “ἕξις” as “a state of possession/having.” But this seems to me a merely apparent solution, based on a misunderstanding of the problem. For what is the difference between a having and a state of having? Of course a héxis in the sense of “having” is a state of having: a having is after all a state. But so is a being-(somehow)-disposed. By the same principle, then, I suppose, “ἕξις” in the sense of “state” should be rendered as “a state of being (somehow) disposed” – or why not “a state of being in a state”? But there is no reason to specify the Aktionsart of the verbal action every time a verb or a deverbal noun is translated: it is simply not informative enough to be worthwhile.


Stephen Menn, “The Origins of Aristotle’s Concept of Ἐνέργεια: Ἐνέργεια and Δύναμις,” Ancient Philosophy 14 (1994): 85.


Menn may well be right to say (“The Origins of Aristotle’s Concept,” 85) that Aristotle’s example of the first type of héxis is of clothes being actually worn rather than tucked away in a closet. Whether Aristotle’s description of the first type of héxis as “a kind of activity of the possessor and the thing possessed” (1022b4–5) fits only clothes that are actually being worn is perhaps more questionable. At least there seems to be no more reason for assuming this to be the case than for thinking that first actualities or states cannot be described as activities. In fact, supposing that Aristotle did want to say that a héxis in one sense is an action involving both the referent of a subject and that of a direct object (i.e., a transitive action), it is hard to envisage how he could have expressed that, with the conceptual apparatus at his disposal, in a better way than this.


Metaph. 5.20, 1022b4–5.


Metaph. 5.20, 1022b10–12.


Metaph. 5.20, 1022b8–10.


EN, ed. I. Bywater (Oxford: Clarendon Press, 1894), 2.5, 1105b25–28; 2.5, 1106a10–12; 2.6, 1106b36–1107a3.


Cf. also EN 6.3–6; de An. 3.3, 428a1–5.


Cf. Cat., ed. L. Minio-Paluello (Oxford: Clarendon Press, 1949), 8, 9a10–13 et passim. See also Alexander, in Metaph., ed. M. Hayduck (Berlin: Georg Reimer, 1891), 417.36–37 (commenting on Metaph. 5.20, 1022b8–10): τῆς γὰρ κατἄλλο σημαινόμενον ἕξεως ἔστιν ἕξις, ὡς τῆς ἀρετῆς καὶ τῆς ἐπιστήμης τε καὶ τέχνης. Cf. ibid. 418.9–12.


E.g., Christopher Kirwan, trans., Aristotle, Metaphysics Γ, Δ, Ε (Oxford: Oxford University Press, 1971), 170.


Cf. Metaph. 10.4, 1055b11–16.


The discussion by Iamblichus apud Simpl., in Cat., ed. K. Kalbfleisch (Berlin: Georg Reimer, 1907), 394.12–395.31 is enlightening on this subject.


Mutatis mutandis, the same may apply to the opposition state – activity, expressed by the term pair “ἕξις”–“ἐνέργεια,” where there is an extra complication in that there is also the term pair “ἕξις”–“χρῆσις,” where we may expect “ἕξις” to mean “having (of a state or other form),” since “χρῆσις” means “use (of a state or other form).”


First actualities are states but not havings. However tempting it may be, for instance, to take the genitive limiting the meaning of “ἕξιν” in Themistius, in de An., 95.30–31 ([ὁ νοῦς] τὴν ἕξιν λέγεται ἔχειν τῶν νοημάτων) as an objective genitive, it seems clear that “[the intellect] is said to have the state of possessing thoughts” (Frederic M. Schroeder and Robert B. Todd, Two Greek Aristotelian Commentators on the Intellect (Toronto: Pontifical Institute of Mediaeval Studies, 1990), 82) is a less satisfying interpretation than “[the intellect] is said to have the first actuality-state of the concepts” (i.e., the concepts in a state of first actuality: cf. id., in de An., 95.21, 115.16).


Examples of the oppositions in definitional contexts are Ammonius, in Cat., ed. A. Busse (Berlin: Georg Reimer, 1895), 84.21–28; Philoponus, in de An., 296.25–26, 296.33–297.4.


Alexander, de An., 85.25–86.6; Alexander(?), de Int., 107.21–28; Themistius, in de An., 95.21–32; Philoponus, de Int., 18.43–44, 19.59–62; Ps.-Philoponus, in de An., 524.28–31. Cf. Priscian, in de An., 219.12–17, 228.28–34, and 229.19–32.


