3.1 Why Distinctions Are Needed

The first few lines of An. 2.5 outline several key conditions with which any successful account of perception must comply. First, it must account for the fact that a perceiver can be activated only by an external perceptual object, and, indeed, that she needs to be affected by the object for as long as she is perceiving it. Furthermore, it must accommodate the salient differences that distinguish perception from processes such as burning, as the capacity for perceiving is clearly not exhausted in the way that the capacity for burning is and the activity of perceiving is essentially object-directed. These criteria cannot be satisfied, I argued, without new conceptual distinctions pertaining, first of all, to the notions of being affected and likeness. Most of An. 2.5 is, indeed, engaged in drawing such distinctions. In the present chapter, we shall concentrate on how the distinctions Aristotle draws at 417a21–418a3 accumulate to form his first general account of perception as a complete passive activity, which lays the groundwork for the rest of his inquiry into perception.

Let us begin, though, with the preliminary distinctions, drawn immediately after the puzzle at 417a2–9, and with briefly reflecting on how these may already be preparing Aristotle’s account:

Aristotle’s observation about the two senses of ‘perceiving’ in (i) may sound trivial, but the wider context shows it need not be. He has, just previously, claimed that that which can perceive ‘does not perceive (οὐκ αἰσθάνεται) [on its own], just as (καθάπερ) the combustible [material] does not burn on its own’ (417a7–8). Now, he seems to be qualifying that claim by saying that, while combustible material does, indeed, in no true sense ‘burn’ on its own (in the absence of fire), we standardly say that a perceiver does ‘perceive’ on her own (e.g. a sleeping person still ‘sees’ and ‘hears’ in the sense of being neither blind nor deaf).

Similarly, we might say, with (ii), that perceivers are endowed with αἴσθησις even when they are not actively engaged in perceiving, whereas the combustible material is hardly endowed with burning until something sets it aflame.Footnote ² This suggests that perceivers relate to perceiving in a very different way than combustible materials relate to burning: there is a much more stable and advanced kind of capacity involved in the former case, which is significantly closer to the exercising of the activity. Such a difference seems directly relevant for the observed contrast between combustible materials, on the one hand, which are progressively consumed by burning, and perceivers, on the other, which are not exhausted by perceiving over time. Although our passage stops short of spelling out this connection, the conceptual distinctions drawn here help a perceptive reader to see where the preceding analogy with burning falls flat. However, further distinctions will be required to articulate the difference between the two kinds of capacities.

More speculatively, one can read Aristotle’s point about the perceptual object in (iii) as preparing the grounds for understanding how the object-directedness of perception distinguishes it from processes such as burning. Aristotle has compared the need, on the side of perceivers, for the presence of external perceptual objects (417a3–4) with the need, on the side of the combustible material, for the presence of a fire in fulfilment (417a9). Now, once the combustible material is set on fire by the activity of a fire in fulfilment – that is, once it starts burning – it does not depend for its burning on the activity of the original fire anymore; indeed, its burning no longer has any occurrent (non-historical) relation to that fire other than being another instantiation of fire in the universe. This is significantly different from the case of perceiving, which is essentially object-directed. In perception, the perceptual object serves not only as the efficient cause of perceiving in the sense of bringing about the episode of perception, but also as the formal cause of perception in the sense that the perceiver’s activity is essentially of that object. My suggestion about the relevance of (iii) is that distinguishing between the fulfilment (ἐντελεχεία) of a perceptual object (e.g. being red) and the activity (ἐνέργεια) of that object (e.g. making itself seen) can help us understand that the latter is intrinsically intertwined with the activity of the perceiver in a way that is absent in the case of fire and burning. Whereas the activity of the agent there (i.e. setting on fire) is over once the respective activity of the patient (i.e. burning) begins, the relevant activity of perceivers (i.e. perceiving) is not only initiated by, but also coincides throughout with, the respective activity of the perceptual objects (i.e. making themselves perceived). Again, when read in this way, the distinction in (iii) provides resources for seeing the limits of the preceding analogy between perceiving and burning; but further distinctions will be required to understand properly how the two kinds of relations to the agent differ.

I argued (in Sections 1.3 and 2.5) that Aristotle’s identification of being affected with being active in the following passage (417a14–17) is intended to prepare us for the idea that, in perceiving, the perceiver is affected by an object to which she has already been assimilated (the element of truth within LAL). However, Aristotle also acknowledges that this provides us with only a provisional understanding of how the assimilation model (spelled out at 417a17–21) can apply to perception. This raises a concern about whether the alleged truth of LAL can be coherently captured at all. What makes it the case that, for perception, being affected passes the tense test together with being active? And how exactly can the two notions be combined in a general account of perception? Only if we can answer these questions will we be in a position to understand properly the disanalogies between perception and processes such as burning that have been sketched out above. Furthermore, only if we can answer these questions will we (as argued in Sections 1.4 and 2.7) be in a position to see how Aristotle can resolve the conflict between LKL and Anaxagoras.

3.2 The Perceptive Capacity as a Fulfilment

In the first set of his distinctions at 417a21–b2, Aristotle lays down what has come to be known as his ‘triple scheme’, or Dreistufenlehre. Here, I limit myself to outlining how these distinctions help to ground the key disanalogies between perception and processes such as burning. We should not be misled by the fact that, at 417a21, all of a sudden, Aristotle turns from considering ‘that which can perceive’ to considering ‘that which knows’ (ἐπιστῆμον). This does not mark a change of topic or a digression; rather, Aristotle seems to think that the case of a knower provides a useful model for understanding other phenomena – in this case, perception. In contrast to approaches that analyse the Dreistufenlehre as a general ontological scheme, I shall emphasize how much the passage is targeted specifically on capturing the essential features of perception.Footnote ³