τί μὲν οὖν τὸ διαφανὲς καὶ τί τὸ φῶς, εἴρηται, ὅτι οὔτε πῦρ οὔθὅλως σῶμα οὐδἀπορροὴ σώματος οὐδενός (εἴη γὰρ ἂν σῶμά τι καὶ οὕτως), ἀλλὰ πυρὸς ἢ τοιούτου τινὸς παρουσία ἐν τῷ διαφανεῖ· οὔτε γὰρ δύο σώματα ἅμα δυνατὸν ἐν τῷ αὐτῷ εἶναι, δοκεῖ τε τὸ φῶς ἐναντίον εἶναι τῷ σκότει· ἔστι δὲ τὸ σκότος στέρησις τῆς τοιαύτης ἕξεως ἐκ διαφανοῦς, ὥστε δῆλον ὅτι καὶ ἡ τούτου παρουσία τὸ φῶς ἐστιν. (De An. 2.7, 418b13–20.)


Mark Eli Kalderon, Form without Matter: Empedocles and Aristotle on Color Perception (Oxford: Oxford University Press, 2015), 43–46, instead concludes that the fire referred to here is incorporeal, but I do not think this is conceivable within an Aristotelian theoretical framework.


Cf. Philoponus, in de An., 341.27–32.


Sens. 3, 439a19–21: ὅταν γὰρ ἐνῇ τι πυρῶδες ἐν διαφανεῖ, ἡ μὲν παρουσία φῶς, ἡ δὲ στέρησίς ἐστι σκότος […].


For transitive “ἔχειν” meaning “to contain,” cf. Metaph. 5.23, 1023a13–17, 1023a23–25.


Philoponus uses “ἕξις” (“habitus” in Moerbeke’s translation) frequently to distinguish the intellect in first actuality from the other states of intellect in his comments on de An. 3.4–8, including once in his discussion of 3.5 (de Int., 56.34).


By and large, Philoponus accepts Aristotle’s definition of light in de An. 2.7, 418b9–10, although he tries to amend it by suggesting that “actuality” in the definiens is used in lieu of “form” and “perfection” (in de An., 324.27–30). Sometimes he seems to make a distinction between light and its actuality (most explicitly at in de An., 153.20–21).


τὸ μὲν γὰρ φῶς ἕξις ἐστὶ τοῦ διαφανοῦς, τὸ δὲ χρῶμα τῆς ἐνεργείας ἐστὶ τῆς κατὰ τὴν ἕξιν τελειωτικόν. παρόντος γὰρ τοῦ χρώματος γίνεται καὶ τὸ διαφανὲς ἐνεργείᾳ διαπορθμευτικὸν τῶν χρωμάτων. (Philoponus, in de An. 349.25–28.) See also id., in de An. 322.2–11.


καὶ ἔστι τῷ ὄντι τελεία ἐνέργεια ἡ ἀθρόα προβολὴ τῆς ἕξεως […] οἵα ἐστὶν ἡ τοῦ φωτὸς προβολή· ἅμα γὰρ τῷ φανῆναι τὸ φωτιστικὸν ἀθρόον πᾶν τὸ ἐπιτήδειον καταλάμπεται. (Philoponus, in de An. 297.2–7.) Cf. id., de Aeternitate mundi contra Proclum, ed. H. Rabe (Leipzig: Teubner, 1899), 65.11–18.


[…] καὶ ὥσπερ ἡ ἕξις ἐνέργειαν προβάλλεται καὶ οὐκ οὐσίαν, οὕτω καὶ ὁ ἐν ἡμῖν νοῦς ἐνέργειαν προβάλλεται καὶ οὐκ οὐσίαν· διὸ τούτοις ἀναλογεῖ. εἰ δὲ ταῦτα οὕτως, οὐ περὶ τοῦ θείου νοῦ ἐστιν ὁ λόγος· ἐκεῖνος γὰρ καὶ οὐσίας προβάλλεται. (Ps.-Philoponus, in de An., 539.29–32.)


ὥσπερ γὰρ τὸ φῶς τὰ χρώματα ποιεῖ ὁρατά (οὐ γὰρ χρώματα αὐτὰ ποιεῖ, ἀλλὰ τὴν ἐνέργειαν αὐτῶν προβάλλεται) […] (Ps.-Philoponus, in de An., 539.27–29).


Philoponus’ phrasing at de Int., 56.43–47 (see n79) suggests that he considers héxis rather than light as the real secundum comparatum.


“Habitui proportionari ait actu intellectum, et hinc autem palam quia non de divino dicit, sed de nostro. Non enim dixit habitui proportionari divinum intellectum neque habitum esse, sed autothen, ut in libro Peri Hermeneias dictum est, sine potentia actus est. Deinde habitus proponit exemplum lumen […].” (Philoponus, de Int., 56.43–47, quoted words in italics.)


Ps.-Philoponus, in de An., 537.33–35.