My first observation is that, when Aristotle announces (right after introducing the assimilation model at 417a17–21), that ‘distinctions need to be drawn about capacity and fulfilment, for we have been speaking simply (ἁπλῶς) about them’ (417a21–2), his primary focus seems not to be on the plurality of ways in which each of the two terms can be used. After all, Aristotle has already distinguished between two kinds of fulfilment at An. 2.1, 412a22–8, and so has not been talking ‘simply’ about fulfilment in the sense of having only a single meaning of the term available. Rather, what simplicity seems to mean is a way of speaking about capacity and fulfilment as mutually exclusive notions.Footnote ⁴ Such an exclusivity may seem to have been implied by the analogy with burning at 417a6–9.Footnote ⁵ One could also understand the preliminary distinctions at 417a9–14 as if the perceiving and perception ‘in capacity’ were capturing a pure capacity simply opposed to fulfilment. If that is correct, then the point of the so-called triple scheme is not so much about there being different kinds of fulfilment and different kinds of capacities; rather, the key point is about recognizing that there is a kind of capacity that is already a fulfilment. This matters not just because it is an interesting ontological category, but because Aristotle’s inquiry cannot move forward until we understand that the perceptive capacity is itself already a fulfilment of that which can perceive.Footnote ⁶

This will be crucial for overcoming the way in which being affected and being active were identified at 417a14–17. This identification could only be provisional because the activity of perceiving is not incomplete, and so is not a change in terms of Phys. 3.1–2: it is complete in the sense of passing (at least a weakened form of) the tense test.Footnote ⁷ In the relevant passage from Phys. 3.2 (quoted almost verbatim at 417a16–17), Aristotle goes on to specify the reason why change is only an incomplete activity:

If the perceptive capacity is to be classed with capacities that are fulfilments – that is, complete capacities – then the provisional identification of perception with a change will be undermined.Footnote ⁸

When introducing the notion of a capacity that is already a fulfilment, Aristotle captures a feature that has been, since Plato, taken to characterize a certain class of human capacities. These capacities are usually those acquired by learning and typically arts, such as medicine, housebuilding, or carpentry. Aristotle’s primary intention is, apparently, to extend this class of capacities to include perception (as suggested most clearly at 417b16–18 and b29–32). The art itself lacks nothing, and its exercise does not aim at its own (or the artisan’s own) improvement or good; rather, it aims at the improvement or good of its object.Footnote ⁹ It is, in other words, complete, and its exercise is not a change in the artisan (rather, it is usually productive of a change in the object). Such a capacity, which is already a sort of fulfilment, must surely be first acquired, and the acquisition undoubtedly is a change in the learner; but once the capacity has been acquired, its employment is no longer a change.Footnote ¹⁰ As Aristotle puts it at An. 2.4, 416a34–b3: a carpenter is not affected qua carpenter in any way; what he undergoes is only a transition from inactivity to activity.Footnote ¹¹ This also implies that the exercise of such a capacity cannot exhaust it – unlike other capacities, such as the capacity to heat something. The exercise of the heating capacity effectively exhausts it because (1) what heats can, owing to generic likeness, be itself cooled by what it acts on, and because (2) it standardly is cooled, owing to the fact that the conditions of A heating B coincide with the conditions of B cooling A.Footnote ¹²

At 417a21–b2, Aristotle further develops this point from An. 2.4, particularly in the last sentence where the capacity that is a fulfilment, on the one hand, and the capacity that is not, on the other hand, are contrasted with respect to how each is brought to fulfilment:

The transition to fulfilment, characteristic of artisans, and of knowledgeable persons in general, is here contrasted with another kind of transition to fulfilment characteristic of those who have not yet acquired the knowledge in question, but who are able do so. Clearly, neither of the two fulfilments considered at 417a21–b2 is itself a change; if there is a candidate for being a change, it is the first kind of transition to fulfilment (i.e. learning). However, the present passage also already suggests a contrast between learning and processes like burning. What makes learning special is exactly that it is the transition to a fulfilment that is a capacity. Learning can, accordingly, be described as a change consisting exactly in overcoming the definitional incompleteness of changes; it is a change through which the subject (τὸ δυνάμει ὄν or τὸ δυνατόν) is fulfilled or completed so as to become capable of being active in a non-kinetic, complete way. It is thus no surprise that Aristotle returns to learning at 417b12–16 and suggests that it is a case of ‘being affected’ and ‘being altered’ only in a very special sense. In addition, it is no surprise that, at 417b16–18 (and b29–32), he goes on to claim that this kind of transition has already taken place for the perceptive capacity of a new-born animal: the perceptive capacity is, already upon birth, a fulfilment of its subject (i.e. a complete capacity). This, I argue, is the key claim for which the distinctions at 417a21–b2 serve to prepare us.Footnote ¹⁵

Now, if Aristotle’s aim was simply to confront the assimilation model of perceiving as a kind of being affected by and assimilated to perceptual objects with the perspective on perception as an activity of something complete, in a sort of antinomy, he could stop here. But this is not what he does. Rather, in what follows, he turns to the notion of being affected and draws an additional set of distinctions pertaining to it (as well as to the related notion of being altered). His aim, I shall argue, is to show what the latter perspective (i.e. perceiving as a complete activity) tells us about the former (i.e. perceiving as being affected and assimilated), and how the former needs to be transformed, if it is to be reconciled with the latter.Footnote ¹⁶

3.3 Kinds of Passivity

Aristotle’s distinction between two kinds of being affected at 417b2–5 has often been understood as further developing the preceding contrast between two kinds of transition (417a30–b2).Footnote ¹⁷ The idea has been that preservative πάσχειν simply overlaps with the transition from not being active to being active. However, there are at least two reasons for resisting this reading. First, the notion of preservative πάσχειν seems to be reapplied at 417b12–16 to the first kind of transition (such as learning).Footnote ¹⁸ Second, at 417b2–12, the notion of preservative πάσχειν does not seem to overlap with the idea of transition from not being active to being active. Rather, it contains the idea that something else is the agent (so it cannot apply to productive activities, such as building a house); moreover, it appears to be intended to capture the way in which something is affected throughout the entire time of being active, or so I shall argue. This last point will become important: if it is correct, then the passage turns out to bear directly on the phenomenon of continued perceiving (as introduced in Section 1.3) and the object-directedness of perception (emphasized in Sections 2.4, 2.5, and 3.1), on the question of how the truth of LAL (as interpreted in Section 2.5) can be distilled from traditional errors, and so, finally, on how Aristotle can solve the key puzzle about knowing (as analysed in Sections 1.4 and 2.7).