καὶ κατἄλλο δὲ ἀναλογεῖ ἕξει ὁ ἡμέτερος νοῦς, οὐχ ὅτι ποιότης τίς ἐστιν, ἀλλὥσπερ ἡ ἕξις ἔχει πρὸς τὸ πρῶτον δυνάμει (τελειοῖ γὰρ αὐτό) οὕτω καὶ ὁ κατἐνέργειαν νοῦς τελειοῖ τὸ δυνάμει τοῦ νοῦ. (Ps.-Philoponus, in de An., 539.32–35.)


Cf. Alexander, de An., 15.12–13.


Rodier, Aristote, Traité de l’âme, 460; Hicks, Aristotle, De anima, 501.


[…] ἀναγκαῖον δοκεῖ καὶ ἐπὶ τοῦ νοῦ ταύτας εἶναι τὰς διαφοράς. καὶ ἐπεί ἐστιν ὑλικός τις νοῦς, εἶναί τινα δεῖ καὶ ποιητικὸν νοῦν, ὃς αἴτιος τῆς ἕξεως τῆς τοῦ ὑλικοῦ νοῦ γίνεται. (Alexander, de An., 88.22–24.)


[…] ἀνάγκη ἄρα καὶ ἐν τῇ ψυχῇ ὑπάρχειν ταύτας τὰς διαφοράς, καὶ εἶναι τὸν μέν τινα δυνάμει νοῦν, τὸν δέ τινα ἐνεργείᾳ νοῦν […] ὃς ἐκείνῳ συμπλακεὶς τῷ δυνάμει καὶ προαγαγὼν αὐτὸν εἰς ἐνέργειαν τὸν καθἕξιν νοῦν ἀπεργάζεται. (Themistius, in de An., 98.21–23.)


[…] ὡς γὰρ τὸ φῶς αἴτιον γίνεται τοῖς χρώμασιν τοῦ δυνάμει οὖσιν ὁρατοῖς ἐνεργείᾳ γίνεσθαι τοιούτοις, οὕτως καὶ οὗτος ὁ τρίτος νοῦς τὸν δυνάμει καὶ ὑλικὸν νοῦν ἐνεργείᾳ νοῦν ποιεῖ ἕξιν ἐμποιῶν αὐτῷ τὴν νοητικήν. (Alexander(?), de Int., 107.29–34.) For the testimony of the “C” section, see below.


Pierluigi Donini, “Alessandro di Afrodisia e i metodi dell’esegesi filosofica,” in id., Commentary and Tradition: Aristotelianism, Platonism and Post-Hellenistic Philosophy, ed. M. Bonazzi (Berlin: De Gruyter, 2011), 103–6.


καὶ ἔστιν ὁ μὲν τοιοῦτος νοῦς τῷ πάντα γίνεσθαι ⟨, ὡς ἕξις τις⟩, ὁ δὲ τῷ πάντα ποιεῖν [, ὡς ἕξις τις], οἷον τὸ φῶς. Since οἷον τὸ φῶς in T2’d is not likely to be an adverbial modifier of ποιεῖν in T2’b, it will be preferable to construe ἔστιν in T2’a (and mentally supplied in T2’b) as the copula.


For the claim, see Donini, “Alessandro di Afrodisia,” 104; for the reference, see ibid., 95–96.


Themistius, in de An., 102.24–27. Cf. also Theophrastus fr. 316, briefly discussed in n98 below.


Dimitri Gutas, “Appendix: Themistius on Theophrastus in Arabic (or, What Averroes Read),” in Theophrastus of Eresus: Sources for His Life, Writings, Thought and Influence, Commentary Volume 4: Psychology (Texts 265–327), ed. P. M. Huby (Leiden: Brill, 1999), 213 and n16.


Ps.-Philoponus, in de An., 535.34–38.


[…] καὶ τὴν λέξιν δὲ τὴν ἐν τῷ τρίτῳ Περὶ ψυχῆς τούτοις προσοικειοῦν ἔλεγεν δεῖν καὶ τὴν ἕξιν καὶ τὸ φῶς ἐπὶ τοῦτον φέρειν τὸν πανταχοῦ ὄντα. (Alexander(?), de Int., 113.4–6.)


Alexander(?), de Int., 113.12–24.


Alexander(?), de Int., 112.5–113.2.


Cf. Alexander(?), de Int., 113.2–4.


Alexander(?), de Int., 113.12–18, esp. 16–18.