Here is the relevant passage again, this time supplied with (admittedly excessive) clarifications indicating how I think the text is best construed and understood:

In Section 1.3, I suggested, against the deflationary reading, that we should understand the preposition ὑπό as introducing the efficient cause rather than the terminus ad quem. Moreover, against both the existing non-deflationary readings, I suggested that we should take the likeness in question to be a full-fledged likeness (rather than a potential or generic likeness) resulting from being affected. The upshot is that τὸ δυνάμει ὄν is preserved by the agent precisely in its being assimilated to it. That is: being assimilated to the agent is, for τὸ δυνάμει ὄν, a way of being preserved as what it is. In what follows, I fill in the details of this account, particularly concerning the construal of (iii) and (iv).

Let us begin with a question about (iii) – that is, καί ὁμοίου. Is Aristotle adding a piece of information here about τὸ δυνάμει ὄν, about τὸ ἐντελεχείᾳ ὄν, or about σωτηρία?Footnote ¹⁹ Perhaps it is best to say that the new feature is meant to concern all three of the terms: it is a likeness between the first two that tells us what kind of σωτηρία defines the non-destructive πάσχειν. The mention of likeness thus fits well into what we can call the full grammar of preservation: something is preserved by the agency of something else having a certain effect on the former. Not all cases of preservation, to be sure, need to involve all three of these aspects. In many cases, X will simply be preserved untouched by Y (as when Y wards off some influence on X) and this seems to be the most usual sense of preservation. It is in this sense, for instance, that the inert atmosphere will ‘preserve’ what is in it, or my head will be preserved from the sword’s blow by my shield.Footnote ²⁰ But, in some cases, Y preserves X while having a certain effect on X that needs to be further specified. In this way, for instance, Aristotle says that the tyranny is preserved when being made more like the rule of a king.Footnote ²¹ Similarly, I suggest that καὶ ὁμοίου in our passage adds a third characteristic of the preservation with which the second kind of πάσχειν is identified: τὸ ἐντελεχείᾳ ὄν preserves τὸ δυνάμει ὄν exactly when making it like itself.

It is not difficult to see why the third characteristic needs to be spelled out. If ὑπό introduces the efficient cause, Aristotle is clearly saying that the second meaning of πάσχειν is a σωτηρία (rather than a destruction) of X by Y. But this, on its own, could well mean that X is preserved by Y entirely untouched, just as X would be preserved without encountering anything at all (as my shield preserves my head from the sword’s blow). This clearly cannot be the meaning of σωτηρία in our passage if it is true that the πάσχειν qua σωτηρία is intended to capture how an activity (such as the activity of perceiving) is produced by an agent (such as the perceptual object) in something (such as the perceiver) in which it was not before. Accordingly, saying that τὸ δυνάμει ὄν is preserved in the sense of remaining entirely untouched (in line with Anaxagoras’ ἀπάθεια) and just as it was before would be to no avail, because this would not be an instance of πάσχειν whatsoever.

As a result, rather than being preserved entirely untouched, we can see that τὸ δυνάμει ὄν is preserved in receiving the fulfilment that existed previously only in τὸ ἐντελεχείᾳ ὄν that now acts upon it. Thus, it is preserved by τὸ ἐντελεχείᾳ ὄν in being made like it.Footnote ²² Through this third characteristic, the σωτηρία already acquires the specific meaning requisite for capturing the receptive nature of perception: τὸ δυνάμει ὄν is preserved by τὸ ἐντελεχείᾳ ὄν in the very act of being assimilated to it.

But what does Aristotle mean then when he adds (iv) οὕτως ὡς δύναμις ἔχει πρὸς ἐντελέχειαν? It could be explaining the sense in which τὸ δυνάμει ὄν is made like τὸ ἐντελεχείᾳ ὄν: the former is assimilated to the latter in the sense in which a δύναμις is turned into the corresponding ἐντελέχεια. The idea here would apparently be that the result of the acting of τὸ ἐντελεχείᾳ ὄν on τὸ δυνάμει ὄν – characterized in terms of the latter being like the former, that is, being like something else – is in fact the latter’s proper fulfilment (ἐντελέχεια).Footnote ²³ Alternatively, one could take (iv) to depend on σωτηρία. On this reading, Aristotle would be further specifying the notion of preservation. Not only is it a preservation of τὸ δυνάμει ὄν in the very act of its being assimilated to τὸ ἐντελεχείᾳ ὄν; rather than just being preserved as it was, τὸ δυνάμει ὄν is fulfilled as what it ought to be: in being assimilated to something else it reaches its intrinsic goal.

Be that as it may, the contrast with the ‘destructive’ πάσχειν should by now be sufficiently clear. What undergoes this πάσχειν is not preserved, and a fortiori not fulfilled, as what it was; rather, it becomes something different, contrary to what it was before, and this modifies how it can now be affected. Although Aristotle does not explicitly mention perception at 417b2–5, his primary aim in the wider context of An. 2.5 is, apparently, to stress precisely the fact that perception cannot consist in a destructive πάσχειν – not least because if that which can perceive was affected in this way it would become like the perceptual object in a sense that would diminish (or destroy) its ability to engage in perception. It would be brought ‘closer’ to some objects of the given range and ‘distanced’ from other ones. Moreover, it would be prevented from being further affected by the object to which it has been assimilated: in this way, like could not be affected by like, and so no continued perceiving, and indeed no perception at all, could take place. In this respect, that which can perceive must be impassive: resistant to ‘destructive’ πάσχειν.