Cf. Theophrastus, fr. 316, in Priscian, Metaphrasis in Theophrastum, ed. I. Bywater (Berlin: Georg Reimer, 1886), 31.8–13, where a paraphrase of de An. 3.4, 429b5–9 is immediately followed by the questions: (1) what is the efficient cause of the intellect’s becoming each thing? And (2) what is the result of the becoming, a héxis or a substance? Theophrastus’ answer to the second question is: “rather a héxis, and this is like a thing that perfects its nature.” (For this interpretation – which differs from those of Pamela Huby, trans., Priscian, On Theophrastus on Sense-Perception, in Priscian, On Theophrastus on Sense-Perception with “Simplicius,” On Aristotle On the Soul 2.5–12, trans. P. Huby and C. Steel (London: Duckworth, 1997), 40–41, and William W. Fortenbaugh et al., eds. and trans., Theophrastus of Eresus, Sources for His Life, Writings, Thought and Influence, Part Two: Psychology, Human Physiology, Living Creatures, Botany, Ethics, Religion, Politics, Rhetoric and Poetics, Music, Miscellanea (Leiden: Brill, 1992), 87 – cf. Priscian, Metaphrasis, 31.24–32, as well as Pamela M. Huby, Theophrastus of Eresus: Sources for His Life, Writings, Thought and Influence, Commentary Volume 4: Psychology (Texts 265–327) (Leiden: Brill, 1999), 171.)


“[…] hoc autem solummodo lumen facit ipsum esse actu colorem in quantum facit dyaphanum esse in actu ut moueri possit a colore et sic color uideatur.” (Thomas Aquinas, Sent. de An. 3.4, 219b47–50.) See also Brentano, Die Psychologie des Aristoteles, 172–73.


Similarly, when Aristotle says in Sens. 6, 447a11, that light “produces” (ποιεῖ) vision, this can only mean that it enables vision, which is the “action” of colour qua visible object, relayed by the illuminated body and the sense organ, upon the visual sense.


It is hardly by coincidence that Donini’s paraphrase of T2’a–c reads “[‘siffatto intelletto’ …] si realizza infine come abito” (“Alessandro di Afrodisia,” 104: my italics).


Themistius, in de An., 103.32–36. For discussion of this passage, see Frans A. J. de Haas, “Themistius,” in A History of Mind and Body in Late Antiquity, ed. A. Marmodoro and S. Cartwright (Cambridge: Cambridge University Press, 2018), 125–26.


“Et utique si intellectum conditorem in his diceret, rationabilius utique ipsum magis soli assimilaret, non lumini.”


Philoponus, de Int., 5.92–93.


Alexander, de An., 88.24–89.8.


Alexander, de An., 89.9–11.


Moraux, Alexandre d’Aphrodise, 89, 92–93; cf. id., Der Aristotelismus bei den Griechen, 3:389.


εἰ γὰρ μὴ ἦν τι νοητὸν φύσει, οὐδἂν τῶν ἄλλων τι νοητὸν ἐγίνετο […].


Alexander, de An., 87.28–29: τὰ δὲ τῇ αὑτῶν φύσει νοητὰ κατἐνέργειαν νοητά […].


Cf. Alexander, de An., 88.26–89.1: ἐν πᾶσιν γὰρ τὸ μάλιστα καὶ κυρίως τι ὂν καὶ τοῖς ἄλλοις αἴτιον τοῦ εἶναι τοιούτοις.


Alexander, de An., 89.4–5.


Alexander, de An., 87.25–29.


Moraux, Alexandre d’Aphrodise, 90–92.


A. C. Lloyd, “The Principle That the Cause Is Greater than Its Effect,” Phronesis 21 (1976): 150.


Alexander, de An., 87.28–88.2, 88.10–15, and 90.2–11; cf. ibid. 86.23–28; Alexander(?), Quaestio 3.3, 85.7–14; de Int., 108.3–15, 110.16–17, 110.28–30, 111.22–27, and 111.36–112.4.


Alexander, de An., 84.6–9, 84.19–21, 86.29–87.1, 87.24–25, and 88.10–14; Alexander(?), Quaestio 1.1, 4.15–16; 1.25, 39.15–17; de Int., 108.3–7, 108.14–15, and 110.17–20.


τὰ γὰρ καθόλου καὶ κοινὰ τὴν μὲν ὕπαρξιν ἐν τοῖς καθέκαστά τε καὶ ἐνύλοις ἔχει. νοούμενα δὲ χωρὶς ὕλης κοινά τε καὶ καθόλου γίνεται […]. ὥστε χωρισθέντα τοῦ νοοῦντος αὐτὰ νοῦ φθείρεται, εἴ γε ἐν τῷ νοεῖσθαι τὸ εἶναι αὐτοῖς. (Alexander, de An., 90.5–8.)


ὁ γὰρ κατὰ ἕξιν νοῦς ἀποκείμενά πώς ἐστιν ἀθρόα καὶ ἠρεμοῦντα τὰ νοήματα.


τό τε γὰρ μάλιστα ὁρατόν, τοιοῦτον δὲ τὸ φῶς, καὶ τοῖς ἄλλοις τοῖς ὁρατοῖς αἴτιον τοῦ εἶναι ὁρατοῖς, ἀλλὰ καὶ τὸ μάλιστα καὶ πρώτως ἀγαθὸν καὶ τοῖς ἄλλοις ἀγαθοῖς αἴτιον τοῦ εἶναι τοιούτοις· τὰ γὰρ ἄλλα ἀγαθὰ τῇ πρὸς τοῦτο συντελείᾳ κρίνεται. (Alexander, de An., 89.1–4.)


Cf. Lloyd, “The Principle,” 151.


De An. 2.7, 418b4–6. On this count, too, Alexander was criticised by Moraux, Alexandre d’Aphrodise, 89–90. He has been defended by Accattino and Donini, Alessandro di Afrodisia, 185–86, and, more recently, by Victor Caston, trans., Alexander of Aphrodisias, On the Soul, Part 1: Soul as Form of the Body, Parts of the Soul, Nourishment, and Perception (London: Bristol Classical Press, 2012), 163–64n395.


Alexander, de An., 44.13–15; id., in Sens., 43.13–14, 46.21–47.1, and 47.13. Although, as Accattino and Donini point out (Alessandro di Afrodisia, 186), he sometimes reserves this honorific for the source of light (de An., 46.2–3; in Sens., 45.26–46.3).


Moraux, Alexandre d’Aphrodise, 90; Accattino and Donini, Alessandro di Afrodisia, 288–92.


See R. W. Sharples, “Schriften und Problemkomplexe zur Ethik,” in Moraux, Der Aristotelismus bei den Griechen, 3:593–97; Gweltaz Guyomarc’h, “Racine et rejetons: Le pros hen selon Alexandre d’Aphrodise,” Quaestio 13 (2013): 42n14.


De An. 3.4, 429a22–24, 429b30–430a2; contrast 429b5–9.


Ps.-Philoponus, in de An., 533.25–35.


Ps.-Philoponus, in de An., 518.19–32. On Ps.-Philoponus’ reports of Plutarch’s commentary, see p. 40 above.


See T14 below. For Ammonius’ acceptance of extensive harmony between Plato and Aristotle, see Richard Sorabji, “The Ancient Commentators on Aristotle,” in Aristotle Transformed: The Ancient Commentators and Their Influence, ed. R. Sorabji (London: Duckworth, 1990), 3–4.


Philoponus, de Int., 57.63–69.


On this and the other Neoplatonic attempts to read Aristotle as a champion of innate forms, see Frans A. J. de Haas, “Recollection and Potentiality in Philoponus,” in The Winged Chariot: Collected Essays on Plato and Platonism in Honour of L. M. de Rijk, ed. M. Kardaun and J. Spruyt (Leiden: Brill, 2000), 165–84. One reason for thinking that the attempted solution is Philoponus’ idea, even though his commentary is based on Ammonius’ teaching, is that it is introduced by the phrase “Attendere autem oportet in his […]” (de Int., 37.81), corresponding to Ἐπιστῆσαι δἐν τούτοις χρή […] in Sophonias, in de An., 134.38. It is thus the subject of an epístasis, a personal observation (cf. above n32).


Philoponus, de Int., 16.82–96, 37.81–38.98.


Philoponus, de Int., 39.21–27.


De An. 3.4, 429a22–24, 429b29–430a2.


Contrast de An. 3.4, 429b5–9.


“Assimilatur intellectus in generatione proveniens dormienti aut alienato” (Philoponus, de Int., 38.99–39.20, 39.27–40.43; the quoted passage at 40.42–43). It should be noted that, disconcertingly, the theory that forms are in the (newborn) rational soul “sicut sunt in dormiente geometra theoremata, et indigere ad promptum usum theorematum auferente hoc impedimentum” is credited, at de Int., 14.38–45, to Plato and contrasted with Aristotle’s theory that forms are in the soul in first potentiality.


De An. 3.4, 429b5–9.


“[…] sic videlicet et qui actu intellectus perficit eum qui potentia et ducit in actum, non imponens in ipso non entes species, sed immanifestas entes et occultas propter id quod a nativitate nubilum, elucidans.” (Philoponus, de Int., 40.34–37, trans. Charlton, slightly modified.) Cf. ibid., 56.47–57.69.


See Philoponus, in de An., 5.4–5 with context; ibid., 110.31–34; id., de Int., 33.82–91.