The hope seems to be that by classifying perception as a case of ‘preservative’ πάσχειν we can escape – provided that the notion can successfully be fleshed out – both kinds of difficulties raised by Aristotle against his predecessors. The proposed solution appears to be that (a) that which can perceive is not absolutely impassive, as Anaxagoras’ νοῦς, but rather can be affected by and assimilated to the perceptual object acting on it; however, (b) it is assimilated to the object in such a way that its perceptivity is preserved untouched (unbiased), so that it is capable of being further affected by the same external object and thus of continued perceiving. Perceiving is brought about by way of assimilation, but in such a way that it in no way diminishes the perceptivity of the subject.Footnote ²⁴

3.4 Kinds of Alteration

At 417b5–9, further developing the notion of preservative πάσχειν,Footnote ²⁵ Aristotle returns to the model of the knower and explains exactly in what sense the notion of alteration does, and in what sense it does not, apply to the ‘second fulfilment’ of the knower (i.e. the activity of ‘thinking’ or ‘understanding’) and the transition to it. Although Aristotle does not say it in so many words, he seems to be taking for granted here what he will explicitly assert in An. 3.4 (429a13–18, 429b22–6) – namely, that thinking (νοεῖν, or here θεωρεῖν) is, like perceiving, a case of πάσχειν (to wit, preservative πάσχειν).Footnote ²⁶ Aristotle begins from the ‘coming to be’ involved in the transition from having knowledge to thinking (and mutatis mutandis from having perception to perceiving):

The move from preservative πάσχειν to alteration should not surprise us, as we have seen that the notion of preservative πάσχειν essentially involves a notion of assimilation (wherein τὸ δυνάμει ὄν is made like τὸ ἐντελεχείᾳ ὄν), and Aristotle is now bringing this aspect into focus. He does so by formulating a disjunction: when that which has knowledge comes to be actively theorizing, this coming to be (picked up by ὅπερ) is either not a case of being altered, or, if it is to be understood as being altered, the sort of alteration at issue here is of an entirely different kind.Footnote ²⁷

Aristotle’s two disjuncts are unlikely to represent two different options between which we are supposed to decide. Rather, they seem to articulate two different ways of using the word ‘alteration’. We already know that the assimilation involved in preservative πάσχειν cannot be a change between contrary states. Instead, in being assimilated to τὸ ἐντελεχείᾳ ὄν – that is, to something else – τὸ δυνάμει ὄν is fulfilled as what it itself is, without becoming different (i.e. without becoming, for instance, more ‘distant’ from some objects and ‘closer’ to others). The first disjunct is further reinforcing this point by developing the notion that the ‘advance is into itself’ – that is, being assimilated to something else (i.e. τὸ ἐντελεχείᾳ ὄν) is for τὸ δυνάμει ὄν (exemplified now by that which has knowledge) an advance into itself;Footnote ²⁸ and this cannot be an alteration in the standard sense of becoming different. It is, however, essential for the preservative πάσχειν that, in contrast to the case of productive arts, the advance into itself consists here exactly in being assimilated to something else. Hence the need for the second disjunct. We can either say that there is no alteration at all (if we retain the standard notion of alteration), or we can claim that there is a different genus of alteration. The former option would be insufficient for Aristotle’s purposes, but the latter is yet to be developed.Footnote ²⁹

Aristotle’s present aim seems limited to making conceptual room for a kind of assimilation to something else that is an advance into itself and thus needs to be contrasted with the standard notion of alteration as becoming different. That this is, in fact, the case is supported by the reoccurrence of a very similar disjunction at 417b12–16 (further discussed in Section 3.6), and again at the outset of An. 3.4: at 429a14–15, thinking (νοεῖν) is said to be, like perceiving, ‘either a kind of being affected or something else like that’. The fact that, later in An. 3.4, Aristotle characterizes thinking simply as ‘a kind of being affected’ strongly suggests that he does not intend to exclude the first disjunct from consideration.Footnote ³⁰ Rather, the disjunction is aimed at contrasting the intended notion of being affected with ‘destructive’ πάσχειν: were we to stick to this standard notion of being affected, we would have to say that perception and thinking are ‘something else’ and that this is only ‘like’ being affected. However, building on the distinction between different kinds of πάσχειν at 417b2–5, Aristotle can take the bolder line and maintain that thinking is, after all, a kind of being affected. The disjunction at 417b5–7 seems to be making the very same point with respect to ‘being altered’.Footnote ³¹

Accordingly, what is progressively being shown in An. 2.5 is why the meanings of the two key notions must be extended beyond the ways in which we typically understand them as well as how this can be achieved. This task turns out to be essential for capturing the receptive nature of perception (and thinking) as a way of taking things in.Footnote ³² Towards the end of An. 2.5, Aristotle seems to announce that this extension has successfully been performed. He first reminds us that perceivers are endowed with a complete capacity (417b29–32, cf. 417b16–18) and then he adds an explicit reflection on the usage of the two key notions:

By saying that it was determined how the two kinds of capacity differ, Aristotle is apparently referring to 417a21–b12 as a whole, as is suggested by the consequent (ii). Aristotle can legitimately be taking 417b2–12 as a part of determining how the two kinds of capacity differ, because that which can perceive is characterized by a complete passive capacity and, from 417a21–b2, it is still far from obvious that a complete capacity can be passive. This crucial finding is only established at 417b2–12. Accordingly, when Aristotle says that the distinguishing feature of each kind of capacity is nameless, this also involves the recognition that there is a lack of specific terms for ‘preservative’ πάσχειν (as contrasted with ‘destructive’ πάσχειν) and for alteration as an advance into itself (rather than as a way of becoming different).