Cf. Philoponus, de Int., 52.17–29, 59.14–24. Arguably, even if every individual human intellect that ever existed was actualised by a previously actualised individual human intellect, the principle of prior actuality demands that there be a cause that explains why any individual human intellect has been actualised in the first place. Perhaps this is the reason why Philoponus mentions the divine intellect as well as the teacher’s intellect at de Int., 40.29 and 91.49, although he elsewhere (esp. de Int., 56.43–47) criticises the idea that the productive intellect is divine.


Cf. Philoponus, in de An., 110.29–36, 306.31–33.


Similarly, Philoponus’ account of the fall and subsequent ascent of the rational soul in in de An., 306.24–307.1 is in agreement with that in de Int., 38.99–40.43, except for specifying (306.29–31) that the state of the soul at birth is simply first potentiality or suitability: one might be inclined to suspect, then, again, that this reflects Ammonius’ teaching. Still, the anomaly is glaring, since it is hard to see, if this is the state of the soul at birth, (1) to what purpose a pre-natal state of first actuality is assumed (306.27–28) and (2) how the soul could be brought back to a state of first actuality by perceptible objects (306.31–33). Probably, then, one has to give some “latitude” (cf. de Int., 39.6, 39.12) to the meaning of “first potentiality” and “suitability” in this passage, too.


See Ps.-Philoponus, in de An., 469.17–19, 516.24–25, 516.28–29, 516.30–31, 524.12–16, 533.24–25, 552.30–553.1, 564.38–565.6 (cf. also 558.16–17).


ὥσπερ γὰρ τὸ φῶς οὐκ αὐτὸ ποιεῖ τὰ χρώματα, ἀλλὰ τὰ ἤδη ὄντα φανερὰ ποιεῖ, οὕτω καὶ ὁ ἐνεργείᾳ νοῦς οὐ ποιεῖ τὰ πράγματα, ἀλλὰ τὰ ἤδη ὄντα ἐντυποῖ καὶ ἐγχαράττει τῷ δυνάμει νῷ. (Ps.-Philoponus, in de An., 534.28–31; cf. ibid., 537.29–31.)


ὁ γὰρ ἐν ἡμῖν δυνάμει νοῦς πάντων ἔχει τὴν νόησιν, ὥσπερ καὶ ἡ σκιαγραφία ἐν εἰκόνι πάντων ἔχει τοὺς τύπους, εἰ καὶ μὴ φανερούς, ὧν μέλλει δέξασθαι. (Ps.-Philoponus, in de An., 519.8–12.)


Ps.-Philoponus, in de An., 518.21–26.


Ps.-Philoponus, in de An., 538.4–7.


Ps.-Philoponus, in de An., 539.28–29.


This presupposition is made explicit in the paraphrase of the light analogy in the “B” section of Alexander(?)’s De intellectu, 111.32–36.


Berol. Ham. 512, 122r–29v.


Arnzen, Aristoteles’ De anima, 80–139.


Arnzen, Aristoteles’ De anima, 104–7.


Arnzen, Aristoteles’ De anima, 174.


Van Riet, “Fragments de l’original grec,” 37–40.


According to Hicks (Aristotle, De anima, 496), Sophonias’ paraphrase (in de An., 125.37–39) of de An. 3.4, 429b31–430a2 “betrays the influence of Alexander.” And so it does: of Alexander’s De anima commentary as reported by Philoponus (de Int., 15.65–81).


Sophonias, in de An., 125.15–27.


De An. 3.5, 430a15 “φῶς” and 430a16 “τρόπον”; see above p. 36n7.


ἔοικε γὰρ ὁ μὲν δημιουργὸς νοῦς καὶ οὐσιοποιὸς τῷ ἡλίῳ, ὁ δὲ κατἐνέργειαν ἀνθρώπινος καὶ ἡμέτερος τῷ φωτί. (Sophonias, in de An., 125.20–21.)


P. 59 and nn102–4.


Cf. van Riet, “Fragments de l’original grec,” 37.


See above, p. 40 and n33.


[…] δεύτερον μὲν τοῦ πρώτου καὶ θείου, προσεχῶς δὲ καὶ ὑπερκείμενον ἡμῶν καὶ ἐλλάμποντα τῷ ἡμετέρῳ νῷ. πρὸς τοῦτο καὶ τὸ τοῦ φωτὸς αὐτοῖς τείνει παράδειγμα. (Sophonias, in de An., 135.12–14.) Cf. Philoponus, de Int., 44.25–38 (and Ps.-Philoponus, in de An., 535.5–8, 535.31–536.1, who ascribes the view to Marinus).


E.g., Philoponus, in de An., 5.4–5.


See above, p. 39 and n30.