With this reading of the antecedent (i), it becomes much clearer how the consequent (ii) is meant to follow. The consequent contends that, since there are, regrettably, no specific terms, but it is impossible to define perception without the notions of being affected and being altered, it is necessary that we use the only available terms – namely, the generic ones. Moreover, as long as we keep in mind the fact that their meaning has been extended at 417b2–12 beyond the ways in which they are typically understood, we shall be using these terms as proper terms rather than, say, mere metaphors.Footnote ³³ Armed with these distinctions (that is, effectively, with the διορισμός for which Theophrastus calls in Sens. 19),Footnote ³⁴ we shall finally be in a position to understand how ‘perception comes about in being changed and being affected’ and in what sense it can (and cannot) be analysed as ‘an alteration’ (416b32–4). The closing lines (418a3–6) can therefore, finally, fulfil the promise made at the beginning of the chapter.

3.5 Non-Altering Passive Activity

With the aforementioned disjunction in hand, Aristotle proceeds to say more about alteration at 417b8–9. These lines have often been read as supporting the idea that he intended to reject the second disjunct at 417b5–7, and indirectly also as supporting the deflationary reading of preservative πάσχειν:

If a thinker is no more altered when thinking than a housebuilder is when building a house, then it may seem that the notion of alteration is not applicable to thinking at all (for, as we know, in Aristotle’s view a housebuilder undergoes no alteration whatsoever, not even in any refined sense). Moreover, if Aristotle’s mention of housebuilding here is understood as somehow falling under the notion of preservative πάσχειν, then this notion can, apparently, have no other meaning than the deflationary one of a transition from a complete capacity to its fulfilment – for it is only in this way that it can be applied to a productive activity such as housebuilding. I shall argue that the quoted passage does not support either of these interpretative moves. In order to understand it correctly, we need, first of all, to appreciate the perspective from which housebuilding and thinking are being compared here.

This perspective, I contend, is determined by the specific temporal aspect of thinking under consideration in these lines. Aristotle does not seem to be talking about the transition from not thinking to thinking but, rather, about the continued activity of thinking, as is strongly suggested by the repetition of ὅταν with the subjunctive.Footnote ³⁵ If this is correct, then we can understand why it proved so difficult to decide whether the distinctions at 417b2–9 concern transitions or activities.Footnote ³⁶ The reason is that Aristotle speaks of both. Although the point of 417b5–7 apparently concerns the transition to theorizing, the point of 417b8–9 turns out to be rather about the continued activity of thinking. Understanding the reasoning behind this change of perspective seems crucial for successfully coming to grips with Aristotle’s argument at 417b2–9 as a whole.

Let me begin by adding a caveat concerning the standard view that, at 417b5–7, Aristotle talks about the transition to theorizing rather than theorizing itself.Footnote ³⁷ If the proposed interpretation of 417b2–7 is correct and Aristotle defines preservative πάσχειν here as a way of being preserved as what one is in the very act of being assimilated to something else, then this puts the notion of transition being described here in a novel perspective. We should not expect that Aristotle’s account at 417b2–7 is limited to the first moment of the activity in question (say, the opening of one’s eyes); rather, what he describes seems to be taking place throughout the duration of the subject’s being involved in the perceptual activity. This would confirm that being affected and being active cannot be dissociated in any temporal way (as flagged in Section 1.3); but it does not imply that the one cannot be conceptually distinguished from the other, and it provides even less reason to think that Aristotle ‘merges’ ‘the (instantaneous) transition to perceiving and perceiving [itself]’, as Burnyeat would have it.Footnote ³⁸

It thus seems fair to say that, throughout 417b2–7, Aristotle focuses on the transition to activity – so long as we remain open to the idea that τὸ δυνάμει ὄν is making this transition incessantly as long as the activity is going on. At 417b8–9, however, Aristotle switches to a different perspective, concentrating on thinking (and mutatis mutandis perceiving) as an activity exercised continuously over a certain period of time. One could say that, while at 417b2–7 Aristotle is developing the notion of preservative πάσχειν, 417b8–9 is, strictly speaking, no longer about being affected but, rather, about being active: the preservative πάσχειν and the respective ἐνεργεῖν are numerically the same but differ in being and, at 417b2–9, Aristotle effectively captures their difference in being. At 417b5–7, he focuses on coming to be actively theorizing, whereas at 417b8–9 he discusses the activity of thinking as such – and these are clearly not the same thing. As soon as we realize that there is this switch of perspective, we cease to be confronted with the problem that most interpreters have found here – namely, that Aristotle seems to be first admitting (in the second disjunct of 417b5–7) that X is a kind of alteration and then plainly denying (at 417b8–9) that X can be any alteration whatsoever, while, moreover, suggesting that the latter can be somehow inferred from the former.Footnote ³⁹ The inference starts to make much better sense once we take the switch of perspective into account.

In comparison with the content of 417b2–7, Aristotle’s claim at 417b8–9 is, in fact, not that striking. In a way, he is simply spelling out what we already know from 417a21–b2: the fulfilment of a second capacity is not a change. The art of housebuilding is a case in point: it produces changes in the material, but the housebuilder qua housebuilder does not undergo any change; housebuilding does not mean for her becoming different. I argued (in Section 3.2) that Aristotle’s primary aim at 417a21–b2 is to classify the perceptive capacity as being complete. But this seems prima facie difficult given that the perceptive capacity is passive rather than productive. What happens at 417b2–9, as I see it, is roughly this: (a) Aristotle explains at 417b2–7 what kind of passivity is involved; and (b) he insists at 417b8–9 that this kind of passivity is compatible with the capacity in question being complete like the art of housebuilding.

How exactly does it follow, then, from the notion of preservative πάσχειν developed at 417b2–7 that the activity in question is not itself an alteration? Let us recall that what is affected in the preservative way is, in its very assimilation to the agent, preserved and, indeed, fulfilled as what it itself already is; and that this involves no change to a contrary state. While, say, the perceiver perceiving a certain object is incessantly being affected by and assimilated to this object, there is no progressive changing, no acquiring of the quality as a quality of its own. If the patient were to acquire the quality of the agent as a quality of its own (so that it could retain the quality even when the acting is over), it would thereby lose its ability to be further affected by that agent. It can, therefore, have the quality in itself only dynamically, so to speak, and only for as long as the acting – and the activity – is going on. Once the acting (and so the activity) is over, there is no lasting result.Footnote ⁴⁰ Accordingly, it is legitimate to infer that engaging in the activity of perceiving does not mean that the perceiver becomes different – any more than building a house means becoming different for a housebuilder.

Let me conclude this section by commenting on two different ways of contrasting changes with ‘complete’ activities.Footnote ⁴¹ If the proposed reading of 417b2–9 is on the right track, then at 417b8–9 thinking (and mutatis mutandis perceiving) is concluded to be a complete activity in a somewhat different sense than it is claimed to be a complete activity in Metaph. Θ.6 and EN 10.4. Thinking and perceiving are complete activities because their subjects are complete just as the subject of housebuilding is complete – that is: they are understood as being complete activities by the subject criterion, rather than the telic structure criterion, as Andreas Anagnostopoulos calls them.Footnote ⁴² As shown by the example of housebuilding, something can well qualify as a complete activity by the subject criterion, while failing by the telic structure criterion (i.e. by not passing the tense test). At the same time, it has now been confirmed that nothing can pass the tense test without satisfying the subject criterion. On similar grounds, Anagnostopoulos calls the subject criterion ‘more precise’ and ‘more fundamental’.Footnote ⁴³

I agree that this is the case as far as we are concerned with drawing the general line between changes and non-kinetic activities.Footnote ⁴⁴ However, I am less sure that the same assessment also holds when we are concerned with the specific nature of perception (and thinking), as Aristotle is in An. 2.5. Here, the subject criterion is – in an important sense – not sufficiently precise, for it does not allow us to distinguish passive non-kinetic activities from other non-kinetic activities. Furthermore, it is also questionable whether the subject criterion is more fundamental in this context. When the perceptive capacity is simply claimed to be a complete capacity, it is far from obvious how this is compatible with perception being essentially passive. Without the distinctions provided by 417b2–9, it would seem that we are only confronted with an antinomy between perception’s completeness (on the basis of the subject criterion) and its passivity.Footnote ⁴⁵ However, as I have argued, Aristotle does not simply want to confront us with this kind of antinomy. He also wants to indicate how the passivity of perception needs to be conceived if the antinomy is to be overcome. He does so by introducing the notion of preservative πάσχειν (implying alteration qua advance into itself) and by showing how this kind of passivity is compatible with completeness in the sense of the subject criterion. Now, the notion of preservative πάσχειν has appeared to be essentially informed by the telic structure criterion: it is exactly such a kind of being affected that can pass the tense test (and thereby incorporate the truth of LAL). If this is correct, it means that the idea of perception as an activity that is complete according to the subject criterion can certainly be conceived independently from the telic structure criterion. If this is done, though, it can figure only as one horn of an antinomy, for this obviates the question of perception’s passivity. It is only with the aid of the telic structure criterion, extended from ‘being active’ to ‘being affected’, I submit, that activities like perception can be shown to be complete while being passive – that is, to be complete passive activities.

Be that as it may, the crucial point can be made independently from whether the subject criterion is more precise and more fundamental in the case of activities such as perception: the two criteria point jointly, from two complementary perspectives, to what seems to be Aristotle’s main claim in An. 2.5 – namely that there are complete passive activities, distinct from both passive changes and non-passive complete activities, and that perception is one of them.

3.6 Activity and Acquisition

At 417b9–16, Aristotle steps back from complete passive capacities to provide a comparison with another kind of passive capacities that are not complete, although the way they are fulfilled can be contrasted with destructive πάσχειν, too. In the first sentence, Aristotle contrasts activities such as thinking or perceiving with the process of being taught:

Aristotle turns our attention to what he calls ‘that which leads into fulfilment’ (τὸ εἰς ἐντελέχειαν ἄγον). The referent of this phrase can be understood in two ways: either as picking up on ‘that which is in fulfilment’ from 417b4 (i.e. as the agent of the activities in question that leads τὸ δυνάμει ὄν to fulfilment by assimilating it to itself and so producing this fulfilment in it)Footnote ⁴⁷ or as referring to the acting of this agent on the patient.Footnote ⁴⁸ However we understand τὸ ἄγον, what Aristotle is saying must be that the name of the agent should not be derived from ‘teaching’ or that its agency must not be called ‘teaching’ – apparently because what it does is essentially different from teaching.Footnote ⁴⁹ The problem seems to be that, in contrast to the case of teaching, as the ποίησις corresponding to learning, there is no proper name for the ποίησις that corresponds to ‘knowledgeable thinking’ (νοεῖν καὶ φρονεῖν) or to ‘perceiving’ (αἰσθάνεσθαι).Footnote ⁵⁰ It must be emphasized, nonetheless, that this is an essentially different kind of ποίησις.

Aristotle presents this point as something following on from, or summarizing, what has been said so far,Footnote ⁵¹ and we are now in a position to understand what he means. He has already distinguished at 417a30–b2 between learning and the transition to thinking, stressing that the former involves an alteration in the sense of being brought into a different lasting state. This contrasts sharply with what we have just learnt about passive activities like thinking (or perceiving), which have no such lasting result. Accordingly, although we have no name for the respective ποίησις, we must insist that it is something essentially different from teaching.