ἄλλοι τὸν διδασκαλικὸν τοῦτον εἶναι, ὃς τελειοῖ τὸν ἐν ἄλλῳ δυνάμει παρατιθέμενος τὰ τῶν ἐπιστημῶν θεωρήματα καὶ νοήματα, καὶ αὐτὸς ἐκ δυνάμεως εἰς ἐνέργειαν ἀχθείς ποτε. πέμπτη πρὸς τούτοις δόξα, ἣν οἶμαι καὶ μᾶλλον ἀληθῆ, ἣ ἕνα καὶ τὸν αὐτὸν εἶναι λέγει τὸν δυνάμει καὶ ἐνεργείᾳ, καὶ οὐκ ἔξωθεν ἀλλἐν μιᾷ τῇ αὐτῇ ψυχῇ καὶ ἐν ἑνὶ καὶ τῷ αὐτῷ καθἕκαστα ἀνθρώπῳ, διαφέροντα δὲ ἑαυτοῦ τῷ τελείῳ καὶ ἀτελεῖ. (Sophonias, in de An., 136.17–23.)


κατἄλλην δὲ καὶ ἄλλην ἐκδοχὴν καὶ αἱ λοιπαὶ δόξαι προσβιβαζόμεναι ἀληθεύουσι. (Sophonias, in de An., 136.23–24.)


For the date, see Börje Bydén, “The Byzantine Fortuna of Alexander of Aphrodisias’ Commentary on Aristotle’s De sensu et sensibilibus,” Jahrbuch der Österreichischen Byzantinistik 68 (2019), here 101–2n52. For a general discussion of Metochites’ paraphrases, see Martin Borchert, Der paraphrastische Kommentar des Theodoros Metochites zu Aristoteles’ “De generatione et corruptione”: Textkritische Erstedition und deutsche Übersetzung (Berlin: De Gruyter, forthcoming).


For the series Commentaria in Aristotelem Graeca et Byzantina, published by De Gruyter.


Priscian, in de An., 241.35–37, 242.8–9.


See above, p. 36n6.


Themistius, in de An., 98.12–24.


ἀνάγκη […] τὸν δὲ εἶναι τὸν τελειωτικὸν τὸν πάντα ἐν τῷ πρώτῳ ποιοῦντα ἃ δυνάμει ἐστίν, ὃν τῇ πρὸς ἐκεῖνον συμπλοκῇ τελειοποιοῦντα ὡς ἕξιν τινὰ αὐτοῦ καταλαμβάνεσθαι […] (V175r).


See above, p. 53n85.


Priscian, in de An., 242.31–243.6.


[…] μᾶλλον δὲ οἷόν ἐστι, φησί, τὸ φῶς ἐν τοῖς δυνάμει χρώμασι· τοῦτο γὰρ κυριώτερον εἰς τὸ παραδειγματίζειν, ἵνα μή τις αὐτὸν κατὰ τὸ τῆς ἕξεως ἴδιον ἀνούσιόν τε πάμπαν καὶ ἐν ἄλλῳ θεωρούμενον νομίσῃ· τὸ γάρ τοι φῶς αὐτό τι ὄν, φησί, τὰ δυνάμει ὄντα χρώματα ἐπιδημοῦν ἐνεργείᾳ ποιεῖ χρώματα καὶ τρόπον τινὰ ποιητικόν ἐστιν αὐτῶν (V175r).


Themistius, in de An., 99.13–20.


οὕτω δὴ καὶ ὁ νοῦς ὁ ποιητικὸς ἔχει πρὸς τὸν δυνάμει, ὡς φῶς τι συμπλεκόμενον αὐτῷ τελειοποιεῖ καὶ κατασκευάζει αὐτὸν ἃ δυνάμει πρότερον ἦν ἐνεργείᾳ εἶναι, γινόμενος εἷς μετ᾿ ἐκείνου, οὐ κατὰ τὸ τῆς τέχνης ὑπόδειγμα ἔξωθεν ὤν, ὡς ἡ χαλκευτικὴ τῆς οἰκείας ὕλης ἔξω τοῦ χαλκοῦ καὶ ἡ τεκτονικὴ τοῦ ξύλου (V175r). It should be noted that the Greek text of Themistius is corrupt in the passage that Metochites is drawing on here (In De anima, 99.13–14). It was restored by Gerald M. Browne (“Ad Themistium Arabum,” Illinois Classical Studies 11 (1986): 240), by recourse to Isḥāq ibn Ḥunayn’s translation: οὐ γὰρ ἔξωθεντοῦ δυνάμει νοῦ ὁ ποιητικός, ὥσπερ ἔξωθεντῆς ὕλης ἡ τέχνη, ὥσπερ χαλκευτικὴ τοῦ χαλκοῦ καὶ τεκτονικὴ τοῦ ξύλου (the emendation was accepted by Robert B. Todd, trans., Themistius, On Aristotle’s On the Soul (Ithaca, NY: Cornell University Press, 1996), 187n7). Whether Metochites had a better text at his disposal than that offered by the extant Greek manuscripts or supplied the missing words by his own ingenuity is not clear.