However, this leaves the status of learning somewhat unclear. On the one hand, it can be sharply contrasted with the case of perceiving and thinking. But, on the other hand, it is hard to resist the impression that learning cannot be simply identified as a case of destructive πάσχειν.Footnote ⁵² The following sentence seems to be focused on addressing exactly this concern, and it can be understood as subsuming learning under an expanded category of preservative πάσχειν (despite its difference from complete passive activities, which has just been stressed):

Aristotle’s aim seems to be twofold: he is, on the one hand, resisting the identification of learning with destructive πάσχειν, while, on the other hand, emphasizing its difference from the preservative πάσχειν as it was introduced at 417b2–9. Learning does not mean transitioning to a privative state: it is not a case of ‘destruction’. Rather, Aristotle says something very similar about learning to what he has said about preservative πάσχειν at 417b5–7 – namely, that it leads to a positive disposition (ἕξις) in which, apparently, the very nature of the subject finds its fulfilment.Footnote ⁵⁴ However, this very characterization also contrasts with how preservative πάσχειν was analysed at 417b2–9: learning involves acquiring a new quality of one’s own as a lasting result, whereas perceiving and thinking emphatically do not involve the perceiver or thinker acquiring any quality of their own as a lasting result.Footnote ⁵⁵

In other words, learning is not quite as preservative as perceiving and thinking (I can indefinitely and repeatedly perceive the same thing, but once I have learnt something, I cannot learn it again, unless I have forgotten it). However, learning still possesses a preservative feature in the sense that τὸ δυνάμει ὄν is in no way harmed (or ‘destroyed’) by it – rather, its very nature is fulfilled through being affected. Indeed, this is exactly how a complete capacity is acquired: the result of learning is a completion or perfection (τελείωσις) that enables the subject to engage in a complete activity (whether this activity is passive or not).Footnote ⁵⁶ This explains what Aristotle means when, in the immediately following lines (417b16–18), he compares the acquisition of the perceptive capacity in embryogenesis with learning.

3.7 A Programmatic Definition of Perception

The notion of preservative πάσχειν can, accordingly, be conveniently expanded to capture also what happens in learning. However, the comparison with embryogenesis turns our attention back to what has been the main goal behind introducing preservative πάσχειν, and behind An. 2.5 as a whole – namely, capturing, on the most general level, the receptive nature of perception as a way of taking things in. We have seen that the notion of preservative πάσχειν can serve this purpose only once it has been successfully distinguished from both passive changes and non-passive complete activities. Perception itself is a complete passive activity. Establishing this claim is the key result of An. 2.5.Footnote ⁵⁷

We have also seen what role this notion plays in Aristotle’s engagement with his predecessors. Introducing the notion of preservative πάσχειν would be to no avail if Aristotle characterized it in terms of a generic likeness presupposed by any πάσχειν whatsoever or if he deflated it of any genuine passivity. The notion of complete passive activities, in contrast, does exactly what is needed to resolve the main dilemma about knowing raised in An. 1.2.

By introducing this concept, Aristotle is able to capture the sense in which like perceives like, as well as the way in which like is affected by like in perception, while overcoming the difficulties of the traditional LKL position (as exposed, once more, in the puzzle at 417a2–9). Furthermore, this allows him to incorporate the truthful aspects of LKL without having to rely on the provisional identification of ‘being affected’ and ‘being active’ from 417a14–17 – rather, it frees him to show how an activity can be passive without being incomplete. Finally, it prepares us for the closing passage of the chapter (418a3–6), which applies the assimilation model explicitly to perception. The notion of complete passive activities provides the ground for understanding the present and the perfect aspect of being affected in a temporally non-exclusive way. This is possible only as a result of the fact that, for the entire duration in which the subject is being affected by and assimilated to the relevant object, the subject is preserved. In other words, the subject is impassive in the sense of never being turned into a contrary state by the object, and thereby never acquiring a new quality of its own. If the subject is a perceiver, this is how she succeeds in retaining her untainted ability to perceive. By developing the notion of preservative πάσχειν, Aristotle thus allows two seemingly incompatible intuitions about knowing (i.e. the Acquaintance/Contact Requirement and the Unbiasedness/Externality Requirement) to be successfully incorporated into his first general account of perception. On this basis, Aristotle’s account finally shows how we might arrive at grasping the whole that his predecessors managed to glimpse only in isolated parts.

We still require, however, a better understanding of the notion of impassivity (ἀπάθεια) at stake here. It clearly cannot be the absolute impassivity claimed by Anaxagoras for νοῦς as excluding any πάσχειν whatsoever. Such an impassivity would make any account of how thinking, and a fortiori perceiving, comes about impossible (as Aristotle points out at An. 1.2, 405b21–3 and again at 3.4, 429b22–6). However, it does not seem to be the abstract kind of impassivity that Aristotle refers to in An. 1.4 (408b18–30), either. Here, he makes a point, akin to the argument of Metaph. Z.8 (1033b3–1034a8), that the soul itself (whether it is the perceptive or the rational soul) cannot qua form be what, properly speaking, ‘suffers’ when we become incapacitated owing to malady or old age. While the former (Anaxagorean) impassivity seems too strong, the latter seems too weak, because it considers any form whatsoever (including forms of things undergoing destructive πάσχειν and including productive arts) to be impassive.Footnote ⁵⁸

To appreciate more fully just how the notion of preservative πάσχειν can be spelled out in terms of impassivity and thus be presented as overcoming the conflict between LKL and Anaxagoras’ view, it is helpful to recall the way in which Aristotle defines impassivity in Metaph. Δ.12 and Θ.1.Footnote ⁵⁹ Impassivity is not simply a privation or the incapability of being affected; it is rather a power for resisting a certain kind or range of acting, particularly if such an acting would mean for the patient changing towards something worse. This kind of resistance can be a necessary condition for acting in some way, but apparently – and here we return to the main point of An. 2.5 – it can also be a necessary condition for being affected in some way, particularly if the kind of being affected in question is such as to become the patient’s own fulfilment.Footnote ⁶⁰

This, I submit, is exactly how perception is to be understood according to Aristotle and Theophrastus. It presupposes an ἀπάθεια that is the ‘preservative’ aspect of the perceptive πάσχειν seen, as it were, from the other side. That which can perceive is able to resist any change towards a contrary state (i.e. to resist acquiring a colour or an odour as a quality of its own) that would diminish or destroy its perceptivity. This impassivity is what allows the perceiver to be affected by, and assimilated to, the perceptual object in the way that for her means her own fulfilment qua perceiver while preserving her ability to perceive in an untainted manner. The quality received by the perceiver must remain, as we shall see, a quality of the external perceptual object. This, I suggest, is the kind of impassivity that Aristotle feels entitled to infer from the notion of preservative πάσχειν at An. 3.4, 429a14–15.Footnote ⁶¹ Moreover, this aspect of preservative πάσχειν provides a key to understanding why Aristotle appears to oscillate between classifying perception as a case of being affected, as a quality, and as a relatum: perception is exactly such a kind of being affected that the perceiver is, throughout the entire time, like the perceptual object – in a way that makes perception a relatum of it.

If this proposed reconstruction of Aristotle’s first general account of perception is on the right track, it ought to be stressed that An. 2.5 itself leaves many questions open, despite the fact that it has often been taken to be the ultimate source for their answers.Footnote ⁶²

First, it does not contain any definitive statement on whether perception involves a literal assimilation of the sensory organs, and ‘ordinary’ changes in general, or not. The straightforward literalist reading of An. 2.5, which would identify preservative πάσχειν and ἀλλοίωσις at 417b2–9 with physiological processes involved in perception, seems misguided, as Aristotle’s aim here is to capture the very essence of perception, making the stakes much higher. However, it also seems mistaken to take this fact as implying that, according to Aristotle, perception involves no ‘ordinary’ changes whatsoever. There is nothing in An. 2.5 to exclude literal assimilation from being somehow involved in perception – and, indeed, it is hard to resist the impression that at least perceiving temperatures does, according to Aristotle, somehow involve an instance of literal cooling or heating of the perceiver’s body. Yet, An. 2.5 also seems to imply important constraints on how exactly this might be the case.

It is, first of all, hardly possible for such a literal, and hence ‘destructive’, heating or cooling to constitute the essence of perception. Moreover, such a literal cooling or heating can be, at most, local: if I were in toto assimilated to the object cooling or heating my body, then I would obviously lose my ability to be affected by, and so perceive, it.Footnote ⁶³ Moreover, even on a local level, additional work must be done to explain just how a literal assimilation is compatible with the requisite unlikeness that allows the perceiver to continue being affected by the perceptual object in question and, thereby, to perceive it. This, to be sure, raises a difficulty not only for literalism, but also, as we shall see, for the idea that, for the duration of perceiving, the sense organ is like the perceptual object in the sense of embodying the same ratio in a different pair of contraries.Footnote ⁶⁴ What needs to be explained is how the perceiver can be assimilated to the perceptual object while retaining the unlikeness requisite for being further affected by the same or any other object of the given range. This question is yet to be explored. For now, it is sufficient to stress that spiritualism does not provide a viable alternative: if the only thing that counts in perception is a spiritual change, then it is equally unclear how something that has already been ‘spiritually’ assimilated to the perceptual object can continue being ‘spiritually’ affected by it.

A related point concerns the ‘Democritean’ generic likeness. If the proposed reconstruction points in the right direction, then An. 2.5 says nothing about this sort of likeness. That does not mean, however, that generic likeness plays no role in Aristotle’s account of perception. Rather, in An. 2.7–11, Aristotle clearly takes it on board. Each perceptive organ needs to be a specific instantiation of the genus under which the perceptual objects accessible through it fall: it needs to be exactly in the middle between the two extremes. This notion of the mean (μεσότης, introduced in An. 2.11), which has directly to do with the ‘Democritean’ generic likeness, will be crucial for how the notion of preservative πάσχειν should be fleshed out.Footnote ⁶⁵

All of this, finally, relates to the question of how exactly the perceptive soul, as the primary cause of perception, is involved in this activity – a question that has received much less attention than the one concerning the role the body plays in perception, despite the fact that it is no less tricky and no less central for understanding Aristotle’s account. An. 2.5 does not contain any clear statement on this point (any more than An. 2.4, 416a21–b11 contains any clear statement about the role of the nutritive soul).Footnote ⁶⁶ But we shall see that this question becomes crucial for the task of fleshing out the notion of complete passive activities in An. 2.6–3.2.Footnote ⁶⁷ It contains a dilemma that comes to the fore as soon as the result of An. 2.5 is confronted with Aristotle’s argument in An. 1.3–4 against the idea that the soul could itself be undergoing a change. This dilemma will be worked out in Chapters 4 and 5 and addressed in Chapters 6 and 7.

If the proposed reconstruction is on the right track, then An. 2.5 does not provide answers to any of these pressing questions. Aristotle’s first general account of perception is too preliminary to do that. It is pre-causal in the sense of leaving it open exactly how the respective roles of the soul and the body in perception are to be laid out. It captures, in most general terms, the nature of perception, but without saying anything about its causes. The chapter’s relation to the further discussion in An. 2.6–3.2 is rather like the relation of An. 2.1 to An. 2.2–3.11: in a sense, the most general account describes only what the inquiry proper should accomplish.Footnote ⁶⁸ However, it would be a misunderstanding if this led us to disparage the achievement of An. 2.5 (or, for that matter, An. 2.1). The range of open questions may appear daunting, but we have been left with a well-argued, novel, and attractive notion of what exactly we should be looking for in any satisfactory answers to these questions. And this is no small accomplishment.