Pace Michael Frede, “La théorie aristotélicienne de l’intellect agent,” in Corps et âme: Sur le De anima d’Aristote, ed. G. Romeyer Dherbey and C. Viano (Paris: Vrin, 1996), 379–80.


οὐσία γὰρ αὐτοῦ ταυτὸν καὶ ἐνέργεια, ὡς εἴρηται, καὶ οὐ μεταβάλλει ἄλλο τι ὂν τὴν οὐσίαν καὶ ἄλλο τὴν ἐνέργειαν ἐκ τοῦδε εἰς τόδε μεταβατικῶς καὶ διεξοδικῶς χρώμενος πρὸς τὰς νοήσεις, ἀλλὰ ἀθρόον πάντα ἔχων τὰ εἴδη καὶ τὰ ἐπιστητὰ πάντα· οὕτω γὰρ μόνως ἂν εἴη ταυτὸν ἥ τε οὐσία αὐτοῦ καὶ ἡ ἐνέργεια (V175v); cf. Themistius, in de An., 100.5–11.


Plotinus, Enn. 4.7.13; 4.8.8 et alibi, in Plotini opera, ed. P. Henry and H.-R. Schwyzer (Leiden: Brill, 1951–1973). For the reception of the idea in later Platonists, see Richard Sorabji, The Philosophy of the Commentators 200–600 AD, A Sourcebook, vol. 1: Psychology (with Ethics and Religion) (Ithaca, NY: Cornell University Press, 2005), 93–99.


See above, p. 40 and n33.


Ps.-Philoponus, in de An., 535.8–13, 535.29–31, 536.15–17, 536.24–28, 536.34–537.1, and 538.32–539.1. Ps.-Philoponus’ report of Plotinus’ view is discussed by Blumenthal, “Neoplatonic Elements,” 312–15.


For the date, see Franz Tinnefeld, “Georgios Gennadios Scholarios,” in La théologie byzantine et sa tradition, vol. 2 (XIIIe–XIVe s.), ed. C. G. Conticello and V. Conticello (Turnhout: Brepols, 2002), 516.


In Œuvres complètes de Georges (Gennadios) Scholarios, ed. M. Jugie, L. Petit, and X. A. Siderides, vol. 7 (Paris: Maison de la bonne presse, 1936).


[…] οὕτω καὶ ἐπὶ τῆς ψυχῆς ὁ μὲν δυνάμει νοῦς ἀνάλογον ἔχει ὕλῃ καὶ ὥσπερ παθητικός ἐστιν· ὁ δὲ ποιητικὸς νοῦς ἀνάλογον εἴδει, οὐκ ἔξωθεν ἐπιών (George Scholarios, Œuvres complètes, 7:451.21–23).


The dating is suggested by Tinnefeld, “Georgios Gennadios Scholarios,” 517. Cf. Thomas Aquinas, Sent. de An. 3.4, 219a36–39: “Dicendum est ergo quod ‘habitus’ hic accipitur secundum quod Philosophus frequenter consueuit nominare omnem formam et naturam habitum […]”; George Scholarios, Translatio commentarii Thomae Aquinae De anima Aristotelis, 3.10.22–23 (in Œuvres complètes de Georges (Gennadios) Scholarios, ed. M. Jugie, L. Petit, and X. A. Siderides, vol. 6 (Paris: Maison de la bonne presse, 1933)).


[…] ἀλλὥσπερ τι φῶς σύμφυτον μέν, ὑποκεκρυμμένον δέ, ἀναθρῷσκον καὶ συμπλεκόμενον τῷ δυνάμει νῷ καὶ καταλάμπον αὐτόν, καὶ εἷς μεταὐτοῦ (scripsi: αὐτὸν Jugie) τὸ σύμπαν γινόμενος (George Scholarios, Œuvres complètes, 7:451.24–26).


The paraphrase within brackets is of fr. 320A. For the different wording of fr. 320B and the concomitant complications, see Pamela Huby, Theophrastus of Eresus, Sources for His Life, Writings, Thought and Influence, Commentary, vol. 4: Psychology (Texts 265–327) (Leiden: Brill, 1999), 183–90.


Themistius, in de An., 108.25, 102.26–27, and 107.31–32. Metochites does not use “σύμφυτος” in this context. The word is chiefly employed in his De anima paraphrase as a variant of “συμφυής” in speaking of media that are naturally continuous with sense organs.

  • Collapse
  • Expand